Telegram @tubscstds

Telegram @tubscstds
Telegram @tubscstds
Telegram @tubscstds
studies

Telegram @tubscstds
 studies in

John Wiley & Sons, Inc., New York and London

Telegram @tubscstds
Henry N. Andrews, Jr.
Professor of Botany,
Washington University, St. Louis

Paleobotanist, Missouri Botanical Garden

with a chapter on palynology by

Charles J. Felix, Sun Oil Company

Telegram @tubscstds
 COPYRIGHT © 1961 by JOHN WILEY & SONS, Inc.

All rights reserved. This book or any part

thereof must not be reproduced in any form
without the written permission of the publisher.

Library of Congress Catalog Card Number: 61-6768

Printed in the United States of America

Telegram @tubscstds
n appreciation of the counsel
of three great botanists
HUGH HAMSHAW THOMAS
RAY ETHAN TORREY
ROBERT E. WOODSON, Jr.

Telegram @tubscstds
Telegram @tubscstds
 D n preparing this account of the vegetation of the past,
advice of several colleagues was sought as to the most
suitable content for an introductory textbook. I was advised to
write about the things that interest me and to present a story of
fossil plants that would be suitable for advanced botany students
as well as for geology classes in which the students have little
botanical background. These last two demands are not easy to
reconcile, at least in a comparatively small book. A certain amount
of material has been incorporated concerning living plants which
may be helpful for the nonbotanist, but it must be regarded as
introductory. The serious student, whether of geology or botany,
will find it necessary to refer occasionally to textbooks dealing more
fully with corresponding living plant groups.
Real progress in any field of learning depends on a clear under¬
standing of what is known as well as what is not known; there are
several pertinent remarks that may be made in this connection.
Perhaps the greatest contribution that paleobotany has made is not
in filling gaps in our knowledge of the evolution of the plant king¬
dom but in showing us how many gaps exist. The fossil record
reveals the former existence of major groups, such as the pterido-
sperms, that thrived for a time in the past and then became extinct;
fragmentary remains of many strange plants that are not satisfac¬
torily decipherable suggest unique groups about which we may one
day know more. There are many fossil plants in this category of
the problematical, and they are important in elucidating the com¬
plexity of plant evolution in a way that the extant flora cannot do.
This leads to the matter of classification, both in general and in
dealing with the problematical fossils. My procedure has been
based on the belief that a system of classification should serve to

Telegram @tubscstds
VII
VIII PREFACE

create some semblance of order which we must have in dealing with

any accumulation of facts and that it should demonstrate natural
relationships where they are evident or probable. However, the
introduction of an elaborate classification in an introductory text¬
book has the habit of establishing itself as fact in the students’
minds when actually it may be little more than a working hypothe¬
sis. There is a considerable ferment under way at present concern¬
ing the classification of vascular (woody) plants, and this is only
the beginning. Much of this has resulted from the discovery of new
types of plants—strange primitive ones that were among the first
inhabitants of the land, many distinctive fossils that are early rep¬
resentatives of pteridophytic lines, unique seed plants, and so on.
It is my feeling that the cause of plant evolutionary studies will be
more encumbered than aided by immediately confining such plants
in a rigid taxonomic system. Thus in several chapters I have
reserved a section for especially problematical but highly interest¬
ing plants; the entire Chapter 12 is so devoted. Finally, in the
matter of classification, the outlook here is a polyphyletic one;
the reasons for this are explained in the chapters dealing with the
various pteridophytic and gymnospermous groups.
A few comments may be offered to justify the order in which
the subject material is presented, although additional notations will
be found at appropriate points in the text. There is now some evi¬
dence for tentatively accepting an evolutionary sequence beginning
with the psilophytes and following through the coenopterid ferns,
pteridosperms, and angiosperms. There is good reason to indicate
that certain pteridophytic lines, such as the lycopods and articu¬
lates, do not fit into this general trend, and some gymnospermous
groups quite certainly could not have led to the flowering plants.
Thus I have exploited this knowledge in Chapters 2 to 7. It must
be remembered, however, that we are dealing here with a very
broad stream of evolution, and every effort has been made to avoid
a dogmatic presentation. If the student or teacher prefers to deal
first with all of the pteridophytic groups, and then the gymno¬
spermous ones, in a classical approach, I believe that it can be done
with little inconvenience.
Some aspects of paleobotany are most effectively handled by a
floristic or geographical approach. For example, there is a special
fascination about polar floras; here are violent contrasts that make
a strong appeal to human nature, and although fossil plants from
the Arctic zone are brought into many chapters it seemed to me of

Telegram @tubscstds
sufficient interest to compile a separate account on the topic. In
PREFACE
IX

Chapter 16 I have tried to review some of the more important late

Paleozoic and early Mesozoic floras of the Asiatic and southern
regions. Research in this general paleobotanical province is going
ahead very rapidly, and I am sure that textbooks of the future will
devote increasingly greater attention to it. Having had but little
opportunity to study these floras at first hand, I am aware of the
inadequacy of their treatment here.
There is some divergence of opinion among paleobotanists as to
whether a general textbook of this sort should attempt to encom¬
pass the rapidly expanding field of fossil pollen and spore studies.
It is clearly a branch of paleontology in its own right, and the
chapter devoted to it is intended only as a brief introduction. Since
so many paleobotanists are now engaged in palynological studies
exclusively or incidentally, and the field will attract many students
in the future, it seemed essential to include this separate chapter,
although numerous references to pollens and spores will be found
throughout the book. I am grateful to Dr. Charles J. Felix for his
willingness to prepare this account.
Finally, after some hesitation, I decided to include an outline of
certain of the more useful paleobotanical techniques. Great ad¬
vances have been made in recent decades in paleobotany as a result
of improved methods of extracting information from fossils. It is
my hope that this last chapter may serve to encourage younger
workers to add to our still scanty repertoire of study methods.
It is not possible to acknowledge adequately the many sources
of aid that have, over a period of about two decades, contributed
to the development of this book. Information has been drawn from
many places, most important of which is the literature created by
hundreds of paleobotanists over the last century and a half. Thus
reference citations given at the end of each chapter are intended as
both a source of more detailed information and an acknowledgment
to the original investigator. Thanks are due especially to the many
individuals, institutions, and societies for permission to reproduce
illustrations and quotations from their publications.
Certain individuals have been most generous in the matter of
supplying illustrations and assisting me with the study of plant
collections under their charge: Roland W. Brown, Sergius H.
Mamay, and James M. Schopf of the United States Geological
Survey; Robert W. Baxter, University of Kansas; Ralph W. Chaney,
University of California; Rudolf Florin, Hortus Bergianus, Stock¬
holm; Ove Arbo Hoeg, The University, Oslo; Suzanne Leclercq,

Telegram @tubscstds
University of Liege; Maria F. Neuburg, Geological Institute, Mos-
X PREFACE

cow; Olof H. Selling, Swedish Natural History Museum; Wilson N.

Stewart, University of Illinois; Frangois Stockmans, Royal Natural
History Museum of Belgium; F. M. Wonnacott, British Museum of
Natural History. I am also grateful to Mrs. Ellen Kern Lissant
for her assistance with a considerable portion of the drawings and
to my student Mr. Tom Phillips for his aid in many ways.
My thanks are also due the John Simon Guggenheim Memorial
Foundation and the National Science Foundation. Although
grants-in-aid that I have received from these institutions were not
made for the preparation of this book, it has been possible in the
course of research projects to gather much useful information that
could not have been obtained otherwise.

Henry N. Andrews, Jr.

St. Louis, Missouri
December 1960

Telegram @tubscstds
INTRODUCTION—some basic principles

the PSILOPHYTA and certain other early plants

the PTEROPHYTA—early ferns of the Devonian

r ,^
and Carboniferous

the PTEROPHYTA—Marattiales and Filicales

the PTERIDOSPERMOPHYTA

the ANTHOPHYTA (Angiosperms)

Some paleobotanical problems, earliest
angiosperms, and a gymnospermous group,
the Caytoniales

the ANTHOPHYTA (Angiosperms)

Some Tertiary floras of Europe
and of western United States

the LYCOPODOPHYTA

Telegram @tubscstds
xii CONTENTS

the ARTHROPHYTA 256

the CYCADOPHYTA 289

the CONIFEROPHYTA and GINKGOPHYTA 314

some gymnospermous plants of

uncertain affinities 348

heterospory and the evolution of the

pteridosperm seed 364

fossil plants of the Arctic and Antarctic 380

HEPATOPHYTA and BRYOPHYTA-

the bryophytic plants 397

some Paleozoic and Mesozoic floras 408

an introduction to palynology 436

techniques for studying fossil plants;

some basic references 463

Index 477

Telegram @tubscstds
 1
INTRODUCTION
some basic principles

P aleobotany is the story of the preserved vestiges of the

plant life of the past and of the men and women who
have interpreted the remains that we call fossil plants. If the stu¬
dents of these fossil plants are given a somewhat more prominent
status than is usual in a textbook, it is because some acquaintance
with them seems particularly essential to an understanding of the
subject material. When studying living plants it is often possible,
when an obstacle is encountered, to repeat an experiment, to gather
more specimens, or to refer to an herbarium collection for additional
information. In paleobotanical studies the approach is rather dif¬
ferent; there have never been large numbers of investigators, the
preserved plants are usually fragmentary, collections are scattered,
and all too often they have been badly curated, lost, or destroyed.
One is, therefore, frequently forced to depend on published reports
and to place considerable weight on the general calibre of the
worker. It is not implied that the problems of the paleobotanist
are greater than in the various categories of living-plant botany, but
they are certainly unusual.
The paleobotanist deals with plants that have lived on the earth
during the past half billion years. This is a rather arbitrary figure
since we know that plant life has existed for a much longer period,
but the record prior to about that point in time is extremely meager.
The approach to a study of these plants, as presented here, is pri¬
marily a botanical one; that is, we shall be concerned with the kinds
of plants that have existed in the past for their own sake and for
what they can tell us about the origins of modern floras. However,
like any other branch of science the worker in paleobotany cannot
function as an isolationist. A considerable knowledge of living

Telegram @tubscstds
plants is necessary to interpret correctly those of past ages and an
1
2 STUDIES IN PALEOBOTANY

understanding of the geological surroundings in which the fossils

occur is essential. Since many of the students using this book will
not be equally well grounded in botany and geology, certain elemen¬
tary principles will occasionally be brought in which some readers
may skip over.
If this chapter takes on a somewhat miscellaneous aspect, it is due
to an effort to answer certain basic questions that inevitably arise
in the initial phases of any study of plants and animals of the past.
Among these are the measurement of geologic time, the conditions
under which plants may be fossilized, and the ways in which we find
them preserved today; or, in other words, just what is a fossil and
how is its age determined. More purely botanical questions center
around a correct interpretation of the structures that we can ob¬
serve, the classification of living and fossil plants, and apparent
evolutionary trends. The introductory topics that have been se¬
lected are basic to all the chapters that follow and will be enlarged
upon as occasion demands.
It is not possible in a study of this sort to acknowledge adequately
the many sources from which information has been drawn without
constantly inserting reference citations. At the end of each chapter
there is a list of some of the more important sources, and the bibli¬
ographies which in turn may be found in these articles and books
will lead the interested student to as detailed a study as is possible
from the literature alone.

The Geologic Time Table

Rather early in the present century Bertram Boltwood, an Ameri¬

can radiochemist working in Lord Rutherford’s laboratory at Cam¬
bridge, suggested that the newly acquired knowledge that the
radioactive elements uranium and thorium decay into helium and
lead might be used to measure the ages of minerals containing such
elements. Since then tremendous strides have been made in increas¬
ing the accuracy of our methods for measuring geologic time. A
concise and useful summary of the whole field may be found in
Bowen’s recent book The Exploration of Time. This discussion is
necessarily limited to aspects of time that apply most directly to
the study of fossil plants.
Radioisotopes which have proven to be particularly useful in
geologic dating include uranium-238, uranium-235, thorium-232,
potassium-40, and rubidium-87. We may confine our attention for

Telegram @tubscstds
SOME BASIC PRINCIPLES 3

the moment to uranium; the various isotopes of this element have

correspondingly different half-lives, only the longer ones being use¬
ful in geologic dating. In a rock specimen containing uranium-238
and uranium-235 the isotopes of lead (206 and 207) are found which
are the respective end products of decay. The disintegration rates
of these isotopes of uranium are known; for example, the half-life
of uranium-235 is 710 million years.
According to the known rate of decay a uranium mineral will
contain specific quantities of the decay products, lead and helium.
Thus, in a uranium sample one billion years old, for every 1000
atoms of uranium-238 there should be 166 atoms of lead-206 which
is the end product of uranium-238, 11.9 atoms of lead-207 which is
the end product of uranium-235, and 1410 atoms of helium. There
are, then, four ways in which the age of such a rock can be calcu¬
lated: the ratio of uranium-235 to lead-207, the ratio of uranium-
238 to lead-206, the ratio of the two leads, and the ratio of helium
to uranium. Of these, the ratio of the two leads is subject to the
least error. There are complicating factors in the technique such
as the possible presence of ordinary lead, that is, lead that was
present as such from the start, and the fact that helium tends to
leak away.
It may be of interest to cite several specific dates and compare
them with the ages given on the geologic time chart shown on page
6. A Carboniferous magnetite from the Urals has been dated as
260 million years, a Cretaceous magnetite from British Columbia
as 100 m.y. and a Miocene specimen from Utah as 30 m.y. A suffi¬
cient number of such datings have now been made so that the ages
shown on the chart are reasonably reliable, but taking the world as
a whole we have only a brief framework of such determinations.
The method is unfortunately very precarious with sedimentary
rocks where one cannot be certain of the origin of the radioactive
material, if indeed it is found at all.
Developments in radioactive dating techniques have been so rapid
in recent decades as to leave little doubt that they will continue to
become more accurate and that they can be applied ultimately to
a greater number of geological horizons than is presently possible.
One result of these studies in very recent years suggests an appre¬
ciably greater age for the earth than was supposed a few decades
ago. Based in part on analyses of ancient lead from the Rosetta
Mine in South Africa and lead found in meteorites a figure of 4.5
billion years is now suggested for the age of the earth. There is evi-

Telegram @tubscstds
4 STUDIES IN PALEOBOTANY

dence, as we shall see later, that plant life has existed for close to
half this period.
In view of its widespread use with plant materials it seems appro¬
priate to include a few notes on the time-measuring technique
employing carbon-14 that has been developed by W. F. Libby. This
is based on our knowledge that cosmic ray impacts with nitrogen
atoms (most abundant at the 40,000 foot level in the atmosphere)
result in the transmutation of nitrogen atoms into a radioactive
isotope of carbon with a mass of 14, or as it is commonly referred
to, “carbon-14.” When such carbon atoms are combined with oxy¬
gen to form C02 and reach the earth, they may, through photosyn¬
thesis, be incorporated into plant materials. This radioactive car¬
bon-14 has a half-life of 5568 ± 30 years, which means that the
quantity present in any particular plant material is reduced to half
in that number of years. At the time of the death of the plant the
acquisition of the radioactive carbon ceases; thus with a given sam¬
ple of ancient wood or other plant materials the amount of carbon-
14 remaining can be determined and its age calculated. In view of
the rather short half-life the use of this method is limited to speci¬
mens of not more than 50,000 years old. It is, therefore, of httle
significance to us here since we shall be dealing chiefly with scores
or hundreds of millions of years.
It may be pertinent at this point to consider just what we do mean
by a fossil plant in terms of time, that is, how old do plant remains
have to be to fall in this category. One can of course set an arbi¬
trary point in time, but since time and plant evolution are continu¬
ous phenomena it would have no real meaning. Looking to the
Latin origin of the word fossil we find that it refers to anything dug
out of the earth and may be applied to organic as well as inorganic
remains, although modern usage has come to eliminate the latter.
Perhaps the most reasonable basis for a definition, if one feels more
comfortable in establishing one, is the use to which we put vegetable
debris that is found buried in the earth. For example, plants of very
recent origin, not in excess of five or six thousand years old, have
been studied largely by archaeologists and ethnologists and the
center of interest here has focused around the use of these plants
by ancient societies of men. As another example, the study of plant
remains, particularly pollen but not exclusively so, from Pleistocene
deposits has been followed for many years by certain botanists and
geologists with interests in glacial geology and former plant distri¬
bution patterns. They deal with plant species that are, for the most
part, still living and the ages involved range from several thousand

Telegram @tubscstds
SOME BASIC PRINCIPLES 5

to perhaps several hundred thousand years. Such plant remains

may or may not be referred to as “fossils,” but there is obviously
little to be gained from arguing the matter. Prior to the Pleistocene,
plant remains are without question called fossils, perhaps because
of the increased time span, although it is somewhat arbitrary and
not based on any precise number of years.
A few comments are next in order on the geologic time chart
shown on page 6. Fossil plants and animals now play so impor¬
tant a role in many areas of natural science that almost any brand
of biologist must have in mind the correct sequence of the geologic
eras and periods. For the beginning student it would seem unnec¬
essary to commit these to memory immediately, but rather refer
to the chart as the names appear in the text until the sequence is
mastered.
A few of the problems encountered in fixing any given fossil plant
or animal in this time chart may be considered. If one notes for
example that the Triassic period covered a span of some 30 million
years it might be suspected that this would include many rock
formations varying in thickness and physical qualities, and scattered
very irregularly over the surface of the earth. A great many names
have been applied to the numerous rock units which make up a great
system like the Triassic and it is one of the important tasks of the
geologist to establish the correct stratigraphic sequence of these
units in different parts of the world, or, in other words, to correlate
Triassic rocks of, let us say, North America and Africa. As occa¬
sion arises it will be necessary to elaborate on this time chart and
consider categories below that of the Period. It may be worth
noting that the Geological Society of America has, over a period of
some few years, published in its Bulletin numerous special studies
dealing with the correlation of rock formations in North America.
Many of these are available in reprint form.
Since we shall devote a great deal of time to plants of the Car¬
boniferous system it may be well to note that the Pennsylvanian
period is regarded as synonymous with Upper Carboniferous, and
Mississippian as synonymous with Lower Carboniferous. The Car¬
boniferous terms are generally used by European geologists, whereas
Pennsylvanian and Mississippian are used in this country. Strati-
graphically the two do not coincide exactly, but for much of the
discussion here they will be considered as essentially identical.
The figures given in millions of years must of course be taken with
some reservation. Since it is the primary task of the paleobotanist
to aid in bettering our understanding of evolutionary patterns in the

Telegram @tubscstds
6 STUDIES IN PALEOBOTANY

Major Stratigraphic and Time Divisions in Use by the U. S. Geological Survey

Estimated Ages of
Era System or Period Series or Epoch Time Boundaries in
Millions of Years

Recent
Quaternary
Pleistocene
1
Pliocene
10
Cenozoic Miocene
25
Tertiary Oligocene
40
Eocene
Paleocene
Upper (Late)
Cretaceous
Lower (Early)
1°5
Upper (Late)
Jurassic Middle (Middle)
Mesozoic Lower (Early)
1 50
Upper (Late)
Triassic Middle (Middle)
Lower (Early)
180
Permian
one
Carboniferous

Upper (Late)
Pennsylvanian Middle (Middle)
systems

Lower (Early)

Mississippian Upper (Late)

Lower (Early)
oce
Upper (Late)
Devonian Middle (Middle)
Paleozoic Lower (Early)
OlD
Upper (Late)
Silurian Middle (Middle)
Lower (Early)
ooU
Upper (Late)
Ordovician Middle (Middle)
Lower (Early)
A Qfl
Upper (Late)
Cambrian Middle (Middle)
Lower (Early)
OlU
Informal subdivisions
such as upper, middle,
Precambrian and lower, or upper and
lower, or younger
and older may be used
locally.
—3,000 -

Telegram @tubscstds
SOME BASIC PRINCIPLES 7

plant kingdom, it is more essential that the precise stratigraphic

position of a fossil, or collection of fossils, is known, rather than the
exact age in years.
A botanist cannot help noting the similarity of his problems with
those of the geologist. The latter is confronted with thousands of
rock units composing the crust of the earth and deals with problems
such as the limitations of these units in a restricted area and then-
correlation with sequences of presumed similar age in distant re¬
gions. The botanist attempts to understand the boundaries of his
major biological unit, the species, and interpret the complex evolu¬
tionary patterns that he finds over the surface of the earth.

The Preservation of Plants as Fossils

As might be expected from the very wide range of plant structures

and the diverse conditions under which they have been preserved
the modes of preservation are numerous although, as a matter of
convenience, they may be listed under a few fairly distinct catego¬
ries. In this section we shall consider some of the more significant
modes of preservation and the environmental conditions that were
involved. Some examples have also been selected to reveal the re¬
markably fine preservation that is occasionally encountered. Pale¬
ontologists are sometimes given to making excuses for the inadequa¬
cies of the fossil record; it is evident of course that only fragments
of the floras of the past have been preserved and the entire picture
can never be restored. Yet if one takes a thoroughly pessimistic
viewpoint, which can result only from a complete ignorance of the
remarkable discoveries of the past century, there is no point in con¬
sidering the fossil record. An optimistic outlook is adopted here.
We are constantly being surprised at the way plants were preserved
to form fossils and it is safe to assume that many surprises are in
store, and with the discovery of new sources of material, and im¬
proved techniques for dealing with them, the usefulness of the fossil
record will continue to command increased respect.
Many factors are involved between the tree that lived in a Trias-
sic landscape some 160 million years ago and the specimens that lie
on the laboratory bench awaiting study, or perhaps it might be bet¬
ter to end with the published account that arrived in this morning’s
mail. Some of these factors are: the nature of the plant material,
whether thin delicate leaves or tough leathery ones; the kind of en¬
tombing matrix and the rapidity with which the plants were covered

Telegram @tubscstds
8 STUDIES IN PALEOBOTANY

and thus prevented from exposure to decay organisms; the condition

of the plant materials at the time of deposition; and alterations that
the rocks have been subjected to during the thousands or millions
of years that have elapsed since their formation. Time does not
necessarily play a significant role; that is, the better preserved
plants are by no means correlated with recent horizons. The im¬
portant thing is that the plants were originally preserved quickly
and protected against the forces of decay.
In attempting to arrive at an attractive approach to this discus¬
sion of the origin of our raw material it was decided to select a few
fossil plant deposits, scattered through time and space, which have
been a source of both inspiration and fine specimens to the writer.
The account is thus somewhat biased by the limitations of my own
experience, but I believe the fossils considered are representative.
In 1916 E. W. Berry described a clay deposit near the little town
of Puryear in western Tennessee as “the most remarkable leaf¬
bearing clay that I have ever seen at any geologic horizon.” As a
result of several visits to this locality I fully share this appraisal.
The fossil deposit here is a plastic brown clay about 4 feet thick
which is overlain by 10 feet of a lighter colored clay, both being
mined for brick manufacture. Distributed through the brown clay
are innumerable leaves and occasionally seeds and fruits; it is a de¬
posit that is abundant in specimens and numbers of species, over two
hundred having been recorded. This belongs to a series of beds
known as the Wilcox group and is of lower Eocene age. Some 50
million years ago the trees and shrubs of the surrounding forest shed
their leaves which fell into a nearby lagoon and were quickly covered
by a fine sediment. The latter, with the entombed plant debris,
became consolidated and the area was elevated sufficiently to form
dry land. The clay can be dug out with a stiff shovel and then split
along the bedding plane with a knife to reveal beautiful fossil leaves.
These vary in the amount of leaf material that remains; in some
there is essentially no organic matter preserved at all, thus only the
veins and outline of the leaf remain, the fossil being a typical impres¬
sion (Fig. 1-1). With others there is some organic matter present
but little more than a trace. It is interesting to note that abundant
pollen grains are also found in the clay intermingled with the leaves.
If we next visit a portion of the Yorkshire (England) coast a few
miles south of Scarborough a light gray, readily worked shale is en¬
countered that was probably deposited under rather similar circum¬
stances. However, the fossils here appear coal black against the

Telegram @tubscstds
light color of the shale, and it too is a prolific flora, including liver-
 SOME BASIC PRINCIPLES
9

Fig. 1-1. A. The clay pit at Puryear, Tennessee; a source of abundant leaf impressions
of early Eocene age. B. A leaf impression; virtually no organic matter remains.

worts, ferns, ginkgo leaves, cycadophytes, conifers, and other groups;

we shall have occasion to refer several times to this classic Jurassic
locality. Of particular interest at the moment is the fact that much
of the organic matter is preserved so that the fossils, known as
compressions, can in many instances be removed intact from the
rock; this is readily accomplished by simply soaking a specimen in
hydrofluoric acid, thus dissolving the sediment and liberating the

Telegram @tubscstds
plant parts. Unlike an impression fossil in which little or no cell
10 STUDIES IN PALEOBOTANY

Fig. 1-2. A compression fossil. A. Portion of a fertile fern frond (Senftenbergia) from
the Upper Carboniferous of Illinois. B. One sporangium (enlarged) has been treated
to partially dissolve the wall, revealing the enclosed spores. C. Epidermal hairs (en¬
larged) found on the rachis of the frond.

Telegram @tubscstds
SOME BASIC PRINCIPLES 11

structure remains, we may find in compression fossils the cuticular-

ized epidermal cells of various plant organs, spores, epidermal hairs,
and occasionally other cellular details; thus very valuable diagnostic
data may be gleaned from such remains. The Carboniferous com¬
pression shown in Fig. 1-2 is close to three times as old as the Pur-
year fossils and is yet much better preserved.
Numerous examples of compression fossils will be dealt with in
the following chapters, excellent specimens having been found in the
Lower Carboniferous and even the Devonian. This should not be
construed to imply that plant fossils improve with time! It does
mean that time is of less importance than certain other factors, such
as the kinds of plant structure concerned and the geological condi¬
tions responsible for the initial phases in fossilization.
Any discussion of plant preservation would hardly be complete
without some mention of the fossil forests of Yellowstone National
Park, and they will be considered as an example of a petrifaction.
The extensive exposures of petrified stumps and logs in the north
central part of the park and along the Montana boundary in the
northwest corner present a spectacular paleobotanical display that
is probably not excelled in any other part of the world. As one
ascends Big Horn Mountain in the northwest corner of the park the
dense coniferous woods give way to an open slope which is precipi¬
tous in spots and littered with fragments of silicified wood and
stumps standing in the original position in which they grew. In
ascending further it becomes evident that the mountain at this point
consists of a series of fossil forests, one above the other (Fig. 1-4).
Going back in time some 50 million years it is not difficult to visu¬
alize the sequence of events that resulted in this unique geological
formation. A living forest carpeted the lowlands and it was a more
diverse one than is found in the Yellowstone today; there were many
more broad-leaf trees as well as giant redwoods and others that are
not present in that region today. Destruction in the form of ash
and breccias from nearby volcanoes entombed portions of the forest,
stripping the trunks and branches of their bark and covering the
gaunt remains to a depth of many feet. In the ensuing centuries
another forest developed over the grave of the first one and in turn
met the same catastrophic end. These cycles of forest growth and
volcanic destruction continued for many thousands of years and at
one point 27 successive forests can be counted. As siliceous waters
filtered down through the deposits the wood became infiltrated and
replaced to a greater or less degree by silica. The stumps and logs

Telegram @tubscstds
thus ultimately became somewhat harder than the enclosing ash-
12 STUDIES IN PALEOBOTANY

breccia matrix and, being more resistant to erosion, they stand out
on the steep hillside as spectral but magnificent reminders of the
distant past (Fig. 1-3).
Plant deposits referred to as fossil forests have been recorded from
numerous geographical localities scattered through past ages. At
Florissant, Colorado, there is a particularly fine example where great
silicified stumps stand in the position in which they grew and the
enclosing volcanic ash is rich in leaf impressions and animal remains,
particularly insects. In the journal of his Voyage of the Beagle
Charles Darwin mentions a similar forest near Mendoza in Argen¬
tina. In Mesozoic rocks of South Dakota silicified cycads were
discovered about a century ago and several hundred specimens,
gathered through the efforts of G. R. Wieland and others, are now
in the collections of Yale University.
In contrast to what may be termed a true fossil forest, in which
the stumps stand in their place of growth, there are many deposits,
great and small, of logs that were carried by rivers and ultimately
dropped along the stream course, in a delta or elsewhere, to be
covered by sediments and later infiltrated with mineral matter. The

Fig. 1-3. Petrifaction fossils. A. A silicified tree trunk standing in the original posi¬
tion in which it grew; Yellowstone National Park. B. Fungus filaments in the cortical

Telegram @tubscstds
cells of an Upper Carboniferous fern (highly enlarged).
SOME BASIC PRINCIPLES 13
VXo

Lamar Valley

Fig. 1-4. A profile diagram showing the succession of fossil forests in Yellowstone
National Park. (Prepared by Erling Dorf.)

Telegram @tubscstds
14 STUDIES IN PALEOBOTANY

fossil forests of Arizona, as well as numerous others in the western

states, originated in this manner. A comparable deposit of cor-
daitean logs has been described from the Permian of southwest
Africa.
Among the greatest sources of fossil plant materials found thus
far are aggregations of petrified plant materials known as coal balls
which are found in coal seams of Pennsylvanian age. Much has been
learned about the forests of this age from European and North
American coal-ball deposits. Many of the included plants will be
considered in later chapters; at this point a notation is inserted
simply on the origin of this important source of information.
As vast quantities of plant debris accumulated in the Pennsyl¬
vanian swamps small portions were infiltrated, chiefly with calcium
and magnesium carbonates, or with iron sulfide, and were thus pre¬
vented from being altered into coal. The resultant petrifactions are
frequently more or less spherical, varying in size from a centimeter
or two in diameter to specimens weighing several hundred pounds.
The forms assumed, however, are variable in the extreme, so that
the word “ball” is not wholly appropriate. Coal balls have been
found in England, on continental Europe from Belgium to the
Donetz basin, in Spain, and in the central states of this country from
Kentucky and Indiana west to eastern Kansas. At certain locali¬
ties and in some coal seams they are very abundant and elsewhere
they are scarce or never occur; the quality of preservation, although
highly variable, may be exquisite. (As examples see Figs. 3-5, 5-1,
8-6.)
Although some cellular structure (occasionally a great deal) is
preserved in a compression type fossil, it is generally much better
in a petrifaction. Resistant structures such as wood are especially
common as petrifactions, yet delicate plant parts, even nuclei and
chloroplasts, are sometimes encountered. Indeed, thin-walled cel¬
lular structures such as fleshy seed coats and root hairs are some¬
times better preserved than tougher ones apparently because they
were more readily impregnated with mineral-bearing waters.
Very little is actually known about the petrifying process itself.
It was once explained on the basis of molecular interchange, but in
view of the known differences in molecular structure between the
organic materials and the petrifying minerals (usually carbonates,
silica, and iron sulfide) this seems unlikely. The amount and kinds
of organic matter present in a petrifaction vary greatly. In the case
of some fossil woods the mineral matter may be dissolved out, using

Telegram @tubscstds
hydrofluoric acid with silicified specimens, leaving the original wood
intact.
SOME BASIC PRINCIPLES 15

A few studies have been made in recent years which shed a little
light on the durability of certain organic materials. Investigations
of the shell of the clam (Mercenaria mercenaria) have yielded the
following information: specimens approximately 1000 years old
contained protein essentially unchanged; specimens known to be
not less than 100,000 years old showed the protein reduced to a tar¬
like substance with peptide chains consisting of two or more amino
acids; specimens 25 million years old contained only individual
amino acids. As a far more ancient example, the armor of the
Devonian fish Dinichthys terelli, from the Black Shale of Ohio, was
found to contain several amino acids including alanine, glycine,
proline, and valine.
Illustrating another type of preservation, we may look next to the
hills of central Vermont, near the town of Brandon, where there is
a small deposit of low rank brown coal that has attracted consid¬
erable attention since its discovery about a century ago. Although
of low fuel value and too limited in extent to have been of much
commercial importance, it has interested geologists because of its
occurrence in the New England hills area, and paleobotanists be¬
cause of the variety of fossil plants it contains. Under the leader¬
ship of E. S. Barghoorn the deposit was reopened in 1947 and
extensive collections of seeds, fruits, wood, spores, and pollen were
obtained. The fossil remains are not petrified and must be stored
in water or glycerin to prevent them from drying and shattering.
Although the Brandon flora has not yet been studied in full, it
promises to give us an especially complete and accurate picture of
the flora of Vermont in mid-Tertiary times. Since it includes wood,
seeds and fruits, and pollen, a three-way check on identifications is
possible; the spores and pollen give one facet of facts, the lignitic
wood a second set, and the seeds and fruits still a third.
It is well to emphasize that mineralization is not necessarily a pre¬
requisite to good cellular preservation. A striking instance of this
is the occurrence of abundant wood in the Cretaceous clays of
northern New Jersey that contains no petrifying mineral matter yet
many logs and branches are well enough preserved to suggest that
they were buried several decades ago rather than 100 million years.
This is an excellent example of the fact that plant materials may
be preserved for very long periods of time if they are adequately
protected from the action of biological and chemical decay.
In areas where waters are supersaturated with mineral matter,
plant materials may become encrusted to form a sort of armor coat¬

Telegram @tubscstds
ing. This may be observed at Mamouth Hot Springs in Yellowstone
16 STUDIES IN PALEOBOTANY

Park; the hot waters are supersaturated with carbonates which pre¬
cipitate on cooling, and twigs, leaves, insects, and the like that fall
into them quickly become coated to form an encrustation fossil.
Amber, which is fossil resin, probably derived from coniferous
trees, presents another distinctive mode of fossilization. Amber col¬
lected along the shores of the Baltic is famous for its flowers and
insects. A remarkable instance of preservation in amber has been
recorded by Petrunkevitch; he has described a spider that was en¬
trapped in resin when it was actively making threads and it was pos¬
sible to trace these to their respective spinning tubes. Although the
fossils contained in this medium may be objects of great beauty and
well preserved, its occurrence is so rare as to render it of little
importance in paleobotany.
A few outstanding instances of preservation will be described next,
not in an attempt to exaggerate the potentialities of the fossil rec¬
ord but to indicate what is possible.
A Tertiary Azolla. From an early Tertiary chert in India the late
Birbal Sahni and H. S. Rao have described the petrified massulae
of a species of the water-fern Azolla. This is one of several genera
of heterosporous plants, certainly not all closely related, known as
water-ferns. In Azolla several extraordinary structures develop
within the microsporangium from plasmodial substance which en¬
closes several microspores. These are known as massulae, and they
finally develop numerous peripheral appendages called glochidia
which in turn terminate in a harpoonlike tip. In the Tertiary Azolla
intertrappea (Fig. 1-5) beautifully preserved massulae are known
and have been found anchored to the megaspore.
Microorganisms in Coprolites. With the exception of the diatoms,
studies of microscopic plant remains were for the most part initiated
only a few decades ago and it is my impression that we are just be¬
ginning to realize the potentials of the fossil record. An interesting
example is the discovery of certain algae and bacteria in an upper
Eocene coprolite (fossil feces) from the Bridger formation of Wyo¬
ming. Thousands of bacteria, assigned to the genus Escherichia,
have been isolated from the coprolites as well as numerous cells of
the desmid Staurastrum. The latter is representative of a distinc¬
tive group of unicellular algae; the wall as well as central bodies that
are interpreted as shrunken chloroplasts are preserved. Other struc¬
tures were found which are interpreted as conjugating cells and
zygospores of desmids. These and other aquatic organisms con¬
tained in the coprolites were probably taken in by chance when the

Telegram @tubscstds
animal drank water from a pool at low stage that contained a strong
 SOME BASIC PRINCIPLES
17

Fig. 1-5. A remarkable example of fossilization. Reproductive structures of Azolla

intertrappea, an early Tertiary silicified aquatic fern from Chhindwara, India. The
massulae are shown anchored to fibrils of the megaspore. Certain details have been
slightly intensified. (From Sahni, 1941.)

concentration of the microscopic plants. It is assumed that it was

not an herbivore since the desmids, with their cellulose walls, passed
unscathed through the digestive tract. It is also suggested by
W. H. Bradley in his report on these fossils that the feces, after being

Telegram @tubscstds
dropped by the animal, were exposed for several hours to sunlight
18 STUDIES IN PALEOBOTANY

with sufficient heat to harden or “fix” the cell contents of the des-
mids. Shortly thereafter the feces were covered by sediment, prob¬
ably volcanic ash, and from this the silica was derived which resulted
in silicification of the microorganisms.
Cretaceous Flagellates. Another instance of the preservation of
delicate organic structures is found in Wetzel’s account of unicellular
flagellates (Ophiobolus utinensis) from a European Cretaceous flint.
Flagella may be clearly distinguished in one of his photographs and
it is suggested that the organisms were imbedded in a siliceous col¬
loid at the bottom of a Cretaceous sea, the colloid having originated
from the dissolution of the shells and skeletons of diatoms, radio-
larians, and silicisponges.
Fossil Bananas. The discovery of a fossil banana fruit from a
mid- or lower Cretaceous horizon in Colombia offers a complex
problem. It is a cast 16 cm long and a little over 3 cm in diameter
and although no cellular structure is preserved it presents in its ex¬
ternal form a striking resemblance to a small modern banana. Sev¬
eral competent paleobotanists have examined the specimen and
seem inclined to accept the identification as valid. It is remarkable
in that the banana genus (Musa) is considered by many botanists to
be a rather recent, highly evolved one and native only to Asia and
Africa. However, there is some evidence, from early historical rec¬
ords, that the banana was cultivated in Central and South America
before the advent of Europeans. Additional evidence which sug¬
gests that the banana is not a newcomer to the New World comes
from Tertiary (possibly Oligocene) seeds, also found in Colombia,
which compare quite closely with those of the extant Musa ensete,
one of the seed-bearing, inedible bananas of Africa.
The Geisel Valley Fossils. Certain middle Eocene lignite deposits
in the Geisel Valley of Germany stand out as one of the most re¬
markable fossil discoveries in recent years; although many of the
more extraordinary remains are of animals it may not be amiss to
mention them here for the exceptional types of preservation encoun¬
tered. In certain parts of the deposit bones were found to have been
destroyed by humic acid, whereas others are described as being as
soft as butter and were treated with paraffin or special glue before
removing. The brains and spinal cords of frogs were found pre¬
served as calcium compounds of fatty acids. Blue-bottle grubs were
found in skulls and could be studied in much the same way as
modern larvae. Muscular tissues of fishes, salamanders, lizards, and
mammals were preserved and nuclei were observed in the epithelium

Telegram @tubscstds
of frog’s skin. Stomach contents of fishes, birds, beetles, and mam-
 SOME BASIC PRINCIPLES 19

mals were recognized, and chlorophyll was identified, as well as the

coloring of many insects.
In concluding this section a few notations may be appropriate con¬
cerning pseudofossils. Pitfalls that the unwary paleontologist may
fall into are legion and there are few who have escaped entirely un¬
scathed. A wide variety of wholly inorganic structures may resem¬
ble plant organs, whereas outright fakes or artifacts of one sort or

Fig. 1 -6. Pseudofossils. A. Dendrites, natural size. B. Filaments, superficially resem¬

bling those of an alga, found in a “thunder egg,” about 20X. C. Ichthyosaur coprolite
from the Upper Cretaceous of Selma Creek, Alabama, about 2X; this is a “pseudo¬
fossil” in that it bears a false resemblance to an evergreen cone. (From specimens in

Telegram @tubscstds
the U. S. National Museum.)
20 STUDIES IN PALEOBOTANY

another may, without critical examination, appear to represent

plant life of the past.
Dendrites (Fig. 1-6A) are rather common pseudofossils, being
mineral formations formed in rock crevices; although presenting a
superficial resemblance to fern leaves a careful examination reveals
an absence of the characteristic regularity of form of such a plant
organ.
Figure 1-6C shows an Ichthyosaurus coprolite from the Upper
Cretaceous of Alabama that superficially resembles the cone of an
evergreen tree. It would perhaps be more appropriate to call this
a pseudo-plant fossil.
R. W. Brown has recently written an instructive account of the
plantlike structures that are found in “thunder-eggs”; these are
mineral formations formed in pockets of volcanic rocks and in some
of them filamentous structures have been developed (possibly com¬
parable to the silicate “chemical gardens”) which resemble certain
algae, (Fig. 1-6B). Precise study of such structures reveals details
of form and branching pattern which are quite unlike an alga.
Some years ago what appeared to be a remarkable fossil ear of
corn (Zea) was found in South America and was described as such
by a competent paleobotanist. The specimen created considerable
interest until another paleobotanist, whose suspicions were aroused,
took the trouble to section it and found, indeed, that it was a
cleverly made piece of ceramic artistry.
A specimen resembling a charred corn cob was once sent to my
laboratory for identification; it was alleged to have been found in
coal from a horizon of Pennsylvanian age. A careful examination
left no doubt that it was a corn cob and one which apparently had
been partially burned and, in one way or another, found its way back
into the coal scuttle. I am not sure that the discoverer of the speci¬
men took kindly to this prosaic identification but it was the only
one possible.
It is perhaps evident that the only way to avoid accepting the
unreal for the real is to acquire as broad a knowledge of plant form
and mineral structure as possible and to examine critically any
specimens that may present some reason for suspicion.

The Classification of Plants

As fragmentary as the fossil record may be it has yielded so many

unique plants as to have influenced very strongly our concepts of

Telegram @tubscstds
classification. It has been evident almost since the concept of evo¬
lution became generally accepted that there are tremendous gaps
SOME BASIC PRINCIPLES 21

in the plant kingdom as it exists today. Many species, genera, and

taxa of higher orders stand apart, showing no close relationships to
other plants; such discontinuities imply the extinction in past ages
of large sections of the great stream of plant life. Here and there
the fossil record has served to bridge a few of these gaps but it has
also brought to light many plants that are unlike anything that lives
today, thus adding greatly to the complexity of classification.
A classification should enable us to keep some order among the
approximately one-third of a million species of living plants and,
ideally, it should indicate natural relationships. We have come a
long way in the past century toward achieving the latter objective,
but there is a much longer road ahead. Many recent books cite
what is called an “old classification” and the “new classification”;
this is somewhat misleading in that our understanding of interrela¬
tionships in the plant kingdom is constantly being revised. Occa¬
sionally a botanist, perhaps more courageous than the rest, proposes
somewhat more fundamental revisions; since there has never been
universal agreement on any system, and very likely never will be,
he must be prepared to meet a mixture of praise and ridicule.
It seems necessary at this point to present only a framework of
the major divisions; further details and possible evolutionary se¬
quences will be considered as we go along. The outline given here
is patterned essentially after the classification used by Harold C.
Bold in his recent Morphology of Plants (1957), but since the latter
work deals primarily with living plants we must add a few comments.

1. Division: HEPATOPHYTA—Liverworts
2. Division: BRYOPHYTA—Mosses
3. Division: PSILOPHYTA—Psilophytes
4. (Certain early land vascular plants of unknown affinities)
5. (The preferns or plants intermediate between the earliest
vascular plants on the one hand and the ferns and pteridosperms
on the other; may be tentatively included in Pterophyta.)
6. Division: PTEROPHYTA—Ferns
7. Division: LYCOPODOPHYTA—Lycopods or club-mosses
8. Division: ARTHROPHYTA—Horsetails or articulates
9. Division: PTERIDOSPERMOPHYTA—Seed-ferns
10. Division: CYCADOPHYTA—Cycads and bennettites
11. Division: GINKGOPHYTA—Ginkgo and its relatives
12. Division: CONIFEROPHYTA—Cordaites and conifers
13. Division: GNETOPHYTA-Gnetum, Ephedra, Welwitschia

Telegram @tubscstds
14. (Gymnosperms of uncertain affinities)
15. Division: ANTHOPHYTA—Flowering plants (angiosperms)
22 STUDIES IN PALEOBOTANY

The major groups of plants, exclusive of the algae and fungi

The algae and fungi will be referred to only incidentally and are
therefore not included in the outline. The angiosperms or flower¬
ing plants (the Anthophyta of Bold) are considered in some detail
in Chapters 6 and 7, but for reasons given therein no attempt will
be made to deal with them according to any formal taxonomic sys¬
tem in the way that most of the other groups are treated.
It seems to me that a new classification is justified only in so far
as it reflects an advancement in our understanding of the inter¬
relationships of plants and there is thus no point in abandoning
especially useful features of the older classifications. As an exam¬
ple I would cite the use of the terms pteridophyte and gymnosperm.
The former includes plants now assigned to the psilophyte, lycopod,
horsetail, and fern Divisions. The implication of the system given
here is that these four groups (there are actually probably more than
four as will be explained later) have evolved independently and are
not necessarily closely related. However, they have certain aspects
in common and the name pteridophyte is a convenient collective one
for the assemblage. Correspondingly the term gymnosperm will be
used in referring to the naked-seeded plants (seed-ferns, cycado-
phytes, ginkgos, cordaites-conifers, and gnetophytes) even though
it is evident that this assemblage includes several lines in which the
seed evolved independently.
A few introductory words are especially pertinent in explanation
of the three “Divisions” cited in brackets (Nos. 4, 5, and 14). Paleo-
botanical classification enjoys one especially exceptional problem,
that of dealing with plant remains that are so different from any
previously known that a precise correlation with them is impossible.
Rather than immediately setting up an elaborate system of classifi¬
cation for such novelties it seems to me that it is less misleading,
especially to the beginning student, to admit our lack of under¬
standing. I have thus refrained from classifying some of the more
problematical fossils although suggestions as to their possible affini¬
ties will be given in most instances.
It is very possible that the two major groups of the Cycadophyta,
the Cycadales and Bennettitales, should be recognized as distinct
Divisions. The Gnetophyta is a notoriously difficult assemblage
and since their fossil record is scanty they will be dealt with only
briefly.
A minor departure from Bold’s system appears in the use of the

Telegram @tubscstds
Division name Lycopodophyta instead of Microphyllophyta. It is
SOME BASIC PRINCIPLES 23

true, from the standpoint of the Greek derivation, that the word
lycopod (wolf-foot) has no significant application but it is so well
entrenched in the literature that I have chosen to retain it. It may
be added that the name Microphyllophyta is not entirely satisfac¬
tory since not all of the plants in this group have small leaves (some
Carboniferous forms will be considered in which the leaves are
nearly a meter long) and many plants in other unrelated groups also
have small leaves.

An Introduction to Wood Anatomy

It has been pointed out that compression and petrifaction fossils

are especially significant in that they retain, to a greater or less de¬
gree, the cellular organization of the plant; thus in dealing with such
fossils one is involved in interpreting the anatomy or internal struc¬
ture of plants. It therefore seems desirable to introduce certain
aspects of anatomy that will be of frequent concern. A comprehen¬
sive treatise on plant anatomy should be available for reference, two
of the most useful recent accounts being those of Eames and Mac-
Daniels and of Esau.
All too often it is only the more resistant woody tissues of plants
that are preserved as fossils; thus this aspect of anatomy will be
emphasized in the following pages. Other phases of the subject will
be introduced and elaborated on as the occasion requires.
Most plants grow in length by means of a small zone of actively
dividing cells located at the tips of the stem and root systems. Im¬
mediately back of this apical meristem new cells that are formed
become structurally specialized and ultimately serve certain re¬
stricted functions; the systems thus formed are called primary
tissues. It is a notable feature of the pteridophytic groups that
their shoot system (main stem and branches) increases but little in
diameter throughout the life of the plant, although the arborescent
lycopods and articulates of the Carboniferous are notable excep¬
tions. This is due to the fact that the primary tissue system be¬
comes mature immediately back of the apical growing point and
there is no mechanism which allows continued increase in girth.
In contrast to this mode of growth in the pteridophytic groups
we find in most seed plants (often referred to as the “higher plants”)
that continued lateral or secondary growth is a marked feature of
the plant form. It is readily observed, for example, in a pine tree

Telegram @tubscstds
that the shoot system increases progressively in diameter from the
24 STUDIES IN PALEOBOTANY

terminal twigs down through the main stem, and this increase in
girth goes on throughout the life of the plant. This is accomplished
by means of a layer of actively dividing cells located immediately
external to the wood known as the cambium and the tissue systems
formed are called secondary. Thus while longitudinal or primary
growth is confined to the apices of stems and roots (the older parts
do not increase in length) the secondary growth is general through¬
out these organs. There are of course exceptions and variations to
this general statement; we are dealing with a very large assemblage
of living things and they are often reluctant to conform to man¬
made rules and laws.
The following discussion will deal for the most part with stem
structures since they display more variation than roots and are more
often preserved as fossils. We shall also be concerned only with the
wood or xylem (these two terms are synonymous); the wood is more
resistant to decay than other stem tissues and thus more frequently
preserved and it affords many diagnostic features that are useful in
identification. The term vascular plant will refer to any in which
woody tissues are present; the algae, fungi, mosses, and liverworts
are notably lacking vascular tissue and it is much reduced or absent
in a few of the flowering plants (certain parasites and aquatic forms).
Next, the term stele requires some elaboration as it will probably
be used more often than any other throughout the book. This is
usually defined as consisting of the wood, a central pith when pres¬
ent, and two living tissue systems external to the wood, the phloem
and pericycle. The latter two consist of rather thin-walled cells that
are highly subject to decay and thus not often preserved; conse¬
quently as applied to fossil plants the term stele will usually imply
only the wood and the central pith if present.
In some of the simplest members of the pteridophyte assemblage
the stele consists of a slender cylindrical core of only a few wood
cells (tracheids); this is known as a solidprotostele (Figs. 2-1B, 8-5).
When, in larger axes, parenchyma cells are admixed with the tra¬
cheids, it is known as a mixed protostele. In other primitive plants
the wood may be variously lobed forming an actinostele (Fig. 2-3A).
When a distinct pith of (usually) thin-walled and more or less
isodiametric cells is present the vascular unit is called a siphono-
stele (Fig. 8-7), the wood itself being quite literally siphon-like or
tubular. When the vascular tube is discontinuous, rather than
forming a uniform unbroken cylinder, it is called a dictyostele (Fig.
4-12A), that is, a net-shaped stele.

Telegram @tubscstds
SOME BASIC PRINCIPLES 25

Stemming out from these simple and basic types we find, particu¬
larly in the Pterophyta, a wide range of more complicated stelar
patterns. In the modern tropical fern Matonia the vascular system
in the mature stem consists of several concentric cylinders which
will be termed here a polycyclic stele. In certain Pennsylvanian tree
ferns belonging to the genus Psaronius (Fig. 4-4) a system of many
concentric stelar units (meristeles) forms a highly complex poly¬
cyclic stele. A different degree of complexity is found in the Cre¬
taceous fern genus Tempskya (Fig. 4-13A) where the “trunk” may
be composed of as many as several hundred stems—not simply
numerous steles within a stem but distinct stems which form an
anastomosing system that is held together by thousands of slender
wiry roots.
No system of characters used in identifying and classifying plants
ever seems to work perfectly and vascular structures can be espe¬
cially vexing to deal with. In many pteridophytic plants the mature
stem possesses a stelar system that is quite constant in size and form
but in some there is considerable variation within a single plant.
Moreover, it has been shown that in many living pteridophytes (this
is probably true for all living and fossil plants in this great assem¬
blage) in the earliest stages of development the young stem starts
off with a small solid protostele; as the stem grows upward the
apical growing region may increase rapidly in girth and the stelar
system increases correspondingly in size and complexity until the
adult form is reached.
It will be useful next to examine certain developmental aspects
of the stele, with particular reference to structural details of the
wood cells (Fig. 1-8). In a longitudinal section through the stele of
the living Whisk fern (Psilotum), which is protostelic in the smaller
twigs (Fig. 1-7A) and siphonostelic in the larger ones (Fig. 1-7B), it
will be noted that the outermost wood cells (tracheids) are small in
diameter and there are conspicuous annular or spiral bands of sec¬
ondary wall material laid down on the inner side of the original thin
primary wall of these cells. Immediately within these wood cells
there are somewhat larger ones in which the bands of secondary wall
material are broader and more uniform; they are, moreover, at¬
tached to the primary wall by a slender median strand, so that this
type of wall thickening is actually T-shaped when seen in section;
such cells are called scalariform tracheids. Next in sequence are
cells in which the secondary thickening is uniform over the entire
primary wall with the exception of certain circular or elongate areas

Telegram @tubscstds
26 STUDIES IN PALEOBOTANY

Fig. 1-7. Transverse view of the stem of the extant Psilotum nudum,, about 15X.
A. A small branch showing protostele, s, about 15X. The stele of a larger branch: px,
protoxylem; mx, metaxylem; p, the central fibrous pith. (From Andrews, 1947.)

where the secondary wall forms a dome with a small orifice; these
are known as tracheids with bordered pits.
To complete a series of useful descriptive terms, which correlate
with those given previously, the smaller outermost annular or spiral
tracheids are referred to as the protoxylem (Fig. 1-7B) while the
scalariform and pitted tracheids constitute the metaxylem.
This account can only serve as an introduction to more detailed
considerations of stelar structures that will follow, but a few varia¬
tions may be noted. The entire sequence of tracheid types noted
here is not found in all plants; in some of the psilophytes all of the
wood is composed of annular and spiral cells; in the lycopods there
are a few small annular protoxylem cells and the bulk of the wood
(metaxylem) is scalariform, there being none of the circular bordered
pitted type. Next, the sequence that has been given is somewhat
idealized in that very small steles only a few cells thick, such as
Psilotum itself, may not exhibit the entire series in one section.
There are also transitions among the “typical” types of tracheids,
so that it is not always easy to distinguish sharply between annular
and scalariform cells, and there are frequently transitions between
scalariform and circular bordered tracheids in which one finds cir¬
cular and variously elongated bordered pits in the same cell. The
drawing shown in Fig. 1-8 is taken from the primary wood of the
petiole of one of the cycads; this was selected for illustration since

Telegram @tubscstds
SOME BASIC PRINCIPLES 27

the typical tracheid types as well as intermediates are clearly

defined.
We may next turn to variations in the primary wood based on the
way it matures. The annular or spiral protoxylem cells are first to
mature, and the maturation process continues through the metaxy-
lem. In a plant such as Psilotum, where the protoxylem is outer¬
most, the stele is termed exarch; in a few rare instances, such as
Rhynia, a fossil plant from the Devonian of Scotland, the protoxy¬
lem is centrally located and the stele is termed centrarch (Fig. 2-1B);
when the protoxylem is innermost (but not at the center of the
stem) the stele is termed endarch; and finally when the protoxylem
is located more or less midway in the primary wood and the matu¬
ration process takes place both toward the center and toward the
periphery of the stem it is termed mesarch.
The above discussion has referred exclusively to the primary
wood; a few members of the pteridophytic groups developed sec-

a b c d e f

Fig. 1-8. Pitting transition in tracheids of vascular bundle in petiole of the living
Cycas revoluta: a,b, annular; c, scalariform; d, elongate-bordered pits along left side
of cell and circular to elongate pits to the right; e,f, circular bordered pits.

Telegram @tubscstds
28 STUDIES IN PALEOBOTANY

ondary wood, but it is not until we come to the gymnospermous

plants that this is a constant and conspicuous element of the stem
anatomy. Secondary wood is usually initiated by the division of a
layer of cells, the cambium, around the periphery of the primary
xylem.
By referring to Fig. 8-7, it will be noted that the primary wood
can usually be readily distinguished from secondary by the radial
alignment of the cells of the latter.
In the Carboniferous pteridosperms, cambial activity seems to
have been something of a “new improvement,” there being little
secondary wood, but in some of the Permian members of the group
it was formed in considerable abundance. In the Coniferophyta it
seems to have been a character that was acquired early in the evo¬
lution of the group and, reaching a maximum development for the
entire plant kingdom, it has enabled the sequoias to attain trunk
diameters in excess of 34 feet.
There has been much skepticism cast on the potential of the fossil
record but it seems to me that there is reason to be quite optimistic
about the future contributions of paleobotany. There have been
so few workers in the field until very recently that there has been
hardly adequate opportunity for the science to display its potential.
Techniques have been evolved in recent decades that enable us to
obtain highly significant information from specimens that were dis¬
carded or given only the most cursory consideration fifty years ago
and there is no reason to doubt that improvement in technique will
continue.
It seems especially pertinent, however, to note how little of the
earth’s crust of sedimentary deposits have been studied intensively.
There are large areas of the world in which the surface rocks have
been only casually explored and paleobotanists have been limited
for the most part to natural outcrops and mining operations. I have
no doubt that techniques will develop in the future which will ena¬
ble us to obtain much larger samples of deep-lying sediments.

REFERENCES

Abelson, Philip H. 1956. Paleobiochemistry. Sci. Amer., 195: 83-92.

Andrews, Henry N., Jr. 1947. Ancient Plants and the World They Lived in. Com¬
stock Pub. Co., Ithaca. 279 pp.
Barghoorn, Elso S., and W. Spackman. 1950. Geological and botanical study of the
Brandon lignite and its significance in coal petrology. Econ. Geol., 45: 344-357.

Telegram @tubscstds
Berry, Edward W. 1916. The Lower Eocene floras of southeastern North America.
U. S. Geol. Survey Prof. Paper, 91: 1-481.
SOME BASIC PRINCIPLES 29

-. 1925. A banana in the Tertiary of Colombia. Amer. Journ. Sci., 10: 530-
537.
Bold, Harold C. 1957. Morphology of Plants. Harper and Bros. 669 pp.
Bowen, R. N. C. 1958. The Exploration of Time. Philosophical Library, New York.
Pp. 1-143.
Bower, Frederick 0. 1935. Primitive Land Plants. MacMillan and Co., London,
658 pp.
Bradley, Wilmot H. 1946. Coprolites from the Bridger formation of Wyoming: their
composition and microorganisms. Amer. Journ. Sci., 244: 215-239.
Brown, Harrison. 1957. The age of the solar system. Sci. Amer., 196 (4): 81-94.
Brown, Roland W. 1957. Plantlike features in thunder-eggs and geodes. Ann. Rept.
Smithsonian Institution, 1956. pp. 329-339.
Eames, Arthur J., and L. H. MacDaniels. 1947. An Introduction to Plant Anatomy.
McGraw-Hill Book Co., 427 pp.
Esau, Katherine. 1953. Plant Anatomy. John Wiley and Sons, 735 pp.
Huertas, Gustavo, and Thomas van der Hammen. 1953. Un posible banano {Musa)
fosil del Cretaceo de Colombia. Revista Acad. Colombinan Ciencas Exactas, Fisicas
Nat., 9:115-117.
Hurley, Patrick M. 1959. How Old Is the Earth? Doubleday and Co., New York.
Pp. 1-160.
Jongmans, Willem J. 1949. Het Wissenlend Aspect van het Bos in de Oudere Geolo-
gische Formaties (from: W. Boerhave Beekman: Houtin alle Tijden) Deel, Deventer,
Holland.
Ladd, Harry S. 1957. Chapter I. Introduction to “Treatise on Marine ecology and
paleocology.” Geol. Soc. Amer. Mem., 67: 1-29.
Moore, John A., and Henry N. Andrews, Jr. 1936. Transitional pitting in tracheids
of Psilotum. Ann. Missouri Bot. Gard., 23: 151-156.
Petrunkevitch, Alexander. 1950. Baltic amber spiders in the Museum of Compara¬
tive Zoology. Bull. Mus. Comp. Zool., Harvard Univ. 103: 257-337.
Rodin, Robert. 1951. Petrified forest in south-west Africa. Journ. Paleont.,
25: 18-20.
Sahni, Birbal. 1941. Indian silicified plants. I. Azolla intertrappea Sah. and H. S.
Rao. Proc. Indian Acad. Sci., 14: 489-501.
-, and H. S. Rao. 1943. A silicified flora from the intertrappean cherts
round Sausar in the Deccan. Proc. Nat. Acad. Sci. India, 13: 36-75.
Traverse, Alfred. 1955. Pollen analysis of the Brandon lignite of Vermont. U. S. Bu¬
reau of Mines, Rept. Investigations, 5151: 1-107.
Weigelt, Johannes. 1935. Some remarks on the excavations in the Geisel Valley.
Research and Progress, 1: 155-159.
Wetzel, Otto. 1953. Resume of microfossils from Upper Cretaceous flints and chalks
of Europe. Journ. Paleont., 27: 800-804.
Zeuner, Frederick E. 1950. Dating the Past. Methuen and Co., London. 474 pp.

Telegram @tubscstds
the PSIIOPHYTA
and certain other early plants

Preface to Chapters 2-7

word of explanation is in order concerning the arrange-

a: , ment of the chapters that follow. There seems to be a
broad stream of evolution, with many specialized side branches,
beginning in the psilophyte complex and continuing through the
“preferns” in the Middle and Upper Devonian to true ferns and
seed-ferns in the Carboniferous; and in turn the seed-ferns appear
to be the most likely source of the flowering plants. Imperfect as
our knowledge may be it seems desirable to take advantage of what
is known of the evolutionary sequence in this great assemblage of
plants. A return is then made to other pteridophytic groups
(Lycopodophyta and Arthrophyta); these seem to represent dis¬
tinct lines from their earliest recognizable members and, following
their own independent paths, they expanded into flourishing groups
in the Paleozoic and then rapidly declined. An attempt has been
made, however, to present the subject material throughout so that
the order in which the various chapters are studied is reasonably
flexible.

Introduction—the Contents of This Chapter

The search for the earliest vascular plants constitutes one of the
most exciting chapters in paleobotany. The problem of their
origin and early trends of evolution on the land has long concerned
botanists; it is assumed that they originated from the algae, for
want of a better hypothesis, but we are still very much in the dark
as to how this transition took place. The oldest unquestioned
record of these early vascular plants dates from the upper part of

Telegram @tubscstds
the Silurian, and by Lower Devonian times a rather diverse assem-

30
PSILOPHYTA AND OTHER EARLY PLANTS 31

blage is known to have existed. Progressing upward through the

Devonian one encounters an increasing variety of plants, some of
which are fairly clearly defined as early forms of the lycopod,
articulate, and fern groups, whereas the affinities of others are un¬
certain or wholly obscure.
There is also a considerable accumulation of fossil remains which
suggest that vascular plants existed long before the Silurian; the
evidence is controversial and will probably continue to be so for
some time, but it is such controversy that inspires men to devote
their lives to the study of plant evolution. Finally, one cannot
avoid wondering about the less intricate forms of fife far below the
level of even the simplest vascular plants. Many books and articles
have been written in recent years which speculate on the ultimate
origin of life on the earth; much of this is beyond the scope of
paleontology, but it seems appropriate to note something of what
is known of the earliest records of life at the cellular level.
It is the purpose of this chapter to consider some of these prob¬
lems and, in order to present the subject matter in as clear a
fashion as possible, it will be taken up under several subheadings
as follows:

1. The Psilophyta. These are simple land vascular plants that

display some evidence of being closely related and which may be the
progenitors of later pteridophytic groups.
2. Some Vascular Plants of Uncertain Affinities. The fossils
described here are classified by some authors in the Psilophyta;
however, they show marked divergences from that group and are
certainly not all closely related.
3. Pre-Silurian Plant Records. This section deals with problem¬
atical pre-Silurian reports of vascular plants as well as notations on
some of the earliest known nonvascular plants.
4. A Miscellany of Devonian Plants. Included here are, for the
most part, larger and more complex plants than those dealt with in
the first two sections above; they reveal some of the great diversity
of plant life in the Devonian.

THE PSILOPHYTA

There is considerable disagreement among paleobotanists as to

the limits of the Psilophyta; certainly a number of fossils have been

Telegram @tubscstds
included in the group for want of any other taxonomic category in
32 STUDIES IN PALEOBOTANY

which they might be deposited. For reasons that will be explained

it seems to me that this has led to serious misconceptions concern¬
ing the evolution of early vascular groups; thus the psilophytes are
treated here in a narrow or restricted sense.

The Rhynie Plants

In 1917 Kidston and Lang presented a description of Rhynia

gwynne-vaughani from a Middle Devonian chert bed near Rhynie,
Aberdeenshire, Scotland. There are few plants, fossil or living,
that have been as influential in botanical evolutionary thought.
Unlike previously described early land vascular plants Rhynia and
its chief associates, Horneophyton and Asteroxylon, were preserved
with exquisite perfection and essentially in their entirety. Rhynia
became established in the minds of many botanists as the primitive
vascular plant; it is an interesting plant and a primitive one, but it
is by no means certain that it should be regarded as a focal point
from which all later land plants evolved.
In view of the great importance of the Rhynie deposit a few
remarks may be of interest concerning the site itself. The petrified
remains were first discovered in chert specimens in a stone wall
and when their botanical importance was recognized the rocks were
traced to their source and a considerable amount of the petrified
plant material was excavated. It is worth looking to the original
account for a description of the origin of the chert bed:
The whole history of the formation of the Rhynie Chert Zone, at least of
that portion from which our specimens were taken, can be clearly read. One
can in imagination see a land surface, subject at intervals to inundation,
covered with a dense growth of Rhynia Gwynne-Vaughani. By the decay of
the underground parts of Rhynia and the falling down of withered stems
(for this plant had no leaves) a bed of peat was gradually formed varying
from an inch to a foot in thickness. The peat was then flooded and a layer
of sand deposited on its surface. Again the Rhynia covered the surface, and
this process of the formation of beds of peat, with the deposition of thin
layers of sand, went on until a total thickness of 8 feet had accumulated.
(Kidston and Lang, 1917, p. 764)

Rhynia gwynne-vaughani* was a small plant (Fig. 2-1A) of about

8 inches in height and grew in dense aggregations with Horneophy¬
ton and Asteroxylon. It lacked leaves and the shoot system was
not clearly differentiated into stems and roots although the under-
* Throughout this book species names will be decapitilized regardless of their

Telegram @tubscstds
etymology, other than in direct quotations.
 PSILOPHYTA AND OTHER EARLY PLANTS

Fig. 2-1. A, B. Rhynia gwynne-vaughani; A. restoration; B. transverse sector of stem

showing small protostele and broad cortex. C, D. Horneophyton lignieri; C. restora¬
tion; D. diagrammatic longitudinal section of sporangium. (A and C from Kidston
and Lang, 1921.)

Telegram @tubscstds
34 STUDIES IN PALEOBOTANY

ground rhizome bore numerous rhizoids. The upright, gradually

tapering shoots in some cases bore terminal sporangia.
Anatomically, the upright shoots, which are about 2 mm in
diameter (Fig. 2-1B) consist of an epidermis with rather sparsely
distributed stomata which are lacking in the rhizome, an outer and
inner cortex, and a minute central stele. The outer cortex is a nar¬
row band about four cells thick; the inner cortex is quite broad,
the cells are smaller and rather large intercellular spaces are present.
It extends to the surface in the vicinity of the stomates and is
thought to have been the assimilating tissue of the plant. In the
center is a small cylinder of annular tracheids, the number varying
with the size of the shoot; the innermost tracheids are somewhat
smaller than the surrounding ones, presumably representing the
protoxylem. Surrounding this is a narrow zone of cells that has
been interpreted as phloem by virtue of their position and elongate
nature, although sieve plates have not been observed. No endo-
dermis or pericycle is recognizable. One other feature of some
interest may also be noted: small hemispherical bulges appear on
the upright stems which were the seat of development of lateral
branches; some of their cells developed into rhizoids and the
vascular supply did not connect with the axis on which they were
borne. They appear to have been readily detached and may have
served the function of vegetative propagation.
The sporangia are terminally borne, attained a length of 3 mm
and a diameter of 1.5 mm, although considerable size variation was
found. The wall consists of a cuticularized epidermis followed by
several layers of thin-walled cells and the enclosed chamber con¬
tains a large number of spores averaging about 40 ju in diameter.
No specialized mode of dehiscence has been noted. It is also of
interest that spores were observed in two instances in unmodified
stem tips; since it seems likely that sporangia evolved here by a
modification of the stem apex these may be regarded as primitive
or rudimentary sporangia.
An associated species, R. major, has been recognized in which
the aerial stems attained a diameter of 6 mm; the stele is larger
and the sporangia reach a length of 12 mm, with spores about 65 jit
in diameter. Hemispherical projections and adventitious lateral
branches are lacking.
In the now rather voluminous literature dealing with it, Rhynia
has been referred to as a plant that is simple, primitive, or reduced.
Since these terms will be frequently encountered it may be helpful

Telegram @tubscstds
to consider their useage at this point.
PSILOPHYTA AND OTHER EARLY PLANTS 35

That Rhynia is simple as a land vascular plant stands unchal¬

lenged. The term refers to its construction and in the lack of dif¬
ferentiation into stems and roots, the absence of appendages that
may be called leaves, the uniform dichotomy of its branching, the
vascular system composed of only a slender strand of annular
tracheids, the sporangia which very possibly are only modified
branch tips, it is, by comparison with any other vascular plant,
living or fossil, one of the least complicated.
That Rhynia is primitive has been questioned and this neces¬
sarily involves the third term, reduced. Rhynia comes from a
Middle Devonian horizon and although some geologists have ques¬
tioned this, thinking that it may actually be of Lower Devonian
age, this does not materially affect our discussion. Regardless of
the precise age we have these rather vexing facts to explain: Land
vascular plants of greater complexity are known to have existed in
the Silurian. One plant (Asteroxylon) of significantly more com¬
plex construction lived in association with Rhynia.
There seem to be at least two possible explanations: First,
Rhynia may have attained its status as a simple land vascular
plant at a much earlier date and has evolved no further for many
millions of years; in which case it is truly primitive. We know that
many plants (the pines are a good example) have carried on with
no significant changes for scores of millions of years, so this explan¬
ation is within the realm of possibility. Second, it may be that
Rhynia has been evolved (‘devolved’ may be more appropriate) or
reduced from more complex land plants.
This is an opportune time to insert a few notations on Psilotum,
a living plant that presents some similarity with Rhynia and has
been the subject of much discussion as to whether it is primitive or
reduced. It is rather widely distributed in the tropics, coming as
far north as Florida in the western hemisphere and Hawaii in the
Pacific region. Psilotum is easily grown under warm humid con¬
ditions, and as a botanical oddity it is often found in greenhouses.
It is a densely tufted plant not attaining a height of much over 1
foot; the rhizome system is densely covered with hairs, and both
rhizome and upright shoot system branch dichotomously. The
leaves are very small and certainly play no significant role in pho¬
tosynthesis. The vascular strand of the smaller shoots is a minute
protostele, whereas in the larger ones it is siphonostelic, often with
a fibrous pith; the outermost tracheids are annular or spiral,
whereas those toward the center are scalariform or have elongate

Telegram @tubscstds
to round-bordered pits. The spore-bearing organs are nearly
36 STUDIES IN PALEOBOTANY

globose, three-chambered structures borne in the axil of a two¬

pronged bract. This is usually referred to as a synangium, a term
applied to a sporangiate organ which is thought to have originated
from the fusion of two or more sporangia. The synangium of
Psilotum and its tracheids with bordered pits present a degree of
complexity that is not found in Rhynia.
Psilotum has been regarded by some botanists as a very simple
land plant, possibly a very ancient type that has managed to sur¬
vive for several hundreds of millions of years. Others view its
simple organization as the result of degeneration from a more
complex ancestor. Thus the former viewpoint regards it as primi¬
tive and the latter as reduced. Nothing is known of its fossil
record and there are no related modern plants which offer much
aid in settling the problem.
Returning to the Rhynie plants, we note that Horneophyton
lignieri (Fig. 2-1C) differed from Rhynia in that its rhizome was a
lobed parenchymatous body with numerous rhizoids but with no
continuous vascular strand of its own. It gave rise to upright
shoots rather similar to those of Rhynia; the vascular supply,
which ended blindly in the upper part of the rhizome lobe, con¬
sisted of an inner core of small tracheids and a peripheral zone of
somewhat larger ones.
The sporangia (Fig. 2-1D), borne on the tips of ordinary branches,
were 1 to 2 mm in diameter and 2 or more mm long. Their most
striking feature is the presence of a columella, a central core of
slender, elongate cells continuous with the stele below but not com¬
posed of tracheidal cells. The spores are about 50 ju. in diameter.
The rhizome of Horneophyton has been compared with an early
stage in the development of the sporophyte of certain species of
Lycopodium known as the protocorm; the protocorm itself is not
vascularized but bears on its upper surface leaves that show no
evident orderly relation to the apex, which are called protophylls.
Each protophyll has a vascularized strand which ends in its base.
The third member of the Rhynie trio is Asteroxylon mackiei, a
somewhat larger plant of perhaps a half meter in height (Fig. 2-2).
The leafless underground rhizome gave rise to a more or less
monopodial upright shoot system; that is, in contrast to the equal
dichotomy of Rhynia there is a strong central stem from which
branches depart and these tend to divide by more or less equal
dichotomies. The larger stems are about 1 cm in diameter and the
numerous cuticularized leaves are about 5 mm long.

Telegram @tubscstds
PSILOPHYTA AND OTHER EARLY PLANTS

Fig. 2-2. Restoration of Asteroxylon mackiei. (From Kidston and Lang, 1921.)

Telegram @tubscstds
38 STUDIES IN PALEOBOTANY

The rhizome stele is a simple cylindrical strand of tracheids with

no evident protoxylem and in the transition to the leafy shoot it
becomes stellate, and strongly so in the larger stems (Fig. 2-3A);
thus branches of different sizes display considerable variation in
the complexity of the stele. Traces depart from the arms although
they penetrate only into the base of the leaf. The middle cortex
may possess rather conspicuous, radially elongate, intercellular
spaces, a tissue that is characteristic of plants growing in moist
places.
Associated with these leafy shoots are slender, branched, leafless
axes bearing terminal pear-shaped sporangia which probably repre¬
sent terminal fertile branches. The sporangia are about 1 mm long.

Psilophyton

In 1859 J. W. Dawson described an early land plant, Psilophyton

princeps, from the Lower Devonian of the Gaspe Peninsula in east¬
ern Canada. Since that time other species have been assigned to
the genus from widely scattered localities. In view of its historical
interest, apparent abundance in early Devonian times, and prob¬
lems of interpretation that still enshroud it, Psilophyton is of more
than passing interest.
As a result of later collections Dawson gave a more detailed
account of the plant in 1871. He now had available three kinds of
fossils: leafless dichotomizing shoots which it was thought might
represent the rhizomatous part of the plant; spiny, apparently up¬
right dichotomizing axes; and a third type bearing terminal spor¬
angia and with markedly fewer spines. Since the fertile specimens
have never been found attached to the more profusely spiny ones,
they have been given the name Dawsonites arcuatus Halle (Fig.
2-5D). Recent critical studies of certain specimens that had been
presented by Dawson to several British museums have demon¬
strated spores in the sporangia. The spines, which range from
16 mm to 2 mm in length, sometimes have a dilated tip with dark
contents suggesting the possibility that they were glandular. A
vascular strand has not been found in the spines nor do they con¬
tain stomates although these are present in the stem.
On a recent trip to the Gaspe I had an opportunity, through the
courtesy of several Canadian paleobotanists, to collect at locations
that Dawson had visited. We observed at one place a great abun¬
dance of fine specimens of the spiny type with a few fertile ones;

Telegram @tubscstds
at another place the spiny and spineless shoots were found at the
 PSILOPHYTA AND OTHER EARLY PLANTS
39

Fig. 2-3. A. A stele of Asteroxylon mackiei in transverse section, 30X. B. A portion

of the stele of Schizopodium davidii showing the radially aligned cells in the peripheral
region of an arm of the wood, 15X.

Telegram @tubscstds
40 STUDIES IN PALEOBOTANY

same horizon; and at a third locality a bed several inches thick was
encountered which contained a great mass of the spineless fossils.
There is still no clear-cut evidence as to the full meaning of these
three shoot types although it is my impression that the smooth
ones probably represent rhizomes, whereas the spiny and fertile
ones are the aerial parts. Continued exploration for new collecting
sites and more complete plants should ultimately result in a satis¬
factory restoration of this interesting plant.
It is by no means certain that all of the fossils that have been
assigned to Psilophyton are closely related to the Gaspe species.
One that does admit favorable comparison is P. wyomingense Dorf
from the Lower Devonian of Wyoming which has been selected for
illustration because of its fine preservation (Fig. 2-4). The upright

Fig. 2-4. Psilophyton wyomingense.

A. A rhizome with three upright spiny
shoots. B. A single upright shoot at a
higher magnification. (From Dorf, 1933.)

Telegram @tubscstds
PSILOPHYTA AND OTHER EARLY PLANTS 41

spiny shoots arise from what was apparently a rhizome, the latter
being more sparsely clothed with appendages.
The botanical world seemingly was not ready to accept so strange
a plant in 1859 and little attention was accorded Dawson’s dis¬
covery until the Rhynie fossils stimulated an interest in such early
floras and left no doubt that the plants of that time were quite dif¬
ferent from any modern ones.

Some Other Psilophytes

In view of the variety of plants that have been assigned to the

Psilophyta it seems essential to refer to Kidston and Lang’s origi¬
nal diagnosis of the group which they described as follows:
It is therefore necessary to recognize another group of Pteridophyta, of
equivalent value to those mentioned, to include Rhynia Gwynne-Vaughani
and certain of the specimens described under the name Psilophyton princeps.
This Class is characterized by the sporangia being borne at the ends of cer¬
tain branches of the stems without any relation to leaves or leaf-like organs.
For this Class we propose the name Psilophytales. (1917, p. 779)

In dealing with the earliest land vascular plants we are of course

working along the fringes of the unknown and one would expect that
concepts of the taxonomic entities involved must change as our
knowledge increases; however, for a family, order, or class to have
any significance or serve a useful purpose it must have at least tenta¬
tive limitations. It may be of interest to keep in mind the original
concept as given above as we consider other fossils that seem to be
related.
Among the oldest and smallest of vascular plants is the genus
Cooksonia, known from two species, C. pertoni Lang and C. hemi-
spherica Lang. They come from rocks of Downtonian age which
have been variously considered as uppermost Silurian, basal Devo¬
nian, or transitional. Cooksonia (Fig. 2-5C) consists of slender,
leafless, dichotomizing shoots up to 6.5 cm long and not exceeding
1.5 mm wide. A thin vascular strand of annular tracheids pene¬
trates the delicate branch system; the terminal sporangia in C. per¬
toni are wider than long, whereas they are nearly spherical in
C. hemispherica.
Hedeia corymbosa Cookson (Fig. 2-5E) was originally described
from the Upper Silurian of Australia and more recently specimens
have been found in the Lower Devonian. All that is known of the
plant is a terminal aggregation of more or less dichotomously fork¬

Telegram @tubscstds
ing branches that formed a three dimensional, corymbose system;
42 STUDIES IN PALEOBOTANY

Fig. 2-5. Fertile terminal branchlets of several psilophytes. A. Taeniocrada dechen-

iana. B. Taeniocrada langi. C. Cooksonia sp. D. Dawsonites sp. E. Hedeia coryrn-
bosa. (A from Krausel and Weyland, 1930; B from specimen in Brussels Natural
History Museum; C from Croft and Lang, 1942; D from specimen in British Museum
of Natural History; E from photographs by Cookson.)

the terminations were mostly fertile but a few sterile tips were
associated.
Taeniocrada decheniana (Goeppert) Krausel and Weyland (Fig.
2-5A) is a Lower Devonian plant from the Rhineland with a dichoto-
mously forking, ribbon-shaped shoot system, a central vascular
strand, and terminally borne sporangia. An approximate idea of
the over-all size of this plant may be gained from the fact that the

Telegram @tubscstds
branches are about 1.5 cm broad and the sporangia are 3 to 7 mm
PSILOPHYTA AND OTHER EARLY PLANTS 43

long. Specimens preserved in the Geological Institute at Cologne

display tremendous quantities of the plant at a particular horizon
and suggest that it probably existed as a dense growth like some
modern aquatics such as the cat-tail (Typha). As additional evi¬
dence of its aquatic habit stomates have not been detected although
the epidermis is fairly well preserved.
Other specimens have been described under the name T. langi
Stockmans from the Lower Devonian of Belgium (Fig. 2-5B) in
which the shoots bear short-stalked, laterally arranged, ovoid bodies
about 3 mm long and 2 mm broad. These have been given the pro¬
bationary status of sporangia, but spores have not been found in
them to settle the point; if this should prove to be the case the ques-
ion arises as to whether this plant should be included in the genus
Taeniocrada. The possibility may also be entertained that these
are specialized vegetative reproductive structures, or even food
storage organs such as are found in the living horsetail, Equisetum
arvense.

PLANTS OF UNCERTAIN AFFINITIES—DOUBTFULLY

REFERABLE TO THE PSILOPHYTA

The plants described below have been referred by some authors

to the Psilophyta. It does not seem to me appropriate to do so but,
in any event, they reveal some of the divergence in morphology of
Silurian and early Devonian plants.
Associated with the Gaspe fossils that he described as Psilophyton
princeps, Dawson encountered several of stouter construction to
which he gave the name P. robustius. Recent investigation of a
specimen preserved in the Hunterian Museum collections in Glas¬
gow has revealed a plant of very distinctive branching pattern which
is now known as Trimerophyton robustius (Dawson) Hopping. The
main axis is about 1 cm wide and the portion preserved is almost
9.5 cm long (Fig. 2-7E). Lateral branches depart in a spiral pattern
and almost immediately trifurcate; each branch again divides into
three but this time unequally. Two of these secondaries assume an
ascending position and dichotomize twice, each ultimate branch
being terminated by a closely compacted cluster of two or three
sporangia; the third one continues out a little farther, dichotomizes,
and likewise is terminated by sporangial clusters. Spores have been
found in the sporangia, leaving no doubt as to the identity of these

Telegram @tubscstds
organs.
44 STUDIES IN PALEOBOTANY

Zosterophyllum has been depicted as having a dichotomously

forking, leafless shoot system with a vascular strand of annular
tracheids; it is possible that the plant depended on water for sup¬
port. The sporangia are aggregated into a terminal spike, which,
in Z. rhenanum Krausel and Weyland, from the Middle Devonian
of the Rhineland, attained a length of 5 cm. In Z. fertile Leclercq
from the Lower Devonian of Belgium it is clear that the sporangia,
which were rather few in number on the spike, were borne on a short
stalk which is recurved so that the sporangia tend to point back
toward the axis. In Z. australianum Lang and Cookson (Fig. 2-6)
the spike was in excess of 4.5 cm long and bore several dozens of
sporangia; the reniform sporangia were attached by short stalks and
were rather large, attaining a width of 4 mm and opened by means
of a slit along the distal edge.
Several other early Devonian plants bear a resemblance to Zos¬
terophyllum in having their sporangia aggregated together into some
sort of spike-like cluster. The sporangia of Bucheria ovata Dorf
(Fig. 2-7B) from the Lower Devonian of Wyoming were arranged in
two rows; thus the spike presents a bilaterally symmetrical habit
in contrast to the spiral arrangement in Zosterophyllum; the spo¬
rangia were sessile and apparently dehisced longitudinally.

Fig. 2-6. The fertile spike of Zosterophyllum austra¬

lianum. (From Cookson, 1949.)

Telegram @tubscstds
PSILOPHYTA AND OTHER EARLY PLANTS 45

Fig. 2-7. Spore-bearing organs of certain early vascular plants. A. Gosslingia breco-
nensis. B. Bucheria ovata. C. Pectinophyton bipectinatum. D. Yarravia oblonga.
E. Trimerophyton robustius. (A from specimen in British Museum of Natural History;
B from Dorf, 1934; C from Ananiev, 1957; D from photographs by Lang and Cookson,
1935; E from Hopping, 1956.)

Protobarinophyton obrutschevii Ananiev is known from dichoto-

mously forking, leafless shoots containing a strand of annular tra-
cheids; at the terminations of the shoots are spikes consisting of two
rows of sporangia which are sessile, apparently elongate-oval, and
dehisced along an apical slit. Pectinophyton bipectinatum Ananiev
(Fig. 2-7C) is known from leafless shoots with a rather distinct
monopodial type of branching, there being a prominent central axis
with side branches which present a combination of monopodial and
dichotomous forking. Some of these are terminated by spikes with
two rows of short appendages. Each one tends to fold inward
making a nearly complete ring, the inner margin of which bears a
flat sporangium. The incurved appendage is considered as the stalk
of the sporangium and may have also functioned as a dehiscence aid
by straightening out at the time of maturity. Both these fossils
come from the Lower Devonian of western Siberia and, although it
is hoped that future discoveries will shed further light on the struc¬
ture of the plants as a whole, they are of particular interest in that
they add to the variety of plants with terminally clustered spo-

Telegram @tubscstds
46 STUDIES IN PALEOBOTANY

rangia. It is conjectural as to just how closely related these plants

may be.
Yarravia oblonga Cookson (Fig. 2-7D) is a member of the
Silurian flora of Australia and is based on axis fragments up to 5 cm
long which bear a terminal synangium of five or six sporangia.
One of the most puzzling of the Devonian fossils is Gosslingia
breconensis Heard, a strange plant from the Lower Devonian of
Wales which is known from small, nearly equally dichotomizing
fragments, the largest of which is shown in Fig. 2-1 A.. The main
axis is about 2 mm wide and the ultimate tip was circinately coiled.
The most novel feature of Gosslingia is the distribution of the sup¬
posed sporangia which are scattered along the branch system. Each
one was borne on a short stalk not more than 1 mm long; the spo¬
rangium itself was about 2 mm wide and 1 mm long and apparently
opened into two valves. A fine strand of tracheids is present in the
slender branches.
Quite a number of Devonian fossils are known from petrified stem
remains in which the reproductive organs are lacking; their exact
relationships are thus difficult or impossible to determine. Never¬
theless some of them present anatomical features of considerable
interest. Schizopodium davidi Harris is a stem from the Devonian
(probably Middle) of Queensland with a stele that is similar in its
general form to that of Asteroxylon. Stems up to 15 mm in diame¬
ter have been found and, at least in the portions preserved, they lack
leaves. In the smallest specimens the wood takes the form of a
small cross, whereas in larger ones it is more complex, being rather
profusely lobed; some of the lobes may become separate entities
resulting in a dissected protostele. Of particular interest is the ar¬
rangement of the tracheids. The scalariform protoxylem cells are
located toward the outer part of the arms; the tracheids toward the
center are metaxylem cells with bordered pits; the peripheral tra¬
cheids are, however, unique. In transverse section (Fig. 2-3B) they
are seen to be arranged in regular radial rows suggesting secondary
origin, but several features indicate that this is not the result of cam-
bial growth of the kind found in most higher plants. The tracheids
vary considerably in structure, some being vertically elongate, some
nearly cubical, and others are radially elongate. In well-preserved
stems there is, moreover, no evidence of crushing of the phloem or
cortical tissues. It is therefore likely that these peripheral tracheids
constitute a tissue that is intermediate between typical primary and
secondary wood, having been formed by a general meristematic zone

Telegram @tubscstds
rather than a narrowly defined cambium.
PSILOPHYTA AND OTHER EARLY PLANTS 47

The cambium is one of the greatest achievements of plant evolu¬

tion and a contributing factor to the success of the plants in which
it is present. Schizopodium seems to offer a clue to the origin of
this tissue.

PRE-SILURIAN PLANT RECORDS

In 1935 Cookson and Lang described a flora from the Silurian of

Australia which stands as a landmark in paleobotany, for before that
time only the most fragmentary vascular plant remains had been
found from horizons below the Devonian. Two members of this
flora, Hedeia and Yarravia, have been cited and in Chapter 8 a third
is described, the lycopod Baragwanathia, a plant of respectable size
with stems several centimeters in diameter. Thus it is evident that
vascular plants of some diversity existed at the time, which initiated
considerable speculation as to how much farther back in the Paleo¬
zoic they might be traced. Intriguing clues were not long in making
their appearance.
In recent years several Indian paleobotanists have described
spores and wood fragments from the Middle and Upper Cambrian
of Kashmir, Spiti, and other Indian localities. Several scores of
spore types have been recorded which are thought to represent
pteridophytic groups and even the seed-ferns, although the latter
seems questionable. These discoveries have been supported by re¬
ports of large assemblages of spores from the Lower Cambrian blue
clay of the Estonia-Latvia-Lithuania area by Naumova in 1949.
Dr. Naumova has informed me that she has found many other
Cambrian spores since that time.
This evidence from spores has been supported by a few fragmen¬
tary macrofossil remains. Probably the most significant of these is
Aldanophyton antiquissimum Kryshtofovich, a supposedly lyco-
podiaceous plant from the Middle Cambrian of the Aldan Mountain
range in Siberia. The plant is represented by shoots up to 13 mm
wide and 8.5 cm long which are covered with microphyllous leaves
attaining a length of 9 mm. Although these specimens leave much
to be desired it seems possible that they do represent land vascular
plants, but since reproductive organs are unknown they cannot be
identified with certainty as lycopods. Association of the plant re¬
mains with trilobites, which are considered to be an index fossil for
the horizon, is the principal evidence for their Cambrian age.
Although this discussion is designated as one dealing with pre-

Telegram @tubscstds
48 STUDIES IN PALEOBOTANY

Silurian remains it seems worth mentioning the discovery in Silurian

rocks of minute plant fragments which seem to represent pieces of
tracheid cells. The large bordered pits that characterize the speci¬
mens (Fig. 2-8B) are quite unlike those found in any really primitive
land plant and suggest fragments from the tracheids of large forest
trees such as a pine or sequoia. It is not implied that they do rep¬
resent these plants but rather that they are tracheidal fragments
of arborescent plants.

B
Fig. 2-8. A. The oldest known structurally preserved fossil plants from the Gunflint iron
formation, Ontario. B. A tracheid fragment isolated from Silurian sediments of New

Telegram @tubscstds
York State. (A from Tyler and Barghoorn, reprinted from Science with permission;
B courtesy of William Evitt.)
PSILOPHYTA AND OTHER EARLY PLANTS 49

In summary it may be noted that, combined with the clearly

recognized Silurian vascular plants, the suggestive evidence from
earlier horizons indicates that vascular plants may have occupied
the land as early as the Cambrian. Many paleobotanists are reluc¬
tant to accept this as yet and it is an exciting paleontological fron¬
tier that will be explored with increasing interest in the years to
come.
It seems pertinent to consider at this point certain evidence per¬
taining to the earliest records of plant life of any kind on the earth.
There are many reports of fossils referred to as algae from pre-
Cambrian rocks, but very few show any recognizable cellular struc¬
ture. Thus the discovery of well-preserved plants in rocks in
southern Ontario that are dated as being 1700 million years old
stands as a significant landmark in our knowledge of the earliest
forms of plant life. These fossils are tentatively assigned to the
blue-green algae and “simple fungi” (Fig. 2-8); it has been possible
to separate them from the rock matrix so that there is no question
of their organic nature. Supposed algal remains have been recorded
from Rhodesia in rocks which have been dated as not less than 2600
million years old. These do not, however, display the clear-cut
structural features of the Ontario fossils. Carbon deposits have also
been dealt with in some detail by Rankama who has presented evi¬
dence in support of their organic origin. Some of these, reported
from Finland, take the form of rings or tubes and have been given
the name Corycium enigmaticum Sederholm. It does not seem ap¬
propriate to this writer to assign a binomial to deposits that are
lacking any recognizable plant form. Rankama has also described
carbonaceous slates from Canada, some of which are also believed
to be biogenic, with an age cited as 2.5 billion years.
As a matter of convenience in reference some of this evidence
pertaining to the earliest known land vascular plants and the earliest
evidence of thallophytic plant life is tabulated below:

Earliest thallophytes Earliest vascular plants

1. Structurally preserved fossils 1. “Baragwanathia” flora

from the Gunflint iron forma¬ from the Silurian of Aus¬
tion, Ontario: 1700 m.y. tralia: about 340 m.y.
2. Algae reported from Rhodesia: 2. Aldanophyton from the
2600 m.y. Cambrian of Siberia:
about 520 m.y.

Based on these records, some simple arithmetic reveals a gap of

Telegram @tubscstds
about 1200 million years between the earliest record of the thallo-
50 STUDIES IN PALEOBOTANY

phytes and that of vascular plants. It is of course hardly necessary

to point out that we are dealing in very “round” figures. Perhaps
the most obvious answer to explain this great gap is that land vas¬
cular plants did exist prior to the Cambrian but, if preserved, they
have not been found thus far; in turn it would have to be admitted
that the Ontario fossils represent a significant period of prior evo¬
lution. Using the facts at hand it appears probable that thallophy-
tic plant life existed for a very long time before vascular plants
ventured forth on the land.
Although it is entirely speculative, one may suppose that long
before the appearance of vascular plants the thallophytic forms oc¬
cupied a wide range of habitats such as marine waters of varying
degrees of salinity, fresh waters, as well as the land. The group that
we call the algae today includes an extremely diversified and versa¬
tile assemblage of plants; they may well have been so long before
the Cambrian and were able to indulge in numerous vascular plant
experiments.

A MISCELLANY OF DEVONIAN PLANTS

Thus far certain fossils have been introduced which hold together
as a natural, if somewhat narrowly defined, group, the psilophytes.
Others have been described that might be considered referable to
the psilophytes, but with less confidence. In this final section there
are gathered together some of the real problems; it seems to me mis¬
leading to classify them, other than in a very tentative way, in the
present state of our knowledge yet they speak vividly for plant
evolution in Devonian times. There is no significance to the order
in which they are described.

Nematothallus

One of the least spectacular of early Devonian plants, yet perhaps

deserving more serious consideration than has been conferred on it,
is Nematothallus pseudovasculosa Lang. In several of its charac¬
ters it is a really primitive vascular plant, yet quite unlike any cited
above. The fossils consist of thalloid compressions that range from
minute fragments up to specimens 6.5 by 1 cm and 4.5 by 2.5 cm,
although even these are evidently parts of a larger plant. Lang’s
investigation of the thallus has revealed a reticulum of fine tubes

Telegram @tubscstds
2 to 4 [x in diameter which are associated with larger ones up to 25 /x
PSILOPHYTA AND OTHER EARLY PLANTS 51

in diameter; the latter had annular thickenings and appear to have

been abundant in the interior, forming rudimentary conducting
tracts. Toward the outside the slender tubes form an increasingly
close network and are covered by a cuticularized epidermallike
layer described as having a “pseudocellular pattern.” Scattered
among the tubes are spores which are also believed to have been
cuticularized. These curious plant fragments present some of the
minimum essentials for life on the land: specialized conducting cells
of a sort, a cuticle as protection against excess water loss, and resist¬
ant spores. If one can put aside for a moment the traditional con¬
cept of a primitive vascular plant in which we seek root, stem, leaf,
and a spore-bearing organ we may see in Nematothallus a uniquely
unspecialized plant in the early stages of becoming established in a
land environment.

Prototaxites

The year 1859 stands as a prominent one in the records of prob¬

lematical fossil plants, for in the same publication in which he first
presented Psilophyton, Dawson also gave a brief description of the
unique Prototaxites which he found along the south shore of the
Gaspe. Since that time at least 13 other species have been de¬
scribed from horizons ranging from the Upper Silurian to the Upper
Devonian. If measured in terms of its enigmatic nature and con¬
troversies that have arisen over it Prototaxites occupies a very
special niche in paleobotanical literature. The Gaspe specimens are
trunks up to 3 feet in diameter and 7 feet long, although the total
length of the plant is not known. Since their preservation is medi¬
ocre the description given here is of P. southworthii Arnold from
Ontario. The trunk or stem consists of a matrix of interlaced fila¬
ments (Fig. 2-9) which are aligned essentially lengthwise of the
trunk; these are about 5 to 9 n in diameter and are occasionally sep¬
tate. Associated with them are larger filaments or tubes up to 50 \i
in diameter; these are quite conspicuous and have much thicker
walls than the smaller filaments. Another distinctive feature is the
presence of fairly uniformly spaced spots about 0.5 mm in diameter
which seem to represent areas of specialized filament organization,
but the cellular structure is poorly preserved.
Although the name Prototaxites was given to these plants by
Dawson because of a supposed resemblance to the wood of the yew
(Taxus) it has been clearly demonstrated since that they are in no

Telegram @tubscstds
way related to the conifers, and most paleobotanists have been in-
52 STUDIES IN PALEOBOTANY

A B

Fig. 2-9. Prototaxites southworthii; A. transverse section; B. longitudinal section;

both about 100X.

dined to favor affinities with the algae, especially the large brown
seaweeds. However, the notion of an alga with a stipe 3 feet in
diameter has not been universally accepted with complete confi¬
dence! The cellular structure allows no close comparison with the
woody tissue of true vascular plants and the habitat that it occupied
has met with controversy, opinions ranging from an aquatic to a
semiaquatic, to a land environment.

Protosalvinia and Foerstia

At a rather restricted horizon in the uppermost Devonian black

shales of the east central states there is a dense concentration of
fossil plants that present an unusual assortment of problems. They
have been assigned to two genera, Protosalvinia and Foerstia (Fig.
2-10). The latter consists of dichotomously forking bodies which
superficially resemble the terminal ends of the common seaweed,
Fucus, whereas Protosalvinia is based on more or less disc-shaped
structures which are, however, not infrequently lobed. Foerstia
specimens occasionally bear tetrads of spores in deep cavities along
the inner margin of the forked tip and similar tetrads are found in
the Protosalvinia discs. The spores and at least the outer layer of
cells in both are highly resistant to macerating chemicals, being

Telegram @tubscstds
impregnated with a waxy substance.
PSILOPHYTA AND OTHER EARLY PLANTS 53

These fossils, known collectively as sporocarps, present a number

of unusual and unsolved problems. Although they extend over a
wide geographical area, at least from Tennessee to Michigan, they
are stratigraphically confined to a few meters of shale and, in abun¬
dance, to a little more than one meter. The word abundance is an
understatement, for at certain localities they make up a significant
portion of the shale; there may be several hundreds exposed on a
square foot of rock. The types are intermingled and it is question¬
able as to whether they represent two distinct genera. As to their

Fig. 2-10. A selection of fossils belonging to the Foerstia (upper half of photo) and

Telegram @tubscstds
Protosalvinia complex, about 2X.
54 STUDIES IN PALEOBOTANY

affinities, they have been regarded by different investigators as

referable to the red algae, as an independent class of thallophytes,
or as having reached a low bryophytic level. Their great abundance
seems to be best accounted for by a free floating habit and rapid
vegetative reproduction; yet the tough, waxy nature of the cell walls
suggests an adaptation to a dry land habitat. They show no evi¬
dence of having vascular tissue.

Some Problematical “Megaphyllous” Devonian Plants

Distributed through the Devonian from Lower to Upper and from

widely scattered localities through the northern hemisphere plant
fossils have been found which have been referred to the genera
Platyphyllum, Cyclopteris, Ginkgophyllum, Psygmophyllum, Ger-
manophyton, and Enigmophyton. There is no reason to assume
that these were closely related, for about the only common feature
they share is the presence of leaflike organs that were rather large,
elongate, fan-shaped structures. Although reproductive organs are
known in only a few, the plant form they present is quite in contrast
to that of the psilophytes, nor are they referable to the other more
clearly defined groups that are considered in later chapters.
Germanophyton psygmophylloides (Krausel and Weyland) H0eg
from the Lower Devonian of Germany is a plant with a branching
axis which bore rather deeply dissected, fan-shaped leaves approxi¬
mately 30 to 40 cm long. Its anatomy is described as consisting of
a parenchymatous ground tissue through which tubelike cells rami¬
fied; it was first described as a species of Prototaxites. Chiefly be¬
cause of the parenchyma, which is not found in Prototaxites, and
the lack of evidence that the parts of the fossil that are known were
borne on large trunks, it was transferred to a new genus, Germano¬
phyton.
Psygmophyllum gilkineti Leclercq, from the Middle Devonian of
Belgium, is a branching stem fragment some 21 cm long which bore
22 leaves. Each leaf consisted of an elongate petiole up to 33 cm
long which expanded into a wedge-shaped blade 10 to 14 cm long
and of about the same breadth. Frequently dichotomizing veins
ramified from the petiole out into the blade.
Enigmophyton superbum H0eg is somewhat better known than
the plants considered above but no less problematical. Its geo¬
graphical origin adds to the interest of the plant, for it was found
at Mimerdalen in Spitsbergen, is of Middle or Upper Devonian age,

Telegram @tubscstds
and is representative of the extensive floras that have thrived in the
PSILOPHYTA AND OTHER EARLY PLANTS 55

Arctic in past geological ages. Enigmophyton is thought to have

attained a height of 1 meter or more; the smooth stems were 5 mm
in diameter and bore lateral branches about 1 to 2 mm thick. The
leaf developed from a narrow base into a partially dissected fan¬
shaped structure (Fig. 2-11) some 16 cm long and 12 cm broad.
Although woody tissue has not been observed in the stems the im¬
pressions of regularly spaced lines, which apparently indicate veins
in the leaves, allow little doubt that it was a vascular plant.
Associated with the vegetative remains are fructifications which
are thought to have been borne as shown in the restoration; the two
are not, however, known in organic connection. It is a bifurcating
spike 2 to 3 cm long bearing numerous sporophylls; there is appar¬
ently but one sporangium on each sporophyll, some of which contain
megaspores up to 250 Ju in diameter, whereas others contain much
smaller spores 60 to 85 ju in diameter. If these fructifications and
vegetative remains are correctly associated one may see evidence
of lycopodiaceous affinities in the spore-bearing organs, but cer¬
tainly nothing to suggest such a relationship as far as the rest of
the plant is concerned.

Fig. 2-11. Enigmophyton superbum, a problematical fossil from the Devonian of

Telegram @tubscstds
Spitsbergen. (From Hoeg, 1942.)
56 STUDIES IN PALEOBOTANY

SUMMARY

It is my impression that the only unqualified statement one can

make about the early land floras is that they included a considera¬
ble variety of plant types and that we are just beginning to under¬
stand them. The diversity is actually greater than is indicated by
this chapter partly because it is not possible to consider them all
and partly because the earliest representatives of the lycopods and
articulates are dealt with in later chapters.
The discovery of structurally preserved plant remains, referable
to algal and funguslike plants (as reported by Tyler and Barghoorn)
from rocks in Ontario which have been dated at about 1700 million
years by the potassium-argon technique is one of the great paleon¬
tological advances of this century. To know with reasonable cer¬
tainty that plants have lived on the earth for one and three-quarters
billions of years is exciting in itself and would have been considered
as fantastic only a few decades ago. Certainly, an immense array
of thallophytic plant forms evolved between that early date and the
latter part of the Silurian period when we catch a first glimpse of
reasonably well-preserved land vascular plants. Now we have tan¬
talizing bits of evidence that vascular plants existed as far back as
the Cambrian and perhaps in late pre-Cambrian times. The pres¬
ent intense interest in these discoveries will probably discredit or
confirm them within the next few years, although even the most
optimistic searchers can hope to find only vestiges of such early
plant life. A reasonable explanation lies in the theory that a great
many upland forms lived in the early Paleozoic which were not
preserved, or the fossil record has been destroyed by erosion.
The problem of how to evaluate the Psilophyta is one that will
also be subject to much debate in coming years. Should we reserve
the category ‘psilophyte’ for a few plants which conform closely to
Kidston and Lang’s diagnosis or should it be evaluated in a broader
sense to include plants such as those described in the second section
of this chapter? One writer suggests a ‘psilophytic stage’ through
which all of the major groups of vascular plants evolved. However,
regardless of the particular classification one may see fit to follow
it is clear that there were many novel forms of vascular plant life
in the early Paleozoic; a few of these survived to expand into numer¬
ous pteridophytic lines in the Carboniferous while many apparently
faded out in blind alleys.

Telegram @tubscstds
 PSILOPHYTA AND OTHER EARLY PLANTS 57

The study of fossils confirms what might be suspected from

a study of living plants alone, namely, that evolution is a continual
fanning process; a line of racial development radiates out in many
directions producing families, genera, and species that become
adapted to an almost infinite variety of environments. In some of
the larger groups, such as the ferns and flowering plants, the diver¬
sification becomes so great that authorities cannot agree as to
whether they originated as a single line (monophyletic) or whether
they are polyphyletic, that is, whether several or many independent
lines served as a source.

REFERENCES

Andrews, Henry N., Jr. 1960. Evolutionary trends in early vascular plants. Cold
Spring Harbor Symposia, 24: 217-234.
-, and Karen S. Alt. 1956. A new fossil plant from the New Albany shale
with some comments on the origin of land vascular plants. Ann. Missouri Bot. Gard.,
43: 355-378.
Ananiev, A. R. 1957. Now Lower Devonian fossil plants from the southeast of west¬
ern Siberia. Akad Nauk USSR (Botany), 42: 691-702.
Arnold, Chester A. 1940. Structure and relationships of some Middle Devonian plants
from western New York. Amer. Journ. Bot., 27: 57-63.
-. 1952. Observations on fossil plants from the Devonian of eastern North
America. VI. Xenocladia medullosina Arnold. Univ. Michigan, Contrib. Mus.
Paleont., 9: 297-309.
-. 1952. Fossil sporocarps of the genus Protosalvinia Dawson, with special
reference to P. furcata (Dawson) comb. nov. Svensk. Bot. Tidskrift, 48: 292-300.
---. 1952. A specimen of Prototaxites from the Kettle Point black shale of
Ontario. Palaeontographica, 93B: 45-56.
Axelrod, Daniel I. 1959. Evolution of the Psilophyte Paleoflora. Evolution, 13: 264-
275.
Cookson, Isabel C. 1935. On plant remains from the Silurian of Victoria, Australia,
that extend and connect floras hitherto described. Phil. Trans. Roy. Soc. London,
225B: 127-148.
-. 1949. Yeringian (Lower Devonian) plant remains from Lilydale, Victoria,
with notes on a collection from a new locality in the Siluro-Devonian sequence. Mem.
Nat. Mus. Melbourne, No. 16: 117-130.
Croft, W. N., and W. H. Lang. 1942. The Lower Devonian flora of the Senni Beds of
Monmouthshire and Breconshire. Phil. Trans. Roy. Soc. London, 231B: 131-163.
Dawson, J. William. 1859. On fossil plants from the Devonian rocks of Canada.
Quart. Journ. Geol. Soc. London, 15: 477-488.
-. 1871. The fossil plants of the Devonian and Upper Silurian formations of
Canada. Geol. Survey Canada, pp. 1-92.
Dorf, Erling. 1933. A new occurrence of the oldest known terrestrial vegetation from
Beartooth Butte, Wyoming. Bot. Gaz., 95: 240-256.

Telegram @tubscstds
58 STUDIES IN PALEOBOTANY

1934. Lower Devonian flora from Beartooth Butte, Wyoming. Bull.

Geol. Soc. Amer., 45: 425-440.
Ghosh, A. K., and A. Bose. 1950. Microfossils from the Cambrian strata of the Salt
Range, Punjab. Trans. Bose Research Institute, Calcutta, 18: 71-77.
-. 1955. Did vascular plants exist in Cambrian times? Nat. Inst. Sci. India
Bull., No. VII, Symposium on organic evolution, pp. 298-303.
Harris, Thomas M. 1929. Schizopodium davidi gen. et sp. nov.—a new type of stem
from the Devonian of Australia. Phil. Trans. Roy. Soc. London, 217B: 395-410.
Hpeg, Ove A. 1937. The Devonian floras and their bearings upon the origin of vas¬
cular plants. Bot. Rev., 3: 563-592.
-. 1942. The Downtonian and Devonian flora of Spitsbergen. Norges Sval¬
bard og Ishavs-Undersokelser, Nr., 83: 1-228.
Hopping, C. A. 1956. On a specimen of “Psilophyton robustius” Dawson from the
Lower Devonian of Canada. Proc. Roy. Soc. Edinburgh, 66: 10-28.
Jacob, K., C. Jacob, and R. N. Shrivastava. 1953. Evidence for the existence of vas¬
cular land plants in the Cambrian. Current Science, 22: 34-36.
Kidston, Robert, and William H. Lang. 1917-1921. On Old Red Sandstone plants
showing structure, from the Rhynie chert bed, Aberdeenshire. Parts I-V. Trans.
Roy. Soc. Edinburgh, vols. 51-52.
Krausel, Richard, and Hermann Weyland. 1930. Die Flora des deutschen Unter-
devons. Abh. Preussischen Geologischen Landesanstalt, Berlin, 131: 1-92.
Krishtofovich, African N. 1953. Discovery of Lycopodiaceae in the Cambrian deposits
of eastern Siberia. (Russian) Doklady Acad. Sci. USSR, 91 (6): 1377-1379.
Lang, William H. 1931. On the spines, sporangia, and spores of Psilophyton princeps,
Dawson, shown in specimens from Gaspe. Phil. Trans. Roy. Soc. London, 219B:
421-442.
-. 1937. On the plant remains from the Downtonian of England and Wales.
Phil. Trans. Roy. Soc. London, 227B: 245-291.
-, and I. C. Cookson. 1930. Some fossil plants of early Devonian type from
the Walhalla series, Victoria, Australia. Phil. Trans. Roy. Soc. London, 219B:
133-163.
-. 1935. On a flora, including vascular land plants associated with Mono-
graptus, in rocks of Silurian age, from Victoria, Australia. Phil. Trans. Roy. Soc.
London, 224B: 421-449.
Leclercq, Suzanne. 1954. Are the Psilophytales a starting or a resulting point?
Svensk Bot. Tidskr., 48: 301-315.
Naumova, Sofia N. 1949. Spores of the Lower Cambrian. Acad. Sci. USSR Bull.
Geol., No. 4: 49-56.
Rankama, Kalervo. 1948. New evidence of the origin of Pre-Cambrian carbon.
Geol. Soc. Amer. Bull., 59: 389-416.
-. 1954. Early Pre-Cambrian carbon of biogenic origin from the Canadian
shield. Science, 119: 506-507.
Read, Charles B. 1936. A Devonian flora from Kentucky. Journ. Paleont., 10:
215-227.
-, and G. Campbell. 1939. Preliminary account of the New Albany shale
flora. Amer. Mid. Nat., 21: 435-453.
Stockmans, Francois. 1940. Vegetaux eodevoniens de la Belgique. Mem. Mus. Roy.
Hist. Nat. Belgique, 93: 1-90.
Tyler, Stanley A., and Elso S. Barghoorn. 1954. Occurrence of structurally preserved

Telegram @tubscstds
plants in Pre-Cambrian rocks of the Canadian shield. Science, 119: 606-608.
the PTEROPHYTA
—early ferns of the Devonian
and Carboniferous

The Classification of Early Fernlike Plants

By mid-Devonian times an assemblage of plants had begun to

evolve that, by virtue of their size and complexity, may be inter¬
preted as transitional between some of the psilophytes of an earlier
age and the fern and seed-fern complexes of the Carboniferous.
This chapter is devoted to these plants and they will be referred to
collectively as preferns; the difficulties encountered in their classi¬
fication are more than compensated by their fascinating morphology
and the light they shed on vascular plant evolution.
The preferns, as well as the plants that are considered in the fol¬
lowing two chapters, are based on a variety of fossil remains, both
compressions and petrifactions; some are remarkably well known,
whereas others are fragmentary and offer only a hint as to what
their affinities may be. There are many beautifully preserved com¬
pression fossils which reveal much about the general morphology
of the plant and the spore-bearing organs, whereas others are
known from well-preserved petrifactions with excellent cellular
preservation.
The stratigraphic sequence in which we find the preferns, ferns,
and seed-ferns is of limited use in understanding the various evolu¬
tionary lines, for there is considerable overlapping in time of the
groups, and it is by no means easy to determine which are primi¬
tive and which advanced. The seemingly primitive preferns extend
into the Upper Carboniferous and the generally more complex
seed-ferns are found well down in the Lower Carboniferous. As to
the true ferns, one extant family, the Schizaeaceae, appears to have

Telegram @tubscstds
had its origin in the Lower Carboniferous and two others, the

59
60 STUDIES IN PALEOBOTANY

Marattiaceae and Gleicheniaceae, are well established in the Upper

Carboniferous. This “stratigraphic mixing” is not uncommon and
usually means that certain members of a group carried on essenti¬
ally unchanged for long periods of time, competing successfully for
a place in the sun with the more advanced lines.
A word of caution is important here concerning the use of sterile
foliage in classification. The true ferns are a diverse assemblage
themselves and very possibly stemmed from several distinct lines
of preferns and in this sense are polyphyletic. There can be little
doubt that fernlike foliage has evolved independently several times
from the prefern complex. Moreover, several leaf types, once con¬
sidered to represent ferns, have been found with seeds attached and
are thus members of the Pteridospermophyta.
The following classification will be followed, although it must be
regarded as tentative.

PTEROPHYTA
Protopteridales | ^ c
~ , .j i 1 Preterns
Coenoptendales |
Anachoropteridaceae, Botryopteridaceae, Stauropteridaceae,
Zygopteridaceae
Cladoxylales
Marattiales
Marattiaceae
Filicales—True ferns
Schizaeaceae, Osmundaceae, Gleicheniaceae, Matoniaceae,
Dipteridaceae (several modern families in which the fossil
record is very sparse are not considered)

A few introductory notes on the terminology used in describing

ferns and fernlike plants seem appropriate. The leaves (fronds),
although extremely varied in size and form, are often large and
highly divided. The main axis of the frond is referred to as the
rachis, and the first and second order branches are called primary
and secondary pinnae, and so on, to the ultimate terminations
which are known as pinnules or leaflets. The proximal portion of
the rachis, between the first primary pinna and the stem, is called
the petiole. Leaves of this general type are called megaphylls,
which means that they are large leaves, in contrast to the micro-
phylls or small leaves of certain other groups, notably the lycopods.
In this discussion of the preferns the term frond will be applied to

Telegram @tubscstds
structures that are regularly arranged on a stem and appear to be
homologous with the fronds of the true ferns.
 EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 61

THE PROTOPTERIDALES

This group was established by H0eg and defined as follows:

Plants of a more or less fern-like habit, the foliar shoots having a tendency
to flatten in one plane, but with little differentiation between axes and
leaves. The ramifications of the foliar shoots may resemble pinnules, but no
broad laminae are found. Sporangia borne terminally on pedicels, forming
panicles or clusters. Sporangium, as far as is known, with apical dehiscene.
Homospory, or in some species probably heterospory. (1942, p. 178)

Professor H0eg referred the following genera to the Protopteri-

dales (to which I have added Archaeopteris): Svalbardia, Proto-
pteridium, Aneurophyton, Eospermatopteris, Rhacophyton, and
Archaeopteris.

Svalbardia

Svalbardia polymorpha H0eg is from the upper part of the Mid¬

dle Devonian of Spitsbergen and seems appropriate to begin with
since it might be considered as an advanced psilophyte or a simple
prefern. It is based on a considerable number of fragments of var¬
ious sizes and degrees of preservation, some of which reveal very
clearly the significant details of the general branching pattern as
well as the ultimate sterile subdivisions (leaves) and spore-bearing
parts. The basal portion of the plant is not known; the largest
axis found is 1.5 cm in diameter and the longest branch fragment
45 cm long; it is supposed that the plant attained a height of as
much as 2 meters. Branching throughout, except for the ultimate
foliar organs, is monopodial; that is, there is a strong central or
main axis in contrast to the equal dichotomous branching in
Rhynia. The primary branches often depart in nearly opposite
pairs, but it is possible that as many as three or four branches
departed at a node. It is certain that the branches were not
arranged in one plane; thus the entire shoot system seems to have
been three-dimensional.
The ultimate sterile branches (Fig. 3-1) consist of a fairly distinct
main axis which bears the “leaves”; the latter may consist of simple
undivided appendages but more usually are dichotomous-forking
structures which divide two to four times. As many as six veins
have been observed in the proximal portion (petiole) and the ulti¬
mate divisions seem to have but one. The leaf as a whole is about
2.5 cm long. It is significant to note that there is a certain amount

Telegram @tubscstds
of variation in the form of these leaves; some are nearly filiform
throughout and others are laminated to some degree.
62 STUDIES IN PALEOBOTANY

Fig. 3-1. Svalbardiapolymorpha, probably a transitional type between the psilophytes

and coenopterid ferns. A, fertile branch; B, sterile branch; both somewhat restored.
(From specimens in the Geological Museum, Oslo.)

The fertile branches bear nearly opposite, ultimate divisions on

which the sporangia are arranged. The latter are 1.5 to 2 mm
long, nearly cylindrical, and borne, singly or in pairs, on a short
stalk. Svalbardia, although larger and more complex than psilo¬
phytes of the Rhynia type, is quite clearly a plant in which the dif¬
ferentiation of leaves as distinct organs had not progressed to a
very advanced degree. The general organization of the fertile
branches is very similar to that of Archaeopteris and although
most species of this genus possess distinct laminate leaflets, A. fis-
silis, from the Upper Devonian of Ellesmere Island and the Donetz
region, has leaflets that are finely dissected and closely approach
those of Svalbardia.

Archaeopteris*

Archaeopteris is a characteristic Upper Devonian plant (Fig.

3-2A) and is represented by 15 or more species which have been

Telegram @tubscstds
* See footnote on page 412.
 EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS

Fig. 3-2. A. Archaeopteris hibernica, a pinna with fertile pinnules in the central
region. B-E. Archaeopteris latifolia: B, C, partially restored fertile and sterile frag¬
ments of fronds; D, E, micro- and megasporangial masses. (A from Schimper, 1874;
B-E drawn from photographs by Arnold, 1939.)

Telegram @tubscstds
64 STUDIES IN PALEOBOTANY

recorded from numerous localities in New York, Pennsylvania, and

eastern Canada, from Great Britain, Ireland, Belgium, Germany,
the USSR (Donetz), and from high latitudes in Ellesmere Island
and Bear Island.
Archaeopteris hibernica (Forbes) Dawson, the type species, was
found at Kilkenny, Ireland, a little over a century ago and speci¬
mens from there are rather widely distributed in European muse¬
ums. The fronds attained a length of over 80 cm and are bipinnate;
that is, the strong central rachis bore primary branches on which
the pinnules are arranged; the latter are more or less wedge-shaped
but vary in different species, some having entire margins whereas
others are lobed to fimbriate. The fertile pinnae are found toward
the basal part of the frond; a pinna may be entirely fertile or par¬
tially so, with sterile pinnules toward the proximal end. The
sporangia are borne in clusters which seem to be homologous to
the sterile pinnules.
Archaeopteris latifolia Arnold (Fig. 3-2B-E) from the Upper
Devonian of Port Allegany, Pennsylvania, is of special interest as it
was in all probability a heterosporous species. The fertile pinnae
have been shown to bear two kinds of sporangia, both of which are
about 2 mm long but some are slender (about 0.3 mm in diameter)
and others notably broader (about 0.5 mm in diameter). Although
the sporangia actually attached to the pinna rachis had shed their
spores, it has been demonstrated that the numerous associated ones
contain spores of two kinds. There are 8 to 16 spores, which aver¬
age about 300 jii in diameter, in the larger sporangia, whereas the
others contain a hundred or more spores of about 35 /x diameter.
It is thus virtually certain that this was a heterosporous plant
although all other species, so far as is known, were homosporous.

Rhacophyton

Professor Leclercq’s study of Rhacophyton zygopteroides has

added a notable chapter to our understanding of the protopterids;
it was a plant with a fairly stout stem (Fig. 3-3) and probably
attained a height of a meter. A dozen or more sterile fronds were
borne near the base of the stem and above this were one or more
fertile ones.
The sterile fronds bore two rows of pinnae, and these in turn
bore two rows of “pinnules” which consist essentially of a branch
system that dichotomizes several times, there being no appreciable
development of a lamina.

Telegram @tubscstds
 EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 65

Fig. 3-3. Rhacophyton zygopteroides. A. Restoration of a portion of a plant showing

the basal parts of several sterile fronds and a single distal fertile frond. B. Basal por¬
tion of a fertile pinna showing two sporangial masses. C. Portion of a sterile pinna
with several “pinnules.” (From Leclercq, 1951.)

The fertile frond is larger and presents several points of particu¬

lar interest. The restoration figure is based on a specimen on
which only one fertile frond was found, and since the stem was
broken above its point of departure it is more than likely that there
were others. The fertile frond has pinnae in two rows, but each
forks immediately after its departure from the rachis; thus at first
glance the frond gives the appearance of having four rows of pri-

Telegram @tubscstds
66 STUDIES IN PALEOBOTANY

mary pinnae. Each pinna branch divides to form secondary pinnae

and pinnules of a pattern very similar to that of the sterile leaves.
In the middle to upper portion of the fertile frond each pinna-pair
bears on its under side, close to the rachis, two rather profusely
dichotomizing branch systems which terminate in sporangia. The
sporangia are about 2 mm long and exannulate, there being a
rather large number in each cluster.

Eospermatopteris

By Upper Devonian times the land flora included plants of

rather large size. The Lycopodophyta was represented by respect¬
able forest trees and Callixylon, possibly an early cordaite relative,
is known from trunks in excess of 5 feet in diameter. The Ptero-
phyta, not to be outdone in this advance toward arborescent forms,
is nobly represented by Eospermatopteris from eastern New York
state. As long ago as 1869 large stump casts of this plant were
exposed near the town of Gilboa as the result of a flood; later in¬
tensive search and quarrying operations brought to light several
scores of them and occasional trunk fragments, a few of which bore
the basal portions of petioles. The plant has thus been regarded
as a tall unbranched tree “fern” which attained an estimated
height of as much as 40 feet, with a loose crown of fronds 6 to 9
feet long (Fig. 3-4). The stumps are bulbous in shape and quickly
tapered to a more slender trunk; for example, one of the largest
measures 3.5 feet in diameter and tapers to a breadth of a little less
than 2 feet at a height of 22 inches. Numerous slender roots
radiated out from the base of the stump to a length of 9 feet. The
fronds that are believed to have been borne by these trunks were
fernlike in their gross morphology, but they lacked a distinct
lamina; that is, the ultimate divisions were dichotomously forking
terminations of the branch system. Some of these terminated in
ovoid bodies up to 6 mm long and 3 mm in diameter; although
originally thought to have been seeds it has been shown that they
were sporangiate organs.
The stumps have been found at several horizons indicating a
succession of forests which existed over a considerable period of
time; they were periodically inundated and partially destroyed by
invasions of the sea. Nothing is known of their internal structure
other than that some stumps display a coarse reticulate pattern
suggesting the cortical fiber strands found in the Carboniferous
seed-ferns.

Telegram @tubscstds
 EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS
67

Fig. 3-4. Restoration of Eospermatopteris from the Upper Devonian of eastern New
York. (From Goldring, 1924, courtesy of New York State Museum and Science Service.)

Telegram @tubscstds
68 STUDIES IN PALEOBOTANY

Aneurophyton germanicum Krausel and Weyland from the

Upper Devonian of Germany has been considered to be closely
related to, or possibly identical with, the Gilboa plant. It is
represented by thrice pinnate fronds with ultimate ramifications
that are little more than dichotomizing terminations of the branch
system. Some fronds were apparently wholly sterile whereas
others bore clusters of sporangia in the proximal region. In its
anatomy the rachis or main axis consists of a central triangular
core of primary wood which is surrounded by secondary wood com¬
posed of tracheids with numerous series of circular (probably bor¬
dered) pits and tall, uniseriate rays. The primary pinnae were
probably borne in three rows, or at least departed so as to build up
a three-dimensional branching frond. Suggestions of possible affin¬
ities of this fossil have been drawn with the Lower Carboniferous
Stenomyelon which is considered to be the stem of a pteridosperm,
as well as the fronds of Eospermatopteris.

Tetraxylopteris

A partially petrified fossil plant, Tetraxylopteris schmidtii Beck,

from the Upper Devonian of New York offers several points of
special interest as a transitional type between the psilophytes and
the more advanced vascular plants of the Carboniferous. The
plant as a whole is believed to have attained a height of about 3
meters. The stems reach 2.5 cm in diameter, occasionally dichoto¬
mized, and bore regularly arranged appendages which are inter¬
preted as fronds. These are three-dimensionally branching organs
with the primary divisions decussately arranged; that is, they were
given off in nearly opposite pairs and the successive pairs alternate,
the arrangement being similar to that of the distribution of leaves
on a horsechestnut twig. The ultimate divisions terminate as
short dichotomized branchlets which are thought to have been
terete in cross section; there was certainly very little flattening and
consequently little or no lamina.
The sporangia, which were borne in rather dense clusters toward
the apical portion of certain fronds, are 2.5 to 5 mm long and 0.5
to 0.8 mm in diameter; there is no evidence of an annulus and, al¬
though spores were not found, the plant is tentatively considered
to have been homosporous.
The first and second order branches of the frond have a central
cruciform primary xylem strand, whereas that of the ultimate

Telegram @tubscstds
branchlets was terete; the rachis was not as well preserved but is
EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 69

thought to have been cross-shaped also. The rachis, as well as the

first order branches, and the basal part of the second order ones,
developed secondary xylem in which the tracheids have several
regular rows of crowded, circular-bordered pits. The wood rays
are uni- to multiseriate and vary considerably in height. A narrow
band of elongate, apparently fibrous cells, was present in the outer
cortex.
In the branching pattern of its shoot system Tetraxylopteris dis¬
plays an early but distinct phase in the differentiation of organs
that can be designated leaves; the secondary wood and cortical
fiber cells are characters suggestive of the pteridosperms. It would
be most helpful if more information could be obtained concerning
the reproductive organs.

COENOPTERIDALES

It is with some reluctance that the plants described here under

this ordinal name are separated from the Protopteridales; the lat¬
ter are known largely from compression fossils whereas the coeno-
pterids are known from petrifications. This is obviously an
unsatisfactory basis on which to segregate fossil plants. It is more
than likely that there is overlapping of the plants included in the
two orders but, in the light of our present knowledge, this seems to
be the most expedient way of dealing with them.
The coenopterids are characterized by protostelic stems, fronds
that are three-dimensional in their branching pattern and usually
lack a lamina, and with terminally borne sporangia. There are
exceptions to these features in the plants that are included within
the group. A particular problem lies in the fact that in certain
ways, such as size, complexity of their fronds, and stem anatomy
(where known), they are of somewhat simpler organization than
the Protopteridales, yet the coenopterids are predominantly Car¬
boniferous with some members found in the Permian. It is pos¬
sible that some of them are reduced rather than truly primitive
and the smaller size of many of them may simply mean that we
are dealing with fragments of larger plants.

Botryopteridaceae

Botryopteris is one of the better known genera of coenopterids

Telegram @tubscstds
and is represented by several species ranging from the Lower Car-
70 STUDIES IN PALEOBOTANY

boniferous to the Permian. The stem of Botryopteris trisecta

Mamay and Andrews (Fig. 3-5A) is about 8 mm in diameter, with
a small terete protostele surrounded by a broad cortex, a portion
of which is distinguished by a conspicuous band of dark sclerotic or
secretory cells. The stele produces, at 120° intervals (Fig. 3-13A),
an oval-shaped vascular strand which gives rise to two laterals; the
central strand of these three becomes the trace of the rachis and
the two laterals each trisect again but in a plane that is essentially
at right angles to the previous division. The proximal member of
this trisection assumes a terete form, resembling closely the struc¬
ture of the stem stele.
In the figure the lateral or “primary pinna” at the left shows the
division of the wood into three segments; in the lateral to the right
the strand is just starting to divide; the central rachis trace is
clearly M-shaped. It may also be noted that the next leaf trace
has just departed from the stem and is in an insipient stage of divi¬
sion itself.
The primary pinna continues to divide and apparently termi¬
nated in slender cylindrical branchlets with a very small central
xylem strand. However, some specimens of the main rachis have
been found with subdivisions that terminate in distinct pinnules.
We may then visualize B. trisecta as a small plant, possibly
epiphytic, with three-dimensional fronds (that were at least par¬
tially laminated) arranged on the stem in a close spiral.
Some of the prefems show a tendency to produce their sporangia
in clusters; this is carried to an extreme in Botryopteris which
renders it unique in the plant kingdom. Botryopteris globosa
Darrah is a species based on such a sporangiate cluster; specimens
have been found at the same Pennsylvanian locality in Illinois
from which B. trisecta was derived and it is very possible that the
two represent one plant; other specimens have come from Iowa and
Kansas. B. globosa consists of a massive globose aggregation of
thousands of sporangia, about 5 cm in diameter (Fig. 3-5B, C). A
profusely divided branch system ramifies the fructification with a
central strand containing the characteristic Botryopteris M-shaped
rachis trace. The densely packed, pear-shaped sporangia are at¬
tached in clusters at the terminations of the ultimate ramifications
of this branch system.
Aside from its tremendous size the chief peculiarity in this organ
lies in the modification of the peripheral sporangia (Fig. 3-5B). In
most cases the annulus covered the distal three-quarters of the

Telegram @tubscstds
sporangium with the exception of a narrow longitudinal band of
 EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 71

Fig. 3-5. Botryopteris. A. A cross section of B. trisecta including the stem (s) and basal
portion of a frond; the central rachis (r) presents the characteristic M-shaped trace;
the strand of the primary pinna at the left (pi) has trisected and the one at the right
(pr) is in the initial stages of doing so. B. A small portion of the sporangial aggregate
of B. globosa showing the peripheral sporangia (p) with their large annulus cells and
the interior sporangia (i) which are filled with spores. C. A terminal branchlet of the

Telegram @tubscstds
sporangial aggregate. (A from Mamay and Andrews, 1950.)
72 STUDIES IN PALEOBOTANY

thin-walled cells that functioned as the line of dehiscence. The

outermost two or three layers of sporangia are, however, quite dif¬
ferent in that the annulus cells are tremendously enlarged. Many
years ago Bernard Renault described this transition in a French
species, B. forensis, and came to the conclusion that the peripheral
sporangia were sterile ones that served a protective function.
Recent studies based on the American material described here sug¬
gest that the difference in structure is a maturation process; the
enlargement of the annulus cells is gradual through the peripheral
region of the entire sporangial aggregate and, although the outer¬
most sporangia, with their tremendous annulus cells, have dehisced,
some retain a few spores. If the spores in the vast bulk of the in¬
terior sporangia were to have been dispersed, it would seem as
though the fructification as a whole must have expanded (for
which there is no evidence) or the outermost sporangia progres¬
sively matured, shed their spores, and then fell away. A final
problem lies in the mode of attachment of the massive sporangial
aggregate; the central strand with its M-shaped trace attests to its
correct assignment to Botryopteris but we do not know how it was
borne.
Adding to the complexity of this genus, Surange has described in
three species (B. antiqua, B. ramosa, and B. elliptica) a relation¬
ship in which an organ with a bilaterally symmetrical stele (shal¬
lowly M-shaped) produced appendages with a radially symmetrical
stele (protostele) which in turn gives off spirally arranged struc¬
tures with the M-shaped strands. This is interpreted as a trailing
organ (dorsiventral stem) which gives rise to a radial upright stem
that in turn produces petioles. Regardless of the names one applies
to this sequence of morphological entities, its real significance
seems to lie in the fact that we are dealing with plants in which
the differentiation between stem and leaf was not as clear-cut as in
most modern plants.

Anachoropteridaceae

The plants assigned to this family have protostelic stems and

abaxially turned petiole traces, that is, the trace is curved with the
convex side facing out.

Tubicauiis

This is a genus of about five valid species with terete protosteles

Telegram @tubscstds
and with petioles in which the C-shaped strand presents the unu-
EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 73

sual abaxial orientation; note, for example, the contrast in this

character between Tubicaulis and the osmundas (see page 114).
In Tubicaulis scandens Mamay from the Upper Pennsylvanian
of Illinois the trace departs from the periphery of the stem stele as
a bar-shaped strand which gradually assumes the form of a shallow
C and bears two rows of pinna traces. It grew as an epiphyte on a
Psaronius plant; a Tubicaulis specimen from Chemnitz, Germany,
has been reported which occupied a comparable habitat.
Tubicaulis stewartii Eggert (Fig. 3-6), also an Illinois species, was
probably an upright plant with stems 1 cm in diameter and xylem
in which a considerable amount of parenchyma was associated with
the tracheids. Of particular interest is the conspicuous lacunar
middle cortex of the stems and petioles; the presence of such air
passages suggests rather strongly an aquatic or semiaqua tic habitat
for the plant.
In contrast to the species cited above, which seem to have been
quite diminutive, T. solenites, the type species described by Bern-
hard Cotta in 1832, was a large upright plant, the known fragment
being 40 cm high and with a basal diameter of 14 cm.

Apotropteris

Apotropteris minuta Morgan and Delevoryas from the upper

Pennsylvanian of Illinois has “inverted” traces like those of
Tubicaulis but differs in several significant ways. The stem con¬
tains a minute protostele with a centrally located protoxylem and

Fig. 3-6. Tubicaulis stewartii. The stele of the stem appears at s with several petioles
toward the right, 3X; the characteristic aerenchymatous tissue may be seen in the en¬

Telegram @tubscstds
larged petiole (inset).
74 STUDIES IN PALEOBOTANY

the leaf traces depart at irregular intervals in contrast to the

closely compacted phyllotaxy of Tubicaulis. At the node the stem
stele is bisected, leaving a deep indenture by the departing petiole
strand, but regains its terete form above the node.

Anachoropteris

The indecisive stem-leaf structure of Botryopteris is not confined

to that genus of coenopterids. In Anachoropteris clavata Graham
(Fig. 3-7) from the Upper Carboniferous of Illinois a “petiole” has
been observed to produce on one side of its C-shaped strand a
“stem” with a terete stele which in turn bears appendages with
C-shaped strands, although appreciably smaller than those of the
first order. It has been suggested that some petioles of the plant
may have functioned as stolons producing new stems in somewhat
the same way some modern plants do.

Grammatopteris

A possible connecting link between the coenopterids and the

Osmundaceae is found in Grammatopteris baldaufi (Beck) Hirmer.
The fossil was originally found in a Lower Permian deposit at Hil-
bersdorf, near Chemnitz; it was cut into several pieces and ulti¬
mately scattered among several private and public European collec¬
tions. These were later located through the efforts of Professor
Sahni from whose work the following description is taken. The
plant was a small tree-fern with a stem that was enclosed in a heavy
armour of cylindrical petioles 6 to 8 mm in diameter. The stem has
a protostele about 4 mm in diameter composed chiefly of elongate,
pitted, or scalariform tracheids but with an admixture of broad short
ones in the central portion. The middle cortex is characterized by
crowded nests of sclerotic cells.
Features that suggest affinities with the Osmundaceae are: the
structure and closely compacted orientation of the petiole bases;
“incisions” in the outer part of the stele which may be interpreted
as rudimentary leaf gaps; “parenchymatous” tracheids in the center
of the stele which suggest the earliest origin of a pith (a somewhat
more extensive differentiation is found in Thamnopteris).

Stauropteridaceae

Stauropteris burntislandica P. Bertrand from the Lower Carbon¬

iferous of Pettycur, Scotland is of special importance as the only

Telegram @tubscstds
known heterosporous coenopterid. The central shoot of the plant
 EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 75

Fig. 3-7. Anachoropteris clavata. Restoration (vascular system

only) of a primary petiole with a stem departing at the left; the lat¬
ter is giving off four secondary petiole traces and several root strands.
(From Delevoryas and Morgan, 1954.)

is usually regarded as a rachis and the stem that it is presumed to

have been borne on has not been found; the petiole is small, meas¬
uring about 2 mm in diameter (Fig. 3-8) and its vascular strand is
distinctively four-lobed. Branch traces depart at intervals of about
2 mm from opposite sides of the rachis strand; these almost imme¬
diately divide into two, resulting in a four-ranked branching system,
or perhaps it is more accurately described as consisting of two
double rows.
The discovery of heterospory in this plant by K. R. Surange
constitutes one of the notable advances in our knowledge of the
coenopterids. The megasporangium (Fig. 13-5) is a slightly asym¬
metric, ovoid organ with a tapered tip and measures about 1.3 mm
long and 0.5 mm in maximum diameter. The basal two-thirds is

Telegram @tubscstds
76 STUDIES IN PALEOBOTANY

Fig. 3-8. Stauropteris burntislandica, transverse section of central rachis, 50X.

parenchymatous with a very slender, central vascular strand. The

distal part consists of a cavity containing two relatively large
megaspores.
The history of the investigation of this plant is an interesting one
and the final chapter or chapters remain unwritten. The structures
that have been demonstrated to be megasporangia were first inves¬
tigated by R. Scott in 1908 who suggested that they might be
megasporangia but discarded this concept in favor of a glandular
nature; Surange’s studies revealed the undoubted presence of two
large megaspores and more recently W. G. Chaloner found two
aborted spores associated with them, thus demonstrating the pres¬
ence of a tetrad in which only two spores mature. The latter
author has also discovered closely comparable spore tetrads from
Carboniferous (Dinantian and Namurian) coals in Ireland, Scotland,
and England and more recently still they have been reported by
J. M. Schopf from an Upper Mississippian coal in western Kentucky.

Telegram @tubscstds
A consideration of the possible significance of these curious mega-
EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 77

sporangia in the evolution of the pteridorperm seed is taken up in

Chapter 13.
The microsporangia are also terminally borne and are nearly
spherical structures about three-quarters of a millimeter in diame¬
ter. The numerous microspores measure 40 to 50 n in diameter in
contrast to approximately 170 /x for the megaspores. In the mate¬
rial of S. burntislandica that I have personally had available to
study the microsporangia are quite scarce although the megaspo¬
rangia are abundant.
The Upper Carboniferous Stauropteris oldhamia Binney appears
to have been homosporous since sporangia, closely comparable with
the microsporangia of S. burntislandica, are the only reproductive
organs that have been found on the plant.

Zygopteridaceae

Ankyropteris

Eight species of Ankyropteris have been described, several of

which display features of particular interest. A. glabra Baxter
(Fig. 3-9) from the Pennsylvanian of Indiana is known from a stem
specimen about 29 cm long which gave off four leaves through this
distance; the internodes are thus quite long. The stem is about
12 mm in diameter and contains a lobed, mixed protostele. In the
broad cortex several dozens of minute vascular traces can be ob¬
served passing out to peripheral scale-like appendages which have
been designated as aphlebiae or protective scales.
The nodal anatomy of Ankyropteris is worth considering in some
detail since it seems to give us a clue to the origin of axillary
branches, one of the distinguishing morphological features of
most of the higher plants. The origin of the axillary branch stele
in A. glabra is somewhat variable; at some nodes a peripheral lobe
of the stem stele departs and quickly assumes the characteristic
cross-sectional (—) shape of the petiole (also referred to as a phyllo-
phore). A few millimeters above this point on the stem a terete
strand departs which presumably supplies an axillary branch. In
other specimens from the same locality it has been observed that a
“common trace” departed from the stem stele at the node, and this
quickly divided to form the strand of the petiole (phyllophore) and
axillary branch.
Axillary branching is one of the most characteristic features of
the modern conifers and angiosperms, whereas in the extant ferns

Telegram @tubscstds
it has been reported in the Hymenophyllaceae and the Ophioglos-
78 STUDIES IN PALEOBOTANY

Fig. 3-9. Transverse section through the node of Ankyropteris glabra, 6X; s, stem
stele; b, branch stele; p, strand of petiole. (From Baxter, 1951.)

saceae. In Botrychium lunaria the branch stele departs from the

leaf trace shortly after the latter has left the stem stele, whereas in
Helminthostachys the branch stele originates directly from the stem
stele. In these living ferns the branch develops from vestigal buds
only if the apical meristem of the stem is destroyed.
On the basis of the little evidence that is available it is my sup¬
position that the axillary branch originated as a modification of a
basal “pinna” and probably at a time when the megaphyllous leaf
itself was evolving. It is evident, however, that much more research
is needed to elucidate this important morphological structure.
An early Pennsylvanian species, A. hendricksi Read, has been
found in Oklahoma which was probably a rather massive upright
tree fern. The only known specimen is a fragment of a trunk in¬
cluding a stem about 3 cm in diameter with a heptagonal stellate
stele; several petioles or “phyllophores” are associated with the stem
and are held together in a dense matrix of epidermal hair and adven¬
titious roots. The specimen measures about 12 X 8 cm in cross
section, but the position of the stem on one edge indicates that this

Telegram @tubscstds
is only a portion of the original trunk.
EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 79

Zygopteris

The nomenclature of some of the fossils assigned to this family

of coenopterids is involved because several names have been given
to different parts of the same plant by earlier investigators who had
only fragmentary material available. The plants in question present
several novel features and add appreciably to our knowledge of the
group; it is thus necessary to introduce an unfortunate confusion
of names. In 1832 Cotta described Zygopteris primaria as the
trunk of a small tree-fern from the Lower Permian of Floha, near
Chemnitz. The specimen at his disposal was from the terminal part
of the plant and displayed only leaf bases. As a result of searching
through several European museums and performing a clever bit of
correlation Professor Sahni demonstrated the existence of other por¬
tions of the specimen which gave a more complete picture of the
entire trunk. When the scattered pieces were fitted back together
it was shown to be a trunk about 20 cm in diameter, consisting of
a stem only 1.5 cm in diameter surrounded by a thick armor of leaf
bases. The xylem of the stem is pentagonal in cross section with a
stellate core of primary wood only 1 mm thick which is surrounded
by radially aligned secondary tracheids.
The name Botrychioxylon paradoxum next enters into the story
as one given by D. H. Scott for petrified stems found in the Lower
Coal Measures of Oldham, England. His original description was
based on a dichotomously branching stem, the stele of which con¬
sists of a small mixed protostele surrounded by a zone of secondary
wood; the petiole trace, although not well preserved, was described
as zygopterid and was accompanied at first by secondary wood as
is the case in Z. primaria. In describing the stem structure of
Z. primaria Sahni noted a close similarity with Botrychioxylon and
indicated that the latter genus probably would never have been
created had Z. primaria been fully understood at the time. He also
noted the identity of the petiole genus Etapteris with the petiole
of Z. primaria.
Some years ago the present writer described a stem from an
American coal ball under the name Scleropteris illinoiensis which
has been shown by R. W. Baxter to be very similar to the English
Botrychioxylon, and he has also demonstrated the attachment of
Etapteris-type petioles and has given the fossil assemblage the name
Zygopteris illinoiensis. It may be noted, however, that this species
probably had a trailing stem with widely spaced nodes and was thus

Telegram @tubscstds
quite different from the erect treelike Z. primaria.
80 STUDIES IN PALEOBOTANY

Fig. 3-10. Zygopteris illinoiensis; the stele includes a significant development of sec¬
ondary wood (s), 8X.

Austroclepsis

Another arborescent coenopterid, Austroclepsis australis (Osborn)

Sahni, from the Carboniferous of New South Wales attained a
height of 9 to 12 feet with a diameter of 1 foot. The trunk was con¬
structed of several dichotomously branching stems, each with a stele
similar to that of the English Ankyropteris grayi, as well as numer¬
ous leaf bases and many roots. This plant should perhaps be con¬
sidered as a probable coenopterid; if correctly assigned to that group
it extends the known range into the southern hemisphere.

Telegram @tubscstds
EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 81

Biscalitheca

It has been pointed out that the aggregation of sporangia into

dense masses characterized at least some of the protopterids and
coenopterids. Biscalitheca musata Mamay is based on a sporangial
aggregate which measured 10 X 5 X 2 cm and the unique com¬
plexity of the sporangium wall is quite unlike that found in any
other plant. Each sporangium is banana-shaped, being 4 mm long
by 1 mm in diameter. A conspicuous multiseriate (several cells
wide but one cell in thickness) annulus extends along each side. The
tissue between the annuli consists, on the upper surface, of long,
narrow cells, whereas those of the lower surface are more nearly iso-
diametric; distributed among these “interannular” cells are tiny
groups of smaller and somewhat thicker walled cells. The sporan¬
gium wall is, therefore, constructed of three strikingly different cell
types and is one of the most complex in the entire plant kingdom.
There is a fairly close comparison between Biscalitheca and spo¬
rangia borne on Etapteris lacattei and it is primarily on this evidence
that the plant is included in the Zygopteridaceae.
The sporangium contained large numbers of spores and many
were observed in which gametophyte development had progressed,

Fig. 3-11. Restoration of the sporangium of Biscalitheca musata. (From Mamay,

1957.)

Telegram @tubscstds
82 STUDIES IN PALEOBOTANY

within the spore, up to a 10-cell stage. In a few the spore coat had
ruptured and papillate protrusions indicate the initiation of exo-
sporal growth.

Cladoxylales

With the exception of some of the Carboniferous lycopods and

articulates there are few pteridophytic plants which develop very
much secondary vascular tissue; some, however, attain considerable
size through the proliferation of the primary xylem system. The
fossils described below display considerable variety of stelar form,
and it is very possible that when the plants as a whole become bet¬
ter known they will be revealed as a rather diverse assemblage.

Cladoxylon

Cladoxylon is a genus of numerous species of stems, with complex

stelar system, that range from mid-Devonian through the Carbon¬
iferous. The most comprehensive account is Paul Bertrand’s study
of the cladoxylons of Saalfeld and it is another interesting example
of the problems involved in fitting together isolated plant parts.
Cladoxylon radiatum (Unger) Bertrand (Fig. 3-12) from the basal
part of the Carboniferous is a small stem about 1 cm in diameter
with a stelar system consisting of about 18 slender, sheetlike xylem
segments which tend to radiate out from the center; these consist
of primary wood only. As in all of the cladoxylons there are con¬
spicuous peripheral “loops” near the distal end of each stelar seg¬
ment (meristele); these were probably occupied by the protoxylem
elements admixed with thin-walled parenchymatous cells. Rather
large branches were initiated by the division of the distal part of
several of the meristems. C. taeniatum (Unger) Bertrand (Fig. 3-12)
is a larger stem, up to 5 cm in diameter. The stelar system consists
of five small, central meristeles that are cylindrical in cross section;
they are surrounded by numerous peripheral ones that are radially
elongated to various degrees. In contrast to C. radiatum, each of
the primary steles is surrounded by a band of secondary wood; its
tracheids are scalariform and interspersed with parenchymatous
rays. Appendages are given off in which the stelar segments tend
to be arranged in a bilateral symmetry; these had been described
previously from fragmentary material under the generic names
Hierogramma and Syncardia.
A Middle Devonian species, C. scoparium, from Germany, has

Telegram @tubscstds
been described as having a stelar system that was apparently simi-
 EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 83

Fig. 3-12.Diagrams of the stems showing the stelar systems of: A. Cladoxylon taen-
iatum (about 2X) and B. C. radiatum (about 6X); solid black indicates primary wood.
(From Bertrand, 1935.)

lar to that of C. radiatum. This description was based on compres¬

sion fossils, and in their restoration Krausel and Weyland show it
as an irregularly branching plant with two kinds of small append¬
ages, both sterile ones that were several times forked and probably
photosynthetic as well as others that were dissected, fan-shaped
structures which probably bore sporangia along the margin. Better
preserved specimens are needed to substantiate fully the restoration
of this plant and Bertrand held the opinion that it was not closely
related to the Saalfeld fossils.
It seems likely that these fossils represent more than one genus
and their curious and varied appendages cannot be closely identified
with the usual leaf-stem categories of the higher plants.

Xenocladia

Xenocladia medullosina Arnold from the upper Middle Devonian

of Erie County, New York, is based on a stem fragment measuring
1X5 cm, although it is apparently only a portion of the original.
It is polystelic with a central core of units that are more or less circu¬
lar in cross section, and peripheral to these are more or less radially
elongated ones. Each stele consists of a small central mass of pri-

Telegram @tubscstds
84 STUDIES IN PALEOBOTANY

mary tracheids surrounded by secondary xylem in which the tra-

cheids have two rows of bordered pits.

Steloxylon

This is another Upper Devonian genus of polystelic stems, some¬

what similar to Xenocladia, but the numerous anastomosing steles,
which are surrounded by a parenchymatous tissue, are all nearly
cylindrical. In S. irvingense Read and Campbell the individual
strands range from 0.8 to 1.7 mm in diameter, whereas in S. sanctae-
crucis R. and C. they reach 3.8 mm in diameter.

Pietzschia

Pietzschia polyupsilon Read and Campbell from the Upper De¬

vonian of Kentucky is known from stems about 25 mm in diameter
in which the vascular system consists of a single peripheral ring of
about 54 radially elongate strands of primary wood only. In the
development of the vascular supply of an appendage four of the
strands pinch off a segment of xylem and these fuse to form a strand
with the form of an inverted U.

THE EVOLUTION OF THE MEGAPHYLLOUS LEAF-

THE TELOME CONCEPT

Thus far we have dealt with plants of a very ancient lineage

which in many cases cannot be described with the same terminology
that is used with extant ones. Prior to the discovery of the Rhynie
fossils it had long been accepted that vascular plants were composed
of roots, stems, leaves, as well as reproductive organs which were
generally regarded as modified leaves of one sort or another. This
older philosophy of plant form was based largely on studies of the
flowering plants, which did not make their appearance in any abun¬
dance until mid-Mesozoic times, and grew out of the ideas expressed
by the German poet-philosopher Goethe in his treatise on the meta¬
morphosis of plants or, as Sporne aptly expresses it, as “a slight mis¬
representation of those ideas.”
If the concept of organic evolution is to be applied to the higher
plants we might expect that in earlier geologic periods we should
find traces of the origin of the root-stem-leaf differentiation that
does characterize most of the living pteridophytes and seed plants.
This is precisely the contribution that the psilophyte and prefern

Telegram @tubscstds
assemblages of fossils are beginning to make. Although the infor-
EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 85

mation at hand is only a beginning, significant gaps have already

been filled in and we are able to catch glimpses of how modern
plants evolved. An especially significant aspect of this evolution is
the light that is shed on the development of the megaphyllous leaf,
that is, the leaf type that is found in most true ferns and seed-ferns.
The sketches given in Fig. 3-13 illustrate some of the apparent
stages in the evolution of such a leaf. As a starting point a simple,
dichotomously forking shoot system such as Rhynia (Fig. 2-1) may
be taken. Here there is no distinction between stem and leaf and
the ultimate units of the system are now generally referred to as
sterile or fertile telomes, depending on whether or not a terminal
sporangium is present. This morphological concept originated from
the work of the able German botanist, Walter Zimmermann; a brief
and readable account is found in the reference cited at the end of
the chapter. It is a most useful term with the primitive vascular
cryptogams where the terminology of living plants proves inade¬
quate; its application to the higher plants seems to me unfortunate
since it is by no means clear as to what telomic units are in all cases;
consequently, their homology with telomes in the lower groups is
obscure.
One of the first steps in the evolution of the leaf seems to have
been the differentiation of the shoot system into a central axis with
secondary units that are arranged in a regular fashion. These units
or appendages were at first three dimensional in their branching pat¬
tern and probably nonlaminate; Botryopteris trisecta may be taken
as an example (A).
Confining our attention to a secondary unit or appendage the next
stage that may be recognized is known as planation. The three-
dimensional branch system is now flattened into a two-dimensional
one and as an example a portion of the frond of Telangium affine
is shown in diagram B. (It will be evident that some liberty is
taken with the strict rules of geometry in calling a flat leaf two-
dimensional since it of course does have some thickness!) It will
be noted here that there is still little evidence of the development
of a blade or lamina.
Most modern fern fronds are monopodial in their branching; that
is, they possess a strong, straight, central rachis from which the
primary branches depart. This originated through a modification
known as overtopping; instead of equal dichotomies one fork of each
branch is conspicuously stronger than the other. This may be ob¬
served occasionally in certain living fern leaves. As an example a

Telegram @tubscstds
fragment of the Carboniferous Lonchopteris bricei is given in Fig. C.
86 STUDIES IN PALEOBOTANY

Fig. 3-13. Some phases in the evolution of the megaphyllous leaf. A. A portion of the
stem of Botryopteris trisecta with the basal parts of three fronds. B. A representative
part of a leaf of Telangium affine; note two-dimensional form, dichotomous branching,
and lack of distinct pinnules. C. A frond fragment of Lonchopteris bricei with unequal
dichotomies. D-G. Pinnules of Rhacopteris species: D, R. geikiei; E, R. inaequilatera;
F, R. flabellata; G, R. lindseaeformis. (B from Miller, 1857; C from Kidston, 1911;
D-F from specimens in the British Museum of Natural History; G from Kidston, 1923.)

Telegram @tubscstds
EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 87

The development of a distinct blade (pinnule) appears to have

resulted from the webbing together (called syngenesis) of a simple
telomic system. As an example the pinnules of four species of
Rhacopteris are shown in Figs. D-G.
It should be emphasized that this evolutionary “series” from a
completely undifferentiated shoot system such as that of Rhynia
to a fern or seed-fern in which the fronds are monopodial, flat, lami¬
nated organs arranged in a regular pattern on an axis (stem) is out¬
lined only in a very general way. Many other examples might,
however, be given; for example, in Rhacophyton and Tetraxylopteris
there is a distinction between stem and leaf in that we find morpho¬
logical units arranged in a regular fashion along a main axis. The
“leaves,” however, are best described as telomic branch systems
since they are three-dimensional and nonlaminate. This type of
leaf-stem relationship seems to have been characteristic of many of
the protopterids and coenopterids.
There is now no reasonable doubt that megaphyllous leaves
evolved along this general path but it was a kind of evolution that
was going on in many different subgroups in slightly different ways
and at different rates. The examples selected reveal only the trend
as a whole.

REFERENCES

Arber, Agnes. 1946. Goethes’Botany. The Metamorphosis of Plants 1790. Chronica

Botanica, 10: 67-115.
Arnold, Chester A. 1939. Observations on fossil plants from the Devonian of eastern
North America. IV. Plant remains from the Catskill delta deposits of northern
Pennsylvania and southern New York. Univ. Michigan, Contrib. Mus. Paleont.,
5: 271-314.
_. 1952. Observations on fossil plants from the Devonian of eastern North
America. VI. Xenocladia medullosina Arnold. Univ. Michigan, Contrib. Mus.
Paleont., 9: 297-309.
Baxter, Robert W. 1951. Ankyropteris glabra, a new American species of the Zygop-
teridaceae. Amer. Journ. Bot., 38: 440-452.
_. 1952. The coal-age flora of Kansas. II. On the relationships among the
genera Etapteris, Scleropteris and Botrychioxylon. Amer. Journ. Bot., 39: 263-274.
Eggert, Donald A. 1959. Studies of Paleozoic ferns. The morphology, anatomy and
taxonomy of Ankyropteris glabra. Amer. Journ. Bot., 46: 510-520.
_. 1959. Studies of Paleozoic ferns. Tubicaulis stewartii sp. nov. and evolu¬
tionary trends in the genus. Amer. Journ. Bot., 46: 594-602.
Beck, Charles B. 1957. Tetraxylopteris schmidtii gen. et sp. nov., a probable pterido-
sperm precursor from the Devonian of New York. Amer. Journ. Bot., 44. 350-367.

Telegram @tubscstds
88 STUDIES IN PALEOBOTANY

Bertrand, Paul. 1935. Contribution a l’etude des Cladoxylees de Saalfeld. Palaeon-

tographica, 80B: 101-170.
Chaloner, William G. 1958. Isolated megaspore tetrads of Stauropteris bumtislandica.
Ann. Bot.. n. s., 22: 197-204.
Delevoryas, Theodore, and Jeanne Morgan. 1952. Tubicaulis multiscalariformis: a
new American coenopterid. Amer. Journ. Bot., 39: 160-166.
-. 1954. An anatomical study of a new coenopterid and its bearing on the
morphology of certain coenopterid petioles. Amer. Journ. Bot., 41: 198-203.
-. 1954. A further investigation of the morphology of Anachoropteris cla-
vata. Amer. Journ. Bot., 41: 192-198.
-. 1954. Observations on petiolar branching and foliage of an American
Botryopteris. Amer. Mid. Nat., 52: 374-387.
Goldring, Winifred. 1924. The Upper Devonian forest of seed ferns in eastern New
York. N. Y. State Mus. Bull., 251: 50-72.
Holden, H. S. 1930. On the structure and affinities of Ankyropteris corregata. Phil.
Trans. Roy. Soc. London, 218B: 79-114.
Lang, William H. 1913. Studies in the morphology and anatomy of the Ophioglos-
saceae I. Ann. Bot., 27: 203-242.
-. 1915. Studies in the morphology and anatomy of the Ophioglossaceae III.
Ann. Bot., 29: 1-54.
Leclercq, Suzanne. 1951. Etude morphologique et anatomique d’une fougere du
Devonien Superieur. Ann. Soc. Geol. Belgique, Mem., 40, 9: 1-62.
Mamay, Sergius H. 1952. An epiphytic American species of Tubicaulis Cotta. Ann.
Bot., n. s., 16:145-163.
-. 1957. Biscalitheca, a new genus of Pennsylvanian coenopterids, based on
its fructification. Amer. Journ. Bot., 44: 229-239.
-, and Henry N. Andrews, Jr. 1950. A contribution to our knowledge of the
anatomy of Botryopteris. Bull. Torrey Bot. Club., 77: 462-494.
Read, Charles B. 1935. An occurrence of the genus Cladoxylon Unger in North
America. Journ. Washington Acad. Sci., 25: 493-497.
-. 1938. A new fern from the Johns Valley shale of Oklahoma. Amer. Journ.
Bot., 25: 335-338.
Renault, Bernard. 1875. Recherches sur les vegetaux silicifies d’Autun et de Saint-
Etienne, etude du genre Botryopteris. Ann. Sci. Nat., Botanique, Paris, ser. 6,
1: 220-240.
Sahni, Birbal. 1928. On Clepsydropsis australis, a zygopterid tree-fern with a
Tempskya-like false stem, from the Carboniferous rocks of Australia. Phil. Trans.
Roy. Soc. London, 217B: 1-37.
-. 1932. On the genera Clepsydropsis and Cladoxylon of Unger, and on a
new genus Austroclepsis. New Phyt., 31: 270-278.
-. 1932. On a Palaeozoic tree-fern, Grammatopteris baldaufi, (Beck) Hirmer,
a link between the Zygopterideae and the Osmundaceae. Ann. Bot., 46: 863-877.
-. 1932. On the structure of Zygopteris primaria (Cotta) and on the rela¬
tions between the genera Zygopteris, Etapteris and Botrychioxylon. Phil. Trans.
Roy. Soc. London, 222B: 29-46.
Scott, Rita. 1908. On Bensonites fusiformis, sp. nov. a fossil associated with Staurop¬
teris bumtislandica, P. Bertrand and on the sporangia of the latter. Ann. Bot.,
22: 683-687.
Scott, Dukinfield H. 1905. The sporangia of Stauropteris oldhamia. New Phyt.,

Telegram @tubscstds
4: 114-120.
EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 89

Sporne, Kenneth R. 1959. On the phylogenetic classification of plants. Amer. Journ.

Bot., 46: 385-394.
Surange, K. R. 1952. The morphology of Stauropteris burntislandica P. Bertrand
and its megasporangium Bensonites fusiformis R. Scott. Phil. Trans. Roy. Soc.
London 237B: 73-91.
-. 1952. The morphology of Botryopteris antiqua with some observations
on Botryopteris ramosa. The Palaeobotanist 1: 420-434.
-. 1954. Botryopteris elliptica sp. nov. from the Upper Carboniferous of
England. The Palaeobotanist, 3: 79-86.
Zimmermann, Walter. 1952. Main results of the “Telome Theory.” The Palaeobota¬
nist, 1: 456-470.

Appendix—Carboniferous Fernlike Foliage

Fern—like foliage constitutes one of the most conspicuous fea¬

tures of Carboniferous plant-bearing rocks. Numerous genera and
hundreds of species have been recognized since plants of that age
began to be collected and studied seriously at the start of the nine¬
teenth century. In the earlier days efforts were made to identify
such fossils with living ferns, but it soon became apparent to the
more critical investigators that they were dealing with extinct
plants and in the latter decades of the last century the added sus¬
picion began to take form that many of them were indeed not
ferns at all.
The problem of identification of these fossils centers around the
facts that a vast majority of specimens are sterile, that is, they
bear no reproductive organs, sporangia or seeds, and some species
have been founded on fragmentary or poorly preserved specimens.
Many of these “fossil ferns” do have a certain esthetic attraction,
with the result that museum collections have been swollen with
quantities of material that is of little biological importance. One
good fertile specimen of a given species will tell far more than any
quantity of sterile ones. It is true, however, that a few workers
have become sufficiently clever in distinguishing the various types
to use them effectively in stratigraphic studies.
Regardless of one’s opinions concerning the relative importance
of such fossils their abundance makes it impossible to ignore them.
As a result numerous form-genera have been established; in the
case of the femlike foliage a form-genus may be defined as one
including species that are similar in their gross morphological fea¬
tures but in which reproductive organs are lacking and consequently
the natural affinities are unknown. This definition should perhaps
be qualified in that some of the genera, notably Neuropteris and

Telegram @tubscstds
Alethopteris, include species which are known to have borne seeds
90 STUDIES IN PALEOBOTANY

and it is now highly probable that all of them are leaves of

pteridospermous plants; the genus Pecopteris includes several
species that have been found to bear marattiaceous fructifications,
and are thus true ferns, whereas seeds have been found on other
pecopterids.
The general problem is complicated by the fact that the form-
genera are not in all cases sharply defined as far as their gross
morphological characters are concerned. A similar intricacy is
encountered in the identification of cycadophyte and ginkgophyte
foliage (see Chapters 10, 11) in Mesozoic rocks. Studies of cuticu-
lar characters of the leaves of these groups have aided greatly in
their classification; very few such investigations have been under¬
taken with the Carboniferous fernlike leaves.
It is the objective in this brief treatment to simply introduce a
few of the more common and representative form-genera, particu¬
larly ones that must be referred to in later chapters on true ferns
and pteridosperms.
Adiantities. The fronds are three to four times pinnate with
slender rachises; the pinnules are obovate or wedge-shaped with a
truncate or rounded apex.
Alethopteris. The fronds are several times pinnate, with pin¬
nules that are attached by a flaring base and usually make an acute
angle with the pinna midrib. Most, if not all, of the Alethopteris
species were borne on stems of the medullosan pteridosperms (see
Chapter 5.
Alloiopteris. These are presumed to have been fronds of con¬
siderable size in which the ultimate pinnae are elongate, nearly
linear, and divided to varying degrees in different species. The
linear pinnae and characteristic mode of lobing seem to distinguish
this group quite clearly.
Diplotmema. The rachis dichotomizes into two equal branches
below which no pinnules are attached; the latter are deeply and
finely divided and merge into pinnules of the Sphenopteris type,
although the fronds as a whole were much smaller than in that
form-genus.
Linopteris. This is a rather uncommon but distinctive type.
The pinnules are similar in shape to those of Neuropteris but have
a fine-meshed net venation in contrast to the open dichotomous
venation of Neuropteris.
Lonchopteris. This is another uncommon type in which the pin¬
nules are closely comparable to those of Alethopteris but differ in

Telegram @tubscstds
their net venation.
 EARLY FERNS OF THE DEVONIAN AND CARBONIFEROUS 91

Fig. 3-14. Pinnule morphol¬

ogy of some Carboniferous
fronds. A. Alethopteris lon-
chitica Schlotheim; B, Neu-
ropteris heterophylla Brong-
niart; C. N. scheuchzeri
Hoffmann; D. Mariopteris
bellani P. Corsin; E. M.
leharlei P. Corsin; F. Linop-
teris sp.; G. Lonchopteris
rugosa Brongniart; H. Pecop-
teris rarinervis P. Corsin; I.
P. volkmanni Sauveur; J. P.
platoni (Grand’Eury); K. Al-
loiopteris sp.; L. Sphenopter-
idium dissectum (Goeppert);
M. Diplotmema furcatum
Brongniart; N. Rhodea smithi
Kidston; 0. Sphenopteris
flabellifolia Kidston; P. S.
elegantiformis Stur; Q. S.
kayi Arber; R. Adiantites an-
tiquus (Ettingshausen); (D,
E from Corsin, 1932; I, J from
Corsin, 1951; L, 0, P, Q, R
from Kidston, 1923, by per¬
mission of the Controller of
Her Britannic Majesty’s Sta¬
tionery Office.)

Pecopteris. This group consists of many species of large fronds

that were several times pinnate. Many have been found with fern
sporangia and consequently have been segregated to distinct
genera. The pinnules are erect or nearly so, attached along their
entire base and with a sparsely forking vein system.
Neuropteris. This is another common group of many species,
most, or all of which are pteridosperm leaves; several have been
found with seeds attached. The pinnules have a contracted base,
being attached by a very slender stalk or petiolule.
Mariopteris. This includes a large assemblage in which specific
delimitations are especially difficult. The pinnules are attached
along their entire base and it perhaps is not amiss to describe them
as somewhat intermediate between those of Pecopteris and
Sphenopteris; the lobing is similar to the latter but usually not as
pronounced. In some species the fronds divide by uniform

Telegram @tubscstds
92 STUDIES IN PALEOBOTANY

dichotomies comparable to a Gleichenia, and Kidston was of the

opinion that most Mariopteris species were vines with relatively
slender stems and depended on surrounding vegetation for support.
Rhacopteris. The fronds are but once pinnate, with alternate,
often overlapping flabelliform pinnules which range from entire to
very deeply dissected (Fig. 3-13, D-G).
Rhodea. The pinnules are dissected to the point where there is
essentially no lamina present; it seems questionable, however,
whether there is a real generic distinction between the species of
this form-genus and certain ones in Sphenopteris and Diplotmema.
Sphenopteridium. The fronds are twice pinnate and the rachis
dichotomizes in the basal part; the pinnules are more or less wedge-
shaped and variously divided.
Sphenopteris. The fronds are two to four times pinnate; the
pinnules have a contracted base and vary from nearly entire to dis¬
sected types of all sorts, and with blunt or pointed lobes. In his
“Fossil plants of the Carboniferous rocks of Great Britain” Kid¬
ston recognized 77 species; the total number for all areas and ages
would probably amount to several hundred.

REFERENCES

Corsin, Paul. 1951. Bassin Houiller de la Sarre et de la Lorraine. I. Flore Fossile.

4me fasc. Pecopteridees. Etudes des GTtes Miner, de la France, pp. 175-370.
-. 1932. Bassin Houiller de la Sarre et de la Lorraine. I. Flore Fossile.
3me fasc. Mariopteridees. Etudes des Gites Miner, de la France. Pp. 110-173.
Crookall, Robert. 1929. Coal Measure Plants. Edward Arnold Co., London. 80 pp.
Danze-Corsin, Paule. 1953. Contribution a l’etude des Mariopteridees. Les Mario¬
pteris du Nord de la France. Service Geologique des H.B.N.P.C.
Danze, Jacques. 1956. Contribution a l’etude des Sphenopteris. Les fougeres
Sphenopteridiennes du Bassin Houiller du Nord de la France. Service Geologique
des H.B.N.P.C., Lille, 568 pp.
Janssen, Raymond E. 1939. Leaves and stems from fossil forests. Illinois Popular
Science, ser. 1: 1-190. Springfield.
Kidston, Robert. 1923-25. Fossil plants of the Carboniferous rocks of Great Britain.
Mem. Geol. Survey Great Britain., vol. II, pts. 1-6.
Miller, Hugh. 1857. The Testimony of the Rocks. Gould and Lincoln, Boston,
502 pp.
Noe, A. C. 1925. Pennsylvanian flora of northern Illinois. Illinois State Geol. Surv.
Bull. 52: 1-18.
Schimper, W. P. 1874. Traite de Paleontologie vegetale. Atlas. J. B. Bailliere et
Fils, Paris. 110 pp.

Telegram @tubscstds
 4
the PTEROPHYTA
—Marattiales and Filicales

M ost of the plants that are dealt with in this chapter are
i referred to as the true ferns, a large and exceptionally
varied assemblage. Many of them are plants of great beauty and
have long attracted the interest of both amateur and professional
botanists. Numerous systems of classification have been proposed
and recent studies have tended to split them into an increasing
number of taxa. For example, in his excellent three-volume work
The Ferns, F. 0. Bower, in 1923 recognized 14 families (excluding
the “Polypodiaceae”) whereas Pichi-Sermolli, in a recent treatise
on the classification of the pteridophyta, recognizes 17 orders and
44 families. This number of families seems to me excessively high,
but it does point to the clearly evident diversity of the, approxi¬
mately, 10,000 species of extant ferns.
Only those fern families with significant fossil records are
described; there is also included, at the close of the chapter, a con¬
sideration of the Noeggerathiales, a problematical and interesting
group of pteridophytes. The order of subject material is, there¬
fore, as follows:

PTEROPHYTA
Marattiales
Marattiaceae
Filicales
Schizaeaceae, Gleicheniaceae, Matoniaceae, Dipteridaceae,
Osmundaceae, Tempskyaceae
(Some ferns of uncertain affinities, possibly transitional be¬
tween the coenopterid-protopterid complex and the true
ferns)
Noeggerathiales—relationships to other pteridophytes unknown

A few additional introductory comments may be helpful in out¬

Telegram @tubscstds
lining the nature of the fossil records and reasons for the particular
93
94 STUDIES IN PALEOBOTANY

selection of subject material. The Marattiaceae is represented by

many beautifully preserved sporangiate organs in the Upper Car¬
boniferous which are associated with stems of complex stelar anat¬
omy that probably bore the fertile fronds; the group was well
established at that time and undoubtedly had had a long previous
history of which little is known. The Gleicheniaceae likewise come
in rather suddenly at about the same time. The Schizaeaceae has
been traced back to the Lower Carboniferous and there is a par¬
ticularly significant series of compression fossils which sheds light
on the evolution of the sporangia of this family, but virtually noth¬
ing is known of the anatomy of these plants. By contrast the
earliest members of the Osmundaceae, which appear in the Per¬
mian, are known from a series of petrified stems which attained a
rather modern aspect by mid-Mesozoic times. The Tempskyaceae
is described in some detail because of its unparalleled trunk struc¬
ture and the fact that it was an abundant floral element of late
Mesozoic landscapes. Under the heading “ferns of uncertain affin¬
ities” a few fertile ferns are introduced because of their transitional
nature, relative to the position of the sporangia, between the pre¬
ferns and true ferns.

MARATTIALES

Marattiaceae

This family is represented today by seven genera; all are tropical

with four confined to the Malayan region and only one, Marattia,
native to both eastern and western hemispheres. The fronds of
many members of the family are large; particularly distinctive are
the sori, which are usually arranged in two rows, one along either
side of the midrib. The sorus of Angiopteris consists of 20 or more
separate sporangia arranged in two closely aggregated rows,
whereas in Marattia the sorus is made up of a similar number of
sporangia that are fused into a single unit. Christensenia differs
from the other genera in having a radially symmetrical synangium
of about a dozen sporangia. In all, the sporangial walls are thick,
multicellular structures enclosing prodigious numbers of spores, at
least for members of the Pterophyta. Some of the spore numbers
per sporangium are approximately as follows: Angiopteris 1500,
Marattia 2500, and Christensenia 7000. In the genera Angiopteris,

Telegram @tubscstds
Marattia, and Archangiopteris the stem is rather short and mas-
MARATTIALES AND FILICALES 95

sive and like most, if not all, modern ferns the stelar system origi¬
nates in the sporeling as a single small protostele, but soon develops
into a large and complicated system of several series of meristeles
or stelar segments.
The Carboniferous Marattiaceae were tall, graceful tree ferns
with crowns of large pecopterid fronds; the trunks were unbranched,
gradually tapering columns and heavily buttressed with adventi¬
tious roots in the basal region. The leaves, with their sporangia,
and the stems are often remarkably well preserved.
Among the more frequently encountered fossils in American coal
ball petrifactions are leaf fragments bearing sporangiate organs,
often exquisitely preserved, which have been considered to be of ma-
rattiaceous affinities. Cyathotrachus altissimus Mamay (Fig. 4-1) is
a rather large synangium consisting of five to nine sporangia which
are united for about half their length around a central column.
The sporangia are distinct morphological units consisting of a wall,

Fig. 4-1. Cyathotrachus altissimus. A.

Median longitudinal section through a syn-
angium; B. transverse section through
distal part. (From Mamay, 1950.)

Telegram @tubscstds
96 STUDIES IN PALEOBOTANY

probably one cell thick at maturity, enclosing many thousands of

spores; they are, however, enclosed in a synangial envelope several
cells thick. As the fructification approached maturity this enve¬
lope ruptured and breaks occurred between the closely appressed
walls of the sporangia so that they became separate beyond the
column and dehisced by longitudinal slits along their inner wall,
the spores being shed toward the central chamber. This species is
abundant in Iowa coal balls and the genus is also represented by a
species from Illinois and one from the Lower Coal Measures of
England. The synangia of Cyathotrachus, which were probably
borne on foliage of the Pecopteris type, are comparable with those
of the living Christensenia, the chief differences being in the dis¬
charge of spores in the latter by apical pores in contrast to the
longitudinal slits of Cyathotrachus.
Since many of the fossil plants considered in this and several fol¬
lowing chapters come from coal ball petrifications of England and
this country, the correlation chart given on page 126 is included as
a reference to the position of the more important localities.
Scolecopteris is a genus of some nine species and is common in
European and American Upper Carboniferous horizons. The
synangium of S. iowensis Mamay (Fig. 4-2) from the Des Moines
series of Iowa usually consisted of 6 sporangia attached to a com¬
mon pedicel but free beyond this basal point; the sporangia are
nearly 1 mm long and taper to bluntly acute apices. The synangia
were arranged in two rows, one along either side of the midrib of
the pinnule. Although the pinnule structure of some species of
Scolecopteris appears to have been of the Pecopteris type this is
not so in all cases. In Scolecopteris incisifolia Mamay (Fig. 4-3),
also from the Des Moines series of Iowa, there are usually four
sporangia in each sorus and the margins of the pinnule are lobed
and strongly enfolded with a synangium attached at the base of
each lobe.
Eoangiopteris andrewsii Mamay differs from the above in that
the sporangial aggregate is distinctly elongate; there are five to
eight sporangia which are shallowly immersed in a basal receptacle
but are essentially separate; they were borne on Pecopteris type
pinnules (Fig. 4-3). The term sorus is used to designate such
sporangial units; it would perhaps be preferable to use it through¬
out our description of the marattiaceous fructifications, but since
it is important to distinguish between those in which the sporangia
are organically fused (synangia) and those in which they are not,

Telegram @tubscstds
the latter will be called sori.
 MARATTIALES AND FILICALES 97

Fig. 4-2. Scolecopteris iowensis,

a median longitudinal section.
(From Mamay, 1950.)

As usual the problem exists here of correlating petrifaction and

compression specimens. Asterotheca is a common genus of
pecopterid fronds bearing marattiaceous sporangiate organs. In
Asterotheca parallela, Radforth has shown that each fructification
is composed of three to five sporangia joined at their bases to a
common stalk; approximately 1500 to 2000 spores were produced
by each sporangium. Several students of these plants are of the
opinion that Asterotheca and Scolecopteris species should be
referred to as a single genus. The chief difficulty lies in the fact
that the compression fossils are rarely preserved well enough to
allow positive comparison with the petrified ones.
Large pecopterid fronds bearing fructifications such as those
described above are frequently found associated with petrified
trunks referred to the genus Psaronius, which is characterized by
a complex polycyclic stelar system. Numerous species of the genus
have been recognized, chiefly from the European Carboniferous;
although not rare on this side of the Atlantic they have received
but little attention until recently. It has been recognized for some
time that there is much variability displayed in the organization of

Telegram @tubscstds
the numerous primary stelar segments (meristeles) that make up a
98 STUDIES IN PALEOBOTANY

B
Fig. 4-3. Restorations of representative portions of the fertile pinnules of: A. Scole-
copteris incisifolia; B. Eoangiopteris andrewsii, about 20X. (From Mamay, 1950.)

stem and consequently some doubt has been cast on the validity of
many specific identifications. They present a problem rather akin
to that of the medullosan seed-fern stems described in the next

Telegram @tubscstds
chapter, for almost every specimen that turns up differs a little
from any other.
MARATTIALES AND FILICALES 99

A significant advance in our knowledge of the plants has resulted

from Morgan’s recent study of numerous specimens from a locality
in southeastern Illinois. A nearly continuous sequence of speci¬
mens was found which demonstrate clearly that the stelar system
developed obconically, as, years ago, F. O. Bower demonstrated is
the case with most modern ferns. That is, the stele in the basal-
most part of the trunk is small and quite simple; in the case of P.
blicklei Morgan it is about 2 cm in diameter and is described as
dicyclic, the vascular system consisting of a central strand sur¬
rounded by a ring of several others. (It would be most interesting
to have the sporeling stage at hand, but very probably it was a
protostele.) This simple stelar system gradually proliferated as
the trunk increased in height and girth to a polycyclic stele con¬
sisting of ten or more concentric cycles.
The leaf traces follow a similar pattern of increasing complexity;
there are only three vertical rows of traces, representing three rows
of petioles, in the basal portion of the trunk of E. blicklei, whereas

Fig. 4-4. Psaronius blicklei. A. Transverse view of the trunk showing numerous con¬

Telegram @tubscstds
centric stelar cycles; a portion of the inner root zone is intact. B. Junction of inner
and outer root zones, 5X. (A from Morgan, 1959.)
100 STUDIES IN PALEOBOTANY

specimens from the distal part of a trunk display up to 14

orthostichies.
Another characteristic feature of many Psaronius specimens is
the mantle of roots surrounding the stem. The specimen of P.
blicklei with the small stem cited above was found with a mantle
of roots over 30 cm thick; thus the over-all diameter of the speci¬
men (quite clearly a stump) was close to three-quarters of a meter.
The root mantle is composed of two distinct zones: an inner one
of adventitious roots that are embedded in a parenchymatous
matrix; and an outer zone of appreciably larger adventitious, free
roots. These roots, either isolated specimens or tangled clusters of
them, are ubiquitous in certain American coal-ball localities and
often occur to the exclusion of almost any other kind of plant
remains.
The individual root has a central star-shaped stele and a broad
cortex which is essentially a meshwork of chains of parenchyma
cells with large intervening chambers. Such an aerenchymatous
organization is usually found in the roots, stems, and leaves of
aquatic or semiaquatic plants and suggests a swampy habitat for
the Psaronius trees.
Partially petrified specimens, as well as compression or impres¬
sion fossils of tree-fern trunks, have been described under the
generic names Megaphyton, Caulopteris, and others; they are
characterized chiefly by large petiole scars and distinctively shaped
trace scars. Megaphyton was established for a specimen with only
two rows of such scars, one on either side of the stem; Caulopteris
has been used for specimens with several rows arranged in what
appears to be a spiral pattern. In view of the variation that is now
known to exist in a trunk of P. blicklei it may be understood with
what hesitancy one can assign species names to impressions or
poorly preserved petrifactions.
In summary, the marattiaceous tree ferns were one of the domi¬
nating elements of the Upper Carboniferous and Permian forests.
Their unbranched trunks attained a height of probably 50 feet or
more, and although the stem itself tapered obconically the dense
mantle of adventitious roots resulted in a heavily buttressed base
which tapered upward to the crown of leaves which were few in
number but of great size. Some are estimated to have reached a
length of 3 meters, and these in turn bore sporangiate organs quite
similar to those of the living members of the family.
The evidence for such a restoration comes chiefly from the

Telegram @tubscstds
actual attachment of the sporangia to leaf fragments and the asso-
 MARATTIALES AND FILICALES 101

Fig. 4-5. Reconstruction of a psaroniaceous tree fern about 25 feet tall. Leaf scars
are visible on the upper part of the trunk below the crown of fronds; most of the trunk
is enclosed in adventitious roots which increase in thickness toward the base. (From
Morgan, 1959.)

Telegram @tubscstds
102 STUDIES IN PALEOBOTANY

ciation of these fossils with the trunks. However, in 1888 Renault

and Zeiller figured a compression specimen of a trunk with the
basal portion of a frond still attached; the few fragments of the
ultimate frond divisions indicated a Pecopteris leaf. There is thus
much that remains to be learned. The abundance of petrified
frond fragments in American coal balls suggests the possibility of
reconstructing the entire frond with some accuracy, and basal por¬
tions might be correlated with the petiole traces that are frequently
found around the periphery of the trunk.
Records of Mesozoic and Tertiary fossils of the family are not
rare, but the preservation is generally unsatisfactory. Asterotheca
meriani (Brongniart) Stur from the Triassic of Lunz, Austria, was
probably comparable in habit to the Carboniferous plants. Each
pinnule of the frond bore two rows of synangia which consisted of
four sporangia fused in the basal portion. Marattiopsis hoerensis
(Schimper) Thomas from the Rhaetic of Greenland is based on
fertile leaf fragments with synangia 7 mm long and 1 mm broad.
They consisted of 17 or more sporangia, each of which contained a
spore mass 900 \x long and 350 /x wide, the spores being 28 ju, in
diameter. This fossil is closely comparable if not identical with
modern species; in fact, Harris notes that “Marattiopsis is only
known to differ from Marattia in age.” Specimens of Marattiopsis
macrocarpa (Morris) Seward and Sahni, from the Jurassic of the
Rajmahal Hills in India, are sufficiently well preserved to disclose
the presence of the family at that time.
Nathorstia alata Halle, a Lower Cretaceous fossil from Lago
San Martin, Patagonia, is probably marattiaceous; it is an interest¬
ing fern with slender, tapering pinnae 10 to 12 cm long with an
anastomosing vein system. There is a row of sori along each side
of the midrib; they are densely crowded, circular in outline, and
about 1.5 mm in diameter. Each sorus is divided into 12 to
15 chambers.

FILICALES

Schizaeaceae

This family is represented today by four genera of diminutive

plants, but what they lack in size they make up for in their unu¬
sual morphology and fossil history. Although the family favors
warm latitudes Lygodium, the climbing fern, has been reported as

Telegram @tubscstds
far north as southern New Hampshire. Schizaea also invades
MARATTIALES AND FILICALES 103

temperate regions and some species are especially striking with

their small grasslike leaves only a few inches high. Anemia is the
largest genus, with over 60 species, and lives as far north as the
Florida Keys and southern Texas. Mohria is restricted to South
Africa and in its sporangial structure presents a possible link with
the Osmundaceae.
For the most part the sporangia are arranged separately in two
rows, rather than in groups (sori), on the pinnules and the annulus
consists of a single row of cells at the distal end. The sporangia
are erect in Schizaea and Anemia (Fig. 4-6) but curved in Lygo-
dium, lending the effect of being attached at one side.

Telegram @tubscstds
Fig. 4-6. Sporangia of some extant ferns: A. Anemia mexicana; B. Lygodium palma-
tum; C. Gleichenia longissima; D. Osmunda regalis.
104 STUDIES IN PALEOBOTANY

Since petrified remains of the vegetative organs are virtually un¬

known, the anatomy of the living members of the family will not
be considered. The frond structure is, however, highly distinctive
and enables us to trace Lygodium and Anemia far back into the
past.
Anemia is but little changed since mid-Cretaceous times.
Anemia fremonti Knowlton (Fig. 4-7) from the Cretaceous Frontier
formation of southwestern Wyoming is very close to the living A.
adiantifolia found around limestone sink holes of the Florida Keys.
The fine grained shale of the Frontier formation has preserved the
fertile pinnae so well that portions of the compression can be
removed from the rock using the cellulose transfer method (see
Chapter 17). The size and general morphology of the frond with
its two conspicuous fertile pinnae, the arrangement of the sporangia
in two rows on the under side of the pinnules and the sculpturing
of the spores, are almost identical with the corresponding charac¬
ters of A. adiantifolia. Unless the physiology of schizaeaceous
plants has changed appreciably during the past 100 million years
one may suppose that the Cretaceous landscape of what is now
southwestern Wyoming was far more lush, with a warmer and
more equitable climate, than prevails there now.
Anemia poolensis Chandler is based on fragments of fertile pin¬
nae obtained by Miss Chandler by sifting early Tertiary sediments
of southern England. Although the fragments are small they
reveal the characteristic arrangement of the sporangia in two rows
and the conspicuous apical annulus; the spores, however, differ
from most other species, living and fossil, in being smooth. This is
far north of the present range of this tropical to subtropical genus
and fits in with the abundant evidence from angiosperm seeds and
fruits that a warm climate prevailed in southern England during
the early Tertiary. Fertile pinnules of Lygodium were also found
in the same deposits.
Lygodium pumilum Brown from the Upper Cretaceous of Wyo¬
ming is one of the oldest species of this genus; although known
only from sterile leaflets it is quite clearly a Lygodium and is con¬
sidered to be allied with the living L. palmatum, differing from the
latter in having very small pinnules.
Lygodium skottsbergii Halle, from Eocene deposits in southern
Chile, attests to a wider distribution for the genus than is known
at present; Coronel, the locality from which it was obtained, is
some 30° south of the present limits of Lygodium. Macerations of

Telegram @tubscstds
sporangia revealed spore counts of between 120 and 169; some of
 MARATTIALES AND FILICALES 105

Fig. 4-7. Anemia fremonti, an Upper Cretaceous fern from southwestern Wyoming.
A. Restoration of a complete fertile frond; B. portion of a frond; C. part of a fertile
pinna after removal from the rock, showing three clusters of small pinnules; D. one
group of pinnules in side view; E. a single pinnule more highly enlarged with seven

Telegram @tubscstds
sporangia. (From Andrew and Pearsall, 1941.)
106 STUDIES IN PALEOBOTANY

these were fragmentary and the number was thus very possibly
256, which is typical for modern species. Although distinct from
any lygodiums that live today it was probably closest to L.
palmatum.
There are numerous reports in the literature of fertile schizae-
aceous leaf fragments assigned to the genus Klukia, although
rather few reveal clearly defined details. Harris has described
specimens from the Jurassic of Yorkshire under the name Klukia
exilis (Phillips) Raciborski which are at least twice pinnate fronds
with pinnules of the Pecopteris type. Each pinnule bore 6 to 14
sporangia which display the characteristic single-rowed annulus.
The same species has been reported by Endo from the Upper
Jurassic of northern Hondo in Japan.
Using a transfer technique which enabled him to study the
sporangia in detail Radforth has demonstrated a sequence of fossil
forms (Fig. 4-8) that gives a clue to the origin of the unique
sporangium in the Schizaeaceae. The annulus of the Upper Car¬
boniferous Senftenbergia plumosa (Artis) Radforth is apical, as in
the Mesozoic and modern members of the family, but is two to
three cells deep instead of only one. In S. ophiodermatica (Goep-
pert) Stur, also Upper Carboniferous, the annulus is two to three
cells deep, but the arrangement is less regular than in S. plumosa.
In S. sturi (Sterzel) Radforth, from a Lower Carboniferous horizon,
the annulus cells are less clearly distinguished from the others of
the sporangium and tend to form an irregular patch around one
side of the sporangium.
The Senftenbergia fronds are of the Pecopteris type and indicate
plants of larger size than most modern members of the Schizae¬
aceae. If all of the species mentioned above are correctly assigned
to this family, there is a rather strong suggestion that the distinc¬
tive annulus originated from an irregular cluster of cells which be¬
came localized at the distal end of the sporangium; they later
became oriented into two rows and finally reduced to one.
In connection with this apparent sequence of Carboniferous
forms it is of interest to note, briefly, a specimen described by Rad¬
forth and Woods from the Lower Cretaceous of western Canada.
It is probably referable to the genus Klukia, the ovoid, sessile
sporangia being arranged in two rows on the pinnule. Although
there is typically but a single row of annulus cells, occasional
sporangia show a rather indefinite second row; the spore number
per sporangium is approximately 390 which is higher than in most

Telegram @tubscstds
living species of the Schizaeaceae but appreciably lower than in
two of the Carboniferous fossils noted above.
 MARATTIALES AND FILICALES
107

Fig. 4-8. Sporangia of certain Carboniferous species of Senftenbergia showing prob¬

able stages in the evolution of schizaeaceous annulus. A. S. sturi; B. S. ophioder-
matica; C. S. plumosa. (From Radforth, 1938, 1939.)

Gleicheniaceae

The genus Gleichenia is a large one, including, according to the

interpretations of different authorities, 80 to 100 living species, al¬
though certain subgroups of the genus may actually be of generic
rank. In most gleichenias the branching pattern of the leaf is
highly distinctive; it takes the form of a uniform pseudo-dichotomy

Telegram @tubscstds
because of the frequent failure of the leaf apex to develop. In
some species the fronds attain a length of over 20 feet.
108 STUDIES IN PALEOBOTANY

The sori are arranged in two rows along either side of the midrib
of the leaflets. The sorus usually consists of a ring of sporangia
arranged around a central receptacle; the annulus (Fig. 4-6) is a
conspicuous and nearly complete ring of cells oriented in a slightly
oblique equatorial position.
Gleichenia is widely scattered in the tropics and versatile of
habitat; its distribution has been rather colorfully described by
Seward as follows:
Gleichenia occurs in the clearings of tropical forests, on the edge of jungles,
on tropical alpine peaks, in the heath vegetation on the higher slopes of
Ruwenzori, . . . China, India, Australia ... in the rain forests of Mexico,
Costa Rica ... in the West Indies, along the Andes to the Falkland Islands,
. . . from Natal to Table Mountain, in Madagascar, in the Island of Reunion,
on Amsterdam Island, in New Guinea where Dr. Wollaston tells me that he
collected several species, some growing at an altitude of over 12,000 feet. . . .
(1922, pp. 224-225.)

The earliest genus attributed to the family is the Upper Carbon¬

iferous Oligocarpia (Fig. 4-9). It has been restudied recently by
Abbott who recognizes nine species which are rather widely dis¬
tributed through Europe and North America. The foliage is of
the Sphenopteris type with fronds that are several times pinnate.
The sori are rosettelike groups of (usually) four or five sporangia,
although the number may be as high as 17. The sporangia are at¬
tached to a low dome-shaped receptacle and are arranged in a
single ring with occasionally one or two in the center.

Fig. 4-9. Oligocarpia capitata. A. A fertile pinna, about 6X; B. two aspects of
a sporangium showing the equatorial annulus. (From Abbott, 1954.)

Telegram @tubscstds
MARATTIALES AND FILICALES 109

There is sufficient variation in sporangial structure in the ferns

so that a sharp delimitation of families is not always an easy mat¬
ter. However, the manner of grouping of the sporangia combined
with the annulus structure offers strong support of the assignment
of Oligocarpia to this family. Some of the species appear to have
been quite widely distributed; for example, O. brongniarti Stur is
recognized from Upper Carboniferous horizons in Silesia, France,
England, southeastern Canada, and Ohio.
Unlike the Carboniferous Oligocarpia most of the Mesozoic ferns
assigned to the Gleicheniaceae display the apparent dichotomous
pattern of the modern fronds and most authors have referred these
fossils to Gleichenites. G. rostafinskii Raciborski from the Juras¬
sic of Poland is one of the oldest Mesozoic species. In the Creta¬
ceous the gleichenias were abundant and widespread; several
species flourished in western Greenland and others are known
from England, Germany, France, and the Balkans, and as far south
as Patagonia. Gleichenites coloradensis (Knowlton) Andrews was
associated with Anemia fremonti in the Frontier formation of Wyo¬
ming and compares closely with certain living species. Fossil rec¬
ords are rather scarce from the Eocene on, there being little doubt
that the cooling climate of the early Tertiary, through what is today
the north temperate zone, spelled the extinction of this family in
the north.

Matoniaceae

Contrasts in present and past distribution patterns often reveal

some striking stories of plant migration and few of these exceed in
interest that of the Matoniaceae. Because of the unique branching
pattern of the frond, it is possible to recognize the fossil ancestors
with an unusual degree of certainty far back in the Mesozoic.
The family is represented today by only three or four species in
the genera Matonia and Phanerosorus which are of limited distri¬
bution in the Malay region. The frond may attain a height of 8
feet and has been described as “a very perfect example of a cata-
dromic helicoid dichopodium.” In somewhat simpler language this
means that the main rachis of the leaf divides into two parts, each
of which bends down (catadromic) and continues to divide in a
helicoid or spiral fashion. The sori consist of six to nine sporangia
which are covered by a hemispherical flap, the indusium, which
falls away when the sporangia are mature. The structure of the

Telegram @tubscstds
110 STUDIES IN PALEOBOTANY

sporangium is similar to that of Gleichenia although the position

of the annulus is not quite as regular.
During the past the matonias have wandered far beyond their
present confinement. They were rather widely distributed in cen¬
tral Europe during the Mesozoic and have been found as far north
as Greenland; they have been recorded from Australia, India and
Japan, from North Africa, and from several localities in this country.
Two matoniaceous ferns referred to the genus Phlebopteris have
been found in the Triassic of the southwest; P. smithii (Daugherty)
Arnold (Fig. 4-10) comes from the Petrified Forest National Monu¬
ment in Arizona and P. utensis Arnold, with smaller fronds, shorter
pinnae, and a somewhat different vein pattern was found in Utah.
A fine specimen of a fertile frond of Matonidium americanum Berry,
about 30 cm in diameter, has been described from the Upper Cre¬
taceous of Montrose County, Colorado. Matonidium indicum Sahni
from Baroda State, India, is probably of Lower Cretaceous age and
is distinguished by an upward flaring of the petiole into a funnel-

Telegram @tubscstds
Fig. 4-10. Phlebopteris smithii, from the Triassic of Arizona, natural size.
 MARATTIALES AND FILICALES 1 ]i

shaped lamina, presenting a possible transition to the leaf type of

the Dipteridaceae.

Dipteridaceae

This is a family of one living genus, Dipteris, with five species

found in the Indo-Malayan region. They are plants with creeping
rhizomes and long-stalked fronds similar in general appearances to
those of Matonia. There is, in fact, some question as to whether
the dipterids should be recognized as a separate family. The leaf
plan differs from Matonia in being anadromic; that is, the branch¬
ing of the two equal subdivisions of the leaf turn up rather than
down.
The blade of the living Dipteris lobbiana may be almost filiform;
the apex of the petiole dichotomizes into two equal parts and these
in turn divide, resulting in a leaf that is deeply dissected into 24 or
more narrow segments; there is a row of sori on either side of the
midrib of the segments. The blade of D. conjugata (Fig. 4-11B),
although deeply divided into two equal halves is generally much less
deeply dissected.
The number of sporangia per sorus is quite variable and their
orientation lacks the regular radial pattern of Matonia and Glei-
chenia. The annulus is an obliquely arranged, nearly complete ring
of cells; the degree of obliquity is, however, rather slight and Dip¬
teris was originally placed in the Polypodiaceae because of a general
similarity in annulus structure.
In a comprehensive survey of the fossils referred to the Dipteri¬
daceae, Oishi and Yamasita have recognized six genera (Hausman-
nia, Clathropteris, Dictyophyllum, Thaumatopteris, Goeppertella
and Camptosorus) with some 60 species; some of the genera are,
however, separated on rather slender grounds.
Clathropteris meniscoides var. elegans Oishi from the Mesozoic
(Nariwa Series) of Japan must have been a fern of great beauty in
life; the blade was divided into about 30 equal, slightly toothed
segments that are separate almost to the petiole.
The family was represented by numerous species in the Rhaetic
floras of Greenland and some of them were ferns with rather large
fronds; for example, Clathropteris meniscoides Brongniart had a
blade about 35 cm in diameter, the 10 or 12 divisions being about
2.5 cm broad and characterized by a net venation of rather large
rectangular meshes. Several species of Dictyophyllum, Thaumatop¬
teris, and Hausmannia also thrived in these northern climes during

Telegram @tubscstds
the Mesozoic.
112 STUDIES IN PALEOBOTANY

Fig. 4-11. A. Dipteris, Gleichenia

and other ferns growing along a
river bank in central Viti Levu,
Fiji; B. a single frond of Dipteris
conjugata. (Photographs by O. H.
Selling.)

The leaf structure in Hausmannia is rather different from that

of the other genera. In H. dentata Oishi the blade is only dissected
a third of the way to the rachis and in H. nariwaensis Oishi it is
nearly entire, resembling superficially the leaf of the common garden
nasturtium.
The fossil history of the family is briefly summarized by Oishi
and Yamasita as follows:
The conclusion arrived at from the investigation of fossil records of Dip-

Telegram @tubscstds
teridaceae is that it forms a conspicuous group of ferns among the Mesozoic
flora of the world. It has been known from Europe, Eastern Asia, North and
MARATTIALES AND FILICALES 113

South America, Australia, Greenland and Graham Land but there are no satis¬
factory examples from India and Africa. Thus the occurrence of fossil Dip-
teridaceae in these continents is of special interest as indicating that tropical
or subtropical conditions prevailed in the regions where the fossil forms are
now found. The fossil Dipteridaceae reached its maximum development,
numerically and possibly biologically also, in the Rhaetic and the Liassic
epochs, and it began to wane numerically towards the end of the Mesozoic,
though a certain genus, namely Hausmannia, began to flourish in the younger
Mesozoic time. (1936, p. 175)

In summary it may be noted that it seems likely that the three

families Gleicheniaceae, Matoniaceae, and Dipteridaceae stem from
a common ancestor in late Paleozoic times. There appears to be
some blending in the Mesozoic and several authorities have stressed
the difficulties of classification that are involved in working with
fossil representatives. The works of Harris, Oishi and Yamasita,
Hirmer and Hoerhammer, Seward and Dale are particularly worth
consulting.

Osmundaceae

Among modern plants that deserve the name of “living fossil” the
osmundas occupy a place in the foremost ranks. Judging from stem
anatomy the extant species and their immediate ancestors have
been conspicuous and widespread elements of the earth’s vegetation
for well over 100 million years.
Foliage attributed to the Osmundaceae is quite abundant, par¬
ticularly in Mesozoic rocks, and there is an exceptional series of
petrified stems that traces the family back to the Permian. The
stem in the three living species of Osmunda that are so abundant
through the eastern portion of the United States (O. regalis, the
Royal fern; O. cinnamomea, the Cinnamon fern; O. claytoniana,
the Interrupted fern) gives the appearance of being a rather mas¬
sive structure owing to the investing leaf bases which remain for
many years. The stem proper is, however, usually not over a cen¬
timeter or two in diameter and contains a very hard sclerotic cortex
and a cylindrical stele that is divided into numerous segments by
the departing leaf traces. The stems of our American species attain
a height of only a few inches; as to other extant osmundas, Todea
barbara in South Africa may reach a height of several feet and
Leptopteris wilkensiana, found in New Caledonia, is reported to
attain a height of 10 feet.
Our knowledge of the petrified remains has been gathered to¬

Telegram @tubscstds
gether in a series of contributions by Kidston and Gwynne-Vaughan
and much of the following is taken from their work. One of the
114 STUDIES IN PALEOBOTANY

youngest of the fossil stems is Osmundites schemnitzensis from the

Upper Miocene (or Lower Pliocene?) of Hungary which was found
associated with other plant remains including beech and sequioa.
The stem is small with a parenchymatous pith and a dissected xylem
cylinder of 17 or 18 strands and only 3 to 4 mm in diameter; in the
structure of the stele and the form of the departing leaf traces it
corresponds closely with present day osmundas. In fact, as far back
as the Eocene the stem anatomy is essentially the same as in mod¬
ern species; Osmundites chandleri Arnold (Fig. 4-12B) from the
Eocene of Oregon had a stem 15 mm in diameter with a small xylem
cylinder composed of about 34 strands. O. dowkeri Carruthers
from the Eocene of the Isle of Wight (England) was of very similar
size and anatomical dimensions.

C D E
Fig. 4-12. Diagrams of the stelar systems of certain fossil osmundaceous stems.
A. Osmundites braziliensis. B. 0. chandleri. C. O. kolbei; stippling represents scat¬
tered central tracheids. D. O. dunlopi; note that the stele is essentially continuous
(siphonostelic). E. Thamnopteris schlechtendalii; stippling represents inner tracheid
zone and solid black outer zone. (A from Andrews, 1950; B from Arnold, 1952;

Telegram @tubscstds
C-E, from Kidston and Gwynne-Vaughan, 1907-14.)
 MARATTIALES AND FILICALES
115

The range of these plants extended as far north as Spitsbergen

in Tertiary times. O. spetsbergensis Nathorst (probably late Ter¬
tiary) is known from a rhizome in which the stele had been lost, but
the general organization of the leaf bases is indicative of its osmun-
daceous affinity. Associated with it were portions of leaves con¬
sidered to be almost identical with the foliage of the living Osmunda
regalis; preserved sporangia fragments indicate that the annulus
was somewhat more strongly developed than in modem osmundas.
Some late Mesozoic fossils seem to represent the giants of the
osmunda line. Osmundites kolbei Seward from the Wealden of
South Africa is estimated to have attained a height of 90 cm and
the entire “trunk” (stem and leaf bases) measured about 14 cm in
diameter. The stele, although flattened prior to fossilization, meas¬
ured about 19 mm in diameter and consisted of a ring of some 56
strands. Of special interest here is the presence of numerous tra-
cheids scattered through the pith (Fig. 4-12C); these differ from the
tracheids of the main woody cylinder in being irregular in size and
shape but they are clearly enough tracheidal elements. O. brazi-
liensis Andrews (Fig. 4-12A) from Brazil is known from a stem 8 cm
in diameter which is enclosed in a closely compacted root sheath
4.5 cm thick, so that the entire trunk was close to 15 cm in diame¬
ter. It seems safe to assume that this was a tree fern of at least
several feet in height. The exact age of this fossil is not known but
it is probably not older than late Mesozoic.
It seems probable that the larger arborescent osmundas grew in
a manner similar to the Carboniferous Psaronius, with the petiole
bases being gradually sloughed off in the older part of the stem
which was in turn enclosed in a mantle of roots, resulting in a
heavily buttressed base.
A Jurassic species found in New Zealand, O. dunlopi Kidston and
Gwynne-Vaughan, differs from the younger ones in having a stele
that forms a continuous or nearly continuous cylinder.
The oldest petrified stems that are referred to the Osmundaceae
come from the Permian of the southern Urals. Of six genera that
have been described Thamnopteris schlechtendalii (Eichwald)
Brongniart (Fig. 4-12E) is perhaps the most informative. The stem
with its ensheathing leaf bases measures about 12 cm across and
has a xylem cylinder 11 mm in diameter. The stele differs from the
previously considered members of the family in that it consists of
two distinct zones: an outer continuous ring of tracheids of the usual
scalariform type and an inner core of irregularly shaped, somewhat
thinner-walled ones. This inner portion probably formed a solid

Telegram @tubscstds
116 STUDIES IN PALEOBOTANY

core but the central part is missing. The leaf trace buds off from
the periphery of the outer xylem as a small oval-shaped mass and
becomes C-shaped as it passes through the cortex. A sclerotic outer
cortex and abundant, large petiole bases with C-shaped strands bear
out the osmundaceous affinities of the plant.
The evolution of the osmundaceous stem seems, therefore, to have
followed essentially this course: from semiprotostelic plants such
as Thamnopteris, the Mesozoic siphonostelic forms evolved (O. dun-
lopi) and the development of large leaf gaps resulted in dictyostelic
types. There is a retention of some tracheids scattered through the
pith region in O. kolbei but they disappear in the Tertiary repre¬
sentatives.
As noted in the previous chapter, the Lower Permian Gramma-
topteris baldaufi may represent a link between the coenopterids
(Anachoropteridaceae) and the Osmundaceae. Its thick armor of
petioles and crowded nests of sclerotic cells in the middle cortex point
toward the latter family, whereas the presence of a few short, broad
tracheids in the central portion of the stele leads into the partially
medullated stele of Thamnopteris. The general shape of the spo¬
rangia as well as their dense clustered organization in certain living
species of Osmunda could readily have evolved from the less spe¬
cialized coenopterids. In brief, it seems very likely that future
research will confirm this suggested derivation of the Osmundaceae;
it may be worth noting that Engler and Diels recognize 17 living
species in their Syllabus der Pflanzenfamilien and since they have
not all been studied critically significant information may be forth¬
coming from studies of the living plants as well as the fossils.

Tempskyaceae

The Mesozoic tree-fern Tempskya is one of the most bizarre

creations in the plant kingdom and because it was sufficiently abun¬
dant as to have played a significant ecological role in the Cretaceous
landscapes it is considered at some length. It was a plant with an
unbranched columnar trunk which attained a diameter of 16 inches
in the largest specimen known; this tapered slowly upward to a blunt
apex, the largest ones probably attaining a height of 15 to 20 feet.
The upper part of the trunk was covered with a dense mass of
hundreds or perhaps thousands of small leaves (Fig. 4-14).
Fragments of the petrified trunks have been found in rather
widely scattered localities: in the Karaganda river basin of the

Telegram @tubscstds
USSR, Bohemia, England, and in this country in Maryland and
MARATTIALES AND FILICALES 117

several of the northwestern states. Stratigraphically, the genus

ranges from the upper part of the Lower Cretaceous to the mid Up¬
per Cretaceous although there is some evidence to indicate that it
extended back to the late Jurassic. About a dozen species have
been recognized, but many of them are questionable. The basic
organization in all seems to be very similar and the description given
here is based chiefly on a large collection of specimens found in the
Wayan formation of southeastern Idaho which are assigned to
T. wesselii Arnold.
The trunk fragments are often beautifully preserved, and in
a cross section taken from the middle or upper part they are found
to be constructed of many stems held together in a dense matrix of
small adventitious roots. It should be emphasized that these are
complete stems and not simply steles. The trunk section shown in
Fig. 4-13A is a small one 5 inches in diameter and is composed of
about 190 stems; each stem (Fig. 4-13B) is siphonostelic and rather
similar in its anatomy to the rhizome of certain living ferns, such
as Adiantum (Maidenhair fern). The stems dichotomized rather
profusely and gave off small petioles (about 1 mm in diameter) in
two rows. It is thus evident that the trunk bore large numbers of
small leaves which were distributed along the upper half, or perhaps
the upper two-thirds, of its length. This is quite in contrast to most
modern tree-ferns where a few large fronds are borne as a terminal
crown. Although the fossil trunks have not been found with leaves
attached, they occur in Idaho at a horizon in which Anemia fre-
monti fronds are also present; this association presents only a pos¬
sible relationship, but the approximate size of the fronds and then-
distribution on the trunk, as shown in the restoration, are well
established.
The most curious feature of Tempskya is the fact that specimens
taken from the lower part of the trunk are composed exclusively,
or almost exclusively, of roots. The development of the trunk was
apparently somewhat as follows: in the earliest sporeling stage there
was probably a single stem; as this grew and divided the resultant
trunk increased in height and girth; at the same time large num¬
bers of small, wiry roots were produced which held the entire mass
together as a distinct unit, with a texture somewhat like that of a
modern Osmunda; when the trunks reached a height of several feet,
the stems at the base gradually began to decay and the functional
stems above maintained a connection with the soil by means of the
thousands of small roots. These conclusions concerning the on¬
togeny of the trunks were reached by the writer from a study of

Telegram @tubscstds
118 STUDIES IN PALEOBOTANY

Fig. 4-13. Tempskya wesselii from the Upper Cretaceous of Wayan, Idaho. A. Cross
section of a trunk about 5 inches in diameter composed of 190 stems. B. A single stem,
20X; s, siphonostele of stem; t, traces leaving stem stele to supply two petioles. C. A
single root enlarged about 50X showing well-preserved root hairs. (From Andrews

Telegram @tubscstds
and Kern, 1947.)
MARATTIALES AND FILICALES 119

Fig. 4-14. A restoration of Tempskya based

on numerous stems found in southeastern
Idaho. (From Andrews and Kern, 1947.

Telegram @tubscstds
120 STUDIES IN PALEOBOTANY

many hundreds of trunk specimens; the evidence seems reasonably

conclusive, but why they grew in this manner is another question!
As to the affinities of Tempskya, the anatomy of the stems is not
sufficiently distinctive to offer a clue and other bits of evidence are
conflicting. Seward described a specimen of T. knowltoni from
Montana in which he found isolated sporangia among the stems and
roots; they have a characteristic schizaeaceous annulus and this fits
in nicely with the association of trunk specimens of T. wesselii and
the fronds of Anemia fremonti. However, isolated sporangial frag¬
ments have been found in the T. wesselii trunks from Idaho that
are quite clearly not schizaeaceous!
There are a few other ferns, living and fossil, that have a tendency
to develop such compound trunks. The Carboniferous Austroclep-
sis australis was mentioned in the previous chapter, although only
a very few stem branchings are involved. The extant Hemitelia
crenulata which grows in the forests of Java is a tree-fern of con¬
siderable size, attaining a circumference of 20 cm at 30 cm above
the ground. The basal 2 or 3 feet of the trunk is an aggregation of
numerous branches enclosed in a dense matrix of roots, whereas
above this the branches are free. A few other examples might be
cited; it is clearly a mode of growth that has originated independ¬
ently in several unrelated fern groups, but none compares closely
with Tempskya in either detailed anatomy or the huge number of
stems involved. Tempskya may well have been an important source
of food for the larger herbivorous animals of the Cretaceous. Pos¬
sibly future studies will be able to shed some light on its racial
origin and why it disappeared from the scene so suddenly.

SOME FERNS OF UNCERTAIN AFFINITIES

It has been noted earlier that there are conspicuous gaps in our
knowledge between plants of the coenopterid and protopterid types
considered in the last chapter and the ferns of more modern aspect.
Among the more important differences between the two assem¬
blages are the following:

1. The fronds of the ferns are distinctly laminate or “leafy,”

whereas the preferns bore fronds that amounted to little more than
a much divided branch system.
2. The sporangia, whether arranged singly, in clusters, or in
organically fused groups, are found on the under side of the lamina

Telegram @tubscstds
in the ferns in what has been called a “superficial” position, rather
than being borne terminally as is the case in the preferns.
MARATTIALES AND FILICALES 121

As to point number one, evidence has been presented to show how

this transition was accomplished and, at least in a general way, this
is rather well understood. The fossils that are described next offer
at least a sketchy answer to point number two, although it is not
possible to make precise taxonomic correlations.
Acrangiophyllum pendulata (Lesley) Mamay (Fig. 4-15A) from
the Pennsylvanian (Pottsville) of Alabama is known from rather
small fragments of fernlike foliage with sporangia 1.5 mm long

Fig. 4-15. Some Carboniferous ferns showing certain stages in the evolution of the
superficial position of the sporangium. A. Acrangiophyllum pendulata, from the
Pennsylvanian of Alabama; B. Boweria schatzlarensis, from the Westphalian of York¬

Telegram @tubscstds
shire; C. Renaultia gracilis, from Lancashire. (A from Mamay, 1955; B, C from
Kidston, 1923, by permission of the Controller of Her Majesty’s Stationery Office.)
122 STUDIES IN PALEOBOTANY

borne marginally at the tip of each lobe. Although the specimens

are of the compression type, it has been shown that a single spore
mass, consisting of many small spores 20 ju in diameter, is contained
within the sporangium wall, which was several cells thick. It is not
difficult to envisage this as a plant that evolved from one in which
the sporangia were borne terminally (as evidenced by the position
of the sporangia at the tips of veins) on a branch system that has
become partially webbed.
Other ferns of comparable morphology might be mentioned. One
other example will be noted to indicate that this evolution of the
superficial position probably took place in several different groups.
The pinnules of Boweria schlatzlarensis Kidston (from the West¬
phalian division of the Carboniferous) bore several marginal spo¬
rangia, each at the tip of a vein. The sporangia are smaller than
those of Acrangiophyllum, being about 0.5 mm in diameter and
having a distally located annulus of two rows of cells that are con¬
spicuously larger than the others of the sporangium wall. The
larger pinnules bore five to seven sporangia, while only one to three
are found on the smaller distal ones.
Renaultia gracilis (Brongniart) Zeiller (Fig. 4-15) from the West¬
phalian and Lanarkian series of Britain has fertile foliage in which
the sporangia are attached on the under side of the pinnule but close
to the margin. They are broadly oval, being about 0.4 mm long
and 0.3 mm in diameter and are arranged singly or in small groups
of two or three at the end of a vein. No annulus has been observed
and numerous small spores are contained in each sporangium.
Stages in the evolution of an organ in which sporangia shift from
a terminal to a marginal position and finally reach the superficial
one are summarized well in a series of diagrams by Zimmermann
(Fig. 4-16). In Chapter 3 we dealt in some detail with certain
aspects of this, namely, the problem of the evolution of a two-
dimensional, laminate frond. As to the sporangia, there are now
so many plants, both living and fossil, which attest to the general
sequence shown in Figs. I-P as to leave no doubt that this is what
took place.
I have included comparable series that Zimmermann has given
for the lycopods (A-D) and the articulates (E-H). The sequence
for the articulates seems probable, but there is less supporting evi¬
dence than there is for the ferns (Pterophyta); the sequence for the
lycopods (A-D) is, so far as I am aware, purely speculative. The
student may wish to refer back to these figures after studying the

Telegram @tubscstds
chapters on the lycopods and articulates; they constitute brilliant
MARATTIALES AND FILICALES 123

Fig. 4-16. Supposed stages in the development of sporophylls in the lycopods (A-D),

Telegram @tubscstds
in the articulates (E-H), and in the ferns (I-P). (From Zimmermann, 1952.)
124 STUDIES IN PALEOBOTANY

hypotheses which may well be confirmed as our knowledge of the

two groups expands.

NOEGGERATHIALES

In view of the distinctive nature of the spore-bearing organs the

fossils that are included in this group might well merit assignment
to a distinct Division. However, I am not convinced that this step
should be taken until more is known about them; thus, their inclu¬
sion here is admittedly a matter of convenience. The Noeggerathi-
ales is known at present from a small assemblage of foliage and
heterosporous cone specimens found chiefly in continental Europe.
The leaves are best described as cycadlike, but the cones are quite
unlike the reproductive organs of any other group of plants. The
assumption that cones and leaves belong to the same plant, or
closely allied plants, is based on association and the occasional pres¬
ence of leaflets at the base of the cone which compare with those
of the leaves proper.
Although the plants occur as compressions, some of the cones are
well preserved and a recent restoration by Halle, of Noeggerathio-
strobus bohemicus 0. Feistmantel, seems to conform rather well
with the known facts. The cones (Fig. 4-17) appear to have borne
two rows of semicircular scales or half-discs which functioned as
sporophylls; they might be compared with saucers cut in half and
attached to a central axis at the middle of the cut edge. The mar¬
gin of each scale is characteristically fringed and numerous sporan¬
gia, each containing 16 megaspores, are borne on the upper (adaxial
surface).
In Discinites major, another cone genus referred to the group,
Nemejc has found megasporangia with but one large megaspore
(1 mm in diameter) whereas the microsporangia contain numerous
microspores, each about 100 /x in diameter. The cones were at least
10 cm long and 4 cm in diameter. Another species, D. delectus
Arnold, has been found in the Michigan Coal Basin; the mega- and
microsporangia are similar in form, with the former containing 16
spores, each about 700 jti in diameter, and each microsporangium
contains many microspores each about 80 ju in diameter.
These cones, if they may be called such, of the noeggerathias
cannot be compared closely with the spore-bearing organs of any
pteridophytic group. Although they have been referred by vari-

Telegram @tubscstds
MARATTIALES AND FILICALES 125

ous workers to the palms, ferns, and cycads it is evident that there
are no close relationships in these directions. The stand taken by
Halle seems to be the only reasonable one when dealing with such
problematical fossils, and we shall encounter others to which it
applies with equal cogency.

Fig. 4-17. A tentative restoration of Noeggerathiostrobus

bohemicus, about natural size. (From Halle, 1954.)

Telegram @tubscstds
 cc
LU
Q_ U. S. A. EUROPE

Z _i *
<< L
li :
Cl CL l

4 CALHOUN

•4 PARKER (IND.)

^ RADNITZ

-X
DANVILLE
HERRIN ^ (BOHEMIA)

z ◄! HARRISBURG &
< gj PETERSBURG V
> o: <
<0
(IND.) CO

>- o
CO
*2 4 IOWA ?
or
z
4 ROCK ISLAND
O ^BEVIER ?

2 A FLEMING ?
w
q A WEIR-PITTSBURG ? O
CD
CZ
<
o 4 KATHARINA
(RUHR)

- DONETZ
CL < (RUSSIA)
Q_
ZD
FINEFRAU
(RUHR)

BOUXHARMONT COAL BALL

(BELGIUM) HORIZON
rSTALYBRIDGE

< -4 KOKSFLOZ
ce (LOWER — <
CE 5
E SILESIA)

z z
o <.

Fig. 4-18. A chart showing the stratigraphical position of American and European coal
ball horizons. (From Schopf, 1941, modified.)

Halle notes:
The natural tendency to squeeze fossil forms into existing taxonomic groups
has too often led to unfortunate consequences. In the present case it would
seem a safer course—whatever headlines be adopted for practical purposes—

Telegram @tubscstds
to regard the Noeggerathiineae as an isolated group of pterophyta incertae
sedis. (1954, p. 378)

126
MARATTIALES AND FILICALES 127

REFERENCES

Abbott, Maxine L. 1954. Revision of the Paleozoic fern genus Oligocarpia. Palaeon-
tographica, 96B: 39-65.
Andrews, Henry N., Jr. 1950. A fossil osmundaceous tree-fern from Brazil. Bull.
Torrey Bot. Club, 77:29-34.
-, and Kern, Ellen M. 1947. The Idaho Tempskyas and associated fossil
plants. Ann. Missouri Bot. Gard., 34: 119-186.
-, and Cortland S. Pearsall. 1941. On the flora of the Frontier formation of
southwestern Wyoming. Ann. Missouri Bot. Gard., 28: 165-180.
Arnold, Chester A. 1952. Fossil Osmundaceae from the Eocene of Oregon. Palaeon-
tographica, 92B: 63-78.
-. 1956. Fossil ferns of the Matoniaceae from North America. Palaeont.
Soc. India, Lucknow, 1: 118-121.
Bhardwaj, D. C., and Singh, Hari Pall. 1956. Asterotheca meriani (Brongn.) Stur
and its spores from the Upper Triassic of Lunz (Austria). The Palaeobotanist,
5: 51-55.
Bower, Frederick 0. The Ferns (Filicales). Cambridge Univ. Press. Vol. 1. 1923;
Vol. II. 1926; Vol. III. 1928.
Brown, Roland W. 1949. Cretaceous plants from southwestern Colorado. U. S. Geol.
Survey Prof. Paper, 221D: 45-66.
Chandler, Marjorie E. J. 1955. The Schizaeaceae of the south of England in early
Tertiary times. Bull. Brit. Mus. Nat. Hist., Geology, 7: 291-314.
Daugherty, Lyman H., and Howard R. Stagner. 1941. The Upper Triassic flora of
Arizona. Carnegie Inst. Washington Pub., 526: 1-108.
Endo, Seido. 1952. Klukia remains newly found in Japan. The Palaeobotanist,
1: 165-167.
Gillette, Norman J. 1937. Morphology of some American species of Psaronius. Bot.
Gaz., 99: 80-102.
Graham, Roy. 1934. Pennsylvanian flora of Illinois as revealed in coal balls. I. Bot.
Gaz., 95: 453-476.
Halle, Thore G. 1940. A fossil fertile Lygodium from the Tertiary of South Chile.
Svensk Bot. Tid., 34: 257-264.
-. 1954. Notes on the Noeggerathiineae. Svensk Bot. Tidskrift, 48: 360-380.
Harris, Thomas M. 1931. The fossil flora of Scoresby Sound East Greenland. Part
I. Meddelelser om Gronland, 85 (2): 1-102.
-. 1945. Notes on the Jurassic flora of Yorkshire. 19. Klukia exilis (Phil¬
lips) Raciborski. Ann. Mag. Nat. Hist., ser. 11, 12: 257-265.
-. 1947. The problems of Jurassic Paleobotany. Bol. Soc. Geol. Portugal,
6: 5-32.
Hirmer, Max, and Hoerhammer, L. 1936. Morphologie, Systematik and geographi-
sche Verbreitung der fossilen und rezenten Matoniaceen. Palaeontographica, 8IB:
1-67.
Kidston, Robert, and Gwynne-Vaughn, D. T. 1907-1914. On the fossil Osmundaceae.
Trans. Roy. Soc. Edinburgh, Pt. I (1907), 45: 759-780; Pt. II (1908), 46: 213-232;
Pt. Ill (1909), 46: 651-667; pt. IV (1910), 47: 455-477; pt. V (1914), 50: 469-480.
Mamay, Sergius H. 1950. Some American Carboniferous fern fructifications. Ann.
Missouri Bot. Gard., 37: 409-459.

Telegram @tubscstds
-. 1955. Acrangiophyllum, a new genus of Pennsylvanian pteropsida based
on fertile foliage. Amer. Journ. Bot., 42: 177-183.
128 STUDIES IN PALEOBOTANY

Morgan, Jeanne. 1959. The morphology and anatomy of American species of the
genus Psaronius. Illinois Biological Mono. No. 27: 1-108.
Nemejc, Frantisek. 1928. Revise Karbonske a Permske Kveteny Stredoceskych Panvi
Uhelnych. Paleont. Bohemiae, No. 12: 1-82.
Oishi, Saburo, and Yamasita, Kazuo. 1936. On the fossil Dipteridaceae. Joum.
Facutly Sci. Hokkaido Imper. Univ., ser. IV, vol. Ill, No. 2:135-184.
Pichi-Sermolli, Rodolfo, E. G. 1958. The higher taxa of the Pteridophyta and
their classification. Uppsala Universitets Arsskrift, 6: 70-90.
Radforth, Norman W. 1938. An analysis and comparison of the structural features
of Dactylotheca plumosa Artis sp. and Senftenbergia ophio dermatic a Goeppert sp.
Trans. Roy. Soc. Edinburgh, 59: 385-396.
-. 1939. Further contributions to our knowledge of the fossil Schizae-
aceae; genus Senftenbergia. Trans. Roy. Soc. Edinburgh, 59: 745-761.
-. 1942. On the fructification and new taxonomic position of Dactylotheca
parallela Kidston. Canadian Journ. Res., C, 20: 186-195.
-, and Woods, A. B. 1950. An analysis of Cladophlebis (Klukia) dunkeri
Schimper, a Mesozoic fern from western Canada. Canadian Journ. Res., C. 28:
780-787.
Read, Charles B., and Brown, Roland W. 1937. American Cretaceous ferns of the
genus Tempskya. U. S. Geol. Prof. Paper, 186-F: 105-129.
Remy, Winfried. 1953. Untersuchungen fiber einige Fruktifikationen von Farnen
und Pterospermen. Abhandl. Deutschen Akad. Wissenschaften Berlin, (Kl. Math.,
Naturwiss.) Jr., 1952, Nr. 2: 1-38.
Sahni, Birbal, and Rao, A. R. 1933. On some Jurassic plants from the Rajmahal
Hills. Journ. Proc. Asiatic Soc. Bengal, n. s., 27: 183-208.
-, and Sitholey, R. V. 1945. Some Mesozoic ferns from the Salt Range,
Punjab. Proc. Nat. Acad. Sci. India, 15: 61-73.
Seward, Albert C. 1922. The past and present distribution of certain ferns. Linn.
Soc. Journal, Botany (London), 46: 219-240.
-. 1924. On a new species of Tempskya from Montana: Tempskya knowltoni,
sp. nov. Ann. Bot., 38: 485-507.
-, and Dale, Elizabeth. 1901. On the structure and affinities of Dipteris,
with notes on the geological history of the Dipteridinae. Phil. Trans. Roy. Soc.
London, 194B: 487-513.
Zimmermann, Walter. 1952. Main results of the “Telome theory.” The Palaeobota-
nist, 1:456-470.

Telegram @tubscstds
 PER

O ne of the most significant contributions that paleobotany

has made to our knowledge of the plant kingdom is the
discovery and gradual elucidation of the assemblage of fossils that
is known as the pteridosperms or seed-ferns. A few of them are
now sufficiently well known as to reveal a great race of plants that
became extinct many millions of years ago. It is possible to set up
a rudimentary framework of the subgroups that compose the
Pteridospermophyta, and we are beginning to gain some insight
into the evolutionary trends. There are, however, many fragmen¬
tary but interesting remains of the various plant organs which can
with some confidence be called pteridospermous but beyond that
their taxonomic position is uncertain.
Of the features that characterize the pteridosperms the follow¬
ing seem to be most notable: They possessed fernlike foliage and
bore seeds, which are the features that contribute to the name
pteridosperm. The seeds, which were borne in a variety of ways
on the fronds, were in some cases partially enclosed in a distinctive
outer integument or cupule. The stems developed secondary wood
and a prominent outer cortex of longitudinally aligned fiber strands.
The microsporangiate organs were terminally clustered sporangia
which in many cases formed large complex synangial fructifications.
The pteridosperms appear first in the Lower Carboniferous and
are sufficiently advanced as to suggest a Devonian origin, but as of
this writing no undoubted representatives have been found at such
a low horizon. They were abundant through the Upper Carbon¬
iferous and Permian, and in the Triassic and Jurassic we find dis¬
tinct subgroups which most paleobotanists have regarded as later
offshoots of the Paleozoic lines.
The large and complex assemblage of fossil plants that is now
attributed to the Pteridospermophyta will be considered under the

Telegram @tubscstds
following headings:
129
130 STUDIES IN PALEOBOTANY

Lyginopteridaceae
Medullosaceae
(Some seed-bearing pteridospermous foliage)
(Some early fossils attributed to the pteridosperms)
Corystospermaceae
Peltaspermaceae

LYGINOPTERIDACEAE

Although fossils that are attributed to the Pteridospermophyta

range from the Lower Carboniferous to the Jurassic, it seems most
expedient to introduce the group as a whole with the well-known,
even classic example, Lyginopteris oldhamia (Binney) H. Potonie.
This species has been extensively investigated by Williamson,
Scott, and others, chiefly from a single important coal ball horizon
(see chart p. 126) of Yorkshire and Lancashire. Since vegetative
and reproductive characteristics of this species are correlated in
exceptional detail at the level of the Ganister and Halifax Hard
Bed coal, it is appropriate to use it as an example for the family.
The Lyginopteris oldhamia stems (Fig. 5-1) attained a diameter
of about 3 cm; in the center there was usually a rather large pith
of parenchymatous cells and scattered through this were clusters
of somewhat thicker-walled cells with dark contents. The pith is,
however, variable in both size and the relationship of the two cell
types that compose it. Around the periphery are several rather
large, mesarch primary strands (Fig. 5-1C); however, the bulk of
the primary wood is centripetal (develops toward the center of the
stem) so that the bundles are nearly exarch. Secondary wood was
formed in some abundance and consists of large tracheids with
numerous angular bordered pits in the radial walls; the pits are
irregularly arranged in contrast, for example, to the regularly
aligned vertical rows in cordaitean woods. This is a conspicuous
feature of most of the Carboniferous seed-ferns (Fig. 5-8D) although
it is perhaps questionable as to how much importance should be
attributed to it. The rays of the secondary wood vary from one to
a dozen cells wide (tangential dimension) and in height they range
from a small fraction of a millimeter (4 or 5 cells) to over 2 cm. It
would seem that this may be properly regarded as a very unspecial¬
ized type of secondary wood. It is of interest to mention that
varying amounts of internal secondary wood are occasionally found

Telegram @tubscstds
in Lyginopteris stems (Fig. 5-1C). It apparently originated from a
 PTERIDOSPERMOPHYTA 131

Fig. 5-1. Lyginopteris oldhamia. A. Stem in transverse section; p, pith with numer¬
ous sclerotic-secretory cells; t\, a leaf trace in an early stage of its departure; <2, leaf
trace that has passed through the secondary wood but is still undivided; t3, leaf trace
that has divided into two branches (two other traces appear in this divided condition
in the secton); s, secondary wood; oc, outer cortex. B. Tangential longitudinal section
through the outer cortex. C. An enlarged sector of a stele showing: s, normal second¬
ary wood; p, primary vascular bundle; si, internal secondary wood (an occasional anoma¬
lous feature).

Telegram @tubscstds
132 STUDIES IN PALEOBOTANY

cambium just within the primary strands and developed as a mir¬

ror image of the normal secondary wood. In view of the wide
variety of form taken in the formation of secondary tissues in
the Upper Carboniferous and Permian medullosan pteridosperms,
this may have some significance in Lyginopteris other than as a
mere “anomaly.”
A leaf trace originates by the tangential division of a primary
wood strand. It passed out through a gap in the secondary wood,
being accompanied by a small arc of the latter on its outer side.
Shortly after leaving the secondary wood it divides into a pair of
strands which traverse the cortex and ultimately unite again in the
base of the petiole to form a V-shaped leaf trace. Petioles of this
type were first observed and classified under a separate name,
Rachiopteris aspera. Their identity in the northern European
coal balls now seems fully established.
In the best preserved stems the secondary phloem may be dis¬
tinguished immediately outside the secondary wood. This is fol¬
lowed by a narrow band of radially aligned cells which is referred
to as a periderm; it is a distinctive feature of certain pteridosperms,
particularly the medullosas. The outer cortex is especially note¬
worthy (Fig. 5-1B); it consists of radially elongate fiber bands which
form a regular anastomosing network in their longitudinal course
through the stem. The “windows” of this network are filled with
parenchymatous cells. This fibrous outer cortex has been relied
on as a diagnostic feature of pteridosperm stems and petioles. It
is certainly characteristic of many, although the form and organi¬
zation of the strands are not the same in all.
The frond is a primitive sphenopterid type, apparently identical
to that of Sphenopteris hoeninghausi, which attained a length of
about one-half meter; the rachis dichotomized 10 cm from its point
of junction with the stem, and each of the two divisions was three
times pinnate. Some years ago I had occasion to collect some
rather fine stem and leaf impressions, tentatively referable to L.
oldhamia (Fig. 5-3), from a horizon in southern Scotland. The
fossils composed a very nearly “pure stand” in the rock, there being
few other associated plants and I have wondered if, as is supposed,
Lyginopteris actually grew as a vine depending on others for
partial support, or whether it may have formed dense thickets of
its own along the borders of the forests or within them. It should
be noted that these specimens were identified by leaf characters
and the presence of the “dictyoxylon” outer cortex which is so
striking a feature of the Lyginopteris oldhamia stems. However,

Telegram @tubscstds
PTERIDOSPERMOPHYTA 133

the fact that this type of cortical structure is not strictly confined
to L. oldhamia makes is impossible, so far as I am aware, to posi¬
tively identify this species in compression form. Two examples
may be cited to illustrate this rather important point:
In his Fossil Plants of the Carboniferous Rocks of Great Britain,
Part 1, Kidston figures a stem compression about 1.4 cm in diameter
from the Carboniferous Limestone Series (Lower Carboniferous) of
the “dictyoxylon” type that is attributed to Sphenopteris fragilis
(Schlotheim).

Figures 5-3B, C show portions of a stem and leaf from the Car¬
boniferous of Belgium that are attributed to Sphenopteris baumleri
Andra.
In both of these examples the pinnule structure is distinct from
that of the frond of L. oldhamia (although all are probably closely
related), but I do not believe it would be possible to distinguish the
three on the basis of stem impression characters.
The seeds of L. oldhamia, sometimes classified separately as
Lagenostoma lomaxi Oliver and Scott, are barrel-shaped, radially
symmetrical and 5.5 mm long by 4.4 mm in diameter (Fig. 5-2); they
have a rather stout integument, containing nine vascular strands,
which encloses the nucellus; the latter is elongate with a chamber
at the distal end for the reception of pollen grains. Long has
reported remarkable preservation in seeds of a closely related
species, Lagenostoma ouoides Williamson, containing the female
gametophyte; several archegonia are present near the micropylar
end. In all probability the pollen grains germinated in the nucel-
lar chamber and liberated sperms which fertilized the eggs of the
archegonia in much the same fashion as in modern cycads.
The seed was partially enclosed in a lobed husk or cupule (Fig.
6-2A) which bears distinctive glandular emergencies identical with
those present on the petioles and stems. Since most of the seeds
found in English petrifactions are isolated, it is presumed that they
were released from the cupule by a basal abscission mechanism.
Because the seeds present a critical phase in the restoration of
the plant, a few additional notes are appropriate at this point.
The name Lagenostoma lomaxi was first formulated by William¬
son in his catalogue notes but was formally published by Oliver
and Scott in 1904. Although the seeds were not found actually at¬
tached to the Lyginopteris leaves, the following evidence led them
to believe that such was the case: the two are often associated in
the same petrifaction; rachis and cupule bear distinctive, globular-

Telegram @tubscstds
134 STUDIES IN PALEOBOTANY

Fig. 5-2. A nearly median longitudinal section through the seed of Lyginopter-is
oldhamia; the cupule is not present; i, free or apical portion of the integument;
an, apical portion of nucellus; pc, pollen chamber; ga, gametophytic area—seeds have
been found in which this central region is occupied by a female gametophyte.

headed glands which are identical, and these are not known in any
other plant; the vascular bundle of the cupule pedicel compares
closely, even to pitting of the tracheids, with the trace of a small
rachis. This evidence has been accepted by most paleobotanists
as conclusive.
The male or pollen-producing organs of the plant have not been
positively identified, but based on correlation with allied species of

Telegram @tubscstds
PTERIDOSPERMOPHYTA 135

Sphenopteris it is possible that they were of the Crossotheca type.

The specimen shown in Fig. 5-4 is from Illinois and, as is usually
the case, the fertile pinnate branch system apparently terminated
part of a Sphenopteris frond. The ultimate branchlets (fertile
pinnules?) consist of a slightly flattened lamina with a peripheral
row of elongate, pendant sporangia.
The nomenclatorial problems centered around this and related
plants are somewhat more complicated than has been intimated
here, but it has not seemed advisable to discuss them fully since
practically an entire chapter would be required. It may be added

Fig. 5-3. A. A portion of a frond from southern Scotland that was probably borne on
the L. oldhamia stems. B, C. Frond and stem fragments of Sphenopteris baumleri from

Telegram @tubscstds
Belgium. (A from Andrews, 1947.)
136 STUDIES IN PALEOBOTANY

Fig. 5-4. Crossotheca sp.; restora¬

tion based on a specimen from the
Upper Carboniferous of Illinois,
about 2X.

that other species of Lyginopteris, based on stem anatomy, have

been described by Kubart from somewhat older deposits in Lower
Silesia. Furthermore, the multiplicity of species that probably
possessed similar organization is suggested by fossils such as the
one shown in Fig. 5-3B, C. However, it is only in exceptional in¬
stances that diagnostic features of a species of Lyginopteris and a
species of Sphenopteris can be demonstrated in connection. It is
most difficult therefore to establish a precise correlation outside of
deposits like the British coal balls which are actually exceptional
in their occurrence, in the preservation of included plant materials,
and in the extent to which they have been studied. As a conse¬
quence, although the foliar characteristics of Lyginopteris oldhamia

Telegram @tubscstds
appear to be identical with those of Sphenopteris hoeninghausi at
the Ganister coal bed horizon, the history of this correlation of
PTERIDOSPERMOPHYTA 137

diagnostic features is uncertain and debatable in younger and older

deposits.
In summary, Lyginopteris oldhamia was a fairly common plant
in the European Upper Carboniferous which bore sphenopterid
foliage and, in all probability, the seeds called Lagenostoma
lomaxi. It is important as one of the better known pteridosperms
and the pioneering studies devoted to it by Williamson, Oliver, and
Scott, and others form an important chapter in our understanding
of the seed-ferns. However gaps still remain in our knowledge of
the plant.

Some Other Fossils Attributed to the Lyginopteridaceae

Stems

Callistophyton poroxyloides Delevoryas, from upper Pennsylva¬

nian coal balls of Illinois, is a stem that is marked by certain distinc¬
tive features and especially fine preservation. In the center is a
rather large parenchymatous pith with occasional secretory cavi¬
ties; the cells are, in fact, quite thin-walled and in some specimens
this region has collapsed. Several mesarch primary strands are
present around the periphery of the pith and these are followed by
a strong development of secondary wood. In the largest stems,
which attain a diameter of a little over 2 cm, this tissue makes up a
considerable portion of the stem; it is quite like that of Lyginop¬
teris although the rays are usually narrower.
A leaf trace originates from a peripheral primary bundle which
divides to form two, each of which develops a distinct pair of pro-
toxylem groups. Thus at first glance the trace, which is quite con¬
spicuous, gives the impression of being composed of four strands.
It passes out through the wood and cortex, being accompanied on
the outer side by a segment of secondary wood. It may also be
noted that axillary branching seems to have been a well-developed
feature of the plant.
The secondary phloem is composed of rays that are continuous
with those of the wood, parenchyma cells, and sieve cells; the lat¬
ter bear simple sieve plates on their radial walls—an exceptional
feature of preservation. Fiber bands are found in the outer cortex,
although they are less conspicuous than those of Lyginopteris and
they follow a parallel course. Occasional glands are present on the
cortex which are similar to those of the latter genus but smaller.
This fibrous outer cortex tends to be sloughed off in older stems
following the development of periderm in the tissue immediately

Telegram @tubscstds
peripheral to the phloem.
138 STUDIES IN PALEOBOTANY

Fig. 5-5. Callistophyton poroxyloides, a Pennsylvanian seed-fern stem from Illinois,

6X. c, outer primary cortex; p, pith; sp, secondary pholem; sw, secondary wood.
(From a preparation by T. Delevoryas.)

Schopfiastrum decussatum Andrews is a distinctive pteridosperm

stem (Fig. 5-6) from the Des Moines series of Iowa. It is a little
more than 2 cm in diameter and has a mixed protostele surrounded
by secondary wood with narrow rays and tracheids with closely
compacted bordered pits. The outer cortex is composed of
parenchyma and regularly spaced, radially elongate fiber strands;
these run a nearly parallel longitudinal course and thus present a
distinct contrast to the netlike arrangement found in Lyginopteris.
Of special note are the leaf traces which are large, tangentially
elongated strands given off on opposite sides of the primary wood,
with one departing slightly above the other. Although this plant

Telegram @tubscstds
 PTERIDOSPERMOPHYTA 139

Fig. 5-6. Schopfiastrum decussatum, a pteridosperm stem (somewhat crushed) from

the Upper Carboniferous of Iowa; l.t., a large, lobed leaf trace departing from the stele.

Telegram @tubscstds
140 STUDIES IN PALEOBOTANY

is known from only two small fragments, the available material

suggests that successive pairs of traces alternate, resulting in four
rows of leaves. It was either a rare plant or is represented by
specimens that were carried in from a distant region.
Heterangium is a genus of several species from both Lower and
Upper Carboniferous horizons. One of the best known as well as
the oldest is H. grievii Williamson from Pettycur in Scotland. The
stem reached a diameter of 4 cm and apparently branched only
rarely. The center is occupied by a terete, mixed protostele in
which the tracheids occur in groups of six to a dozen scattered
quite uniformly through the parenchyma. A small amount of
secondary wood is present surrounding the protostele. Radially
elongate fiber bands are present in the outer cortex, whereas the
inner cortex is distinguished by horizontal plates of thick-walled
cells. Leaves known as Diplotmema adiantoides are attributed to
the H. grievii stems. In an Upper Carboniferous species from the
McLeansboro formation of Illinois (H. americana Andrews) ap¬
preciably more secondary wood is formed, there being as much as
5 mm in one specimen, although the extent of its development
varies in different specimens. The secondary wood is similar in
tracheidal pitting and ray structure to Lyginopteris. The vascular
supply to a leaf originates as two separate strands at the periphery
of the primary wood; these divide in their passage through the cor¬
tex to form an arc of four bundles in the petiole and the two
central ones soon trifurcate.

Seeds

There are several features of special interest in the recently dis¬

covered seed Tyliosperma orbiculatum Mamay (Fig. 13-3), found in
the Des Moines series of Kansas; it is, in fact, one of the few
pteridosperms in which the cupule is well preserved. The seed is
small, reaching a length of but 3.7 mm and a diameter of 3.5 mm;
it is nearly circular in cross section. The nucellus and integument
present a closely knit unit; a set of vascular strands passes up ap¬
parently in the peripheral region of the nucellus and the tissue that
is identified as integument is not vascularized. Of particular
interest is that the epidermis of the integument is continuous over
the distal end of the nucellus, a relationship that is very difficult to
explain if the integument originated as a distinct tissue and grew
up around the nucellus. The distal portion of the integument is
divided into seven lobes. One cannot avoid the suspicion that

Telegram @tubscstds
PTERIDOSPERMOPHYTA 141

there is no real distinction here between integument and nucellus,

but rather that they compose a unit tissue system. The cupule is
quite deeply lobed, more so than that of Lyginopteris, and it did
not enclose the seed as completely; rather stout, blunt protuber¬
ances occur on its outer surface, but they apparently were not
glandular.
Physostoma elegans Williamson is a beautiful little seed about 6
mm long and 2.3 mm in diameter from the Upper Carboniferous of
England. Like other “lagenostomalian” seeds the integument is
fused with the nucellus except at the apex. The pollen chamber,
formed by nucellar tissue, has a distinctive shape (Fig. 5-7) and
may contain large numbers of pollen. Scott remarked on this in
the first volume af his Studies and I once enjoyed the thrill of cut¬
ting into two specimens, found in an English coal ball, one of
which contained many pollen grains. The distal part of the integu¬
ment is divided into about ten elongate lobes or tentacles and the
entire outer surface of the integument is covered with conspicuous,
fat, unicellular hairs.

Fig. 5-7. Longitudinal aspect of Physo¬

stoma elegans; several pollen grains are
shown in the pollen chamber.

Telegram @tubscstds
142 STUDIES IN PALEOBOTANY

THE MEDULLOSACEAE

The basis of this group of pteridosperms is a large and unique

series of stems that has been found in Upper Carboniferous and
Permian rocks; they are identified chiefly by their polystelic vascu¬
lar systems which, in some species, are of unequalled complexity.
Most of them have been assigned to Medullosa and, although it is
evident that it is a “genus” of extraordinarily broad limits, it seems
to be an expedient taxonomic procedure in the present light of our
knowledge. The fronds borne on these stems were of the Aletho-
pteris and Neuropteris types and in their anatomy the petioles
present a striking superficial resemblance to monocot stems. The
seeds that are associated with the stems and fronds are generally
larger than those of the Lyginopteridaceae and the nucellus and
integuments are separate; the microsporangiate organs are for the
most part large, complex synangia.
Medullosa thompsoni Andrews from Iowa may be considered
first as a representative Upper Carboniferous species. The stem
measures about 4 cm in diameter and contains three steles, each
consisting of a small central primary core of tracheids and
parenchyma surrounded by secondary xylem of large, pitted tra¬
cheids and tall, slender rays. The steles range in over-all size from
7 X 3.5 mm in diameter for the largest to 2 mm in diameter for the
smallest. The stelar system is enclosed by a continuous, narrow
band of thin-walled radially aligned cells, the inner periderm. Ex¬
ternal to this is a broad parenchymatous cortex which is penetrated
by numerous small leaf traces which depart from the steles. The
outer cortex consists of circular to radially elongate fibrous strands
which run longitudinally and occasionally anastomose; a fewr secre¬
tory canals are associated with them.
In Fig. 5-8A a leaf base appears near its point of departure and
Fig. 5-8B shows an isolated petiole at a somewhat higher magnifi¬
cation. The structure of the outer cortex is essentially the same
as the corresponding tissue in the stem although the secretory
canals may be somewhat more abundant. The vascular system,
unlike that of the stem, consists of numerous small, scattered
bundles.
Over 40 species and varieties have been recognized although it is
now evident that some of these actually represent variants within
a species. The genus was established by Cotta as far back as 1832

Telegram @tubscstds
and quite a number of Permian species were described during the
Fig. 5-8. Medullosa thompsoni. A. Transverse section of stem with three central
steles and petiole departing above; B. a petiole; C. stelar system enlarged; D. pitting

Telegram @tubscstds
in secondary tracheids. (From Andrews, 1945.)

143
144 STUDIES IN PALEOBOTANY

latter part of the nineteenth century, a comprehensive account of

them being given by Weber and Sterzel in 1896. Scott described a
well-preserved stem (M. anglica) from the English coal fields in
1899 and this was followed by the discovery of several others.
With the initiation of coal-ball studies in this country a remark¬
able series of specimens have come to light in the Illinois, Kansas,
and Iowa deposits. This work has been summarized and revised
in recent studies by Stewart and Delevoryas.
The species vary tremendously in over-all size and in the nature
of the stelar system. M. endocentrica Baxter from Illinois has a
vascular system of 3 steles, 2 of which are only 5 mm in diameter
and the third one is even smaller; in certain specimens of M. gran-
dis Andrews and Mamay from Kansas the vascular system is com¬
posed of 18 to 20 steles with an over-all diameter of about 12 x
28 cm. In all the medullosan stems the steles tend to anastomose
and divide to a greater or less degree so that the exact number is
not a dependable taxonomic character. Some species vary in the
uniformity of development of the secondary wood; the cambial ac¬
tivity was fairly uniform around the primary wood of certain species,
whereas in others it was strongly endocentric, that is, secondary
wood was developed almost exclusively toward the center of the stem.
A particularly interesting discovery was made by Stewart and
Delevoryas in dealing with a specimen of M. primaeva (originally
described under the name of M. heterostelica) in which there are two
steles in the internodal part of the stem and these suddenly pro¬
liferate to a total of 23 steles at the node; leaf traces depart from
some of them and they quickly reassemble above the node to the
two-stelar condition. A more recent survey of numerous specimens
which are closely related or actually referable to this species sug¬
gest that this nodal-internodal relationship does not always hold
true.
For their bizarre stem anatomy the Permian medullosas of Ger¬
many are deserving of more space than is available here; a few will
be mentioned to indicate some of their diversity. Figure 5-9F is a
cross-sectional diagram of a specimen of M. stellata Cotta that is
preserved in the Swedish Natural History Museum. The stem as
a whole measures approximately 24 x 34 cm and includes about 70
steles in the central portion; each one has a very small central
group of primary tracheids, with perhaps a few associated paren¬
chyma cells. The cells are surrounded by secondary xylem which
makes up the greater portion of the stele. In over-all size they

Telegram @tubscstds
vary from about 3 mm to 25 mm in diameter. This entire central
Fig. 5-9. Diagrams of the stelar systems of several medullosas. A. M. leuckarti,
Permian of Chemnitz; B. M. solmsi, Permian of Chemnitz; C. M. endocentrica, Penn¬
sylvanian of Illinois; D. M. grandis, Pennsylvanian of Kansas; E. M. noei, Pennsyl¬
vanian of Illinois; F. M. stellata, Permian of Chemnitz. (A,F drawn from specimens

Telegram @tubscstds
in the Swedish Natural History Museum.)

145
146 STUDIES IN PALEOBOTANY

complex is, in turn, enclosed within a rather broad and nearly com¬
plete band of secondary xylem.
Several variations of M. stellata have been described, some of
which display only 4 or 5 central steles and with a much thicker
peripheral band of secondary wood than in the specimen described
above. In M. stellata var. typica there are two enclosing bands of
wood which are inversely oriented; that is, there is an inner band
of wood with the phloem located on the inner side; this is
enclosed by a parenchymatous zone with some scattered tracheids,
which is followed by another zone of wood with the phloem located
on its outer side. In the variety gigantea the stems attain a diam¬
eter of 50 cm and, in addition to the stelar system described for
the variety typica, there is a series of several additional concentric
bands of secondary wood. These undoubtedly developed from suc¬
cessive cambia that originated in the cortex as happens in certain
living species of Cycas.
In Medullosa leuckarti Goeppert and Stenzel (Fig. 5-9A) there
are several of the more or less cylindrical central steles, which are
surrounded by a wavy ring of several large, tangentially elongate
steles that have been called a “snake ring.” In M. solmsi Schenk
(Fig. 5-9B) there is a rather large “pith” or central region where a
few inconspicuous small cylindrical steles are found, and this is sur¬
rounded by two series of tangentially flattened and elongate steles.
In the variety lignosa of this species the secondary xylem of the
outermost ring is excentrically developed; in addition there are
several concentric bands of xylem that develop in the “normal”
fashion.
In his Autun and Epinac flora (of France) Renault cites wood
fragments of M. gigas Renault that are 50 cm thick. Judging
from his description and illustrations this is a radial dimension and
implies a stem that was in excess of 1 meter in diameter.
It is quite clear that we have here a mode of stem growth that is
not comparable with that of present-day forest trees. Delevoryas
has shown that in young stems the steles may be more closely ag¬
gregated than in mature ones and that the increase in trunk diam¬
eter is a result of continued growth of the parenchymatous ground
tissue surrounding the steles as well as cambial activity of the
individual steles; thus in older stems the steles may be farther apart
than in younger ones.
The origin and evolution of the medullosan stems have been the
subject of considerable speculation and will undoubtedly continue
to be for some time. Several students of the group have envisaged

Telegram @tubscstds
 PTERIDOSPERMOPHYTA 147

Fig. 5-10. Restoration of a Medullosa about 12 to 15 feet tall; based on specimens of

M. noei. (From Stewart and Delevoryas, 1956.)

Telegram @tubscstds
148 STUDIES IN PALEOBOTANY

lines of increasing complexity, possibly originating in simpler pteri-

dosperms of the Heterangium type; the single stele of the latter with
its mixed primary xylem cylinders surrounded by a moderate devel¬
opment of secondary wood is not unlike a single medullosan stele.
Delevoryas, on the other hand, regards the many-steled forms as
primitive, with a phyletic fusion resulting in the few-steled Carbon¬
iferous species. A second trend, involving tangential fusion may
have led to the M. leuckarti type with continued fusion of the
“snake rings” into the M. stellata and M. gigas types in which
there is a peripheral, continuous ring (or rings) of wood. I am per¬
sonally inclined toward this latter theory, in which case it would
seem as though one may look to late Devonian and early Carbon¬
iferous fossils such as Xenocladia-Steloxylon-Cladoxylon as possible
ancestral types. Quite clearly, however, we do not know enough
about these earlier fossils to render this more than a rough-working
hypothesis. At the other end of the scale, some of the Permian
medullosas begin to approach a type of stem anatomy that is not
distantly removed from that of the cycads.
As a concluding comment on this great array of stems it may be
stated with some certainty that Medullosa is a “genus” of conveni¬
ence. It is difficult to draw sharp lines between the different species
and we are, in fact, just beginning to learn something of the degree
of stelar variation that may be present in the ontogeny of a single
plant, but it seems virtually certain that Medullosa will eventually
be divided into several distinct genera.

Associated Leaves and Reproductive Organs

In American petrifactions in which medullosan stems are abun¬

dant they are constantly associated with Alethopteris and Neurop-
teris type leaflets and the various branch orders of the fronds;
there is little doubt that they constitute the foliage of many Medul¬
losa stems. Other leaf genera which have been suspected of being
medullosan, at least in part, are Odontopteris, Linopteris, Lonchop-
teris, Callipteris, and Taeniopteris.
Among the most interesting fossils that have been referred to the
pteridosperms are the large and complex microsporangiate organs
of the genus Dolerotheca (Fig. 5-1 IB); these are frequently found
associated in American coal balls with medullosan stems and fronds.
They are bell-shaped fructifications consisting of several hundreds
of tubular sporangia embedded in a cellular groundwork.

Telegram @tubscstds
PTERIDOSPERMOPHYTA 149

Fig. 5-11. Pteridosperm microsporangiate organs. A, Aulacotheca elongata, upper

portion cut away to show internal structure, about 6X; B, Dolerotheca formosa, half
of a fructification. (A from Halle, 1933; B from Schopf, 1948, modified.)

Dolerotheca formosa Schopf is nearly 4 cm in diameter and the

individual sporangia are about 14 mm long and 0.8 mm in diame¬
ter; the huge ovoid microspores, of which there are several hundreds
in each sporangium, are about 0.4 mm long. Since they were quite
abundant in certain parts of the Carboniferous forests the big
Dolerotheca “bells” must have presented a most picturesque sight
at the time of spore dispersal.
In 1933 Halle presented a study of several presumed pterido-
spermous microsporangiate organs which is one of the modern clas¬
sics in paleobotany. Although he dealt with compression speci¬
mens, an embedding technique was devised which made it possible
to section the fructifications and prepare restorations which for the
most part seem to be beyond question. The following descriptions
are based largely on his study.

Telegram @tubscstds
150 STUDIES IN PALEOBOTANY

Whittleseya, in its compressed form, is an elongate, shovel-shaped

structure that was originally described from the American coal
fields and was thought to be a ginkgophyte leaf. W. elegans New¬
berry vies in size with Dolerotheca although its morphology is
rather different. According to Halle’s restoration it is a hollow,
campanulate structure with a wall composed of a single ring of long,
tubular concrescent sporangia. The spores are about 0.25 mm long,
which is large but not as large as those of Dolerotheca.
In one of his last contributions the late W. J. Jongmans made a
significant addition to our knowledge of the pteridosperms in dis¬
covering Whittleseya fructifications attached to Neuropteris schle-
hani Stur foliage (Fig. 5-12). These were found in the Limburg
coal field in the southeastern corner of the Netherlands and the
same collection included other specimens of N. schlehani with seed¬
like structures attached which are considered similar to seeds de¬
scribed in this country as Pachytheca vera Hoskins and Cross.

Fig. 5-12. Neuropteris foliage

bearing a Whittleseya media
synangium, about natural
size; a restoration based on
specimens from the Limburg,
Holland, coal field. (From
Stewart and Delevoryas,
1956.)

Telegram @tubscstds
PTERIDOSPERMOPHYTA 151

Aulacotheca elongata (Kidston) Halle is another synangial type

that is known from numerous specimens found in Upper Carbonif¬
erous horizons in Scotland and northern England. Enough have
been found with the spores still intact to leave no doubt as to the
validity of Halle’s restoration (Fig. 5-11 A). The synangium is about
2 cm long and was composed of about nine sporangia.
The synangia or capsules of Goldenbergia glomerata Halle were
less elongate, measuring about 8 mm long and 4 mm broad and con¬
sisting of 12 to 16 sporangia containing spores about 0.4 mm long.
The outer surface is covered with short, pointed spines.
The genus Potoniea is known from several species and numerous
specimens, but its structure is not as precisely understood as in the
genera cited above. As interpreted by Halle it is an open cup-shaped
organ with several dozen sporangia attached to the inside of the
cup. Specimens have been found with the spore masses intact, but
there is still some question as to whether the sporangia were actu¬
ally free or whether there was an enclosing tissue.
Our knowledge of petrified seeds of medullosan affinities is based
largely on association; Pachytesta illinoense (Arnold and Steidt-
mann) Stewart is one (Fig. 5-13) that has been found in some num¬
bers with the stem remains of Medullosa rioei. The seeds vary in
size, some reaching a length of 4.5 cm. They are nearly circular in
a median cross section and become triangular toward the micro-
pylar end. The integument is quite complex, consisting of a thin
inner layer of elongate cells with dark contents. Next there is a
layer of thicker-walled cells which develop numerous conspicuous
longitudinal ribs. The region between the ribs is occupied with
thin-walled cells of rather spongy organization and, as might be ex¬
pected, not often well preserved. The integument is free from the
nucellus except at the base of the seed. Two features of the nucel-
lus are noteworthy: a pendant skirt of tissue at the base and a
doughnut-shaped micropyle at the apex, both of which are apparent
in the restoration. About 24 vascular bundles extend up through
the nucellus and about 40 in the fleshy sarcotesta of the integument
alternating with the sclerotestal ribs.
Stephanospermum is another rather well-known genus of seeds
that may have been borne on the medullosans. S. elongatum Hall
from the upper Pennsylvanian of Illinois is a little under 2 cm long
and about three-quarters of a centimeter in diameter. The integu¬
ment, which is free from the nucellus except at the base, is com¬
posed of the following tissues: a peripheral fleshy layer, traversed

Telegram @tubscstds
by several vascular strands, and very irregularly lobed on the out-
152 STUDIES IN PALEOBOTANY

Fig. 5-13. Restoration of the seed

Pachytesta illinoense; i, integument;
n, nucellus; pc, pollen chamber.
(From Stewart, 1954.)

side; next a fibrous zone, and innermost a single layer of thin-walled

cells. The integument is elongated at the distal end to a length al¬
most equal to that of the main body of the seed; one might suppose
that this elongate, lobed apex could have served as a pollen-catching
mechanism, but the lobing is continuous over the entire integument.
The nucellus, which has an inverted bell-shaped pollen chamber at
its distal end, is distinctive in the manner of its vascularization
which is described as a “trachael mantle” and invests the nucellus
on its inner side. In this species the mantle is two or three cells
wide, the cells being elongate, slender, reticulate to scalariform
tracheids. In another species, S. steward Hall, from Kansas the
mantle is composed of three or four rows of tracheids which are
only two or three times as long as wide.

Some Seed-bearing Pteridosperm Foliage

As the name “seed-fern” implies, this is a group of plants of fern¬

like aspect, yet, unlike the true ferns, they bore seeds. Enough is
now known about them so that distinctive characters, as noted

Telegram @tubscstds
above, are evident in the stems, seeds, and microsporangiate organs.
PTERIDOSPERMOPHYTA 153

However, the most critical feature of the group is the actual attach¬
ment of seeds to the organ on which they were borne. It is there¬
fore pertinent to consider several examples of the attachment of
seeds, or structures that are presumed to be seeds, to the fernlike
fronds.
Sphenopteris tenuis Schenk, from the Permian of Shansi, China,
(Fig. 5-14A) is based on fronds that attained a breadth of some 24
cm. About four seeds are borne on each ultimate pinna, being ar¬
ranged along the midvein on the under side; they are about 4 mm
long by 2 mm broad and have a protracted tip. Seeds have also been
reported on specimens of Sphenopteris alata found in the Somerset
coal field (Radstockian age) of Wales. In this species they were
borne in a terminal position at the tips of apparently specialized
ultimate pinnae in which little lamina was developed; the foliage
otherwise consists of rather finely dissected pinnules verging on the
Diplotmema type. The seeds are 3 to 4 mm long, 2 to 3 mm broad,
and were enveloped in a cupule.

Fig. 5-14. Fertile (seed-bearing) pteridosperm foliage from the Permian of China.
A. Sphenopteris tenuis. B. Alethopteris norinii, C. Emplectopteris triangularis. (A,B

Telegram @tubscstds
from Halle, 1929; C from a drawing supplied by Professor Halle, after Andrews, 1947.)
154 STUDIES IN PALEOBOTANY

Several examples are now known of Alethopteris and Neuropteris

type foliage that bore seeds. Alethopteris norinii Halle, from the
Lower Permian of Shansi (Fig. 5-14B), is known from a fragment of
a leaf with alethopterid pinnules and a seed 40 mm long and 12 mm
broad attached to the rachis or midrib of the pinna, probably on
the upper side. The portion of the frond on which the seeds are
found is a normal one, with pinnules both above and below the
point of attachment of the seed. Arnold has described abundant
seeds associated with the foliage known as Alethopteris grandifolia
Newberry from the Upper Carboniferous of Summit County, Ohio,
and reports one specimen with the basal portion of a seed attached,
apparently in much the same manner as the Shansi Alethopteris.
As long ago as 1904, in the same year that Oliver and Scott demon¬
strated the association of Lagenostoma lomaxi with Lyginopteris
oldhamia, Kidston described a specimen of Neuropteris hetero-
phylla bearing a seed about 3 cm long. It is not quite clear as to
just how the seed was attached, but it was probably joined directly
to the rachis of the ultimate pinna, and enough of a cast is present
to indicate that it was of the radiospermic type, that is, a seed that
was radially symmetrical in cross section.
In 1911 Kidston and Jongmans described seeds attached to a frag¬
ment of a frond of Neuropteris obliqua; these present two points
of special interest. First, the seeds are quite large, being in excess
of 4.5 cm long (the tips were missing so that the entire length could
not be determined) and over 2 cm in diameter and, second, they
were borne on special peduncles about 1 cm long.
Specimens of Neuropteris tenuifolia from the Upper Carbonifer¬
ous of Yorkshire are known in which the presumed seed occupies a
terminal position on the pinna. It is slightly parted at the apex,
suggesting the possibility that it may represent a cupule with or
without its enclosed seed.
Emplectopteris triangularis Halle, from the Permian of China
(Fig. 5-14C), presents several variations on the patterns described
above. The proximal portion of the pinnules is net-veined and the
winged seeds are attached close to the midrib of the pinna, probably
on the upper surface.
One other plant from Professor Halle’s collections of Chinese
Permian fossils will be mentioned briefly, although its proper inclu¬
sion in the pteridosperms is by no means certain. Nystroemia pec-
tiniformis Halle (Fig. 5-15) is based on portions of a branch system
that has the general organization of a fern frond. The ultimate

Telegram @tubscstds
subdivisions (pinnules?) divide several times, each branchlet termi-
PTERIDOSPERMOPHYTA 155

Fig. 5-15. Nystroemia pectiniformis; A. seed¬

bearing and B. microsporangiate portions of
frond. (From Halle, 1929.)

nating in a small, two-horned body about 3 mm long that is pre¬

sumed to be a seed. It is possible that the fossil is a specialized
reproductive portion of a frond. Associated with them were slightly
curved branches bearing clusters of elongate structures that may
represent the microsporangiate organs.
Finally, in Pecopteris pluckenetii, the seeds are attached along
the margins of the pinnules which are essentially normal (pecop-
terid) or but slightly reduced; the position of the seed also corre¬
sponds to the termination of a lateral vein.

The problem of the position of the seed

Aside from finding seeds actually attached to the fernlike foliage

of the pteridosperms, the problem of how they attained the variety
of positions in which they have been reported is hardly of less im¬
portance. In a summary discussion dealing with their position
Halle was shown that the earlier members of the group, such as
Lyginopteris oldhamia (lower Upper Carboniferous), bore the seeds
on specialized segments of the frond; in Pecopteris pluckenetii (mid¬
dle Upper Carboniferous) they are marginal on essentially normal
pinnules; and in certain Permian representatives such as Alethop-
teris norinii, Emplectopteris triangularis, and Sphenopteris tenuis
they are found on the surface of normal foliage leaves—in the first

Telegram @tubscstds
two, on the upper side and on the lower side in the third. There
156 STUDIES IN PALEOBOTANY

is thus a suggestion that this surface position on a laminated leaf

may have come about by the “migration” of the seed from a termi¬
nal, to marginal, to surface position comparable to transitions that
seem to have taken place with respect to the sori in the ferns. There
is a notable difference in that the seed tends to occupy a place on
the upper surface of the frond rather than the lower. Halle points
out that the location of sporangia on the under surface presents an
advantage for spore dispersal, whereas the position of seeds on the
upper surface would be more effective for the reception of wind-
borne pollen.
There are now enough examples known of seeds attached to fern¬
like foliage to leave no doubt that they were borne in a variety of
positions, but a great deal more information is needed before we will
have a satisfactory understanding as to how they evolved.

Some Early Pteridosperm Fossils

The pteridosperms were well established in the Lower Carbonif¬

erous although much less is known about them from that period.
The fossils described here seem to be of considerable interest in that
they display both primitive and advanced characters and it is pos¬
sible that they all belong to a closely related group.
In 1938, W. T. Gordon described petrified stems from Lower
Carboniferous rocks of East Lothian, in southern Scotland, under
the name Tetrastichia bupatides. They are about 1 cm in diameter
and contain a cruciform protostele of reticulately thickened tra-
cheids; in some specimens there is a weak development of secondary
wood, the secondary tracheids being pitted on both radial and tan¬
gential walls. The vascular traces that supply the petioles depart
in opposite pairs, one on each side of the stem, with one slightly
above the other, and successive pairs alternate, resulting in a four-
ranked arrangement as in Schopfiastrum. Fiber strands are present
in the outer cortex of the stem and petioles.
Leaves known as Telangium affine (Fig. 3-13B) also occur in the
Lower Carboniferous of southern Scotland and, in their size and
cortical features of the petiole, there are reasons to believe that they
may have been borne on the Tetrastichia stems. They are common
fossils in the oil shales and may be so well preserved that pieces of
the leaf can be lifted from the rock without any special treatment.
A complete frond would measure about 45 cm long; the “main
rachis” dichotomized two or three times and, as noted previously,

Telegram @tubscstds
the frond is essentially a flattened telomic branch system, there
PTERIDOSPERMOPHYTA 157

being only a feeble development of leaflets as such. The fertile

pinnae of T. affine bear terminal groups of six partially united spo¬
rangia, each about 3 mm long, which contain numerous spores; these
were probably attached at the base of the frond as is known to be
the case in an allied species, T. bifidum.
Some years ago the writer found a large tulip-shaped fossil
(Megatheca thomasi) associated with the Telangium foliage in the
oil shales of Midlothian and at about the same time Professor
Walton discovered a partially petrified seed-bearing organ, Calatho-
spermum scoticum (Fig. 5-16A) in a Lower Carboniferous horizon
of the Kilpatrick Hills of Scotland. It is probable that the two are
identical, but since Calathospermum is more completely preserved
the following description is based on it. It is a tulip-shaped cupule
45 mm long and 20 mm broad and is partially divided into six lobes,
each of which is supplied by several vascular strands. As many as
60 seeds are found within the cupule, being borne terminally on
slender stalks that emanate from the basal region. The seeds are
distinguished by a pollen chamber surmounted by an elongate nucel-
lar tube which is enclosed by prolongations (usually nine) of the
integument which are actually longer than the main body of the
seed. It is suggested that the long stalks bearing the seeds served
to extrude them for more effective pollination and that dispersal
was thus facilitated.

Summary—Paleozoic Pteridosperms

The occurrence of fossils like Calathospermum in the Lower Car¬

boniferous would seem to imply that the pteridosperms had already
undergone a considerable period of evolution. The seeds contained
in the large cupules appear to be highly specialized with respect to
the elongate distal lobes of the integument. It is of interest here
to comment briefly on Gnetopsis elliptica Renault (Fig. 5-16B), a
cupulate organ from the Upper Carboniferous of St. Etienne,
France; it was considerably smaller, being 6.4 mm long, not as deeply
lobed and contained only two seeds. The latter are distinguished
by several elongate, slender, hair-covered processes that extend out
from the distal end of the integument and may have functioned as
pollen-trapping mechanisms.
One cannot avoid the suspicion, on finding fossils like Calatho¬
spermum in the Lower Carboniferous, that pteridosperms will ulti¬
mately be recovered from Devonian horizons. Structures that

Telegram @tubscstds
might be interpreted as cupules have, in fact, been found in the
158 STUDIES IN PALEOBOTANY

Fig. 5-16. A. Calathospermum scoticum; longitudinal section through a cupule with

some of the enclosed seeds. B. Gnetopsis elliptica; a cupule from the Upper Carbon¬
iferous of France, with two seeds. (A from Walton, 1940; B from Renault and Zeiller,
1888 and Oliver and Salisbury, 1911.)

Devonian, but the evidence they afford is not conclusive; they will
be considered further in Chapter 13. The stem and leaf remains
(Tetrastichia, Telangium) seem less advanced than Calathosper¬
mum, although our knowledge of the pteridosperms of this age is
admittedly too scanty to allow one to be at all dogmatic about evo¬
lutionary trends. Tetrastichia is similar in its anatomy to the

Telegram @tubscstds
Upper Devonian Tetraxylopteris which is tentatively placed in the
PTERIDOSPERMOPHYTA 159

protopterids, and Telangium is little more than a two-dimensional

protopterid “leaf.”
A seed-bearing organ as complex as Calathospermum does, how¬
ever, afford conclusive evidence that the pteridosperms are not
closely allied to the true ferns. With the exception of sporangia
that are doubtfully referable to the Schizaeaceae the true ferns have
not been found below the Upper Carboniferous. Thus, the evidence
available at present indicates that the pteridosperms and ferns origi¬
nated as independent lines (probably several independent lines in
each group) from the coenopterid-protopterid complex. The term
seed-fern is thus an unfortunate one since these plants are certainly
not “ferns” that evolved into seed plants.
The pteridosperms came into their own in the Upper Carbonif¬
erous and Permian as a dominant and highly diversified assemblage,
and I have no doubt that ultimately several orders and families will
be recognized. The medullosas seem to form a clearly defined line
and it is one that has commanded considerable attention partly be¬
cause of the intrinsic peculiarity of the vascular anatomy and partly
because some of the Permian species approach the kind of stem
anatomy found in the cycads. There is, however, a very wide gap
between the morphology of the seed and microsporangiate organs
of the medullosas and the comparable structures of any Mesozoic
or living cycadophytes.
It is perhaps evident that the pteridosperms, like many other
fossil groups, present more riddles than solutions. Since they offer
the most likely source that we have as a fountain head of angio-
sperm ancestry the most pressing problem centers around the seed
and its enclosing cupule. In Chapter 13 a few speculations are
offered concerning the nature of the cupule, the integument, and
the nucellus. We do have some intriguing evidence which at least
offers an evasive target for future investigations. Apparently not
all of the pteridosperms had their seeds invested in a cupule—but
is this really true? The positions that the seeds occupied are quite
diverse; but to date there are somewhat less than a score of speci¬
mens known with the seeds actually attached to their fronds, and
in most of these there is little cellular structure preserved.

Corystospermaceae

Several groups of fossils have been found at Triassic and Jurassic

horizons that have been classified as “Mesozoic pteridosperms.”

Telegram @tubscstds
There is sound evidence for so regarding them, but they are suffi-
160 STUDIES IN PALEOBOTANY

ciently distinct in many details as to allow no close correlation with

the Carboniferous pteridosperms.
The Corystospermaceae (Fig. 5-17) is a unique assemblage of
fossil plants, including seed-bearing and microsporangiate organs
and foliage from Triassic beds of Natal, South Africa. They are

Fig. 5-17. Fossils attributed to the Corystospermaceae, from the Triassic of South
Africa. A. Umkomasia macleani, inflorescence. B. Pilophorosperma granulation,
cupule showing micropyle of seed. C. Pilophorosperma sp., cupule with seed.
D. Pteruchus africanus, microsporangiate inflorescence. E. Umkomasia macleani,
portion of a main axis with base of lateral branch and subtending bract and bracteoles.

Telegram @tubscstds
F. Dicroidium sp. G. Stenopteris densifolia. (From Thomas, 1933.)
PTERIDOSPERMOPHYTA 161

regarded as referable to a single family by virtue of close association,

similarity in cuticle structure, and in one case pollen grains charac¬
teristic of a male organ have been found in one of the seeds.
Two genera and several species of seed-bearing inflorescences are
known. Umkomasia macleani Thomas is an inflorescence 3.4 cm
long which produces lateral branches, in one plane, in the axils of
bracts. Each of these branches in turn bears a pair of small scale¬
like structures (bracteoles) and several terminal cupules, each en¬
closing one seed. The cupule is partially divided to form a bivalved
structure which is distinctly helmet-shaped; this feature is especially
pronounced in Pilophorosperma granulatum Thomas in which the
curved micropyle with its bifid tip is shown projecting from the
helmet or cupule. The generic name Pilophorosperma implies that
the helmet is lined with hairs, this being the chief distinguishing
character which separates it from Umkomasia.
Pteruchus africanus Thomas may be considered as typical of the
male or pollen-bearing organs. It is a small branch system which
divides by equal or unequal dichotomies, the branches terminating
in a circular or elliptical lamina. Numerous sporangia were attached
over the entire under surface of the lamina and were probably pend¬
ant in life. The sporangia vary from 1 to 4 mm long and contain
winged pollen.
Leaves known as Dicroidium and Stenopteris are thought to rep¬
resent the foliage of these plants.
The beds from which the corystosperms were obtained are de¬
scribed by H. H. Thomas as having a rich fossil flora and should
ultimately shed considerable light on our knowledge of seed plants
and possibly those approaching the angiosperm stage. It seems to
me that the “Corystospermaceae” should be regarded at present as
an assemblage of pollen and seed-bearing organs that probably bore
fernlike foliage. It was a fairly large and varied group; Thomas has
described two species of Umkomasia, eight species of Pilophoro¬
sperma, a third seed genus Spermatocodon and eight species of
Pteruchus in addition to the foliage genera Dicroidium, Stenopteris,
and perhaps Pachypteris.
The helmet-shaped structure enclosing the seeds appears to be
homologous with the cupule of the Carboniferous pteridosperms and
the male organs are similar in form to earlier ones such as
Crossotheca.
Another distinctive group of fossils [Zuberia zuberi (Szajnocha)
Frenguelli] apparently of pteridospermous affinity, have been de¬

Telegram @tubscstds
scribed from the Triassic of Argentina (Fig. 5-18). The dichoto-
 Telegram @tubscstds
Fig. 5-18. Zuberia zuberi.
male fructification.
A. A portion of a sterile frond.
C. Cupulate organs.
B. Reconstruction of the
(From Frenguelli, 1944.)
PTERIDOSPERMOPHYTA 163

mously forking sterile frond bears pinnules of the Odontopteris

type. The microsporangiate organ is a dichotomously forking
branch bearing numerous appendages that consist of a pedicel be¬
yond which many sporangia are attached at right angles. Associ¬
ated with them are cupulate organs which invite some comparison
with Umkomasia. The preservation of these fossils is by no means
as perfect as one would wish for, but enough evidence is available
to indicate a distinct group of plants that may tentatively be
assigned to this family.

The Peltaspermaceae

This family (Fig. 5-19) is designated for seed-bearing and micro¬

sporangiate organs, as well as foliage, which have been found in
Rhaetic rocks of Greenland and the late Triassic of Natal. The
leaves are typically fernlike, attained a length of 30 cm, and are
widely distributed and abundant in the Greenland beds. Stem

Fig. 5-19. The Peltaspermaceae. A. Lepidopteris natalensis, a frond about !4 natural

size. B,C. Lepidopteris ottonis, from East Greenland; B. restoration of cupulate disc

Telegram @tubscstds
showing three seeds still intact; C. diagrammatic longitudinal section through disc and
one seed. (A from Thomas, 1933; B,C from Harris, 1932.)
164 STUDIES IN PALEOBOTANY

fragments have been found with them which measure 5.5 cm in di¬
ameter and 10 cm long. The microsporangiate organ is a pinnately
branching structure about half the size of the leaves and was prob¬
ably three dimensional in its branching pattern. The sporangia or
pollen sacs were borne in pairs at the tips of the ultimate branchlets;
each pollen sac is about 2 mm long and 1 mm wide and dehisced
longitudinally.
The seed organ is described as a cupulate disc; it was umbrella¬
shaped, consisting of a stalk terminated by a disc 1.5 cm in diameter.
About 20 cavities are found forming a circle around the under side
of the disc and are believed to represent scars where seeds were at¬
tached. In the Natal collections the cupulate discs were found
attached, in a spiral arrangement, to a central axis which attained
a length of 27 cm. They are only about one-third the size of the
Greenland specimens and apparently only one seed per disc reached
maturity.
In the Greenland specimens the seed is ovate, being about 7 mm
long and 4 mm broad with a curved micropylar beak. They were
sufficiently well preserved to allow observation of the inner and
outer cuticle of the integument, the nucellar cuticle, and the mega¬
spore membrane.
The stem fragment, leaf rachis, stalk of the disc, and microspo¬
rangiate organs bear characteristic swellings that are generally
similar and there is a close comparison in the stomatal structure of
all these organs as well as the seeds. This evidence, combined with
their association in localities as widely separated as Greenland and
Natal, leaves little doubt that they may be regarded as representing
a single genus or a group of closely related ones. Their taxonomic
position is less certain. The foliage and microsporangiate organs
are pteridospermous, but the seed-bearing appendages present a
striking departure.

REFERENCES

Andrews, Henry N., Jr. 1940. A new cupule from the Lower Carboniferous of Scot¬
land. Bull. Torrey Bot. Club., 67: 595-601.
-. 1945. Contributions to our knowledge of American Carboniferous floras.
VII. Some pteridosperm stems from Iowa. Ann. Missouri Bot. Gard., 32: 323-360.
-. 1947. Ancient Plants and the World They Lived In. Comstock Pub. Co.,
Ithaca. 279 pp.
Corsin, Paul. 1931 Fructifications de Sphenopteris (Diplotemema) alata (Brgt)
Kidston. Ann. Soc. Geol. Nord, Lille, 56: 25-33.

Telegram @tubscstds
PTERIDOSPERMOPHYTA 165

Delevoryas, Theodore. 1955. The Medullosae—structure and relationships. Palae-

ontographica, 97B: 113-167.
-, and Jeanne Morgan. 1954. A new pteridosperm from Upper Pennsylva¬
nian deposits of North America. Palaeontographica, 96B: 11-23.
Frenguelli, Joaquin. 1944. Las especies des genero Zuberia en la Argentina. Anales
Mus. de la Plata, Paleontologia: n.s. sec. B, Paleobotanica, No. 1: 1-30.
Gordon, William T. 1938. On Tetrastichia bupatides: a Carboniferous pteridosperm
from East Lothian. Trans. Roy. Soc. Edinburgh, 59: 351-370.
-. 1941. On Salpingostoma dasu: a new Carboniferous seed from East
Lothian. Trans. Roy. Soc. Edinburgh, 60: 427-464.
Hall, John. 1954. The genus Stephanospermum in American coal balls. Bot. Gaz.,
115: 346-360.
Halle, Thore G. 1929. Some seed-bearing pteridosperms from the Permian of China.
Kungl. Svenska Vetenskapsakad. Handl., 6: (No. 8) 1-24.
-. 1933. The structure of certain fossil spore-bearing organs believed to be¬
long to pteridosperms. Kungl. Svenska Vetenskapsakad. Handl., 12: (No. 6) 1-103.
-. 1942. Some specimens of Potoniea from the Carboniferous (Westphalien)
of Belgium. Mus. Roy. Histoire Nat. Belgique, Bruxelles, 18 (No. 42): 1-6.
Harris, Thomas M. 1932. The fossil flora of Scoresby Sound, East Greenland. Part
3. Caytoniales and Bennettitales. Meddelelser om Gronland, 85 (5): 1-133.
Hemingway, W. 1941. Neuropteris tenuifolia with carpons attached. Ann. Bot., 5:
193-196.
Jongmans, Willem J. 1930. On the fructification of Sphenopteris hoeninghausi and
its relations with Lyginodendron oldhamiam and Crossotheca schatzlarensis. Geol.
Bureau Nederland Mijngebied, 1930.
-———. 1954. Contributions to the knowledge of the flora of the seam Girondelle
(Lower Part of the Westphalien A). Mededel. Geol. Stichting, ser., CIII No. 4: 1-16.
Kidston, Robert. 1904. On the fructification of Neuropteris heterophylla, Brongniart.
Phil. Trans. Roy. Soc. London, 197B: 1-5.
-and Jongmans, W. J. 1911. Sur la fructification de Neuropteris obliqua
Bgt. Archives Neerland. Sci. Exactes Nat., ser. IIIB, 1: 25-26.
Kubart, Bruno. 1914. Uber die Cycadofilicineen Heterangium und Lyginodendron
aus dem Ostrauer Kohlenbecken. Osterreichische Bot. Zeit., 64: 8-17.
Long, Albert G. 1944. On the prothallus of Lagenostoma ovoides Will. Ann. Bot.,
8: 105-117.
Oliver, F. W., and Scott, D. H. 1904. On the structure of the Palaeozoic seed
Lagenostoma lomaxi, with a statement of the evidence upon which it is referred to
Lyginodendron. Phil. Trans. Roy. Soc. London, 197B: 193-247.
-, and E. J. Salisbury. 1911. On the structure and affinities of the Palaeo¬
zoic seeds of the Conostoma group. Ann. Bot., 25:1-50.
Renault, Bernard, and Rene Zeiller. 1888. Etudes sur le terrain houiller de Com-
mentry, Atlas. St. Etienne, 75 pp.
Schopf, James M. 1948. Pteridosperm male fructifications: American species of
Dolerotheca, with notes regarding certain allied forms. Journ. Paleont. 22: 681-724.
Stewart, Wilson N. 1954. The structure and affinities of Pachytesta illinoense comb,
nov. Amer. Journ. Bot., 41: 500-508.
-and Delevoryas, T. 1952. Bases for determining relationships among the
Medulloseae. Amer. Journ. Bot., 39: 505-516.
-. 1956. The Medullosan pteridosperms. Bot. Rev., 22: 45-80.

Telegram @tubscstds
166 STUDIES IN PALEOBOTANY

Thomas, H. Hamshaw. 1933. On some pteridospermous plants from the Mesozoic

rocks of South Africa. Phil. Trans. Roy. Soc. London, 222B: 193-265.
Walton, John. 1940. An Introduction to the Study of Fossil Plants. Adam and
Charles Black, London. 188 pp.
-. 1949. Calathospermum scoticum—an ovuliferous fructification of Lower
Carboniferous age from Dunbartonshire. Trans. Roy. Soc. Edinburgh, 61: 719-728.
Weber, 0. and Sterzel, J. T. 1896. Beitrage zur Kenntnis der Medulloseae. Natur-
wiss. Gesell. Chemitz, 13: 44-143.

Telegram @tubscstds
the ANTIrHOPIHYTA
(Angiosperms)
Some paleobotanical problems; earliest
Angiosperms; and a Gymnospermous
Group, the Caytoniales

Introduction

The dominance of the angiosperms or flowering plants in the liv¬

ing flora, their use in an almost endless number of ways in our
economy, and the great botanical interest in these plants have all
contributed to the intense desire of many naturalists to know more
about their origins and past distributions. As to origins we still
know very little, but some tantalizing discoveries of pre-Cretaceous
angiosperms have been made in recent years and significant studies
dealing with the problem have appeared; as to former distributions
there have been numerous excellent contributions and some of
these will be presented in detail.
It seems desirable first to enumerate the problems that are asso¬
ciated with angiosperm studies so that the present approach will
be clear and, perhaps of more importance, so that the student may
be aware of avenues of research that may be explored to enhance
our all-too-scanty knowledge.
The angiosperms are preserved in the form of leaf impression
(and occasionally compressions), seeds and fruits that may be petri¬
fied or in a lignitic state, petrified or lignitic wood, pollen grains,
and very rarely flowers are found as impressions or entombed in
amber; in some instances various combinations of these are found
together in the same deposit. As to the value of these remains,
the following may be considered as an introduction which will be

Telegram @tubscstds
elaborated upon later:
167
168 STUDIES IN PALEOBOTANY

1. There has been much criticism of the general validity of

studies based on leaf impressions, some botanists holding that the
available characters are not adequate for significant identifications.
It is true that many of the earlier studies include questionable
identifications and the more recent ones are admittedly not perfect.
However, some of the modern works, which are based on careful
comparisons with herbarium materials and extensive studies of
possibly related living floras, are dependable; quite obviously in all
“leaf floras” some species have more distinctive characters than
others.
2. Recent investigations of fossil seeds and fruits have supplied
what seems to be especially good evidence. Such studies require
the availability of extensive collections of modern seeds and fruits
for comparison and a knowledge of both the external form and
internal structure of these organs; there is much to be done in this
area.
3. Fossil angiosperm woods are abundant from Upper Creta¬
ceous times on and have received relatively little study. Like
seeds, fruits, and leaves, studies of wood must be based on a knowl¬
edge of living plants. There are still thousands of living trees and
shrubs whose wood anatomy is poorly known, if at all; there are
very few collections of woods that are at all comprehensive in the
botanical institutions of the world, and finally, the preparation of
this material for study, living and fossil, is quite time-consuming.
Thus it seems fair to say that studies of fossil angiosperm woods
are hardly beyond the pioneering stage.
4. Finally, fossil pollens are now receiving a tremendous amount
of attention, but these studies are little more than two decades old.
Ultimately they may be expected to contribute a great deal toward
our knowledge of past distribution patterns.
A few introductory comments may be offered concerning the time
element. In general, angiosperm remains from late Tertiary hori¬
zons can be identified with modern genera and species with a con¬
siderable degree of confidence; by “late Tertiary” is implied, rather
roughly, from the upper Miocene to the present. In floras found
within this time range we are dealing largely with plants whose
modern equivalents may be found in the immediate vicinity or at
most a few hundred miles distant. Going down through mid-to
early Tertiary horizons, we find that the comparison with adjacent
floras, or indeed with living species and genera, becomes progres¬
sively less distinct. As examples, the upper Oligocene Bridge

Telegram @tubscstds
Creek flora of Oregon bears close comparison with certain elements
EARLY ANGIOSPERMS 169

in the coastal Redwood forests; the mid-Tertiary flora of Brandon,

Vermont, displays affinities with the modern flora of the extreme
southeastern United States; the Eocene Goshen flora of Oregon
appears to have its strongest bonds with certain modern Central
American areas; the Eocene London Clay flora carries far afield,
with its modern equivalents centered in the Indo-Malayan region.
As might be expected, angiosperms of the Upper Cretaceous are
often difficult or impossible to correlate with modern genera or
even families. Very few studies of Cretaceous angiosperm floras
have been made in recent decades and most of the older ones must
be revised.
In attempting to sort out the vast literature dealing with angio¬
sperm origins and past distribution patterns, the topics outlined
below seem to be of particular interest and importance.

Chapter 6

Earliest fossil evidence of the angiosperms

What is a primitive angiosperm?
Some presumably primitive living angiosperms
The Caytoniales
Early Cretaceous flowering plants
Some Upper Cretaceous angiosperm floras

Chapter 7

Tertiary floras of England and western continental Europe

Late Cretaceous and Tertiary floras of western North America

Earliest Fossil Evidence of Angiosperms

Fossil leaf impressions that represent undoubted angiosperms

appear in mid Lower Cretaceous horizons or possibly a little earlier
in the stratigraphic column, and by early Upper Cretaceous times the
group appears in abundance and great diversity. In explanation of
this apparently sudden appearance of the angiosperms most bota¬
nists have supposed that they actually originated much earlier
than the record indicates, but for one reason or another the fore¬
runners of the group were not preserved. Scattered scraps of
evidence have accumulated which suggest rather strongly that the
angiosperms did exist in the Jurassic or even the Triassic; the more
important of these pre-Cretaceous records will now be considered.

Telegram @tubscstds
170 STUDIES IN PALEOBOTANY

As long ago as 1904 a leaf impression was found in the mid-

Jurassic of Yorkshire and given the rather noncommital name
Phyllites sp. It is a leaf 3.7 cm long and 2.5 cm broad, with three
main veins. It has been compared with Cercidophyllum, and
Seward, in his original account, apparently considered it sufficiently
convincing to write the following:
Had the specimen been found in rocks known to contain the remains of
Angiosperms, there would be no hesitation in identifying it as the leaf of a
Dicotyledon. (1904, p. 153)

In 1955 Kuhn described what appears to be a typical dicot leaf

from the lower Jurassic of Sassendorf, near Bamberg under the
name Sassendorfites benkerti. Of still older age, Furcula granu-
lifer Harris is from the upper Triassic of Greenland; the leaves are
Y-shaped or sometimes unbranched, up to 15 cm long and with a
distinct network of veins similar to that in the dicots.
It must of course be kept in mind that leaves or leaf fragments
with parallel or netted venation do not positively imply monocot
and dicot affinities. Several ferns as well as the gymnosperm
Gnetum have net-veined leaves; others, such as Caytonia which is
gymnospermous, and Glossopteris (the affinities of which are un¬
certain but it is probably a gymnosperm) have leaves that fall within
the range of the dicots as far as venation is concerned.
There are several reports of supposed monocot foliage from pre-
Cretaceous horizons recorded in the older paleobotanical works.
Most of these are based on small fragments of foliage showing
parallel venation and are either unidentifiable or quite clearly mis-
identified. Two accounts of presumed palm leaves, however, seem
to be authentic. Early in the century Lignier described impression
specimens from the lower Jurassic (Lias) of France which appar¬
ently represent the basal portion of a leaf blade. Although this
is a critical and distinctive portion of the leaf and several compe¬
tent paleobotanists have seen fit to accept his identification of the
fossils as a palm (Propalmophyllum liasinum), little serious atten¬
tion has been given them. It was, therefore, an event of the
utmost interest when palmlike leaves (Fig. 6-1), much more com¬
pletely preserved, were reported by Brown from the mid-to upper
Triassic of Colorado. The simple, elliptic, pleated leaves of this
plant (Sanmiguelia lewisi) attained a length of 40 cm and a width
of 25 cm and were borne in an alternate arrangement at the tip of
a rapidly tapering stem. They were collected from a reddish cal-
carious sandstone in the Dolores formation.

Telegram @tubscstds
EARLY ANGIOSPERMS 171

Pleistocene

Pliocene

Miocene
Brandon, Vt. flora (?)

Oligocene

Eocene Goshen flora

Wilcox flora at Puryear, Tenn.
London clay flora

Upper Lance fm., Medicine Bow fm.

Cretaceous Fox Hills fm.

Raritan fm.

Albian stage

Patapsco fm.
Lower Aptian stage
Cretaceous
Patuxent fm.
Greenland flora (?)

Jurassic Caytoniales

Sassendorfites, Propalmophyllum

Furcula
Triassic
Sanmiguelia

Mesozoic-Tertiary time chart showing approximate position of certain

plants, floras, and geological horizons cited in Chapters 6 and 7.

Telegram @tubscstds
172 STUDIES IN PALEOBOTANY

Fig. 6-1. Sanmiguelia lewisi, one leaf of a presumed Triassic palm from southwestern
Colorado. (From Brown, 1956.)

Palm leaves are not uncommon in the Upper Cretaceous and had
Sanmiguelia been found at such a horizon it would have been
accepted without question as a palm. There is certainly no other
known group of plants with which it can be compared at all closely
and I see no reason to question the author’s comment that it
should be “regarded tentatively but credibly as a primitive palm.”
Many botanists have been inclined to regard the monocots as

Telegram @tubscstds
derived from the dicots and thus of more recent origin. If this is
EARLY ANGIOSPERMS 173

true the presence of a palm in the Triassic would push the origin
of the angiosperms back in time much farther than has previously
been supposed, consequently additional evidence bearing on this
exciting discovery is anxiously awaited.
Pollen grains presumed to have been derived from plants belong¬
ing to the water lily family have been reported from a Jurassic coal
in Sutherlandshire, Scotland, and from the early Jurassic at Palsjo
in Sweden. Recently, a restudy of the Scottish pollens by Hughes
and Couper indicates that these are gymnospermous rather than
angiospermous.

What is a Primitive Angiosperm?

In attempting to elucidate the evolution of a group of plants,

whether one is working with living or fossil forms, it is necessary
to have some notion as to what we are looking for; it is, however,
equally important that we should not be rigidly bound by a mental
concept of what the primitive ancestral form should be and ignore
everything that does not fit the preconceived concept.
Some bitter words have been exchanged in the past over the
question of what constitutes a primitive angiosperm. The student
of botany, whether beginning or advanced, who assumes that this
problem is solved is certainly mistaken. Although it is not possible
to consider the problem in all its aspects, I shall briefly discuss cer¬
tain recent morphological studies of living plants in order to lend
as much meaning as possible to the fossil evidence. In order to
emphasize the magnitude of the task of understanding early angio¬
sperm evolution, the following comments are offered from a recent
study by Dr. H. H. Thomas, a leading student of angiosperm
evolution:
In trying to distinguish a relatively primitive plant we look for one in
which each character is primitive. In seeking a fossil ancestor we hope to
find one which has all the characters of a flowering plant. But one of the
earliest discoveries in genetics was that characters can be inherited sepa¬
rately, or rather in groups which are independent. We must therefore not
place too much stress on the occurrence of a single character, such as the
possession of free petals. It is likely that in the course of the evolution of a
genus the different characters, floral and vegetative, will tend to keep in step
to some extent, but we cannot rely on this unless supported by approved
statistical correlation. ... If plants and plant structures are to be studied
from the evolutionary point of view it is not enough to arrange them in a
sequence with reference to their forms. The causal aspect must be studied,
and consideration given to what is known of the physiology of reproduction.

Telegram @tubscstds
(1957, p.132-133)
174 STUDIES IN PALEOBOTANY

It has been known for some time that several plants in the rana-
lian complex (the common buttercup and magnolia are representa¬
tive), long considered as occupying a primitive position, lack vessels
in their wood. Vessel elements are dead cells at maturity and serve
to conduct water and dissolved materials; they are tube-shaped
and differ from tracheids in having one or more holes at each end.
They are a characteristic feature of the wood of both monocots
and dicots and are lacking in most pteridophytes and gymnosperms.
There are exceptions; they have been found in a few ferns, in the
lycopod genus Selaginella, in Equisetum and in the Gnetales, but
it is likely that they evolved independently in those groups.
The known number of species of vesselless dicots is now about
100 (in ten genera) and includes Trochodendron (Trochodendraceae),
Tetracentron (Tetracentraceae), Drimys, Bubbia, Belliolum, Exo-
spemum, and others in the Winteraceae. Even a few species of
vesselless plants in the ranalian group would support its supposed
primitive positions, but the occurence of a hundred presents more
than a suspicion. This information correlates, though not neces¬
sarily species by species, with a seemingly primitive carpel mor¬
phology. Although studies of the nature and evolution of the
carpel have long held a foremost place in botanical research, they
were greatly stimulated by A. C. Smith’s discovery of a new family
with especially primitive floral characters (Degeneriaceae) on the
Fijian island of Vanua Levu. Although vessels are present in
Degeneria vitiensis the anthers and carpels are remarkable. The
former, lacking any differentiation into anther and filament, have
been described as “Broad microsporophylls.” The single carpel of
the flower (Fig. 6-2D-F) is referred to as conduplicate, implying
that it is a leaflike structure folded lengthwise; it bears three main
veins and two rows of ovules which are attached some distance with¬
in the margin. The adjacent inner surfaces are covered with hairs
and it is not until fruit development sets in that these surfaces
actually become concrescent. The outwardly flaring “lips” are
stigmatic throughout the length of the carpel and the pollen tubes
develop as shown in Fig. 6-2F. Comparable conduplicate carpels
are known in certain genera of the Winteraceae and in the
Himantandraceae.
Returning for a moment to the matter of vesselless dicot woods,
we may add a brief comment on the fossil game Homoxylon. * This

* Probably most of the species of Homoxylon should be transferred to Sahnioxylon

of Bose and Sah since Homoxylon was originally established by Hartig in 1848 for a

Telegram @tubscstds
conifer. Bose and Sah (1954) discuss the problems of correctly classifying woods of
this type.
EARLY ANGIOSPERMS 175

Fig. 6-2. A. Lagenostoma lomaxi, the lobed glandular cupule partially encloses a seed.
B. Hypothetical stage in the evolution of a carpel, showing two partially fused cupu-
late organs. C. Pilophorosperma; two cupules, each containing a single seed. D. A
primitive conduplicate carpel in side view. E. Same, unfolded showing vascular
system. F. Same, showing placentation, distribution of glandular hairs (stippling),

Telegram @tubscstds
pollen grains and developing tubes. (A from Scott, 1923; B,C from Thomas, 1934;
D,E,F from Bailey and Swamy, 1951.)
176 STUDIES IN PALEOBOTANY

includes several species from Mesozoic horizons, some of which are

considered to be angiosperms. Boureau has recently described
two species from New Caledonia: H. aviasii from the basal Jurassic
and H. neocaledonicum from the upper Triassic. They are woods
with annual rings, they lack vessels, and the tracheids are scaiari-
form with a tendency toward circular-bordered pitting in the latter
part of the annual ring; the two are considered together here since
they appear to be quite similar. Woods of this type, referred to
Homoxylon, have been compared with certain living vesselless
angiosperms, and there is the added interest in these two species
in that they come from a region that is the center of primitive ves-
selless-living angiosperms. This type of wood is, however, quite
close to that of some bennettitalean woods and so far as I am
aware there is no positive assurance that any of the species assigned
to Homoxylon are indubitably flowering plants.
At this point in our search for angiosperm ancestors I have
chosen to consider a controversial and highly interesting group of
plants, the Caytoniales, which were first described by H. H.
Thomas from the rich mid-Jurassic plant bed at Cayton Bay in
Yorkshire. It should be noted very clearly that the Caytoniales
are not angiosperms but in the nature of their seed-bearing organs
they come very close to being so. Nor is it implied that they are
close ancestors of the modern ranalian plants considered above, yet
the two groups do seem to offer some clues as to how angiospermy
may have arisen. In so large and diversified a group (there are
some 250,000 living species) it is possible that they originated along
several distinct lines, but a discussion of this point is beyond the
scope of the present account.
In most works on fossil plants the Caytoniales are assigned to
the Pteridospermopsida or it is suggested that they are an offshoot
of that group. The leaves, seed-bearing and male organs are as¬
signed to the genera Sagenopteris, Caytonia, and Caytonanthus
respectively. Several species of each are known and although not
found in organic connection there is good reason to suppose that
the association is significant. The leaves have been known for well
over a century and are widely scattered geographically and strati-
graphically, ranging from Upper Triassic to Lower Cretaceous.
Caytonia, associated with Sagenopteris, has been found in Green¬
land, Sardinia, western Canada and probably Scania (Sweden), as
well as in Yorkshire.
Caytonia sewardi Thomas consists of a central axis about 4 cm
long with two rows of structures that will be referred to as fruits.

Telegram @tubscstds
 EARLY ANGIOSPERMS 177

The fruit is about 4.5 mm in diameter, is recurved with a liplike

structure adjacent to the pedicel, and it enclosed a U-shaped row
of seeds. Although the fossils are preserved as compressions the
cellular organization of the seeds is known in some detail. The
fruits were originally thought to have been angiospermous, but a

Fig. 6-3. The Caytoniales. A. Caytonia nathorsti, central axis bearing two rows of
cupules. B. A cupule or “fruit” in longitudinal section showing four seeds. C,D.
Sagenopteris phillipsi; C. a leaf with four leaflets; D. a portion of a leaflet showing
net venation. E,F. Caytonanthus kochi, from East Greenland; E. restoration of a por¬
tion of a microsporophyll; F. two anthers, one with four chambers and the other with
three. (A-D from Thomas, 1925; E,F from Harris 1937.)

Telegram @tubscstds
178 STUDIES IN PALEOBOTANY

later study by Harris revealed pollen grains within the micropyles;

however, the opening must have been very small and it seems pos¬
sible that the pollen reached the micropyles by means of a pollen
drop mechanism. The axes are commonly found showing elliptic
scars where the fruits were attached, the latter apparently being
dropped by a dehiscence mechanism. Both seeds and fruits may
be recovered in abundance by macerating the shale.
Caytonanthus arberi (Thomas) Harris, the microsporangiate
organ, is known from central axes with short lateral branches
which are subdivided and bear four-chambered “anthers.” The
pollen grains, unlike those of the angiosperms, are winged. Based
on studies of well-preserved specimens from Scoresby Sound,
Greenland (Fig. 6-3E), Harris notes that the anther is radially
symmetrical, thus differing from the bilateral symmetry in extant
flowering plants. The way in which the anthers are arranged on
their branch system is also quite different from that of any modem
flowering plant.
The leaves, Sagenopteris phillipsi (Brongiart), were known long
before the reproductive parts were discovered. Although varying
considerably in form they may be described as palmately compound
with lanceolate to oval leaflets which are net-veined (Fig. 6-3C).
The known relationship of these plant organs goes beyond mere
association which, however, is significant in view of the geograph¬
ical distribution. In his original account of the Cayton Bay plants
Thomas demonstrated a similarity between the epidermal cell
structure of Sagenopteris and that of the fruit axes, and Harris has
done the same for Caytonia thomasi and Sagenopteris nilssoniana
from Greenland. But perhaps the most convincing evidence is the
discovery of numerous seeds in the Yorkshire locality with pollen
of the Caytonanthus type in the micropyles.
It has been pointed out many times that the Caytoniales do not
compare closely with any living primitive flowering plants; how¬
ever, they were very close to the angiospermous state and suggest
one way in which the carpel may have evolved. The flowering
plants are a varied assemblage and it is reasonable to suppose that
they are polyphyletic. It may require several keys to unlock the
mystery of their origins, and the problem centers around the carpel
with its enclosed seed or seeds. It may therefore be well to add a
few theoretical notes which are intended to serve as an introduc¬
tion to the literature on the subject for students who may wish to
pursue the matter in more detail.

Telegram @tubscstds
EARLY ANGIOSPERMS 179

Two concepts of carpel origin have attracted interest in recent

years. Thomas has interpreted the apparent inverted Caytonia as
having evolved from earlier pteridospermous plants. Beginning
with the Upper Carboniferous Lagenostoma, in which the single
seed was contained in a lobed cupule (Fig. 6-2A), one may regard
the Triassic Pilophorosperma with its cupule lined with hairs as a
later type, and in Caytonia the seeds are very nearly enclosed.
According to this view the carpel found in primitive modern angio-
sperms, such as Degeneria and Drimys, is the result of a lateral
fusion of two carpels of the Pilophorosperma (Fig. 6-2B, C) or
Caytonia types.
In support of the concept of the carpel of the Ranales as a con-
duplicate or longitudinally folded structure Bailey and Swamy cite
the following evidence: the three main vascular strands, the exten¬
sive stigmatic surface, the attachment of the ovules in a laminal
position midway between the margin and center line, and their
open character (being closed at the time of pollination only by the
loose contact of the stigmatic hairs).
The fused cupule concept is admittedly speculative since the
later stages are not known, whereas the concept of the carpel as an
enfolded leaflike structure, although long favored by many botan¬
ists, finds no supporting evidence in the form of gymnospermous
plants, living or fossil, which may be regarded as ancestral.
Continued studies of living plants will certainly shed light on
this great problem, and I am not inclined to be pessimistic about
possible contributions from the fossil record; many areas of the
earth are still virtually unexplored paleobotanically. In conclud¬
ing their discussion Bailey and Swamy note that:

The most promising possibility of solving the riddle of the origin of the
angiosperms appears to be an intensified search for paleobotanical evidence
in parts of the earth’s surface that have been inadequately explored, par¬
ticularly Austro-Malayan and Indo-Malayan regions. (1951, p. 378)

Early Cretaceous Angiosperms

One of the most disturbing features of the paleobotanical record

is the appearance, in approximately mid-Lower Cretaceous times,
of angiosperm leaf impressions that are distinctly modern in
aspect. Some of the more important floras of this age will be cited
and a possible explanation offered for their sudden appearance.
Nothing is known of the flowers of these plants, but the leaves do

Telegram @tubscstds
180 STUDIES IN PALEOBOTANY

not suggest anything especially primitive. In general the floras are

ones in which ferns and gymnosperms are the dominant elements.
In 1894 Saporta described several leaf fragments from rocks of
Aptian age in Portugal some with parallel venation, others netted,
which were interpreted as angiosperms. From a somewhat higher
horizon in Portugal, considered to be Albian, a significant number
and variety are recorded, including the following genera: Salix
(willow), Aralia, Braseniopsis, Myrica (sweet gale), Laurus (laurel),
Viburnum, Eucalyptus, Magnolia and Sassafras.* The veracity
of these generic identifications may be questioned, but there is no
reason to doubt their angiospermous nature. Certain revisions of
these floras have been given in a more recent study by Teixeira.
Seward has described an interesting assemblage of fossil plants
from Cretaceous horizons in western Greenland which include the
fern families Gleicheniaceae and Matoniaceae, as well as cycado-
phytes, conifers, and Ginkgo foliage; associated with these are
some questionable monocot leaf fragments and several undoubted
dicots, including Artocarpus (breadfruit), Quercus (oak), Meni-
spermites (moonseed), Platanus (sycamore), and others. The exact
age of this flora is uncertain, but it is possibly derived from an
early Lower Cretaceous horizon. Seward has commented on it as
follows:
My view is that the Greenland Cretaceous flora represents more fully than
the floras of other countries the early stages in the transitional period from
an older Jurassic-Wealden vegetation to the type of flora which continued
into the Tertiary period, and still persists in regions remote from its original
home. . . . The point I wish to make is that in the Cretaceous floras of Green¬
land Dicotyledons, which are surprisingly modern in the form of the leaves,
occur in association with Ferns and Gymnosperms, which in other parts of
the world are characteristic of floras distinguished by the absence of any
recognizable examples of modern angiosperms. This fact lends support to
the view that it was within the Arctic circle that the evolution of deciduous
angiosperms progressed with greater rapidity and energy than in more south¬
ern latitudes. (1926. p. 155)

Several angiosperm leaves have been described from the Patuxent

beds of Virginia and have been described under the names Roger-
sia, Proteaephyllum, and Ficophyllum. These also appear as a
very minor element in a predominantly pteridophyte-gymnosperm
flora and at best one can say that they are leaves of dicots. In
contrast, beds of Patapsco age are quite rich in plant fossils that
are unquestionable angiosperms although the generic assignments
*See comment under “Notes” on page 187 concerning use of scientific and common

Telegram @tubscstds
names.
EARLY ANGIOSPERMS 181

may in many cases be questioned. A listing of the species in these

members of the Potomac group and of later Cretaceous horizons
of the Atlantic coastal plain is given in a recent contribution by
Dorf (1952).
In a comprehensive study of plants from many localities of Lower
Cretaceous age in western Canada, Bell described the following
genera from the Blairmore group which is considered to be of Albian
age: Populites (poplar), Ficus (fig), Trochodendroides, Cinnamomo-
ides (cinnamon), Celastrophyllum, Sapindopsis, Fontainea and
Araliaephyllum, and a species of Sapindopsis from a somewhat
lower horizon which may be upper Aptian.
Our knowledge of Lower Cretaceous angiosperms is not based
wholly on leaves. Some years ago Marie Stopes described several
dicot woods from Aptian horizons in England. Five genera were
recognized (Aptiana, Woburnia, Hythia, Sabulia, Cantia), none
of which can be said to display especially primitive characters. All
have vessels, the first three have multiseriate rays, Cantia has
uniseriate rays and vessels in which the pitting varies from round
to oval scalariform, and in Hythia the vessel pitting ranges from
round-bordered to scalariform.
Quite recently Samylina has reported a Lower Cretaceous flora
(Aptian-Albian) from the Zyrianka river, a tributary of the
Kolyma river, in eastern Siberia. Although the flora is predom¬
inantly one of ferns and gymnosperms, about 25% of the species
are angiosperms. A total of 22 have been defined including the
following: Ranunculaecarpus (Ranunculaceae?); Sassafras (Laur-
aceae); Cercidophyllum (Cercidophyllaceae); Crataegites, two
species (Rosaceae); Dalbergites (Leguminosae); Celastrophyllum
(Celastraceae?); Zizyphoides (Rhamnaceae); Araliaecarpum
(Araliaceae); and several species are assigned to the indefinite
Dicotylophyllum. Most of these are plants with small leaves, usually
not exceeding 4 cm in length.

Summary Comments on Early Flowering Plants

An attempt has been made in the preceding pages to gather to¬

gether the essence of what is known about early angiosperms and
their origins, and it is quite clearly a very unsatisfactory story. We
now have several tantalizing pieces of evidence which suggest that
angiosperms existed far below the base of the Lower Cretaceous, of
which Sanmiguelia is perhaps the most difficult to refute; then fol¬
lows the much-discussed “sudden” appearance of angiosperms in

Telegram @tubscstds
182 STUDIES IN PALEOBOTANY

modern aspect in the Lower Cretaceous. Several paleobotanists

have suspected that the sudden appearance is an illusion and it has
been suggested recently that the origins of the group should be
sought in early Mesozoic or even Permian horizons. If Sanmiguelia
lewisi is the authentic palm that it appears to be and the strati¬
graphic origin is correctly given as Triassic, then the Permian seems
likely. But if this is the case why is so little of the earlier record
preserved? In a recent discussion of the problem Axelrod notes that
the early representatives of a group might be expected to be rare
as fossils, and he suggests that the angiosperms originated in the
diverse upland environments of the tropical zone during Permo-
TYiassic times. There is reason to believe that tropical latitudes of
that age were appreciably broader than at present and in all prob¬
ability included a very wide range of habitats. Unfortunately we
find fewer and fewer upland sediments due to continued erosion as
we go back in time.
In another useful review of the problem Takhtajan suggests that
the rapid expansion of the angiosperms is due to the great plastic¬
ity of the group: “In no other group of plants do we observe such
colossal differences as we see, for example, between magnolia and
the cereal grasses, between the orchid and Haloxylon.” (The last
is a small xerophytic tree found in the arid steppes of Central Asia
and the Middle East.) He also places considerable importance on
vessels and insect pollination as factors which influenced successful
evolution.
Directly interwoven with the problem of angiosperm origin as a
whole is the question of a single or multiple derivation. That is,
do the 250,000 extant species really constitute a single evolutionary
line or several? Many botanists favor the former, but a possible
multiple origin should not be discarded. Cheadle has shown that
vessels originated independently in the monocots although he sug¬
gests that the group evolved from ancient (probably long extinct)
members of the Nymphaeales which in turn were derived from the
Magnoliales.
In such a stormy sea of controversy where the sailing is so haz¬
ardous it is cheering to encounter a fresh approach to the old prob¬
lem and as an interesting example I conclude this section with a
few notations on E. J. H. Corner’s Durian theory. The durian
(Durio zibethinus of the Bombacaceae) is a large forest tree native
to the Malayan area and is characterized by slender twigs, simple
leaves, and fruits that are huge, spiny, loculicidal capsules with

Telegram @tubscstds
one to five large seeds in each of its five cavities; the seeds are
EARLY ANGIOSPERMS 183

covered with a thick, white or yellow aril which may be regarded

as a third integument. The fruits are a highly prized food of many
forest animals from elephants to beetles.
The aril is often brilliant red and is found in some 45 families of
flowering plants although the degree of its development varies from
species in which the seed is completely enclosed to others in which
it exists as an indistinct rudiment. It is present in many families
of flowering plants that are not closely related, whereas in groups
of related genera it is possible to trace apparent evolutionary
sequences from the large, colorful, food-rich, arillate capsules to
plants in which the fruits are dry capsules, often with winged
seeds, or drupes, berries or nuts—and in general smaller. Such
sequences are approximately correlated with increased power of
dormancy (in many arillate species the seed must germinate im¬
mediately if they are to survive), and greater dissemination poten¬
tialities. The presence of the aril in numerous unrelated families
in which these reduction series may be observed leads to the
supposition that the arillate fruit is a primitive type.
The following structural correlations define a primitive angio-
sperm in terms of the durian theory; the large fruits must have
been borne on stout branches; thus the ancestral form is postulated
as a massive tree, possibly of cycadlike habit. The lack of dor¬
mancy power and inability to withstand desiccation suggests that
the plants could not have survived outside of tropical rain forests.
In summary it is argued by Corner that “the primitive angiosperm
must have been a mesophytic, tropical, Cycad-like monocaulous
tree with large pinnate leaves and peltate scales, probably mono-
carpic, and producing a terminal cluster of large arillate follicles.”
(1949, p. 414)
There are several attractive features to the theory: it fits in with
other evidence suggesting a tropical origin of the angiosperms; it
allows an early origin of plant types found in both monocot and
dicot groups; it introduces the fruit as a neglected organ of plant
evolutionary studies and one that can be correlated with paleo-
botanical studies more readily than the flower can. The general
organization of the presumed primitive angiosperm (that is, in large
compound leaves and sparsely branched, soft-wooded types) com¬
pares in some way with early seed plants such as the cycads and
seed-ferns.
The theory is not offered as a cure-all or final answer to the prob¬
lem of angiosperm origins but rather as a field of investigation
which may lead to a better understanding of the problem as

Telegram @tubscstds
184 STUDIES IN PALEOBOTANY

a whole. Any serious student of plant morphology would do well

to read Corner’s article.

Some Upper Cretaceous Angiosperm Floras

The difficulties of correctly identifying Cretaceous angiosperms

have been noted; however, it seems worthwhile to introduce a few
of the better known floras. Although all the names may not be
valid, the abundance of genera and the families present in the
Upper Cretaceous at least indicates the diversity of the flowering
plants at that time.
In his account of the flora of the Amboy clays (Raritan forma¬
tion) of New Jersey, Newberry listed 156 species, of which about
four-fifths are dicotyledons; a few ferns are represented and several
conifers. Later studies have revealed a richer assemblage of coni¬
fers but of particular interest at the moment is the diverse aggrega¬
tion of dicots with 28 families recognized:
Juglandaceae Magnoliaceae Aceraceae Cornaceae
Myricaceae Menispermaceae Rhamnaceae Ericaceae
Fagaceae Lauraceae Vitaceae Myrsinaceae
Ulmaceae Rosaceae Tiliaceae Sapotaceae
Moraceae Leguminosae Passifloraceae Ebenaceae
Proteaceae Aquifoliaceae Myrtaceae Caprifoliaceae
Salicaceae Celastraceae Araliaceae Asclepiadaceae
One of the largest fossil floras from any horizon is the lower
Upper Cretaceous flora of the Dakota sandstone; the greater part
of the collections were made by Charles H. Sternberg in the latter
part of the past century in Ellsworth County, Kansas, although
others have been made in Nebraska and Minnesota. Some 460
species were recognized by Lesquereux in his work which appeared
in 1891. No comprehensive revision has been undertaken since
that time and many of the indentifications may be questioned;
there are, however, two points of lasting interest:
1. Aside from a few ferns, cycads, and conifers it is almost exclu¬
sively an angiospermous flora. Most of these are dicots with only
a small sprinkling of questionable monocots.
2. It is a very diverse flora; the figure of 460 species may be in
error, but the number is certainly high.
In a more recent study of a small flora of comparable age in

Telegram @tubscstds
southwestern Colorado, Brown has recognized Ficus (fig), Nelum-
EARLY ANGIOSPERMS 185

biura, Mahonia (Oregon grape), Sassafras, Platanus (sycamore),

Celastrophyllum (staff tree), Juglans (walnut), and Sterculia.
The Lower Medicine Bow flora from northwestern Colorado
totals 64 recognizable forms of which 58 are assigned to species.
They are mostly dicots, with a few monocots, the dominant ele¬
ments being:

Ficus planicostata Dryophyllum subfalcatum

Magnoliophyllum cordatum Rhamnus salicifolius
Trochodendron nebrascensis Quercus viburnifolia
Cinnamomum affine Viburnum marginatum
Myrica torreyi Rhamnus cleburni

In addition to these 10 dominants the flora includes a second

species of Cinnamomum (cinnamon), several other species of Ficus
(fig), two of Magnolia (magnolia), and two of Sabalites (sabal
palm).
The region in Colorado where these fossils were obtained is
presently a high, barren intermontane basin of the Rocky Moun¬
tain system with a mean temperature of 40 to 45° F; there is sage¬
brush-grassland on the ridges and a poplar-willow association along
the flood plains of permanent streams. A few other plants are
found but it is in general a very sparse vegetation. The fossil flora
was much more diverse than the living one and the palm leaves
(Sabalites montana and S. eocenica) alone leave no doubt as to the
great difference in the general climatic relations of this region dur¬
ing the Upper Cretaceous. Dorf concludes that it was intermediate
between warm temperate and subtropical and probably closer to
the latter.
The Lance flora is another warm climate assemblage based on
collections from several uppermost Cretaceous localities in east
central Wyoming. Seventy recognizable forms have been found,
with the following as the dominant elements:

Aracarites longifolia (araucaria) Dryophyllum subfalcatum

Platanophyllum montanum (sycamore) Fraxinus leii (ash)
Salix lancensis Viburnum marginatum
Sequoia dakotensis Cornophyllum wardii
Vitis stantoni (grape) Pistia corrugata

The presence of palm leaves, several members of the Lauraceae,

and a few other tropical or subtropical plants suggest a Cretaceous
climate that was more uniformly warm and moist than the present

Telegram @tubscstds
one of that area. It may also be noted that the flora probably
186 STUDIES IN PALEOBOTANY

represents several different plant communities. Judging from the

requirements of modern equivalents it is unlikely that conifers
(Araucarites longifolia and Sequoia dakotensis), a ginkgo (G.
adiantoides), palms (Sabalites), and several members of the laurel
family would all have been growing in close association.
Even farther to the north Bell has described a rather large
Upper Cretaceous flora from Vancouver Island with warm climate
elements such as: a species of the fern genus Anemia, large palm
leaves (Geonomites), Artocarpus and Ficus (Moraceae), Cinna-
momum and Laurus (Lauraceae), Ginkgo, and several cycad leaves.

REFERENCES

The references for this Chapter appear at the end of Chapter 7.

Telegram @tubscstds
(Angiosperms)
Some Tertiary Floras of Europe
and of Western United States

/Miy attempt to present a comprehensive summary of

lr~\ Tertiary floras is quite beyond the scope of this book.
A great many have been described from various parts of the world
during the last 150 years; the size of the floras varies from a few
species to several hundred and the preservation ranges from excel¬
lent to very fragmentary. There are, however, some studies, or
series of studies, that have been conducted in recent decades that
are based on the best techniques available and which tell intrigu¬
ing stories of changing climates and floral migrations.

Notes

1. In view of the large numbers of plant names that must of neces¬

sity be introduced in this chapter, the usual practice of citing the
authority after the binomial has not been followed. In most cases
the sources of information are evident and they are listed at the
end of the chapter.
2. Since numerous Tertiary floras are considered here, it is neces¬
sary to introduce a rather large number of plant names and I am
aware that many of them may be unfamiliar to the reader. The
scientific names used in the original publications must serve as a
basis, but whenever possible corresponding common names are also
given; however, since many of the fossil species are extinct, it
should be remembered that the common names may not be exactly
equivalent to their usage with living plants. Moreover, with some
species, genera, and families it is not possible to cite common names

Telegram @tubscstds
that are meaningful.
187
188 STUDIES IN PALEOBOTANY

3. The problems of correctly identifying the foliage, seeds, and

fruits of Cretaceous and Tertiary angiosperms are unique and con¬
siderable attention has been devoted to study techniques by the
more competent modern workers. For discussions of this phase of
the work the following references are especially recommended:
Chaney and Axelrod, 1959; Dorf, 1938; Reid and Chandler, 1933;
Szafer, 1946.

TERTIARY FLORAS OF ENGLAND AND WESTERN

CONTINENTAL EUROPE

A series of fossil seed-fruit floras, ranging through the Tertiary,

is known from western Europe which reveals a striking change in
composition and climate. As a matter of reference the floras to be
considered are listed below:

Flora Age Location

London Clay Lower Eocene Isle of Sheppey, southeastern
England
Hordle Upper Eocene Hordle, Hampshire, southern
England
Bembridge Oligocene Isle of Wight, southern England
Pont-de-Gail Pliocene, basal Pont-de-Gail, Cantal, France
Reuverian Lower Pliocene Limburg Province, Netherlands
Kroscienko Pliocene Kroscienko, southern Poland
Teglian Upper Pliocene, Limburg Province, Netherlands
base of
Cromerian Pliocene, top of Cromer, Norfolk, England
(or Pleistocene?)

The London Clay Flora

Early in the last century naturalists began to gather pyritized

seeds and fruits from the beaches of the Island of Sheppey, near
the mouth of the Thames River, and other localities in southeast¬
ern England. In 1840, after several years of study, J. S. Bower-
bank produced a book dealing with the fossils and, although an
excellent work from the standpoint of accuracy of description, suf¬
ficient modern comparative material was not available to render
significant identifications possible. Bowerbank’s interest in the

Telegram @tubscstds
flora was sufficient to have led him to establish “The London Clay
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 189

Club” to facilitate study of the fossils. Although one may assume

that it never ranked among London’s most popular Clubs, an inter¬
est in the seeds and fruits has continued to the present day and in
1933 Reid and Chandler’s The London Clay Flora appeared, pre¬
senting in the finest tradition of British paleobotany a comprehen¬
sive study of this great and unique assemblage of fossil plants.
The London Clay flora, of early Eocene age, includes 314 species
of seeds and fruits; of this number 234 have been identified,
whereas the affinities of the remainder are considered doubtful. It
is almost exclusively an angiosperm flora, there being but 7 coni¬
fers. Of the 100 genera, only 28 are still extant; thus it is family
relationships that will primarily occupy our attention. The pres¬
ent-day distribution of the families which make up the London
Clay flora are: 5 are entirely tropical (Nipaceae, Burseraceae,
Icacinaceae, Bixaceae and Sapotaceae); 14 are almost exclusively
tropical (Palmae, Olacaeae, Menispermaceae, Anonaceae, Laur-
aceae, Meliaceae, Anacardiaceae, Sapindaceae, Sapiaceae, Elaeo-
carpaceae, Sterculiaceae, Dilleniaceae, Myrsinaceae, and Apocyna-
ceae); 21 families are equally tropical and extratropical and 5 are
chiefly temperate. The Lauraceae and Icacinaceae, almost exclu¬
sively tropical families, are represented by 40 and 21 species
respectively. Of modern floras the closest comparison lies with the
forests of Indo-Malaya and particularly the Malay Islands.
A few genera may be mentioned because of their abundance or
particular interest. The most frequently encountered fossil is the
fruit of Nipa burtini which is mentioned as being “strewn in great
abundance wherever the dark patches of pyrites nodules occur” at
Sheppey. The modern Nipa is a brackish water plant found at sea
level in the tropics of the Indo-Malay region. Also abundant are
Wetherellia variabilis (Linaceae) and Hightea elliptica (Myrtaceae
?); the former is believed to be related to the modern Hugonia
which is found in the tropics of Asia, Africa, Australia, and New
Caledonia; the latter, represented in the London Clay collections
by over 500 specimens, has proved difficult to identify and its posi¬
tion in this family is somewhat doubtful. Three members of the
Icacinaceae are abundantly represented: lodes multireticulata,
Stizocarya communis, and Icacinicarya platycarpa. Fossils refer¬
able to modern temperate climate families are not absent from the
flora, one of the very abundant species being Petrophiloides
richardsonii of the Juglandaceae.
The London Clay flora presents many questions that are by no
means alien to the paleobotany of other areas and ages. A few

Telegram @tubscstds
190 STUDIES IN PALEOBOTANY

H J

Telegram @tubscstds
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 191

that immediately arise are: Why are fruits and seeds present in the
clays to the exclusion of leaves? How can one account for the
relationship with a distant tropical flora in this northerly latitude?
Is it possible that the fossils were transported by ocean currents
perhaps hundreds or thousands of miles from their original point
of origin? Is the implication of a tropical climate erroneous in
that the plants concerned, although of tropical affinities, may in
fact be genera that became acclimatized to a temperate climate?
Mrs. Reid and Miss Chandler have presented one of the most com¬
prehensive and penetrating studies of these recurrent problems and
it seems opportune to draw upon their results. A serious student
of fossil botany would do well to read the Introduction to The
London Clay Flora.
As to the seed-fruit composition of the flora Reid and Chandler
refer to H. N. Moseley’s account of the Challenger Expedition per¬
taining to observations that were made some 70 miles northeast of
Point D’Urville, New Guinea, where the Ambernoh River, the
largest in New Guinea, empties into the sea.
We passed through long lines of drift-wood disposed in curves at right
angles to the direction in which lay the river’s mouth, . . . The logs had
evidently not been very long in the water, being covered only by a few young
Barnacles (Balanus) and Hydroids. . . . Various fruits of trees and other frag¬
ments were abundant, usually floating, confined in the midst of the small
aggregations into which the floating timber was almost everywhere gathered.
. . . Very small seeds were as abundant as large ones, the surface scum being
full of them, so that they could be scooped up in quantities with a fine net.
... I observed an entire absence of leaves, excepting those of the Palm, on
the midribs of which some of the pinnae were still present. The leaves
evidently dropt first to the bottom, whilst vegetation drift is floating from a
shore. Thus, as the debris sinks in the sea water a deposit abounding in
leaves, but with few fruits and little or no wood will be found near shore,
whilst the wood and fruits will sink to the bottom farther off land. (1892,
p. 373)

Fig. 7-1. Seeds and fruits from the Eocene London Clay. A. Minsterocarpum alatum,
worn fruit showing seeds, 3X. B. Melicarya variabilis, a nine loculed fruit, 3X.
C. Euphorbiotheca sheppeyensis, longitudinal section of fruit, 3X. D. Nip a burtini,
internal cast of seed, IX. E. lodes multireticulata, endocarp, 3X. F. Anonaspermum
ovale, internal cast of seed showing branching and anastomosing ruminations, 3X.
G. Ampelopsis rotundata, internal cast of seed, 7X. H. Protocommiphora europaea,
a two-loculed endocarp, 5.5X. I. Microtinomiscium foveolatum, locule cast, 6.6X.
J. Wetherellia variabilis, seed with remains of the testa, 8X. (From Reid and Chand¬
ler, 1933.)

Telegram @tubscstds
192 STUDIES IN PALEOBOTANY

Fig. 7-2. The beach at Sheppey; Eocene fruits and seeds come from the clay bank to
the left.

Of the 73 genera that it has been possible to relate in some

degree to modern ones 33 are lowland tropical forms, 29 may be
lowland tropical, montane tropical, or extra tropical and 8 are
montane tropical or extra tropical. Nipa is considered to be a key
plant in that it is a lowland tropical form occupying brackish
waters, whereas a significant number of other species may repre¬
sent plants that grew at a higher altitude and were carried down
to the point of deposition by a large river or were carried a much
greater distance by ocean currents. It is known from the works of
H. B. Guppy and others that seeds and fruits may be carried great
distances, for example, from the West Indies to the shores of
Britain and Scandinavia by the Gulf Stream. Several points mili¬
tate against a long ocean voyage for the London Clay fruits. It is
unlikely that they would be found in abundance along the shore
at Sheppey if they had been carried for thousands of miles; very
few, moreover, display any specialized structural features that
would enable them to remain afloat for the long period of time
necessary; finally, it has been shown that beach drift, even though

Telegram @tubscstds
it may include seeds and fruits that have come long distances, con-
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 193

tains a predominance of local debris, and we have seen that the

London Clay flora is strongly tropical in both numbers of speci¬
mens and species.
It was suggested by Bowerbank in 1840 that the fossils may
have been transported by a river “which probably flowed from
near the Equator towards the spot where the interesting remains
are now so abundantly deposited.” There seems to be little doubt
that river transport is involved but not to the extent suggested.
A direct route from the African tropics, which in itself would have
involved a very considerable distance, was cut off by the Tethys
Sea which occupied the present Mediterranean basin as well as
much of south-central Europe, part of northern Africa, and much
of the present Near East; this sea moreover communicated with
the Indian Ocean. In this connection it is of some interest to men¬
tion a small flora described by Miss Chandler from the Danian
(uppermost Cretaceous) of Egypt in which the fossils are closely
related or identical with those of the London Clay.
An argument that is often used against plants as a reliable index
of climate is the assumption that they may change in the course
of time in their reaction to climate. That some plants are remark¬
ably versatile in their ability to survive under a wide variety of
climatic conditions is evident, and Reid and Chandler note clearly
that conclusions as to climate must be drawn from a study of the
bulk of the fossil flora concerned. One composed of few species
and few specimens might well lead to erroneous concepts, but
when dealing with several hundreds of species and thousands of
specimens this objection is hardly tenable.
A study of later Tertiary floras shows that the bulk of the Lon¬
don Clay plants did not survive under the cooler climatic conditions
that developed even in late Eocene times. Reid and Chandler also
present evidence from Pleistocene floras to indicate that, under
changing climatic conditions, plants are much more apt to migrate
to a favorable locale than to adapt in their current position.
Paleoclimatological problems may be complicated and not infre¬
quently meet with bitter argument. It is not implied that the
authors of the London Clay flora have given a final and faultless
answer, but their study is extraordinarily well documented.

Some Later Tertiary Floras of Europe

The Hordle flora of Hampshire, England, is one of dominantly

Telegram @tubscstds
southern Asiatic relationships, of the south China and Burma
194 STUDIES IN PALEOBOTANY

areas, appreciably to the north of the Malay Islands. There are

also secondary affinities with the flora of Japan, north China, and
the Himalayas and, as compared with the London Clay flora, there
is a noticeable increase in northern elements with European affini¬
ties. In this category may be cited genera such as Salvinia
(Salviniaceae), Pinus (Pinaceae), Potamogeton, and Limnocarpus of
the Potamogetonaceae, Stratiotes (Hydrocharitaceae), Nuphar
(Nymphaeaceae), Corydalis (Papaveraceae), Rubus (Rosaceae),
and Omphalodes (Boraginaceae). Of the dominant lowland trop¬
ical plants of the London Clay only three of these, Nipa (Nipaceae),
lodes (Icacinaceae), and Tetrastigma (Vitaceae) survived until the
end of the Eocene in Europe.
The Bembridge flora from the Oligocene of the Isle of Wright is con¬
sidered as being a warm temperate one. Among the genera which
particularly suggest such a climate are Acrostichum (Leather fern,
based on a leaf fragment), which is found today in southern Florida
and the Keys in fresh water marshes, brackish swamps, and man¬
grove swamps; an Acanthus (Acanthaceae) which is allied with
modern species found in the mangrove swamps of eastern Asia and
Australia; Phyllanthera vectensis (Asclepiadaceae) is compared
with the modern P. perakensis, a liane of the Malay Peninsula;
several palms (Sabal, Palmophyllum, and Palaeothrinax) are rep¬
resented by leaf fragments. The flora includes plants of more
northerly affinities, indicating a change from the climate of London
Clay times; a few of these are Pinus, Typha (Cat-tail), Engelhard-
tia (Juglandaceae), Papaver (Papaveraceae), and Abelia
(Caprifoliaceae).
The Pliocene floras, such as the Pont-de-Gail, Reuverian, Kros-
cienko, Teglian, and Cromerian, show a progressive shift from an
eastern Asiatic affinity to one with western Europe. The Pont-de-
Gail flora still retains some of the Indo-Malayan elements; al¬
though having a dominant relationship with eastern Asia, the west
European relationships of the Reuverian and Kroscienko floras are
becoming pronounced; in the Teglian, the East and West relation¬
ships are about equal, whereas the Cromerian is definitely a west
European flora. It is not possible to consider all of these in detail
and I have, therefore, chosen to discuss two, the Reuverian and
Kroscienko floras, which are large and quite well known.
The Reuverian flora of Limburg is an especially large one; nearly
300 species are known, of which somewhat more than one-half
have been identified with considerable certainty. So far as I am

Telegram @tubscstds
aware this is one of the largest of Tertiary floras; in fact, the
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 195

authors (Reid and Reid) note that it was probably much richer in
species of trees and shrubs (both absolutely and proportionally)
than the modern wooded areas of that region. The highest con¬
centration of modern equivalents is found in the mountains of
western China; in some cases the fossils are so similar that they
cannot be distinguished from the Chinese ones, notably: Gnetum
scandens, Stewartia pseudo-camellia, Magnolia kobus, Zelcova
keaki, Pyrularia edulis, and Prunus maximoviczii. Somewhat
fewer genera are represented by closely allied species in Japan,
eastern Tibet, and the Malay Peninsula, and fewer still have rela¬
tionships with the flora of Europe.

Not only is the alliance marked by the number of species belonging to

Chinese and Japanese genera which are not represented in Europe, but by
the curious fact that when the genus lives in both Europe and China, it is
often the Chinese or Japanese species that most resembles the Reuverian
plant. Thus in the genus Pterocarya (wingnut), P. limburgensis is more
closely allied to P. hupehensis than to P. caucasica. In Styrax (snowball)
the alliance is with S. japonica and S. obassia. In the genus Betula (birch)
we have B. digitata, an ally of the Chinese B. ulmifolia and belonging to a
section of the genus no longer found in Europe. In Cornus (dogwood) we
have either C. controversa or a closely allied species. The only Clematis
(virgin’s bower) is the Chinese C. grata, not the nearly related C. vitalba.
But perhaps the most remarkable of these cases is in the genus Eupatorium;
for not only is the alliance of the Reuverian plant closer to E. japonicum,
but it is closest to an unnamed variety of E. japonicum, collected by Pere
Faurie from a single mountain in Japan. . . . The alliance is shown almost
entirely by species whose northernmost geographical range now lies much
farther south than Limburg; but they are mountain plants in China, and are
not found anywhere on the plains. In other words, though now living in
southern latitudes they are temperate forms, and belong to the moist and
temperate forest-belt found only on the Chinese mountains, and in the
similar moist regions of the Himalaya and Japan. (1915, p. 16)

Some of the Reuverian plants that are still found in Europe, and
in fact ranging considerably to the north of Limburg, are: Picea
excelsa (Norway spruce), Quercus robur (English oak), Carpinus
betulus (European hornbeam), Corylus avellana (European hazel),
Prunus spinosa (blackthorn), Ilex aquifolium (English holly), Vitis
vinifera (wine grape). In view of their ability to withstand con¬
siderable cold the authors suggest they could have survived
through the rigors of the Pleistocene by moving to the south of
France or the Caucasus.
Another large Pliocene flora, with about 40 species in common
with the Reuverian, has been described by Szafer from Kroscienko

Telegram @tubscstds
in southern Poland. The fossil seeds and fruits were obtained
196 STUDIES IN PALEOBOTANY

from an examination of about 100 cubic meters of the sediments

in which they occur; it is a study that will probably take a fore¬
most place in the classics of paleobotanical literature.
About 18% of the species of the flora are most closely related
with living modern European ones, 37% with eastern Asiatic ones,
and 19% with eastern American ones. Szafer concludes that two
distinct elevations are represented in the flora: a lower one in
which the following prevailed: Carpinus (hornbeam), Pterocarya
(wingnut), Alnus (alder), Liriodendron (tulip tree), Ceanothus,
Vitis (grape) and Styrax (snowball), and an upper elevation in
which the following were predominant: Picea (spruce, 3 species),
Tsuga (hemlock, 2 species), Abies (fir), and Pinus (pine).
Taken in total the Kroscienko flora is regarded as indicating a
climate with a mean temperature some 9° C higher than at present,
with less contrast in summer-winter extremes, and a rainfall about
twice that in the region today.
In summary, we have in this sequence from early Eocene London
Clay times to the modern British flora, extending over a period of
about 60 million years, a change from a lowland Malayan flora
which changed rather sharply at the end of the Eocene to a flora
still of eastern Asiatic affinities but more cosmopolitan—a gradual
shift through the Pliocene to a dominantly west European flora.
The change included a pronounced decrease in the percentage of
woody genera and a progressive increase in genera of wide range.
A reversal of this later trend occurs from the Cromerian to the
present and is accounted for by the extinction of many genera
during the Pleistocene.

LATE CRETACEOUS AND TERTIARY FLORAS

OF WESTERN NORTH AMERICA

During the past few decades a series of intensive studies have been
undertaken dealing with the Tertiary floras of the Pacific Coast
states. These investigations have been carried on for the most part
by R. W. Chaney, D. I. Axelrod, and several others and much that
follows is taken from their published accounts. Since the Tertiary
floral succession has not been studied elsewhere in as great detail
as in this area, and the story is a particularly interesting one from
the standpoint of shifting floras and climates, it has seemed to me
more effective to consider it in some detail at the expense of neglect¬

Telegram @tubscstds
ing Tertiary floras elsewhere in the world. In order to render the
account more complete, consideration is also given to some of the
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 197

late Cretaceous floras of the region. There are now hundreds of

publications dealing with the fossil plants of this age and area,
many of which are rather weighty volumes, and in attempting to
extract the essence of this voluminous literature I am aware of the
shortcoming of the following account.

Upper Cretaceous Floras

The Upper Cretaceous floras of the western states flourished

under conditions of more equitable and abundant rainfall and
milder temperatures than prevail today; warm temperate to sub¬
tropical forests grew over much of the present desert and high
mountain terrain.
The following plants have been found in the Medicine Bow forma¬
tion in southwestern Wyoming and adjacent Colorado: Cercidophyl-
lum (Katsura tree), Cinnamomum, Dryophyllum (evergreen oak),
Ficus (fig), Laurus, Lindera, Magnolia, Platanophyllum, Rhamnus
(buckthorn), Sabalites, Viburnum, and Metasequoia. Such an as¬
semblage suggests rather heavy rain throughout the year and mild
temperatures without much frost, in striking contrast to the low
rainfall and great variation in temperature that prevails in that
area now.
In the somewhat older floras of the Frontier formation and Aspen
shales of southwestern Wyoming we encounter several ferns includ¬
ing Anemia, Asplenium, Cladophlebis, Microtaenia, and Tempskya,
with such angiosperms as Cinnamomum, Ficus, Laurus, Liquid-
ambar (sweet gum), Prunus, Quercus (oak), Salix (willow), and
Sassafras. The distribution of the tree-fern Tempskya through
the northwestern states has been mentioned previously and in itself
is an important key to the Upper Cretaceous climate.
The Fruitland and Kirtland formations of the San Juan Basin of
northwest New Mexico have yielded the following ferns: Anemia,
Asplenium, Onoclea, and angiosperms: Cornus, Ficus, Laurus,
Quercus, Sabal, and Salix. The late Upper Cretaceous Vermejo
flora of southern Colorado and adjacent New Mexico includes the
following: Artocarpus, Cissites, Credneria, Ficus, Laurophyllum,
Magnolia, Platanus, and Sabalites.
Looking far to the northward the Upper Cretaceous of Alaska
has yielded a flora that has a considerable assemblage of plants that
do not occur there now. Over 200 species have been described by
Hollick which include 15 genera of gymnosperms and 73 genera of
angiosperms. Ginkgo was common, being represented by several

Telegram @tubscstds
species; five species of the cycadophyte Nilssonia have been de-
198 STUDIES IN PALEOBOTANY

scribed; Sequoia and Metasequoia are present, and recently Arnold

has recorded a new genus Parataxodium which is thought may be
the ancestral stock from which Metasequoia and Taxodium (bald
cypress) were derived. Among the numerous dicots are : Juglans
(walnut), Populus (poplar), Hickoria (hickory), Betula (birch),
Quercus (oak), Ficus (fig), Aristolochia (birthwort), Nymphaeites,
Castaliites, Menispermites, Magnolia, Laurus, Platanus (syca¬
more), Credneria, Sorbus, Cassia, Acer (maple), Rhamnus, Vitis
(grape), Aralia, Hedera (ivy), and Viburnum.
The Cretaceous floras are not as well known as the Tertiary ones
in western North America and the problems of their identification
are much greater. Yet allowing for numerous mistakes it seems
evident that they reflect a warmer climate through the New Mexico
to Alaska area, with a heavier and more equitable rainfall through¬
out the year than prevails at present.

Tertiary Floras

In a recent review of certain Tertiary floras of North America

Axelrod notes that over 200 are now known which include, in total,
some 800 genera and several thousand species. The succession of
these floras that have been excavated from Eocene to Pliocene hori¬
zons in the Rocky Mountain and Pacific Coast states tells a story
of cooling climates accompanied by an increase in winter-summer
extremes and a lowering of the annual rainfall to the point of ad¬
vanced desert conditions in the intermontane areas. In attempting
to summarize the vast amount of evidence now available I have been
particularly concerned with the difficulty of understanding the
vegetation of this great area at any one time during the Tertiary,
to say nothing of grasping the whole sequence of changes during the
past 60 or 70 million years. The present tremendous variation in
habitats, and consequently plant communities, varies not only in
going from north to south but, in many parts of the region, may be
altered drastically in a few miles. Because of the several mountain
ranges that have arisen during the Tertiary it is likely that such
abrupt floristic shifts are more pronounced now than at any other
time since the late Cretaceous; nevertheless there have been innu¬
merable local floristic differences.

Eocene

During the middle Eocene a vast lake extended across south¬

western Wyoming and adjacent parts of Colorado and Utah. The

Telegram @tubscstds
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 199

sediments that accumulated in this body of water, partly as a result

of volcanic action, are especially famous for the beautifully pre¬
served fossil fish that are found a few miles west of Kemmerer, Wyo¬
ming. Plants are also abundant at certain horizons and in a review
of the flora Brown has recognized 135 species (exclusive of thallo-
phytes) of macrofossils; in addition, many species have been estab¬
lished on fossil pollen.
The Green River flora includes: a fern (Lygodium kaulfussi);
two cycads based on pollen (Cycadipites, Dioonipites); three pines
known from pollen and a three-needle pine known from foliage; two
species of Potamogeton, an aquatic monocot—one from pollen and
another from a fruiting head; two palms have been recorded from
pollen and three from leaves. A partial list of the dicots known from
leaves, pollen, or both includes: Salix (willow), Juglans (walnut),
Hickoria (hickory), Tilia (linden), Alnus (alder), Betula (birch),
Carpinus (hornbeam), Rhus (sumac), Schmalzia, Sapindus (soap¬
berry), Koelreuteria (golden rain-tree), Liriodendron (tulip tree),
Oreodaphne, Liquidambar (sweet gum), and Acer (maple).
The Green River flora is one of the few that have been studied
thus far from both pollen and macrofossils (leaf impressions); this
advantage of checking one against the other offers a better chance
for correct identifications and will undoubtedly be pursued to that
end in future investigations. A few discrepancies appear in the flora
that cause one to wonder; for example, three species of oaks have
been identified from leaves but no oak pollen has been found. In
view of the abundance with which the oaks discharge pollen one’s
first reaction is to question the identification of the leaves. Brown
also cautions that in dual studies of this sort it is desirable to obtain
both pollen and macrofossils from the same horizon or, if separated,
this must be taken into account. In conclusion he suggests a warm
temperate climate that was sometimes well watered, sometimes dry,
and also offers the following interesting comment on the altitude:
The ecologic group of species comprising Hickoria, Juglans, Liquidambar,
Ailanthus, and Tilia, if judged by its modern equivalent, suggests that the
Green River flora, with the exception of some conifers, grew at an altitude
above the sea level of probably not more than 3000 feet and very likely at a
considerably lower level. The Green River formation, now from 5000 to
10,000 feet above the sea, must therefore have been uplifted several thousand
feet since the deposition of the sediments that contain the fossil flora. (1934,
p. 51)

The late Eocene Goshen flora of west-central Oregon includes,

among 45 species, the following as dominant elements: Quercus

Telegram @tubscstds
200 STUDIES IN PALEOBOTANY

(Fagaceae), Ficus (Moraceae), Aristolochia (Aristolochiaceae),

Drimys and Magnolia (Magnoliaceae), Anona (Anonaceae), Chry-
sobalanus (Rosaceae), Inga (Leguminoseae), Ilex (Aquifoliaceae),
Cupania and Allophyllus (Sapindaceae), Meliosma (Sabiaceae),
Nectandra and Ocotea (Lauraceae), Tetracera (Dilleniaceae), Calyp-
tranthes (Myrtaceae), Diospyros (Ebenaceae), Cordia (Boragi-
naceae).
The present distribution of the genera of the Goshen flora is far to
the south of Oregon. What seem to be the closest modem equiva¬
lents are found in the tropical rain forest of the Pacific slope of
Panama and the temperate rain forest of Costa Rica. On the basis
of this comparison it is postulated that the late Eocene climate of
the region had a mean annual temperature of 65 to 70° F and a
rainfall of 70 inches or more.
The Chalk Bluffs flora, of Middle Eocene age, is known from
several localities in north central California and presents problems
of interpretation that seem worth considering. The most abundant
species in the flora are:

Thouinopsis myricaefolia Mallotus riparius

Cercidophyllum elongatum Chaetoptelea pseudo-fulva
Per sea pseudo- car olinensis Gordonia efregia
Platanophyllum whitneyi Liquidambar californicum
Platanus appendiculata Rhus mixta

Chaetoptelea, known from both foliage and fruit, is one of the

most characteristic fossils of the flora and is compared closely with
C. mexicana which is found at subtropical elevations in southeast¬
ern Mexico and Central America. Cercidophyllum, also with foli¬
age and fruits preserved, is abundant in the early Tertiary floras of
North America; the modern C. japonicum (Katsura tree) is a large
forest tree of southwestern China and is found north to Hokkaido
in Japan. Persea pseudo-carolinensis is compared with the swamp
red bay of southeastern United States which is often associated with
Platanus, Magnolia, Gordonia, and Liquidambar. The character¬
istic fruits, seeds, and leaves of Liquidambar are well represented.
The leaves are of both the three-lobed and five-lobed forms, the
former being compared with Liquidambar from southern Mexico.
The climate in which the flora existed was probably similar to
that of the Goshen plants, but there are some obvious inconsisten¬
cies which cannot be overlooked. For example, Artocarpus (bread¬
fruit), Rhamnidium, and Tabernaemontana which are tropical

Telegram @tubscstds
genera are associated with temperate climate Carya (hickory),
 TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 201

Fig. 7-3. Some plants from the Middle Eocene Chalk Bluffs flora of Nevada County,
California. A, Laurophyllum litseaefolia MacGinitie; B, Platanophyllum whitneyi
(Lesquereux) MacGinitie; C, Mallotus riparius MacGinitie. (From MacGinitie, 1941.)

Telegram @tubscstds
202 STUDIES IN PALEOBOTANY

Acer (maple), and Fraxinus (ash). This occurrence of climatically

divergent elements in a fossil flora is not an uncommon problem and
has been discussed at some length by MacGinitie. He points out
that there are several possible explanations: the identifications may
be in error; the requirements of the modern equivalent plants
may have changed or we may have an inadequate understand¬
ing of their present distribution and climatic requirements; the
elements of the fossil flora may include upland plants as well as
lowland ones.
The remarkable succession of fossil forests in the Yellowstone
Park area has been mentioned in Chapter 1 where our attention was
focused on the mode of origin of the deposits which include silicified
stumps and logs as well as leaf-bearing beds. Knowlton’s original
study of the plants appeared over 60 years ago and their age has
been regarded as Miocene; Professor Dorf has recently initiated an
intensive restudy of the fossils and believes the age to be Eocene.
This seems much more acceptable in view of the plant assemblage
and what we now know about western Tertiary floras in general.
As to the composition of the flora he notes:

The buried forests were apparently a mixture of warm temperate forms,

such as Sequoia, Pinus, Platanus, Magnolia, Juglans, Cercidophyllum, and
Castanea (chestnut), and of subtropical forms, such as Persea, Ficus, Colum¬
bia, and Laurus. The assemblage is interpreted to indicate a warm lowland
region with annual rainfall of 50 to 60 inches. (1959, p. 95)

The abundance of dicot genera is in striking contrast to the

modern forests. Today 84% of the Park is forested and three-
quarters of it is lodgepole pine; a large portion of the remaining
quarter is forested with other species of pine, as well as fir, spruce,
and Douglas trees. The only deciduous trees that are at all abun¬
dant are Populus tremuloides (quaking aspen), P. angustifolia
(narrow-leaved cottonwood), Betula glandulosa (birch), and Ame-
lanchier alnifolia (shadbush). To this writer one of the most fabu¬
lous features of the fossil forests is the presence of great silicified
Sequoia stumps in excess of 14 feet in diameter.
Summer at present in the Yellowstone country is pleasant but
fleeting. I have not had the pleasure of being there in mid-winter,
but visits in late May, when a light snowfall was experienced each
night, and to the Idaho country just to the south in September,
when the weather behaved comparably, gives one the impression
of a sub-Arctic climate! It was certainly milder in western Wyoming

Telegram @tubscstds
during the Eocene.
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 203

Oligocene

The Florissant beds of the high Rockies to the west of Colorado

Springs consist of lake sediments resulting from volcanic action.
They are noted for the abundance of leaf impressions, insects, and
associated sihcified stumps and they attracted attention as early as
1874; numerous publications have appeared from the studies of
Lesquereux, Knowlton, Kirchner, Cockerell, and others. The pres¬
ent account is taken from the recent revision of the flora by
MacGinitie.
The flora, probably of early Oligocene age, is a diverse one of well
over 100 species. A few mosses are present, as well as Equisetum,
and several conifers including A bies (fir), Pinus (pine), Picea (spruce),
Torreya, Chamaecyparis (cypress), and Sequoia. There are a few
monocots, with Typha (cat-tail) being quite abundant. The most
abundant fossil is Fagopsis longifolia (Betulaceae) which is thought
to have occupied a dominant place along the ravine banks compara¬
ble to that of the birches and alders today. The elm family is
represented by Celtis (hackberry), Ulmus (elm), and Zelkova; the
last, an Asiatic genus, is one of the dominant plants in the flora.
Others occurring in abundance are a poplar related to the western
black cottonwood (Populus trichocarpa); seven genera of the Rosa-
ceae including Amelanchier (shadbush), Cercocarpus (mountain
mahogany), Crataegus (hawthorn), Malus (apple), and Prunus;
Ptelea (Rutaceae) is recognized by its characteristic winged fruits;
several members of the flora represent genera now confined to Asia
such as Ailanthus (Simarubaceae) and Koelreuteria (Sapindaceae)
—the beautiful golden rain tree (Koelreuteria paniculata) is culti¬
vated as far north as Missouri today.
Many of the Florissant species have their closest living counter¬
parts in the modern flora of southwestern United States and parts
of Mexico. In brief, the environs of the Florissant Oligocene flora
is thought to have been one with an absolute minimum temperature
of not below 20° F and an average annual temperature of not less
than 65°, with an annual rainfall of about 20 inches. It was a
mesophytic forest rich in species growing along stream and lake
shores at an elevation of almost 3000 feet. The living forest is com¬
posed largely of yellow pine and some Douglas trees, Engelmann
spruce, Colorado blue spruce, and limber pine; the only arborescent
dicot of importance is aspen, whereas along the streams are found
alder, birch, several willows, wild roses, chokecherry, shrubby poten-

Telegram @tubscstds
tilla, serviceberry, and a few others.
204 STUDIES IN PALEOBOTANY

In his consideration of the climatic relations of the Florissant

fossils MacGinitie offers a comment that is worth noting as one of
very general application:

There is another aspect of the Tertiary environment which has a bearing

on the presence of apparently contradictory groups of plants in the general
ecological picture of a fossil flora. All evidence points to higher average
temperatures in western or southern states in pre-Pliocene times than obtain
today. This appears to have been the result, not of higher maxima, but of
higher minima of temperature. The extreme cold waves of winter, which
bring freezing weather today into the Gulf littoral and below zero tempera¬
tures to the central states, must have been greatly mitigated in Middle and
Lower Tertiary time. (1953, pp. 39-40)

He demonstrates the importance of temperature minima by a

comparison of the California coast from Monterey north to Fort
Bragg, with northeastern Kansas and north central Missouri. Both
regions have an annual average temperature of 55° F, but the aver¬
age January temperature at Columbia, Missouri, is 31° and that of
the California area 48°; the absolute minimum in the former is
about —26° and in the latter about 21°. The much higher mini¬
mum temperature permits the cultivation of some subtropical trees
along the California coastal strip.
By Upper Oligocene times the tropical elements in the latitude
of Oregon had largely disappeared, for the Bridge Creek flora of the
John Day Basin in the eastern part of the state presents a distinctly
cooler climatic assemblage. The dominant elements are Quercus
consimilis, Metasequoia occidentalis, Alnus carpinoides, and Um-
bellularia. A comparison has been drawn between this Oligocene
flora and the modern redwood forest, other genera in common
being Pteridium (bracken fern), Equisetum (horsetail rush), and
Asarum, whereas elements which occur today near the redwood
forest but on more open slopes and valleys are Philadelphus (mock
orange), Crataegus (hawthorn), and Fraxinus (ash).
In discussing certain European floras it was noted that the mid-
Tertiary brought in not only temperate climate genera but ones that
had a wider range, and a comparable trend appears in the western
American floras. For example, we also find in the Bridge Creek
flora Platanus (sycamore), Juglans (walnut), and Celtis (hackberry)
which are found today in other parts of the west than the redwood
forest as well as some that are not represented in western North
America at present, as: Carpinus (hornbeam), Fagus (beech), Ulmus
(elm), and Tilia (linden).

Telegram @tubscstds
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 205

Fig. 7-4.Plants of the Bridge Creek flora. Leaf and seed catkin of Alnus carpinoides.
(From specimens in the U. S. National Museum.)

Miocene

Two distinct Miocene floras are recognized by Axelrod:

1. The Arcto-Tertiary flora of the northern Great Basin and

Columbia Plateau which is typified by temperate hardwood de¬
ciduous and conifer forests living under climatic conditions of 35 to
50 inches of rain distributed quite evenly throughout the year.
2. The Madro-Tertiary flora, occupying the area from southern
California east into Mexico, composed of semiarid, live-oak wood¬
land chaparral, and thorn forest with only 15 to 25 inches of rain
distributed biseasonally; the winters were mild and the summers
hot.

Vast outpourings of lava took place in Miocene times east of the

Cascade range, resulting in the area known as the Columbia Plateau;
some concept of the magnitude of this volcanic action may be gained

Telegram @tubscstds
206 STUDIES IN PALEOBOTANY

Fig. 7-5. Plants of the Bridge Creek flora. A. Metasequoia occidentalis. B,C. Leaf
and fruit of Ostrya oregoniana. D. Quercus sp. E. Zelkova hesperia. (From speci¬
mens in the U. S. National Museum.)

Telegram @tubscstds
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 207

from the fact that the total accumulation of lava is estimated at

about 100,000 cubic miles. New lakes and swamps developed as a
result of the damming of streams and valleys, and numerous fossil
localities were formed as plant remains, and ash from local volca¬
noes, accumulated in these newly formed bodies of water. The
geological action of the time must have been violent in the extreme;
this and other aspects of Tertiary paleobotany are vividly described
in Chaney’s The Ancient Forests of Oregon.
The Mascall flora which has been excavated from sediments in
east central Oregon between Dayville and Mount Vernon includes
nearly 70 species of which the following 10 (78% of the flora) are
most abundant:

Mascall flora plants Most closely related living species

Taxodium dubium T. distichum (bald cypress) of south¬
eastern United States
Quercus pseudo-lyrata Q. borealis (red oak) of eastern United
States, and Q. kelloggii (Californian
black oak) found in California to
central Oregon on the west side of
the mountain ranges
Carya bendirei (Hickories are found today in eastern
North America and extend into
Texas)
Platanus dissecta Resembles P. racemosa of western
United States
Quercus merriami Q. rubra of eastern United States
Acer bolanderi A. grandidentatum (maple) of the
Rocky Mountains and A. leuco-
derme (maple) of southeastern
United States
Metasequoia occidentalis M. glyptostroboides, the dawn red¬
wood
Ginkgo adiantoides
Acer negundoides A. negundo, the box elder, which
ranges across the U. S., and A.
henryi of Asia
Ulmus speciosa U. fulva (the slippery elm) of eastern
North America

Thirty-four of the Mascall species (more than half the total) have

Telegram @tubscstds
modern equivalents in eastern North America; most of the trees in
208 STUDIES IN PALEOBOTANY

the fossil flora were deciduous including the two most numerous
gymnosperms, Taxodium and Metasequoia. The climate in which
these plants lived is postulated as being comparable with that of
Ohio or of Szechuan, China, at middle altitudes.
The Miocene Tehachapi flora far to the south is composed of a
very different assemblage of plants. It has been obtained from a
thick, white ash bed in Kern County, California, and is approxi¬
mately 800 miles south of the Columbia Plateau. The closest rela¬
tionship with a living flora lies in northern Mexico where 65% of
the Tehachapi fossils have modern equivalents. The dominant
elements, with their Sonoran (Mexico) relatives, are:

Tehachapi Miocene flora Modern equivalents; Sonora, Mexico

Quercus convexa Q. oblongifolia
Q. browni Q. chrysolepis (Californian live oak)
Sabal miocenica S. uresana
Dodonaea californica D. viscosa
Amorpha oblongifolia
Robinia californica R. neo-mexicana
Populus prefremonti P. wislizeni
Prunus prefasciculata P. fasciculata (Desert almond)

The fossil leaves represent two distinct vegetational units, an

arid, subtropical lowland group including the palm Sabal (both
leaves and petrified stems are known), Bursera (Burseraceae),
Colubrina (Rhamnaceae, buckthorn family), Euphorbia (Euphor-
biaceae, spurge family), Ficus (Moraceae), and Pithecolobium
(Leguminosae, pea family), whereas the remainder are from upland
floras. This floristic comparison with northern Mexico and south¬
west United States indicates a Miocene climate of hot summers
ranging up to 105° F and possibly as low as 25° in the winter; at
the lower elevations rainfall was in the vicinity of 12 inches, increas¬
ing to about 25 in the adjacent uplands.
Extensive uplift of the Cascade-Sierra Nevada mountain range
in early Pliocene times resulted in still greater aridity of the Great
Basin area. This allowed a more northerly extension of the xero-
phytic elements of the north Mexican vegetation noted above. The
Weiser flora from southwestern Idaho is made up of 32 genera in
which oak, maple, pine, juniper, ash, sycamore, Arbutus (Ericaceae,
heath family), Pseudotsuga (Douglas tree), and Castanopsis (Faga-
ceae, beech family) are dominant elements. The inferred climate
is one with warm, dry summers and a rainfall of 20 to 30 inches.

Telegram @tubscstds
Although more temperate and xeric than the earlier Tertiary floras
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 209

it is not as strongly so as later Pliocene vegetation. The immediate

vicinity from which the Weiser flora was excavated is an arid region
(less than 13 inches rain) of semidesert shrubs in which black sage
and salt-bush are dominant.
By later Pliocene times the vegetation had approached a compo-

Fig. 7-6. Late Tertiary plants from Nevada. A, Populus pliotremuloides Axelrod;
B, Populus alexanderi Dorf, both Mio-Pliocene. C, Sassafras sp.; D, Sequoiadendron
chaneyi Axelrod; E, Quercus hannibali Dorf; F, Alnus smithiana Axelrod, all late

Telegram @tubscstds
Miocene. (From Axelrod, 1956.)
210 STUDIES IN PALEOBOTANY

sition close to that of the present time. For the northern area, for
example, Chaney notes:
The Pliocene floras of Oregon give us, for the first time in the records of
past vegetation, a sense of modern trees—the sorts of trees which are still liv¬
ing in western America. While the gap in time is still several million years,
it seems clear that during this epoch the climate and topography of Oregon
were much as we know them today. The Pliocene is a record of an immediate
yesterday, not of remote ages. (1956, p. 43)

REFERENCES

Arnold, Chester A. 1955. A new Cretaceous conifer from northern Alaska. Amer.
Journ. Bot., 42: 522-528.
Axelrod, Daniel I. 1939. A Miocene flora from the Western Border of the Mohave
Desert. Carnegie Inst. Washington Pub., 516: 1-129.
-. 1940. Late Tertiary floras of the Great Basin and Border Areas. Bull.
Torrey Bot. Club, 67: 477-487.
-. 1950. Evolution of desert vegetation in western North America. Car¬
negie Inst. Washington Pub., 590.
--—. 1952. A theory of angiosperm evolution. Evolution, 6: 29-60.
-—-. 1956. Mio-Pliocene floras from West-Central Nevada. Univ. California,
Pub. Geol. Sci., 33: 1-322.
-. 1958. Evolution of the Madro-Tertiary Geoflora. Bot. Rev., 24: 433-509.
Bailey, Irving W. 1944. The development of vessels in angiosperms and its signifi¬
cance in morphological research. Amer. Jour. Bot., 31: 421-428.
-and Nast, C. G. 1943. The comparative morphology of the Winteraceae
II. Carpels. Journ. Arnold Arb., 24: 472-481.
-and Smith, A. C. 1943. The family Himantandraceae. Journ. Arnold
Arb., 24: 190-206.
-and Smith, A. C. 1942. Degeneriaceae a new family of flowering plants
from Fiji. Journ. Arnold Arb., 23: 356-375.
-and Swamy, B. G. L. 1948. Amborella trichopoda Beill., a new morpho¬
logical type of vesselless dicotyledon. Journ. Arnold Arb., 29: 245-254.
-. 1951. The conduplicate carpel of dicotyledons and its initial
trends of specialization. Amer. Journ. Bot., 38: 373-379.
Bell, W. A. 1956. Lower Cretaceous floras of western Canada. Geol. Survey Canada
Mem., 285: 1-331.
-. 1957. Flora of the Upper Cretaceous Nanaimo group of Vancouver Island,
British Columbia. Geol. Surv. Canada Mem., 293:1-84.
Berry, Edward W. (et al.) 1911. The Lower Cretaceous deposits of Maryland. Mary¬
land Geol. Survey, 1911. Pp. 1-622.
-. 1919. Upper Cretaceous floras of the eastern Gulf region in Tennessee,
Mississippi, Alabama, and Georgia. U. S. Geol. Survey Prof. Paper, 112: 1-177.
Bose, M. N., and S. C. D. Sah. 1954. On Sahnioxylon rajmahalense Sahni, and S.
andrewsii, a new species of Sahnioxylon from Amrapara in the Rajmahal Hills
Bihar. The Palaeobotanist, 3: 1-8.

Telegram @tubscstds
 TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 211

Boureau, Edouard. 1954. Decouverte du genre “Homoxylon Sahni dan les terrains
secondaires de la Nouvelle-Caledonie. Mem. Mus. National Hist. Nat., Paris ser C
3: 129-143.
Brown, Roland W. 1934. The recognizable species of the Green River flora. U. S.
Geol. Survey Prof. Paper, 185-C: 45-68.
-. 1950. Cretaceous plants from southwestern Colorado. U. S. Geol. Survey
Prof. Paper, 221-D: 45-66.
-. 1956. Palmlike plants from the Dolores formation (Triassic) of south¬
western Colorado. U. S. Geol. Survey Prof. Paper, 274-H: 205-209.
Chandler, Marjorie E. J. 1954. Some Upper Cretaceous and Eocene fruits from
Egypt. Bull. Brit. Mus. Nat. Hist., Geology, 2: 149-187.
Chaney, Ralph W. 1925. A comparative study of the Bridge Creek flora and the
modern redwood forest. Carnegie Inst. Washington Pub., 349: 1-22.
-. 1925. The Mascall flora—its distribution and climatic relation. Carnegie
Inst. Washington Pub., 349: 23-48.
-. 1947. Tertiary centers and migration routes. Ecological Monographs,
17: 139-148.
-. 1956. The ancient forests of Oregon. Condon Lecture, Oregon State
System Higher Education, Eugene. 56 pp.
-and Axelrod, D. I. 1959. Miocene floras of the Columbia Plateau. Carne¬
gie Inst. Washington Pub., 617: 1-229.
-and Sanborn, E. I. 1933. The Goshen flora of west central Oregon. Car¬
negie Inst. Washington Pub., 439: 1-103.
Cheadle, Vernon I. 1953. Independent origin of vessels in the Monocotyledons and
Dicotyledons. Phytomorphology, 3: 23-44.
Corner, E. J. H. 1949. The Durian theory or the origin of the modern tree. Ann.
Bot., 13: 367-414.
Dorf, Erling. 1936. A Late Tertiary flora from southwestern Idaho. Carnegie Inst.
Washington Pub., 476: 75-124.
-. 1938. Upper Cretaceous floras of the Rocky Mountain Region. I. Strati¬
graphy and paleontology of the Fox Hills and Lower Medicine Bow formations of
southern Wyoming and northwestern Colorado. Carnegie Inst. Washington Pub.,
508: 1-78.
-. 1942. Upper Cretaceous floras of the Rocky Mountain Region II. Flora
of the Lance formation at its type locality, Niobrara County, Wyoming. Carnegie
Inst. Washington Pub., 508: 79-159.
-. 1952. Critical analysis of Cretaceous stratigraphy and paleobotany of the
Atlantic coastal plain. Bull. Amer. Assoc. Petrol. Geol., 36: 2161-2184.
-. 1959. The fossil forests of Yellowstone Park, Wyoming. Internat. Bot.
Congress, Montreal, Abstracts II: 95.
Erdtman, Gunnar. 1948. Did dicotyledonous plants exist in early Jurassic times?
Geol. Stockholm For hand. March-April: 265-271.
Fontaine, William M. 1889. The Potomac or Younger Mesozoic flora. Mon. U. S.
Geol. Survey, pp. 1-377.
Guppy, H. B. 1917. Plants, Seeds and Currents in the West Indies and Azores.
William and Norgate, London. 531 pp.
Harris, Thomas M. 1932. The fossil flora of Scoresby Sound, East Greenland. Part
2. Description of seed plants incertae sedis, etc. Medd. om Gronland, 85 (3): 1-112.
-. 1933. A new member of the Caytoniales. New Phyt. 32: 97-114.

Telegram @tubscstds
212 STUDIES IN PALEOBOTANY

-. 1937. The fossil flora of Scoresby Sound, East Greenland, pt. 5. Med-
delel. om Gronland, 112 (2): 1-112.
-. 1940. On Caytonia Thomas. Ann. Bot., n.s. 4: 713-734.
-. 1941. Caytonanthus, the microsporophyll of Caytonia. Ann. Bot., n.s.
5: 47-58.
Hollick, Arthur and Martin, G. C. 1930. The Upper Cretaceous floras of Alaska.
U. S. Geol. Survey Prof. Paper, 159: 1-116.
Hughes, N. F. and Couper, R. A. 1958. Palynology of the Brora coal of the Scottish
Middle Jurassic. Nature, 181: 1482-1483.
Knowlton, Frank H. 1916. Flora of the Fruitland and Kirtland formations. U. S.
Geol. Survey Prof. Paper, 98-S: 327-344.
-. 1899. Fossil flora of the Yellowstone National Park. U. S. Geol. Survey
Mon., 32, pt. 2: 651-882.
Kryshtofovich, African N. 1918. On the Cretaceous flora of Russian Sakhalin. Joum.
College Sci. Imper. Univ. Tokyo, 40 (8): 1-73.
Kuhn, Oskar. 1955. Das erste Dicotylenblatt aus dem Jura. Orion, 10 Jahrgang,
No. 19-20: 802-803.
Lesquereux, Leo. 1891. (Ed. post, by F. H. Knowlton). The flora of the Dakota
group. U. S. Geol. Survey Mon., 17: 1-400.
Lignier, O. 1908. Vegetaux fossiles de Normandie. V. Nouvelles recherches sur le
Propalmophyllum liasinum Lignier. Mem. Soc. Linn. Normandie, Caen, 23: 1-14.
MacGinitie, Harry D. 1941. A Middle Eocene flora from the central Sierra Nevada.
Carnegie Inst. Washington pub., 534: 1-178.
-. 1953. Fossil plants of the Florissant beds, Colorado. Carnegie Inst.
Washington Pub., 599.
Moseley, H. N. 1892. Notes by a naturalist on the “Challenges,” 2nd ed. London.
540 pp.
Newberry, J. S. 1895. (Ed. post, by A. Hollick). The flora of the Amboy clays.
U. S. Geol. Survey Mon., 26: 1-260
Reid, Clement, and Reid, Eleanor M. 1915. The Pliocene floras of the Dutch-Prus-
sian border. Meded. van de Rijksopsporing van Delfstoffen, No. 6, Gravenhage, pp.
1-177.
Reid, Eleanor M. 1921. A comparative review of Pliocene floras, based on the study
of fossil seeds. Quart. Journ. Geol. Soc. London, 76: 145-159.
-and Chandler, M. E. J. 1926. Catalogue of Cainozoic plants in the Depart¬
ment of Geology. Vol. I. The Bembridge flora. Brit. Mus. Nat. Hist., pp. 1-206.
-. 1933. The London Clay flora. Brit. Mus. Nat. Hist., pp. 1-561.
Samylina, V. A.. 1959. New occurrences of angiosperms from the Lower Cretaceous
of the Kolyma basin. Acad. Nauk USSR (Botany), 44 (4): 483-491.
Saporta, Gaston. 1894. Flore fossile du Portugal. Travaux Geol. du Portugal, Lis¬
bon, pp. 1-286.
Scott, Dukinfield H. 1923. Studies in Fossil Botany. Adam and Charles Black,
Ltd., London. 446 pp.
Seward, Albert C. 1904. The Jurassic flora. II. Liassic and Oolitic floras of England.
Cat. Mes. Plants Brit. Mus., pp. 1-183.
-. 1926. The Cretaceous plant-bearing rocks of western Greenland. Phil.
Trans. Roy. Soc. London, 215B: 57-175.
Simpson, J. B. 1937. Fossil pollen in Scottish Jurassic coal. Nature 139: 673.
Stopes, Marie C. 1912. Petrifactions of the earliest European angiosperms. Phil.

Telegram @tubscstds
Trans. Roy. Soc. London, 203B: 75-100.
TERTIARY FLORAS OF EUROPE AND WESTERN UNITED STATES 213

-. 1915. The Cretaceous flora. Part II. Lower Greensand (Aptian) plants
of Britain. Cat. Mes. Plants Brit. Mus., pp. 1-360.
Szafer, Wladyslaw. 1946. The Pliocene flora of Kroscienko in Poland. Polsk. Akad.
Umiet., Krakow, 72: 91-162.
Swamy, B. G. L. 1949. Further contributions to the morphology of the Degneriaceae.
Journ. Arnold Arb., 30: 10-38.
Takhtajan, Armen L. 1954. Origins of angiospermous plants. Soviet Science press,
1954. Amer. Institute Biol. Sci., transl. 1958, pp. 1-69.
--. 1957. On the origin of temperate flora of Eurasia. Acad. Nauk. U.S.S.R.,
42: 1635-1653.
Teixeira, Carlos. 1948. Flora Mesozoica Portuguesa. Servicos Cool, de Portugal.,
pt. I. Pp. 1-118.
Thomas, H. Hamshaw. 1925. The Caytoniales, a new group of angiospermous plants
from the Jurassic rocks of Yorkshire. Phil. Trans. Roy. Soc. London, 213B: 299-363.
-. 1931. The early evolution of the angiosperms. Ann. Bot. 45: 647-672.
-. 1934. The nature and origin of the stigma. New Phytologist, 33: 173-
198.
-■. 1957. Plant morphology and the evolution of the flowering plants. Proc.
Linn. Soc. London, 168 Session, 1955-56, pts. 1 and 2: 125-133.
--. 1958. Palaeobotany and the evolution of the flowering plants. Proc.
Linn. Soc. London, 169 Session, 1956-57, pts. 1 and 2: 134-143.
-. 1958. Fossil plants and evolution. Journ. Linn. Soc. London, Botany,
56: 123-135.
Traverse, Alfred. 1955. Pollen analysis of the Brandon lignite of Vermont. Bureau
Mines, Rept. Invest., 5151: 1-107.

Telegram @tubscstds
the LTCOPODOP1H1YTA

Introduction

The lycopods have held a prominent position in paleobotany

since the earliest days of the science. This interest has centered
largely in the Carboniferous members of the group because of the
abundance of fossils, the diversity of form that they exhibit, and
their arborescent habit. Some of them vie in size with all except
the largest of modern forest trees and the fossil record indicates
that in places they grew in more or less pure stands. As a result,
they enter into our economy today as an important contributor of
the raw material that became coal.
The lycopods were, however, well established by late Silurian
times and possibly earlier. The record indicates that several lines
of development evolved during Middle and Upper Devonian times
and the group was established as one of the dominant elements in
the vast, low-lying Upper Carboniferous swamps of North America
and Europe. The later record of the lycopods is relatively scanty;
the arborescent forms disappeared with the close of the Paleozoic.
The living members of the group are represented chiefly by the
genera Lycopodium and Selaginella, plants that rarely attain a
height of more than a foot or two. They are typical microphyllous
plants with a dichotomous branching system or exhibit a combina¬
tion of monopodial and dichotomous branching.
The spores are produced in sporangia which are borne singly at
the base of a scalelike sporophyll; numerous sporophylls are usu¬
ally aggregated into small but distinct cones. In some species of
Lycopodium the sporangia are borne on ordinary leaves, and groups
of these fertile leaves alternate with sterile ones along the length
of the stem. They are homosporous, that is, the spores in any one
species are all of essentially the same size.
Selaginella is heterosporous; there is a striking difference be¬
tween the small microspores (which develop into the male gameto-
phytes) and the megaspores (which develop into the female gameto-

Telegram @tubscstds
214
LYCOPODOPHYTA 215

phytes). An individual sporangium contains only one type, but

both microsporangia and megasporangia may be found in the same
cone. It is suggested that the student review the life cycle and
general morphology of these plants before studying the fossil forms.
As usual there are conspicuous gaps in our knowledge of racial
development in the group and there are a few particularly vexing
fossils which suggest that it was a more diverse assemblage than has
been supposed until quite recently. A wholly satisfactory classifica¬
tion for the lycopods is thus not possible; the following outline is
intended more as a guide to the better known fossils than as a
distinct system of classification.

1. Early lycopods of the Silurian to mid-Devonian

2. Lepidodendraceae
3. Sigillariaceae
4. Some herbaceous lycopods, Selaginellaceae
5. Pleuromeiaceae

EARLY LYCOPODS OF THE SILURIAN-DEVONIAN

It is possible that Aldanophyton antiquissimum (see Chapter 2)

from the Middle Cambrian of Siberia is a lycopod, but since the
spore-bearing organs are not known it can be accepted as only a
suggestion that the group existed at that early date.
The oldest fossil that we can assign with confidence to the
lycopods is Baragwanathia longifolia Lang and Cookson (Fig. 8-1)
from the Silurian of Australia. The plant probably exceeded
somewhat in size our largest species of Lycopodium; fragments of the
dichotomously branching stems were found up to 28 cm long and
although most specimens are 1 to 2 cm in diameter, they are
known to have reached 6 cm. These stems bore slender leaves 1
mm wide and 4 cm long and were apparently rather lax in texture.
Reniform sporangia containing spores are present in certain areas
of the stem associated with ordinary leaves; the general organiza¬
tion thus appears to be quite like that of the living Lycopodium
lucidulum although it has not been established for certain as to
whether the sporangia in Baragwanathia were actually attached
to the leaf on its upper surface near the base or directly to the
stem. The woody tissue of the stem consists of a primary stele,
with annular tracheids, similar to that of Asteroxylon but some¬
what more complex, there being at least 12 rays of xylem.

Telegram @tubscstds
216 STUDIES IN PALEOBOTANY

Fig. 8-1. Baragwanathia longifolia. A. Portion of a leafy shoot. B. Specimen show¬

ing sporangia. (A photograph courtesy Isabel Cookson; B from Lang and Cookson.)

It is a little disturbing to find in the Silurian a plant that com¬

pares so closely with modern lycopodiums; one may hesitate in
claiming that species like L. lucidulum are direct descendants of
this ancient Australian plant but it seems possible. It was a plant
of respectable size and insofar as our knowledge goes indicates that
the lycopods were among the forerunners of early land vascular
vegetation.
Drepanophycus is a genus that deserves particular attention as
an early to mid-Devonian lycopod. Based on German specimens,
it is depicted by Krausel and Weyland (Fig. 8-2) as a plant of
Lycopodium-like habit with a creeping stem which produced up¬
right dichotomously forking shoots attaining a height of about 45
cm; the stems were clothed with rather stout spiny appendages,
some of which had a single sporangium on the upper surface.
It was rather widely distributed as specimens have been reported
from the Gaspe in Canada, Norway, Germany, Scotland, Wales,
Belgium, and China. Thus far only specimens from German and
Welsh deposits have been found fertile, but there seems to be little
doubt that most of those from other localities are referable to

Telegram @tubscstds
LYCOPODOPHYTA 217

Drepanophycus. Some of the Belgian fossils from the Emsien

(upper part of the Lower Devonian) attained a diameter of about
2.5 cm and it seems likely that these may have exceeded the
estimated height of 45 cm. The spines were penetrated by a single,
slender vascular strand and have a distinctive broad, expanded
base on the larger branches. A single sporangium is borne on the
upper surface of the leaf and occupies a position about midway be¬
tween the proximal and distal ends; there is a suggestion from Welsh
specimens that it may occasionally have been borne at or near the
tip. The Lower Devonian Sugambrophyton pilgeri Schmidt was
similar to Drepanophycus, the branches of the dichotomously
forking shoot system attaining a breadth of 2.6 cm. The leaves,
however, are distinctive in that the ones on the basal part of the
plant are simple, whereas those on the upper parts are smaller and
dichotomize two or three times.
Protolepidodendron (Fig. 8-3) is best known from the Middle
Devonian of Germany and is characterized by leaves that fork near
the tip. Radially elongate sporangia are occasionally found on the
upper surface of the leaves. The shoots reached a diameter of

Fig. 8-2. Restoration of Drepanophycus

spinaeformis from the Lower Devonian
of Germany. (From Krausel and Wey-
land, 1935.)

Telegram @tubscstds
218 STUDIES IN PALEOBOTANY

about 2 cm and were supplied with a triangular protostele which is

not typically lycopodiaceous. Protolepidodendron has also been
reported from the district of Chanyi in China (probably Middle
Devonian); the stems are up to 6 mm broad and bear leaves 3.5
mm long. In describing these Halle notes that distinct leaf scars
are not present, suggesting that the plants were herbaceous and
did not shed their foliage as did the larger Carboniferous lycopods.
Although the plants considered above are mid-Devonian or older
it seems expedient to consider one Upper Devonian lycopod, Col-
podexylon trifurcatum Banks, at this point. It was a plant with
dichotomously forking stems up to 2.5 cm in diameter, but frag¬
ments up to 2 feet long suggest that it may have been taller than
the earlier Devonian lycopods. The leaves were three-forked and
persistent on the stem; the stele is composed of primary, scalari-
form tracheids and varies in cross-sectional shape from only slightly
to quite deeply lobed.

LEPIDODENDRACEAE

The arborescent lycopods of the Carboniferous were among the

dominant elements in the earth’s vegetation of this age and they
present unique aspects of morphology and growth that are unpar-

Fig. 8-3. Restoration of Protolepidoden¬

dron scharyanum. (From Krausel and
Weyland, 1935.)

Telegram @tubscstds
LYCOPODOPHYTA 219

alleled among plants of the present day. The fossil remains are
especially abundant as compressions in the roof shales of Upper
Carboniferous coal mines and as petrifactions in coal balls. Per¬
haps the most commonly encountered genus is Lepidodendron
which includes many scores of species. As is always the case with
larger plants our information comes from scattered remains of
roots, stems, leaves, and reproductive organs. Although all these
organs have never been correlated for a single species, our under¬
standing of Lepidodendron as a genus is remarkably comprehen¬
sive and it is in many ways representative of the arborescent
lycopods. The following description is a composite one in that
fossil remains representing several species and certain closely allied
genera are included.
Mature trees (Fig. 8-4) are known to have exceeded 100 feet in
height and 3 feet in diameter. The branching pattern was dichoto¬
mous although the equality of the two branches at each division
varied considerably; the root system consisted of four major
branches which likewise dichotomize several times, the ultimate
subdivisions bearing numerous spirally arranged rootlets. Linear
grasslike leaves, which varied tremendously in length in different
species (and possibly on a single plant), were borne toward the ter¬
minal portions of the ultimate branchlets and, upon falling from
the stem, left a characteristic scar which retained its identity even
on the oldest parts of the trunk. The spore-bearing cones are
basically similar to those of the living lycopods, but usually are
much larger.

Stem Anatomy

Unless otherwise indicated the following description is based on

Lepidodendron scleroticum Pannell, a species that is abundant in
certain Illinois coal balls. It has been selected for initial study be¬
cause it is known from numerous specimens which represent
various branch orders of the plant. The smallest twigs (Fig. 8-5)
measure about 4 mm; in the center is a minute, solid protostele
and this is surrounded by phloem, a rather uniform cortex of
parenchyma cells, and outermost are several conspicuous leaf bases.
The larger branches display a more complex anatomy and reveal
typical aspects of the plant. Figure 8-6 is a representative sector
of a branch about 4 cm in diameter; the central pith, which is
partially destroyed in this specimen, is enclosed by the primary
wood (Fig. 8-7) which is in turn surrounded by a band of secondary

Telegram @tubscstds
220 STUDIES IN PALEOBOTANY

Fig. 8-4. A. Restoration of an arborescent lycopod of the Lepidophloios type. B. A

young plant prior to the first dichotomy of the trunk, based in part on Lepidophloios
pachydermatikos. (B from Murdy and Andrews.)

wood (Fig. 8-8). The latter consists of scalariform tracheids and

rays which are one cell broad and vary from one to a score or more
cells high. In the nature of the wood we find the first of several

Telegram @tubscstds
characters that contrast with the structure of modern forest trees;
LYCOPODOPHYTA 221

it composes a small portion of the total volume of the stem. The

large thin-walled tracheids were probably efficient water conduc¬
tors, but as a supporting tissue they were quite inadequate. The
cambium and secondary phloem lie immediately outside the
secondary wood, but these tissues are rarely preserved.
The cortex in L. scleroticum (Fig. 8-6) consists of two rather
well-defined zones. The inner part is a groundwork of roundish
thin-walled cells and scattered through it are numerous clusters or
“nests” of cells, with somewhat thicker walls, filled with a dark
material which renders them quite conspicuous. The nests, a dis¬
tinctive feature of this particular species, are more or less isodia-
metric and are composed of at most a few dozen cells. The outer
cortex also consists of a groundwork of thin-walled parenchyma
and dispersed through it is an anastomosing network of longitud¬
inally elongate fiber strands.
Outside the cortex is an extensive development of tissue that has
been called periderm or cork. This is a unique feature of the
arborescent lycopods, both in its structure and relative abundance,
and will be described in some detail. Rather early in the growth

Telegram @tubscstds
Fig. 8-5. Lepidodendron scleroticum, a small protostelic twig, 20X.
222 STUDIES IN PALEOBOTANY

Fig. 8-6. Lepidodendron scleroticum. Portion of a branch about 4 cm in diameter;

pw, primary wood; sw, secondary wood; ic, inner cortex; oc, outer cortex; pd, periderm;
Ic, leaf cushions. (From Pannell, 1942.)

of the stem a periderm cambium was initiated in the outermost

part of the cortex; this is quite distinct from the wood cambium
and it continued to divide, probably throughout the life of the
plant, resulting in the formation of an extensive periderm. In L.
scleroticum the periderm cells are radially aligned, six to ten times
longer than broad, and with sharply tapered ends. The conspic¬
uous holes scattered through this tissue as it appears in Fig. 8-6
may represent areas of less resistant cells that served a secretory
function.
The periderm was more complex in some other species; Lepido¬
dendron johnsonii Arnold, from the Pennsylvanian of Chaffee
County, Colorado, presents some especially interesting features. It

Telegram @tubscstds
consists chiefly of the usual radially aligned fibrous cells and in this
LYCOPODOPHYTA 223

case they are twenty to thirty times as long as they are broad. In then-
equal length and regular arrangement they resemble the storied
structure found in the wood of certain dicots. In the inner region
the fibrous cells are interrupted by radial rows of tangentially
elongate ones. In tangential section they appear oval-shaped and
some are divided into two cells by a transverse wall. It has been
suggested that they may have afforded elasticity to the stem, per¬
mitting movement within the tissue during growth, or in response
to bending of the trunk and branches. A second interesting feature
of this Colorado species is the presence of periodic tangential rows
of glands. These are of complex organization; each gland extends
vertically for an undetermined distance and is composed of thin-
walled, brick-shaped cells, within which are oval masses of smaller
ones with dark contents presumably of a waxy or gummy nature.
Although the periderm usually seems to have been initiated in
the extreme periphery of the cortex, this is not always the case; it
may start 15 or 20 cells deep and, in a related genus, Lepidophloios,

Fig. 8-7. Lepidodendron scleroticum; this specimen shows an early stage of an unequal

Telegram @tubscstds
dichotomy, 14X; p, pith; px, primary wood; sw, secondary wood.
224 STUDIES IN PALEOBOTANY

Fig. 8-8. Lepidodendron scleroticum, portion of the stele of a large branch showing
extensive development of secondary wood, 5X. (From Pannell, 1942.)

even deeper. Periodicity of growth in the periderm is suggested in

some of the lycopods; in Sigillaria scutellata it is banded due to
variation in cell size, lending a superficial appearance to the struc¬
ture of the wood of a modern conifer. In a Lower Carboniferous
lycopod from Scotland, Levicaulis arranensis Beck, there is evi¬
dence that the periderm developed from a series of cambia or meri-
stematic regions rather than a single one. As to magnitude of
development, Professor Baxter has in his collections at the Univer¬
sity of Kansas, a slab of petrified periderm from an Iowa locality
with the extraordinary radial dimension of 40 cm; by the most
conservative estimates this implies a trunk close to 4 feet in
diameter.
It is thus evident that the periderm was a complex tissue in the
arborescent lepidodendrons and their immediate relatives and it
made up a considerable part of the trunk. As a result its function
has been the subject of some controversy. Several investigators
have concluded that it should not be compared structurally or func¬
tionally with the cork of modern trees. In his recent study of this
tissue in Levicaulis, Beck suggests that it be termed secondary
cortex rather than periderm and he also points out that it may well
have been a living tissue. There can be no doubt that it was the
chief supporting element of the arborescent lycopods, and it is also
apparent that further study is needed before we will fully under¬
stand it.

Telegram @tubscstds
LYCOPODOPHYTA 225

Leaves

Terminal branchlets of Lepidodendron and other genera are occa¬

sionally preserved with the foliage attached (Fig. 8-9), and isolated
leaves are common in petrifactions associated with the stems. The
generic name Lepidophyllum has long been used for isolated speci¬
mens of foliage, but is invalid since it was originally used for a living
South American flowering plant. Miss Snigirevskaya has recently
proposed the name Lepidophylloides, one that is appropriate and
correct.
In the roof shales immediately overlying the coal from which
Lepidodendron scleroticum is found leaves occur which are about
3 to 4 mm wide and nearly 30 cm long. The anatomical structure

Fig. 8-9. Terminal leafy branch-

lets of a Lepidodendron from the
Upper Carboniferous of Holland,
about 0.5X. (From a specimen in
the Swedish Natural History
Museum.)

Telegram @tubscstds
226 STUDIES IN PALEOBOTANY

of these lycopod leaves is of particular interest and several species

have been described from the American Pennsylvanian. The type
shown in Fig. 8-10C is constantly associated in coal balls with L. scle-
roticum, there being no reasonable doubt as to the identity of the
two organs.
An elongate group of tracheids composes the single central strand,
which is surrounded by thin-walled cells which are interpreted as
phloem. There is next a more or less continuous sheath referred
to as transfusion tissue; these are somewhat elongate cells with
transverse end walls and reticulate wall thickenings. Referring to
the lamina of the leaf, it is distinctive by virtue of the abundance
of hypodermal tissue; that is, immediately beneath the epidermis
there are several rows of thick-walled, elongate fibrous cells arranged

Fig. 8-10. A. Underside of a portion of the leaf of an arborescent lycopod from Kansas
showing one of the stomatal bands. B. Stomatal band at a higher magnification.
C. Cross section of a leaf (one side of lamina omitted) found associated with Lepido-
dendron scleroticum (mesophyll not actually preserved); h, fibrous hypodermal tissue;
m, mesophyll; vb, tracheids; ph, phloem; t, transfusion cells. (A,B from a preparation

Telegram @tubscstds
by Gilbert A. Leisman.)
LYCOPODOPHYTA 227

in distinct rows. Between the upper and lower hypodermal layers

there was a loosely compacted tissue which presumably was photo¬
synthetic. Such hypodermal tissue is generally characteristic of
plants found growing under extreme xerophytic conditions. The
stomates were confined to two longitudinal bands on the under side.
These show exceptionally well in maceration preparations of the
epidermis.
The length of the leaves borne by the arborescent lycopods varies
greatly and there is some evidence to suggest that there may be
some variation correlated with growth stages of an individual plant.
In the collections of the British Museum of Natural History there
are several specimens attributed to Lepidodendron lycopodioides
with leaves that measure 13, 15, and 55 mm long respectively; a
specimen of L. sternbergii with leaves 6 cm long and 3.5 mm broad;
a specimen of L. acutum with leaves 20 mm long and about 2.5 mm
broad. Zeiller has given some exceptionally interesting data for
lycopods from the Valencienne coal basin of France: in L. lycopo¬
dioides the leaves were found to be 1 cm long on the small branch-
lets and 3 to 5 cm long on the larger branches; in L. obovatum the
leaves on small branchlets are 4 to 5 cm long while those on larger
ones attain a length of 60 to 80 cm; and in L. dichotomum branch
specimens bear leaves 2 to 6 cm long, while on fossils that presum¬
ably represented the trunk, the leaves reach a length of 40 cm. A
possible explanation for this variation within a species will be given
in a later paragraph dealing with the mode of growth of these plants.
A Lepidodendron specimen has been found in the West Virginia
coal fields with leaves 76 cm long and 5 mm wide and there are re¬
ports in the literature of lycopod leaves up to 1 meter in length
which I am inclined to believe are not exaggerated.
It was noted in the description of the stem anatomy that the peri¬
derm cambium usually lies near the periphery of the stem immedi¬
ately within the leaf bases. This is the opposite of the relationship
that we find in modern (dicot) trees and shrubs in which the peri¬
derm cambium is innermost and produces cork cells toward the
outside. The result is that in Lepidodendron and related genera
the external pattern of leaf cushions remains intact and distinct
throughout the life of the plant (Fig. 8-11). A great many species
have been established on compressions or impressions of the exterior
of the shoot system; some of these are certainly invalid and we still
have inadequate data on the variations within a single plant.
A well-preserved leaf cushion shows the area where the leaf was
actually attached and within this are three scars, a central one rep-

Telegram @tubscstds
228 STUDIES IN PALEOBOTANY

Fig. 8-11. A. Exterior of a large Lepidodendron branch or trunk showing character¬

istic leaf cushion pattern, about 0.5X. B. Diagram of a single leaf cushion; l, leaf scar;
vb, vascular trace scar; up, upper parichnos scars; Ip, lower parichnos scars.

resenting the vascular strand, and two lateral parichnos scars. The
latter represent channels of thin-walled, rather loosely arranged
parenchyma cells which were continuous from the mesophyll of the
leaf through the cushion to the cortical tissues of the stem. In some
lepidodendrons there is another pair of larger scars just below the
leaf scar; these appear to represent the outer surface of channels
of tissue very similar to the parichnos and, like the latter, probably
served an aerating function.

Cones

A great many species of cones have been described which were

borne on the lepidodendrons and related plants; most of these have
been given the name Lepidostrobus. A description of Lepidostrobus
diversus Felix from the middle Pennsylvanian of Indiana will illus¬

Telegram @tubscstds
trate most of the distinctive features of the lepidostrobi. They
LYCOPODOPHYTA 229

(Fig. 8-12) are slender, attaining a maximum diameter of 1 cm and

a length of a little over 11 cm. Many closely imbricated, spirally
arranged appendages (sporophylls) are borne on the central axis.
Each sporophyll consists of a pedicel attached at about 90° to the

Fig. 8-12. Longitudinal section showing representative portions of the cone of

Telegram @tubscstds
Lepidostrobus diversus; distal part of cone is microsporangiate, basal part is mega-
sporangiate, and central region is mixed. (Adapted fiom Felix, 1954.)
230 STUDIES IN PALEOBOTANY

cone axis; it is rather broadly flared toward the distal end and turns
up sharply to form the lamina which is long enough to overlap one
or two sporophylls above. Also at the junction of pedicel and
lamina there is a downward projection known as the heel; although
not present in all lycopod cones it is a conspicuous feature of many
and probably served to enclose the sporangium more effectively
during its maturation process. A single vascular strand, which
originates from the cone stele, passes out through the pedicel and
up into the lamina.
A single radially elongated sporangium is attached to the upper
surface of the pedicel. In the upper portion of the cone each spo¬
rangium contains large numbers (at least several thousands) of
microspores measuring about 20 fi in diameter; the sporangia in the
basal part of the cone each contain 16 megaspores which are over
700 /x in diameter. The central region was transitional with mega-
and microsporangia irregularly mixed.
Immediately distal to the sporangium, there is in many of the
heterosporous lycopods (Fig. 8-20A), including the living Selaginella,
a small tongue-shaped structure called a ligule that is sunken in a
ligular pit. This is also associated with the vegetative leaves and
its former presence is sometimes shown in branch compression speci¬
mens as a minute dot just above the leaf scar. The ligule,
function and origin of which are unknown, is considered to be char¬
acteristic of the heterosporous lycopods and lacking in the homo-
sporous ones; however, its presence has not been demonstrated in
L. diversus and it is my feeling that the attention that has been
devoted to it is quite out of proportion to its importance.
One of the problems encountered in studying cones is illustrated
by this species; if, instead of having a complete one, only basal and
terminal fragments were available several interpretations might be
possible: the two might be taken as parts of a single cone; they
might be interpreted as representing a single species in which mega-
and microsporangia were borne in separate cones; the microspo-
rangiate part might be interpreted as the cone of a homosporous
species and unrelated to the other. The evidence that we have to
date indicates that a great many, if not most, of the arborescent
lycopods of the Carboniferous bore bisporangiate cones that were
generally similar in their organization to Lepidostrobus diversus.
There was, however, a great deal of size variation and some
attained a magnitude of slender baseball bats; where extensive for¬
ests of the plants were established the showers of spores must have

Telegram @tubscstds
LYCOPODOPHYTA 231

been quite impressive at times! The size that some attained is

indicated by the following measurements, based on cones attributed
to Lepidostrobus that are preserved in the collections of the British
Museum and the Geological Survey in London; all are from the
Carboniferous of Britain:

A specimen 13 mm in diameter X 15.5 cm long; complete.

A specimen 25 mm in diameter X 32.5 cm long; diameter very
uniform and broken at one end, thus not complete.
A specimen 30 mm in diameter (with an axis 5 mm in diameter)
X 16 cm long; not complete.
A specimen 50 mm in diameter (with an axis 10 mm in diameter)
X 19 cm long; a fragment only.
A specimen 22 mm in diameter with an axis 4 mm in diameter;
length 28 cm and probably complete.

Nemejc has described lycopod cone compressions from the Car¬

boniferous of Bohemia under the name Sporangiostrobus which
attained a diameter of 6 cm. Macerations revealed microspores
50 jx in diameter and the megaspores measured 2.5 mm. The affini¬
ties of these cones have not been established and it is not necessarily
implied that they belong in the Lepidodendraceae; they are noted
here to indicate the dimensions attained in the lycopods in both cone
and megaspore size.
Among the numerous species of heterosporous lycopod cones in
the Carboniferous there is a distinct evolutionary sequence involv¬
ing: reduction in the number of megaspores per sporangium; increase
in size of the megaspores that remain; enclosure of the megaspo¬
rangium by the sporophyll to form a seed. Our knowledge is not
sufficiently complete so that we can cite an unbroken evolutionary
line, but the general tendency is illustrated by the following
examples:
Lepidostrobus noei Mathews from the lower Mississippian of
Kentucky is a cone in which each megasporangium contains several
hundred spores, each about 350 n in diameter, whereas the micro¬
spores, produced in huge quantities, are 50 [x in diameter.
L. diversus, described above, has 16 megaspores per sporangium.
L. foliaceus Maslen from the Pennsylvanian is reported to have
not more than four megaspores per sporangium.
L. braidwoodensis Arnold is based on a compression specimen
from Illinois; maceration of the sporangia revealed, in each one, a
single large megaspore about 2 mm in diameter. Three much

Telegram @tubscstds
232 STUDIES IN PALEOBOTANY

smaller undeveloped spores are found at the apex of the large one;
thus, of a single tetrad within each megasporangium, only one spore
matured.
Lepidocarpon is a genus established by D. H. Scott in 1901 for
megasporangiate cones that are similar in their basic organization
to Lepidostrobus but in which the sporangium is almost completely
enclosed by the sporophyll. In the course of development of the
cone the sporophyll grew laterally up around the sporangium which
is a radially elongate sac but differs from Lepidostrobus in that it
is broad at the base and tapers upward to an acute apex. In a sec¬
tion tangential (Fig. 8-13) to the axis of the cone the unique mor¬
phology of Lepidocarpon is most clearly revealed; the sporophyll
encloses the sporangium except for a radially elongate slit or “micro-

Fig. 8-13. Tangential sections of: A. sporophyll of a Lepidostrobus (microsporangiate)

cone, 20X; B. a seed of Lepidocarpon magnificum, 9X; s, sporophyll; sp, sporangium
wall; g, gametophyte. B is a composite photograph, the gametophyte having been

Telegram @tubscstds
taken from another specimen.
LYCOPODOPHYTA 233

Fig. 8-14. A portion of the branch system of an arborescent lycopod (probably

Lepidophloios sp.) from West Virginia; the cones on the short side branches at the
left are approximately 10 cm long.

pyle” at the apex. A single megaspore matured in each sporangium

and specimens have been found in English and American coal balls
with the gametophyte preserved.
So far as I am aware Scott’s original specimen is the only one on
record with the seeds attached to the cone axis; they are abundant
enough as isolated specimens and apparently were shed from the
cone axis by an abscission mechanism. As to the corresponding
microsporangia, the available evidence suggest that they were borne
on separate cones. Large microsporangiate cones have been found
associated with an Illinois species, Lepidocarpon magnificum An¬
drews and Pannell; they are 5 cm in diameter and in excess of 16
cm long. Spores, identical with those in the microsporangia, were
also found in the seeds. It is thus likely that many of the purely
microsporangiate cones found in the Upper Carboniferous belong
to Lepidocarpon.
In summary, Lepidocarpon is a seed in the sense that it is an
integumented megasporangium and, at least with respect to this

Telegram @tubscstds
234 STUDIES IN PALEOBOTANY

organ, it represents the peak of evolution attained by the lycopods.

However, it is quite apparent that the “integument” here is not
homologous with the integument of the seed plants considered in
later chapters. There is no reason to doubt that Lepidocarpon is
terminal in a line of lycopod evolution which became extinct toward
the close of the Paleozoic.
The arborescent lycopod stem compressions, some of which are
definitely referable to Lepidophloios (Fig. 8-15), occasionally have
conspicuous round scars (Fig. 8-16) the significance of which has
been much debated. Some investigators have held that they rep¬
resent the scars of vegetative branches that were sloughed off by
an abscission mechanism, possibly similar to the living cottonwoods,
whereas others have been of the opinion that they represent spe¬
cialized cone-bearing branches. Unfortunately, distinct generic
names have been established for such specimens; Halonia being
applied to ones in which the scars are arranged in a spiral pattern
and Ulodendron for those in which the scars are arranged on two
opposite sides of the branch.
There are actually very few specimens of Carboniferous lycopods
known with the cones attached and I feel that any generalized state¬
ment concerning the halonial or ulodendroid scars is premature; the

Fig. 8-15. Lepidophloios scoticus from

the Lower Carboniferous of Scotland,
about 0.5X. (Kidston collection, Geo¬
logical Survey Great Britain.)

Telegram @tubscstds
 LYCOPODOPHYTA 235

Fig. 8-16. Lepidophloios scoticus. The regularly spaced round scars probably repre¬
sent points where small cone branches were attached. (British Crown copyright, by
permission of the Controller of her Britannic Majesty’s Stationery Office.)

following comments will be confined to a few specimens in which

the evidence seems clear-cut. Some years ago I obtained a fine
specimen of the terminal part of a Lepidophloios branch system
from the West Virginia coal fields. The dichotomizing branches are
close to 2 cm in diameter and bear leaves at their tips that were
probably several decimeters long. Pendant side branches about 9

Telegram @tubscstds
mm in diameter are borne, about 8 cm apart, and each one is ter-
236 STUDIES IN PALEOBOTANY

minated by a cone, the largest being 3.5 cm in diameter and 15 cm

long. Petrified Lepidophloios stems found in Kansas coal balls shed
some light on the anatomy of these fertile branches. Small terete
vascular strands depart from the stele of the main branch (appar¬
ently a very unequal dichotomy of the latter) and pass out to the
periphery where they terminate in the ulodendroid scars. It is
therefore established that in some Lepidophloios species cones were
borne on relatively small special branches and these were probably
shed by an abscission mechanism after discharge of the spores.

Root System

The underground organs of Lepidodendron and closely allied

plants are assigned to the genus Stigmaria; they will be referred to
as a “root” system and although there is no doubt that they func¬
tioned as such, in their gross morphology and anatomy they are
quite unlike the roots of other plants.
Stump casts are not uncommon in Upper Carboniferous rocks and
perhaps the finest example is an exhibit in Victoria Park, Glasgow
(Fig. 8-17). In the course of developing the park a path was cut
through what had once been a quarry and several stumps were dis¬
covered; subsequent excavation was carried out to expose more and
a building erected for protection against the weather. The stumps
clearly reveal a sequence of events beginning with a lycopod forest
that flourished in the region in Carboniferous times; it was a low-
lying area and subject to flooding, and probably as a result of land
subsidence several feet of mud accumulated, bringing about the
death of the trees. The remaining stumps eventually decayed but
not until the surrounding matrix had consolidated somewhat; then
the cavities were filled with mud, resulting in casts of the original
stems and roots. More sediments accumulated later and this was
followed by an uplift that raised the area to become dry land.
Finally, through the curiosity of man the remnants of the long
buried forest have been revealed.
In most cases the root system originated at the base of the trunk
as four massive primary roots, the largest known being 80 cm in
diameter; these radiate out horizontally or dipped down slightly into
the ground attaining a length of 14 meters. They dichotomized at
least twice, but the divisions took place within a few feet of the
trunk. This stigmarian rootstock in turn bore appendages (root¬
lets) which were arranged in a very regular (quincuncial) pattern

Telegram @tubscstds
(Fig. 8-18) which may be interpreted as a spiral in which the root-
LYCOPODOPHYTA 237

lets of successive spirals form nearly longitudinal rows. The rootlets

occasionally dichotomized and reached a length of 37 cm.
The main rootstock of Stigmaria ficoides contained a stele (Fig.
8-19) of radially aligned tracheids. A few of the innermost cells are
spiral tracheids, whereas the rest are scalariform; thus the wood is
almost exclusively secondary. It may be added that several species
of Stigmaria have been recognized and in one or two of them some
primary centripetal wood (as in the stems) was present. The wood
was followed by a broad cortical region which is usually decayed
and finally there is, in larger specimens, a conspicuous band of peri¬
derm. The vascular trace to the rootlets departs from the inner-

Fig. 8-17. Stump casts of arborescent lycopods excavated from Carboniferous rocks

Telegram @tubscstds
in Victoria Park, Glasgow.
238 STUDIES IN PALEOBOTANY

Fig. 8-18. Stigmaria ficoides, from the Coal Measures of Newcastle, England; portion
of a rootstock with rootlets. Length of longest rootlet shown is 21 cm. (British Mu¬
seum Natural History.)

most part of the root stele and passes horizontally through a broad
ray in the wood.
Close to their point of attachment to the rootstock the rootlets
appear in cross section as follows: there is a broad outer cortex fol¬
lowed by a rather clearly delimited middle one; the cells of the latter
are smaller, less regular in shape, and there are some air spaces be¬
tween them. There is next an inner cortex enclosing the vascular
strand which is monarch, that is, there is one protoxylem point, and
occasionally a few rows of secondary wood cells are present.
Within a few millimeters of their departure from the rootstock
(Fig. 8-19B) the rootlets lose their middle cortex; a conspicuous gap
is thus evident between the outer and inner cortical layers. Occa¬
sionally a connective may be observed which is a remnant of the
middle cortex and joins the inner cortex, with the vascular strand,
to the outer cortex.
The stigmarian root system presents several unusual features
which are difficult of interpretation. The stele differs from that of
the stem in the lack of any appreciable amount of centripetal pri¬

Telegram @tubscstds
mary wood (it would be interesting to find a petrified specimen
LYCOPODOPHYTA 239

showing the transition between the base of the trunk and the root-
stock). The regular arrangement of the rootlets is quite unlike the
irregular distribution of branch rootlets in most plants. A feature
of special physiological interest is the lack of root hairs; in view of
the abundance and excellent preservation of these fossils in English
coal balls it seems certain that specimens with root hairs would have
been found if they did exist. A rough comparison with the root sys-

B C

Fig. 8-19. A. The stele of a Stigmaria rootstock, 2X. B. A rootlet in cross section,
7X; ic, inner cortex; oc, outer cortex; the vascular strand is the minute cluster of cells
within the inner cortex. C. Inner cortex and vascular strand highly enlarged; a few

Telegram @tubscstds
secondary tracheids are evident.
240 STUDIES IN PALEOBOTANY

tem of modern seed plants of comparable size suggests that the

stigmarian root system had a very low absorption area; this, coupled
with the small amount of wood, seems to imply a conducting system
of low efficiency and explains at least in part the xerophytic nature
of the leaves which are constructed to lower drastically the tran¬
spiration rate.
The lycopods had reached arborescent dimensions by Upper De¬
vonian times although none are known that were as large as the
Carboniferous forest giants. In searching through museum collec¬
tions for a representative or two that might bridge the gap between
the small plants of the early to mid-Devonian and the great forest
trees of the Carboniferous it seemed to me that Bothrodendron
kiltorkense in some way meets the need. This plant was first found
in the Upper Devonian of Kiltorcan Hill, County Kilkenny, Ireland,
and has also been reported from Bear Island. In its gross habit it
seems to have been similar to the later lepidodendrons; the main
stem is known to have attained a diameter of 1 foot and rose from
a stigmarian rootstock; equal dichotomies characterized the branch¬
ing pattern of the aerial shoot system and the plant attained a
height of 25 feet. The branches seem to have been exceptionally
slender, to a point that the plant probably presented a delicate,
almost filmy appearance in life. The leaf scars are small, nearly
round, and in many specimens they are arranged in a perfectly
whorled pattern, whereas in others it appears to be a very low angle
spiral. The leaves measure 2 to 15 cm long and 1 mm wide.
The cones are imperfectly known from compressions but present
a few features that are interesting enough to make one wish for bet¬
ter preserved material. They were borne terminally and consisted
of whorls of sporophylls, each with an elongate sporangium on the
upper surface, whereas the distal portion of the sporophyll, accord¬
ing to Johnson, attained a length of 20 cm. There is a rather
remarkable specimen in the Museum in Stockholm labeled Lepidos-
trobus bailyanus from Kiltorcan which is probably a cone of B.
kiltorkense; its most striking feature is the long slender sporophyll
which reaches a length of 9 cm. There is thus a resemblance in size
between the sporophylls and ordinary leaves and, in life, there
probably was no conspicuous distinction between the sterile and
fertile branch tips. Johnson was able to observe 10 to 20 mega¬
spores in some of the sporangia and, in others, smaller bodies which
are thought to have been microspores.
Some years ago Watson described a cone from the Upper Carbon¬

Telegram @tubscstds
iferous of England that was believed to have been borne on the stems
LYCOPODOPHYTA 241

of Bothrodendron mundum; since the cone was not actually found

attached it was given the name Bothrostrobus. It is a heterospor-
ous cone about 8 mm in diameter (Fig. 8-20); the sporangia are
nearly globose, being slightly taller than broad and are further char¬
acterized by large ligules; not more than four megaspores were
present in each sporangium.

Mode of Growth in the Arborescent Lycopods

In the foregoing discussion I have tried to point out some of the

more important features of the lycopods as well as some of the un¬
solved problems; one of the most intriguing of these is the mode of
growth of these strange trees.
Restorations of plants assigned to Lepidodendron and Lepi-
dophloios show them as trees of great height with a columnar trunk
that dichotomized to form a dense crown above. Scott, in his
Studies in Fossil Botany, cites a trunk found in the Lancashire coal
fields that was 114 feet long from the base up to the first branch¬
ing. This is probably an exceptionally tall specimen but it empha¬
sizes a distinctive feature of the plants, that is, the great height that
was reached prior to the initiation of branching. Some light is shed
on this by a recent discovery of Lepidophloios stems from Kansas

Fig. 8-20. A. Bothrostrobus mundus, diagrammatic median section of cone; l, ligule;

me, megasporangium; mi, microsporangium. B. Spencerites insignis, portion of a cone
showing one sporophyll with distally attached sporangium. (A from Watson, 1907;

Telegram @tubscstds
B from Berridge, 1905.)
242 STUDIES IN PALEOBOTANY

coal balls which are approximately 12 cm in diameter and composed

almost wholly of primary tissues. There is no secondary wood and
only a narrow band of periderm 3 to 5 mm thick which was formed
deep within the cortex. It is evident that these stem or branch
specimens grew by means of a rather large apical meristem and, as¬
suming a mode of growth comparable with that in the modem
cycads and tree ferns, a very tall unbranched trunk could have been
formed. But in the case of the lycopods an obvious problem cor¬
related with this mode of growth is the matter of the photosynthetic
tissue available to the plant. It is difficult to understand how a
trunk could have grown to a height of many feet, actually well over
100 in some cases, with only the small leaves a few centimeters long
that seem to have been borne by many of these plants (Fig. 8-9).
If, however, the trees in this initial stage of growth were clothed
with leaves up to a meter in length it is more readily understood.
There is evidence cited above from Zeiller’s Valencienne speci¬
mens that some of the lycopods may have started with such long
leaves and they became progressively smaller as the branch orders
decreased in size.
In summary, it has been suggested by Andrews and Murdy that
in at least some of the lycopods the trunk attained a fairly large
diameter at an early stage in growth and developed as a uniform
columnar trunk up to 100 feet or more (Fig. 8-4B). The apical
meristem then divided to form two branches which may have been
equal or unequal, but in either case successive branch orders became
progressively smaller, reaching a minimum diameter, at which time
growth ceased. There was probably considerable variation in dif¬
ferent species, particularly in relation to the amount of secondary
growth.

SIGILLARIACEAE

The sigillarias were Carboniferous lycopods with certain basic

points of similarity to the lepidodendrons; the two diverge, however,
in several ways, including features of the leaf cushion pattern and
cone structure.
The arrangement of the leaf cushions in Sigillaria (Fig. 8-21)
appears to be a spiral one as in Lepidodendron, but the development
of distinct longitudinal ridges results in an alignment of the cush¬
ions in vertical rows. The gross outline of the leaf cushion is

Telegram @tubscstds
hexagonal, round, or oval.
LYCOPODOPHYTA 243

A B
Fig. 8-21. A. Sigillaria mammillaris Brongniart, portion of a trunk, about 0.3X.
B. Sigillaria coriacea Kidston, fragment of stem showing details of leaf scars. (British
Museum Natural History.)

Sigillaria was certainly not uncommon, for stem impressions are

widely distributed geographically and stratigraphically through the
Upper Carboniferous; great trunk casts are exposed in the cliffs
along the Nova Scotia coast at Joggins and I had occasion several
summers ago to visit an open-pit coal mine in northern Pennsylvania
where at one point we found vast quantities of sigillarian stem
impressions of various branch orders. Thus, although the plants
are abundant as compressions or casts, petrified stems are conspicu¬
ous by their scarcity. The great English paleobotanist Williamson
remarked on this rather forcefully in 1872 and his impression holds
good today:
Considering the abundance of Sigillariae in the coal-measures, it is mar¬
velous that indisputable specimens displaying their internal organization
should be so rare; but such is the case. After years of search I have only met
with three specimens of the Sigillarian character of which there can be no
doubt.

The lack of petrified stems is perhaps, in part, apparent rather

than real; so far as I am aware there are no reliable anatomical cri¬
teria for distinguishing sigillarian stems from those of the lepidoden-
drons when the characteristic leaf bases are not preserved.
Adding to the uncertainty of stem identification is the problem

Telegram @tubscstds
of correlating these organs with isolated petrified cones. The most
244 STUDIES IN PALEOBOTANY

informative specimens that are considered referable to Sigillaria are

those of Mazocarpon oedipternum Schopf from the Calhoun coal-
ball horizon (upper Pennsylvanian) of Illinois. It is a heterosporous
species with the megaspores and microspores borne in separate
cones.
The sporophylls of the megasporangiate cones (Fig. 8-22) were
arranged in a low spiral or verticillate order and they are distin¬
guished in this species by a conspicuous heel. There are eight
megaspores in each sporangium, being arranged in a peripheral posi¬
tion and surrounded by parenchymatous tissue. Several mega¬
spores have been found with fully developed female gametophytes,
each with a single archegonium. Large numbers of microspores
were produced in each sporangium of the microsporangiate cones;
the unique parenchymatous tissue of the megasporangia is not
present here.

Fig. 8-22. Mazocarpon oedipternum. A. Radial

longitudinal section through a portion of mega¬
sporangiate cone; B. a single megasporophyll in
tangential view, about 4X. (From Schopf, 1941.)

Telegram @tubscstds
LYCOPODOPHYTA 245

Another genus of cones, Sigillariostrobus, known from compres¬

sion specimens is also referred to the Sigillariaceae. These vary in
different species from 15 to 25 cm long and from 1.5 to 2.5 cm in
diameter. They were borne singly at the end of a pedicel or
specialized branch. One of the reasons for regarding these cones
as sigillarian is the discovery by Zeiller, in 1884, of specimens in
which the pedicel displayed a leaf cushion structure comparable
with that of the stems.
They are heterosporous with mega- and microsporangia borne in
separate cones. In three Polish species described by Bochenski it
was found that 12 megaspores were present in each megasporang¬
ium and the megaspores were large, exceeding 2 mm in diameter.
Thousands of microspores were present in each microsporangium.
The sporangia were apparently shed from the cone axis as they
approached maturity as many specimens have been described
which retain their sporangia only in the basal region. The mega-
sporangiate cones sometimes present a very distinctive appearance
as compression fossils in that the general outline of the cone may
be retained perfectly, but apparently the only structures preserved
are the large megaspores. Thus, at first glance the cones give the
appearance of being simply a large mass of megaspores (Fig. 8-23).
This unique mode of preservation has been explained as a result of
the reduction, to a thin, clear film, of the parenchymatous central
cushion around which the megaspores are clustered. Thus the
more resistant megaspores are held together in very nearly their
original position, whereas the remains of the parenchymatous
tissue are not visible except by special treatment.

HERBACEOUS LYCOPODS

Not all of the Carboniferous lycopods were large trees. Fossil

specimens characterized by frequent branching, leaves only a few
millimeters long, and minute cones are known from Paleozoic, as
well as Mesozoic rocks, that were small, low-growing herbaceous
plants. Some of these were so similar in appearance to the extant
Selaginella that they have been referred to that genus; others have
been described under the names Selaginellites and Lycopodites.
Owing to considerable variation in the quality of preservation, the
fact that some specimens are fertile and others not, and to the
divergent views of different investigators, the nomenclatural his-

Telegram @tubscstds
246 STUDIES IN PALEOBOTANY

Fig. 8-23. Compression specimens of part of a

megasporangiate Sigillariostrobus from the Upper
Carboniferous of Pennsylvania, about 2X.

tory of the fossil herbaceous lycopods offers a state of confusion

that is impressive even for paleontological tangles. To the best of
this writer’s understanding the following is a brief summary of the
usage of the two names cited above:

Telegram @tubscstds
Selaginellites was instituted by Zeiller in 1906 for small, appar¬
ently herbaceous plants bearing cones that are clearly heterosporous.
LYCOPODOPHYTA 247

Lycopodites has been employed by most authors for plants of

similar diminutive size, with minute leaves resembling those of the
modern Lycopodium, and which presumably were homosporous.
So far as I am aware, however, no fossil specimens referred to
Lycopodites have actually been proven to be homosporous. Thus
in practice it has been used simply for presumably herbaceous
lycopods that have not been clearly established as representing
homo- or heterosporous plants.

In the following paragraphs a few of the better known fossils are

introduced.
Selaginellites crassicinctus Hoskins and Abbott, from the Des
Moines series (Pennsylvanian) of Kansas, is based on a cone with
an over-all diameter of 5 mm. The sporophylls are about 3 mm
long and bear megasporangia toward the basal part of the cone
and both megasporangia and microsporangia in the upper region.
The megasporangia usually contain four spores which have a
characteristic equatorial flange and a maximum diameter of about
800 /x; large numbers of microspores, each about 80 /x in diameter,
are contained in the microsporangia. A small ligule not more than
225 /x long is present.
Selaginellites canonbiensis Chaloner is a cone from the Upper
Carboniferous of Scotland and is 3.5 mm in diameter; the lower
portion is megasporangiate and the upper microsporangiate. There
were four or possibly more megaspores, each 630 ft in diameter, in
a megasporangium. The microspores measured about 43 /x in
diameter.
Selaginellites polaris Lundblad from the Triassic of East Green¬
land is known from rather large cones measuring about 7 mm in
diameter; the megasporangia contained numerous spores each
412 ft in diameter, whereas the microspores average about 40 /x.
As noted above, where the comparison has seemed sufficiently
close to justify it, some authors have assigned these heterosporous
herbaceous lycopods to the genus Selaginella. Miss Lundblad has
described specimens showing both cones and vegetative parts from
the Rhaetic of Scania, Sweden as Selaginella hallei. The foliage
is dimorphic, the leaves on the lower side of the stems being 1 to 2
mm long, and those on the upper side are closely adpressed and
about half that length. The cone is only 5 mm long and 1 mm in
diameter and apparently very similar in construction to that of a
living Selaginella. The megasporangia contained four spores, each
about 426 ft in diameter and the microspores are 50 ft in diameter.

Telegram @tubscstds
Selaginella amesiana Darrah is a Pennsylvanian specimen that
is remarkable for the preservation of the female gametophytes. It
248 STUDIES IN PALEOBOTANY

is a partially petrified cone found in a nodule from the famous

Mazon Creek area of northern Illinois. The cone is 3.5 mm in
diameter and the portion preserved includes many megasporangia;
these contain four megaspores each and in many of them the
female gametophyte is well preserved. The nuclei are clearly
defined and in some cells structures can be discerned which appear
to be chromosomes.
Thus the evidence seems to suggest rather strongly that the
modern Selaginella and possibly Lycopodium are plants of very
ancient lineage. The similarity between some of the fossil species
and the modern selaginellas implies close affinity, so close in fact
that it seems reasonable to regard the latter as stemming directly
from Carboniferous forms such as have been described.

PLEUROMEIACEAE

Pleuromeia sternbergi (Munster) Cor da represents a curious line

of Lower Triassic lycopods possibly related to the living Isoetes.
The stem (Fig. 8-24) is unbranched and attained a height of 2
meters and a maximum diameter of 9 cm. The leaves, which are
not as densely arranged on the stem as in the Carboniferous
lycopods, are 11 cm long and about 1.5 cm broad except for the
broadly flaring basal portion. A ligule was present and two vascu¬
lar bundles traverse the length of the leaf. As the trunk increased
in height the leaves were shed from the basal region.
The base of the plant flared out into a four-lobed root system
with the four points turned upward; numerous rootlets are borne
over the outer surface.
The stem is terminated at the top by a massive spike of sporo-
phylls which may be termed a cone. Although complete specimens
have not been found, the fragments bear either microspores or
megaspores; thus the plant was heterosporous and apparently
dioecious. The megasporangia are large, being about 2 cm broad
and contained megaspores 500 to 700 ji in diameter; the micro¬
sporangia are about 1.2 cm broad and the microspores measure 15
to 25 n in diameter.
According to Seward the original specimen on which Munster’s
description was based came from a split stone taken from the tower
of Magdeburg Cathedral, an unusual paleobotanical collecting site!
The sporangia are reported as borne on the under (abaxial) side

Telegram @tubscstds
of the sporophylls, although this has been questioned by certain
 LYCOPODOPHYTA 249

Fig. 8-24. A. Restoration of Pleuromeia sternbergi. B. Diagrammatic sketch of

Stylites gemmifera. (A from Hirmer, 1933; B from Rauh and Falk, 1959.)

authors. The abaxial position has been explained as possibly

resulting from an ancestral type in which sporangia were peltately
arranged, those on the upper (adaxial) surface having been lost.
Some support for this theory may be found in the Upper Car¬
boniferous Spencerites insignis (Williamson) Scott; this seems
properly regarded as a lycopod cone but it is distinctive in certain
characters. The axis contains a typical lycopod stele and the

Telegram @tubscstds
sporophylls are arranged in alternating verticils of ten to a whorl.
250 STUDIES IN PALEOBOTANY

Each sporophyll is constructed of a rather slender pedicel and an

upturned lamina (Fig. 8-20B). The distal part of the pedicel ex¬
pands into upper and lower fleshy lobes with a single sporangium
attached to the upper one. This presents a contrast to other
lycopod cones in which the sporangium is attached either near the
proximal end of the pedicel or along the greater length of it. No
ligule has been observed. The spores have a characteristic equa¬
torial wing and measure about 280 n in diameter, a size that is
problematical in that most microspores are much smaller than this
and megaspores are usually somewhat larger. It seems most likely
that the portion known is simply the megasporangiate portion of
the plant. The systematic position of Spencerites is uncertain and
it was introduced here because of the pads or lobes of the pedicel;
although only the upper one bears a sporangium, the symmetry
suggests that formerly one might have been borne on the under
side as well.
The theory of an ancestral type with peltate sporangiophores
from which forms like Pleuromeia and Spencerites might have
evolved is not, however, supported by any pre-Carboniferous
evidence.

SUMMARY

The lycopods are an ancient line of plants in which the distinc¬

tive characters were acquired in the Silurian, or possibly pre-
Silurian times. The record seems to indicate that the group
gained evolutionary momentum in the mid-Devonian and was
already quite diverse by the close of that geologic period. Arbores¬
cent forms had then made their appearance and the forest trees of
the Carboniferous were huge by any standard. There is sufficient
evidence to indicate that small herbs have existed throughout the
racial life span of the group; Selaginella-like plants are known
from the Carboniferous on up through the Mesozoic to the present
and there seems to be no reason to doubt that this genus, as we
know it today, has existed with little change for some 250 million
years. It also seems to me more than likely that Lycopodium may
stem back directly to the Silurian Baragwanathia, but the fossil
record of plants that can be closely compared with Lycopodium is
a scanty one.
Several lines of arborescent lycopods flourished in the Carboni¬

Telegram @tubscstds
ferous and these are probably derived from a common stock in the
LYCOPODOPHYTA 251

Devonian. Some evidence for this is found in Archaeosigillaria

primaeva, described by David White in 1907, from the Upper
Devonian of New York. When originally excavated the specimen
measured some 5 meters in length and consisted of a rather bulb¬
ous base 38.5 cm in diameter and shortly above this the trunk
tapered gradually to a diameter of 12 cm at the upper end. Its
chief point of interest lies in the nature of the leaf cushions which,
in the lower third of the trunk are sigillarian in form, whereas in
the central part they become more lepidodendroid and distinctly so in
the upper portion. It thus appears to combine features of Sigil-
laria and Lepidodendron as these genera are known in typical
Carboniferous specimens. The leaves were narrow, not over 3 cm
long and apparently persistent in their attachment since specimens
are found as low as 70 cm from the base.
An abundance of stem remains in coal ball petrifactions attests
to the dominance of the arborescent lycopods in the Upper Car¬
boniferous; at one locality in Illinois approximately two-thirds of
the plant material is Lepidodendron and in the southeastern
Kansas coal fields the petrifactions contain large quantities of
steles and periderm. An unusual bit of evidence bearing on the
abundance of the lycopods, as well as a distinctive type of plant
preservation, is recorded by Scott (1920, p. 182) in a description of
the Russian paper-coal:

At Tovarkovo, in the province of Toula, in Central Russia, beds of a

peculiar kind of coal, called leaf-coal or paper-coal, have long been known.
The seams are about 8 inches thick, and lie near the surface of the ground,
only covered by sand. The so-called coal has all the appearance of a bed of
excessively thin dead leaves, intermixed only with structureless organic mat¬
ter of the nature of humic acid. The leafy films, on investigation, have
proved to consist entirely of layers of cuticles, belonging to the stems of
ancient plants, from which all other tissues had rotted away, ages before.
The cuticles are perfectly fresh and pliable, and not in any way fossilized,
although geologists are agreed that the bed (which covers an area of many
square miles) belongs to the Carboniferous Limestone horizon, in the lower
part of the Carboniferous formation. The cuticle is in some cases complete,
corresponding to the whole circumference of the stem which it once enclosed;
it is perforated by numerous small round holes, regularly disposed, and cor¬
responding in arrangement with the leaf-traces of a Bothrodendron.

As to the acquisition of the seed habit, Lepidocarpon was not

the only one to achieve this level of development. A similar organ
has been described for Miadesmia membranacea Bertrand although
this was probably a plant of much smaller habit; the seed here dis¬

Telegram @tubscstds
played a characteristic fringed integument and a more attenuate
252 STUDIES IN PALEOBOTANY

micropyle. It seems to me that Miadesmia and Lepidocarpon

are sufficiently distinct as to suggest that they evolved independ¬
ently as seeds from separate lines of lycopods. There is of course
no reason to suppose that these plants were at all closely related
to the “true” gymnospermous groups such as the pteridosperms,
cordaites, and cycadophytes; the seed integument here is certainly
not homologous with that of the two lycopod seeds.
The arborescent representatives seem to have declined rapidly
toward the end of the Paleozoic. This decline, however, may not
have been as sudden as the macrofossil record has indicated. For
example, Harris has described 13 new species of Triletes from the
Rhaetic rocks of East Greenland; this is a genus of spores generally
attributed to the lycopods and the Greenland megaspores range in
size from 250 ja to 1400 /a.
Pleuromeia has been the subject of considerable speculation as
to its position in the lycopod clan. Some interesting information
bearing on this problem has come from the discovery of a new
genus of living lycopods, Stylites, assigned to the Isoetaceae. Be¬
fore dealing with this directly it seems appropriate to add a few
notations on the fossil history of Isoetes (quillwort).
The quillworts are among the most distinctive of all pterido-
phytes; there are several scores of species found in aquatic environ¬
ments or places that are moist for part of the year. The axis of
the plant is extremely short, bearing abundant roots on the under
side, and a dense rosette of long, strap-shaped leaves above. Other
than those formed in the early part of the growing season, most of
the leaves are fertile, bearing a large sporangium on the upper sur¬
face near the base. Both mega- and microsporangia occur on the
same plant although the two may be developed in different seasons,
according to Pfeiffer.
Isoetes is a genus of some antiquity; R. W. Brown has reviewed
the American specimens and recognizes two species:
I. horridus (Dawson) Brown from the Paleocene of Wyoming
seems identical in habit with modern species; the dense rosette of
sporophylls and casts of the sporangia and spores leave no doubt
as to the identity of the fossil. Other specimens attributed to it
have been found in Lower and Upper Cretaceous horizons.
Another species, I. serratus Brown, although not as well pre¬
served, was found in the Frontier formation (Upper Cretaceous) of
southwestern Wyoming. The implication is thus quite clear that
Isoetes had evolved to essentially its present morphology by mid-

Telegram @tubscstds
Mesozoic times.
LYCOPODOPHYTA 253

Returning to Pleuromeia, one is tempted to interpret it as a

diminutive descendent of the arborescent lycopods of the Carboni¬
ferous, one in which branching of the subterranean system is much
reduced. Isoetes might correspondingly be regarded as a lycopod in
which the entire shoot and root system is reduced to a point
where there is very little of either left. Fitting into this picture is
a Lower Cretaceous plant Nathorstiana arborea Richter (see
Magdefrau, 1932) which has been described as an elongated Isoetes.
It is a small plant with an erect stem a few centimeters long which
is unbranched.
A few years ago a new genus of living isoetaceous plants was dis¬
covered which adds a significant fragment to this particular picture
puzzle of lycopod evolution. Two species have been described,
Stylites andicola Amstutz and S. gemmifera Rauh, both of which
were found growing around the boggy margins of a lake in central
Peru at an elevation of 4750 meters. Although Isoetes rarely
branches, the plants of Stylites gemmifera dichotomize at least
three times (Fig. 8-24B). Rosettes of leaves, apparently very like
those of Isoetes, are found at the apex of each branch whereas the
“caudex” bears fleshy roots along one side only. Thus, although
Pleuromeia-Nathorstiana-Stylites-Isoetes do pot present an un¬
broken sequence they present significant links in what appears to
have been a fine of evolution in which the axis became progressively
reduced to the extreme state found in Isoetes.
The lycopods are an ancient group of plants including a diverse
assemblage in respect to size and general morphology; growth
mechanics was unique and is still poorly understood in the arbores¬
cent forms. Like most of the other major divisions of plants, we
are only beginning to unravel their past history.

REFERENCES

Amstutz, Erika. 1957. Stylites, a new genus of Isoetaceae. Ann. Missouri Bot.
Gard., 44: 121-123.
Andrews, Henry N., Jr., and Murdy, William H. 1958. Lepidophloios—and ontogeny
in arborescent lycopods. Amer. Journ. Bot., 45: 552-560.
-and Panned, E. 1942. Contributions to our knowledge of American Car¬
boniferous floras II. Lepidocarpon. Ann. Missouri Bot. Gard., 29: 19-28.
Arnold, Chester A. 1938. Note on a lepidophyte strobilus containing large spores,
from Braidwood, Illinois. Amer. Mid. Nat. 20: 709-712.
_. 1940. Lepidodendron johnsonii, sp. nov., from the Lower Pennsylvanian

Telegram @tubscstds
of Central Colorado. Univ. Michigan, Contrib. Mus. Paleont., 6: 21-52.
254 STUDIES IN PALEOBOTANY

Banks, Harlan P. 1944. A new Devonian lycopod genus from southeastern New York.
Amer. Journ. Bot., 31: 649-659.
Beck, Charles B. 1958. “Levicaulis arranensis,” gen. et sp. nov., a lycopsid axis from
the Lower Carboniferous of Scotland. Trans. Roy. Soc. Edinburgh, 63: 445-456.
Benson, Margaret. 1908. Miadesmia membranacea, Bertrand; a new Palaeozoic
lycopod with a seed-like structure. Phil. Trans. Roy. Soc. London, 199B: 409-425.
Berridge, E. M. 1905. On two new specimens of Spencerites insignis. Ann. Bot.,
19: 273-279.
Bochenski, Tadeusz A. 1939. On the structure of Sigillarian cones and the mode of
their association with their stems. Polish Acad. Sci., 7: 1-16.
Brown, Roland W. 1939. Some American fossil plants belonging to the Isoetales.
Journ. Washington Acad. Sci., 29: 261-269.
Chaloner, William G. 1958. A Carboniferous Selaginellites with Densosporites micro¬
spores. Palaeontology, 1: 245-253.
Darrah, William C. 1938. A remarkable fossil Selaginella with preserved female
gametophytes. Harvard Univ. Bot. Mus. Leaf., 6: 113-135.
Dijkstra, S. J. 1956. Lower Carboniferous megaspores. Mededelingen Geol. Sticht-
ing (Heerlen), n.s., No. 10: 1-18.
-and Pierart, P. 1957. Lower Carboniferous megaspores from the Moscow
Basin. Mededelingen Geol. Stichting (Heerlen), n.s. No. 11: 5-19.
Felix, Charles J. 1952. A study of the arborescent lycopods of southeastern Kansas.
Ann. Missouri Bot. Card., 39: 263-288.
-. 1954. Some American arborescent lycopod fructifications. Ann. Mis¬
souri Bot. Card., 41: 351-394.
Graham, Roy. 1935. Pennsylvanian flora of Illinois as revealed in coal balls. II. Bot.
Gaz., 97: 156-168.
Hirmer, Max. 1933. Rekonstruktion von Pleuromeia sternbergi Corda, nebst bemer-
kungen zur Morphologie der Lycopodiales. Palaeontographica, 78B: 47-56.
Hoskins, John H. and Abott, Maxine L. 1956. Selaginellites crassicinctus, a new
species from the Desmoinsian series of Kansas. Amer. Journ. Bot., 43: 36-46.
Johnson, Thomas. 1913. On Bothrodendron (Cyclostigma) Kiltorkense Haughton sp.
Sci. Proceed. Roy. Dublin Soc., n.s. 13: 500-528.
Kisch, Mabel H. 1913. The physiological anatomy of the periderm of fossil lycopod¬
iales. Ann. Bot., 27: 281-320.
Krausel, Richard, and Hermann Weyland. 1935a. Neue Pflanzenfunds in Rheinis-
chen Unterdevon. Palaeontographica, 80B: 171-190.
-. 19355. Pflanzenreste aus dem Devon. IV. Protolepidodendron Krejci.
Senckenbergiana, 14: 391-403.
Lang, William H., and Isabel C. Cookson. 1935. On a flora, including vascular land
plants, associated with Monograptus, in rocks of Silurian age, from Victoria,
Australia. Phil. Trans. Roy. Soc. London, 224B: 421-449.
Lundblad, Britta. 1948. A Selaginelloid strobilus from East Greenland (Triassic).
Meddeleser Dansk Geol. Foren., 11: 351-363.
-. 1950. On a fossil Selaginella from the Rhaetic of Hyllinge, Scania.
Svensk Bot. Tidskrift, 44: 477-486.
Macgregor, Murray and Walton, John. 1948. The Story of the Fossil Grove. Pub¬
lished by City of Glasgow Public Parks, 32 pp.
Magdefrau, Karl. 1931. Zur Morphologie und phylogenetischen Bedeutung der fos-
silen Pflanzengattung Pleuromeia. Beih. Bot. Centralbl., 48: 119-140.

Telegram @tubscstds
LYCOPODOPHYTA 255

Mathews, G. B. 1940. New Lepidostrobi from central United States. Bot. Gaz.
102: 26-49.
Nathorst, Alfred G. 1902. Zur Oberdevonischen Flora der Baren- Insel. Kongl.
Svenska Vetensk.-Akad. Handl., 36 (3): 1-60.
Pannell, Eloise. 1942. Contributions to our knowledge of American Carboniferous
floras. IV. A new species of Lepidodendron. Ann. Missouri Bot. Gard., 29: 245-274.
Pfeiffer, Norma E. 1922. Monograph of the Isoetaceae. Ann. Missouri Bot. Gard.,
9: 79-232.
Rauh, Werner and Falk, Heinz. 1959. Stylites E. Amstutz, eine neue Isoetacee aus
den Hochanden Perus. Sitzungsberichte Heidelberger Akad. Wissen. Jr., 1959,
pp. 1-83.
Schmidt, Wolfgang. 1955. Pflanzen-reste aus der Tonachiefer-Gruppe (Unteres
Siegen) des Siegerlandes. Sugambrophyton pilgeri, n.g., n. sp., eine Protolepidoden-
dracee aus den Hamberg-Schichten. Palaeontographica, 97B: 1-22.
Schopf, James M. 1941. Contributions to Pennsylvanian paleobotany Mazocarpon
oedipternum, sp. nov. and Sigillarian relationships. Illinois State Geol. Surv., Rept.
Investig., No. 57: 1-40.
Scott, Dukenfield H. 1898. On Spenserites, a new genus of lycopodiaceous cones
from the Coal Measures, founded on the Lepidodendron spenceri of Williamson.
Phil. Trans. Roy. Soc. London, 189B: 83-106.
-. 1901. On the structure and affinities of fossil plants from the Palaeozoic
Rocks. IV. The seed-like fructifications of Lepidocarpon, a genus of lycopodiaceous
cones from the Carboniferous formation. Phil. Trans. Roy. Soc. London, 194B:
291-333.
-. 1920. Studies in Fossil Botony. A. and C. Black, Ltd., London, 434 pp.
Snigirevskaya, Natalie S. 1958. An anatomical study of the leaves (phylloids) of some
lycopsids in the Donets basin coal balls. Acad. Nauk USSR (Botany), 43: 106-112.
Watson, D. M. S. 1908. The cone of Bothrodendron mundum. Mem. Manchester
Lit. Phil. Soc., 52(1): 1-15.
White, David. 1907. A remarkable fossil trunk from the middle Devonic of New
York. N. Y State Museum Bull., 107: 327-340.
Williamson, William C. 1887. A monograph on the morphology of Stigmaria ficoides.
Palaeontographical Soc. London, 1886 vol.: 1-62.
Zeiller, Rene. 1888. Etudes des Gites Mineraux de la France; Bassin Houiller de
Valenciennes. Paris.

Telegram @tubscstds
the

Although problems of classification are not lacking, most

lr\ of the plants that have been assigned to this division
stand apart quite clearly from all other groups. The ecological re¬
quirements of the arthrophytes and general evolutionary pattern
have followed rather closely those of the lycopods as far as geologic
time is concerned. Several early and mid-Devonian fossils of con¬
siderable morphological interest have been referred to the group
although it will be evident that their exact relationships to later
forms is still uncertain; but by late Devonian times the die was cast
and a burst of evolution in the Carboniferous resulted in some of
the most unique plants that have ever lived. It was a diverse and
plentiful assemblage, yet certain basic characters hold them together
as a unit. Apparently decline set in rapidly at the close of the
Paleozoic and a single genus, Equisetum, survives today. A few
botanists have devoted serious consideration to possible relation¬
ships with other major groups and it has been suggested that they
may have been progenitors of the seed plants; such speculations are
certainly in error.
The terms arthrophyte, sphenopsid, and articulate have been
used as common names; the last seems particularly apt because of
the characteristic jointed nature of the stems and whorled arrange¬
ment of the leaves and sporangia-bearing organs which correlate
with a comparably unique internal organization. Plants in other
groups are jointed or articulated but are not readily confused with
the articulates.
The student who is not well acquainted with Equisetum, the lone
generic survivor of this great race of plants, would do well to famil¬
iarize himself with its numerous distinctive features. Briefly, the
various species range from 1 to 4 or 5 feet high (a few tropical ones
have been reported much higher, but their stems are at most 2 or
3 cm in diameter) and the shoot system consists of underground
rhizomes which may produce a great profusion of upright shoots;

Telegram @tubscstds
in some species the latter are sparsely branched and in others they

256
ARTHROPHYTA 257

are regularly and lavishly so. Two types of upright shoots may be
present, fertile ones that carry on but little photosynthesis and bear
a cone at the apex, and sterile shoots. Other species bear only one
type of shoot which is both fertile and photosynthetic. In all cases
the leaves are inconspicuous scalelike structures borne in a sheath¬
ing whorl at the node. Internally, through the internodal region
there is a large central pith canal surrounded by a ring of only a
few tracheids which are associated with minute protoxylary
canals; in this character they present a striking divergence from
the larger Carboniferous forms in which several inches of second¬
ary wood may be present. Another series of canals is present in
the cortex.
The cone of Equisetum, which is terminal on the shoot system,
consists of whorls of stalks which flare out into a disc, making an
umbrellalike organ, and around the rim of the disc there are several
pendant sporangia.

Classification

There are a few problems in particular that seem to me to have

a direct bearing on any attempt at classifying the articulates. The
earlier Devonian fossils are in part fragmentary and in other cases
not as distinctly “articulate” as the Carboniferous plants; they are
not sufficiently well known to enable us to understand their inter¬
relationships or affinities with the later members, and they are
therefore treated under the heading “Certain Early Articulates.”
A surprising number of new cone types have turned up recently,
indicating that the Sphenophyllales and Equisetales are more com¬
plex assemblages than has been supposed. Two especially complex
sporangiate organs, Eviostachya and Cheirostrobus, have been
found in the Devonian and Lower Carboniferous respectively; aside
from their intrinsic interest they suggest a far more diverse Devo¬
nian assemblage than is presently known and I have chosen to con¬
sider them at the close of the chapter simply as “Problematical
Articulates.”

Certain early articulates

Equisetales
Calamitaceae
Equisetaceae
Sphenophyllales

Telegram @tubscstds
Some problematical articulates
258 STUDIES IN PALEOBOTANY

SOME EARLY ARTICULATES

Rather than follow a strict chronological order it will perhaps

lend greater clarity if the two best known Devonian articulates,
Calamophyton and Hyenia, are described first.
Some especially fine specimens of Calamophyton (C. bicephalum
Leclercq and Andrews) have been discovered in Middle Devonian
rocks near the village of Goe in eastern Belgium. The plants (Fig.
9-1) attained a height of at least 2 feet and consisted of a “main
axis” which tended to divide in a more or less digitate fashion, as
many as seven branches being formed at very nearly the same point.
These in turn branched monopodially or with a mixture of mono-
podial and dichotomous forking. It is not known how the structure
referred to as the main axis was borne and most of the specimens
are predominantly sterile or fertile.
The ultimate branches bear sterile appendages (“leaves”) that are
about 10 mm long (Fig. 9-2A). They were slender, dichotomous
structures, probably terete in cross section, which divided two to
four times. Successive divisions were at right angles to each other
so that the appendage as a whole was clearly three-dimensional.
They exhibit a weak tendency toward a whorled arrangement.
The ultimate fertile appendages are quite complicated for so
ancient a plant. Each one (Fig. 9-2B) is also about 10 mm long;
the primary stalk divided into two equal branches or “heads,” from
which the specific name bicephalum is taken. Each of these gave
rise to three short side branches which terminate in a pair of slender
sporangia.
There is sufficient similarity between the sterile and fertile ap¬
pendages so that they may be regarded as homologous organs. They
were probably rather rigid in life, especially the sterile ones since
their three-dimensional branching pattern is retained in fossiliza-
tion, that is, they dip down into the rock matrix and much tedious
work with fine steel needles was required to expose them.
Hyenia is probably closely related to Calamophyton and has been
found associated with it; in fact, the only clear distinction between
the two lies in the general habit. In Hyenia elegans Leclercq a
stout rhizome (Fig. 9-3B) bore numerous upright shoots that were
rarely branched. Hyenia vogtii Hdeg from the Middle Devonian
of Spitsbergen offers an interesting problem in branch morphology.
The plant is known from axes up to 9 mm thick which presumably
correspond to the upright leafy shoots of other hyenias; however,

Telegram @tubscstds
ARTHROPHYTA 259

Fig. 9-1. Calamophyton bicephalum, a specimen from the Middle Devonian of eastern
Belgium, about 0.5X. (From Leclercq and Andrews, 1960.)

Telegram @tubscstds
260 STUDIES IN PALEOBOTANY

Fig. 9-2. Calamophyton bicephalum. A. Two sterile appendages (leaves) showing

their three-dimensional form, 5X. B. A fertile appendage; the sporangiophore divides
into two parts, each of which bears three short branches that terminate in paired
sporangia, 4X. (From Leclercq and Andrews, 1960.)

these axes produced lateral branches which resemble the main one
but are somewhat more slender. Leaves 1 to 1.5 cm long are ar¬
ranged in fairly regular verticils on both primary and secondary
branches. Of special note is the position that the branches occupy.
According to Professor H0eg who described these fossils:
In our new species, branching is frequent and regular; it is of considerable
interest to note that the branches are strictly lateral, and that they take the
place of leaves, thus showing that they are homologous with those organs.
We have a parallel case of development within another group of primitive
plants, still living, the bulbils of Lycopodium selago and other bulbilliferous
species of Lycopodium being inserted in the place of leaves . . .

A few Lower Devonian fossils have been reported which lack the
quality of preservation and completeness of the Middle Devonian
plants but may be regarded tentatively as early articulates.
Protohyenia janovii Anan’ev from western Siberia (Fig. 9-3A) is
known from short unbranched shoot fragments, some of which bear
sterile appendages that forked several times and probably served

Telegram @tubscstds
ARTHROPHYTA 261

as leaves. Other shoots bore appendages of essentially the same

form, but they were terminated by oval sporangia which measure
about 3x2 mm. It is thought that these shoots were attached to
a rhizomatous stem like that of Hyenia, but this has not been con¬
firmed. There is an unmistakable similarity here between the
sterile and fertile appendages and it is not difficult to accept the
supposition that the latter might have evolved into the more com¬
plex organs of the Middle Devonian fossils.

EQUISETALES

Calamitaceae

This family includes the arborescent articulates, commonly

known as the calamites. The larger members of the group were

Fig. 9-3. A. Restoration of Protohyenia janovii; the upright shoot at the extreme right
is fertile, the side branches terminating in sporangia. B. Restoration of a portion of
a plant of Hyenia elegans; the left-hand shoot is fertile, whereas the other two are
sterile, about 'AX. (A from Ananiev, 1957; B from Leclercq, 1940.)

Telegram @tubscstds
262 STUDIES IN PALEOBOTANY

trees that probably attained a height of 50 feet or more; conse¬

quently we are confronted with the usual assortment of vegetative
and reproductive organs. Although there are several generic names
and numerous species under each category (leaf, stem, cone, etc.),
they belong for the most part to a rather clearly defined group of
plants. Thus the objective in the following pages will be to convey
an understanding of a “generalized” calamite, following the ap¬
proach used with the arborescent lycopods.

Stem Structure

In general habit the calamites were jointed, profusely branched

plants; if one can imagine the vegetative shoots of the common
Equisetum arvense enlarged about thirty or forty times some con¬
cept may be gained of the way these plants appeared in the Car¬
boniferous forests.
Several genera have been recognized on the basis of differences
in the structure of the secondary wood; with small twig specimens,
in which this tissue is present only in the initial stages of develop¬
ment, such generic distinctions cannot be made.
Figure 9-4A shows a small twig from an Illinois coal ball. A pith
is present and bordering its periphery is a ring of small protoxylary
canals, so-called because the first formed wood cells are found along
the lateral and outer sides of each canal. There is very little pri¬
mary wood in any of the calamite stems, cambial activity having
been initiated very early. The phloem, formed on the outside of
the wood, is rarely preserved. There is next a broad cortical region
of essentially uniform, isodiametric cells.
Somewhat larger branch specimens bring in two characteristic
features of the shoot system, a pith cavity and strongly developed
secondary wood. The pith cavity is not the result of decay but is
rather a natural canal due to the failure of the apical meristem to
form this tissue other than as a thin peripheral band. Three dis¬
tinct genera have been established on the basis of marked differ¬
ences in the secondary wood;
Arthropitys is the most frequently encountered wood type being
represented by numerous specimens and species; delimitations of
the latter are difficult, thus our discussion here will be at the generic
level. Figure 9-5A shows a sector of an Arthropitys stem that was
probably about 12 cm in over-all diameter, only the peripheral pith
and wood being shown. The xylem is composed of wood sectors

Telegram @tubscstds
(ws) which radiate out from the protoxylary canals as a focal point,
ARTHROPHYTA 263

Fig. 9-4. A. A small Calamites twig, 30X; p, pith; pr, protoxylary canal; c, cortex.
B. A calamite root, 5X. C. A leaf, probably referable to Asterophyllites.

and large interfascicular rays (ir) which lie between them. The
relationship between the two varies greatly, even among different
specimens that apparently were derived from a single species. The
rays may be broadly wedge-shaped masses of tissue that quickly
lose their identity in the radial direction or they may extend undi¬
minished in breadth to the outer periphery of the wood.
The wood sectors are composed of tracheids and small wood rays.
The tracheids, although apparently constant within a species, dis¬
play pitting on their radial walls that is scalariform, circular-
bordered, or more or less intermediate. The wood rays are generally
one to three cells wide and vary greatly in height.
In Calamodendron the wood sectors are flanked on either side

Telegram @tubscstds
by bands of vertically aligned fibrous cells; thus the radial sequence
264 STUDIES IN PALEOBOTANY

ws

ir

ws

ir

A B
Fig. 9-5. The wood of a calamite stem of the Arthropitys type. A. Transverse section;
B. tangential longitudinal section showing part of three nodes; ir, inter-fascicular ray;
ws, wood sector. (Both about 5X.)

is wood sector, fiber band, interfascicular ray, fiber band, wood

sector, etc.
In the third genus, Arthroxylon, the cells of the interfascicular
rays are vertically elongated almost to the point of being fibrous.
The wood sector is similar to that of the other genera although in
A. williamsonii Reed the ray cells in this region are also vertically
elongate; thus in a tangential section it is quite difficult to differen¬
tiate between interfascicular ray and wood sector.
Very little is known about the tissues outside the xylem other
than in the small twigs. Many years ago W. C. Williamson described
an English Arthropitys specimen in which the wood attained a
radial thickness of 5 cm and this was partially surrounded by a
fibrous periderm of radially aligned cells 5 cm thick which is similar
in general aspect to the periderm of the lycopods. A short time ago
I had occasion to examine a slide of this specimen preserved in the
British Museum and it seemed to me rather likely that the periderm
is correctly associated with the wood, but since the two are not in

Telegram @tubscstds
organic connection one cannot be sure.
ARTHROPHYTA 265

We have recently discovered Arthropitys specimens in considera¬

ble abundance in an Illinois coal ball locality which reveal two novel
features. Some are exceedingly large, the radial dimension of the
secondary wood exceeding 12 cm; so far as I am aware this is ap¬
preciably larger than any previously known petrified articulate stem
and suggests an over-all diameter in life of at least 35 cm (allowing
for a large pith canal and broad cortex).
Some of these large stelar fragments retain a cortex several centi¬
meters thick of apparently roundish parenchymatous cells, there
being no evidence of periderm tissue.
It is thus evident that additional information is needed concern¬
ing the cortical anatomy of these unique plants and it will almost
certainly be forthcoming in the near future.
Pith casts of the calamitean stems (Fig. 9-6) are among the most
frequently encountered elements of the roof shale floras of Carbon¬
iferous coal mines. When the trunks and branches fell into the
swamps, the large pith canal was soon filled with mud which hard-

Fig. 9-6. A pith cast of Cata¬

mites cruciatus; the round scars
represent points where branches
were attached, slightly reduced.
(Swedish Natural History
Museum.)

Telegram @tubscstds
266 STUDIES IN PALEOBOTANY

ened to form a cast; occasionally they are found in their original

cylindrical form, but more often they were flattened as the load of
sediments above increased. As the stem tissue around this mud
core decayed, the primary rays tended to break down more rapidly
than the wood sectors; thus the casts display a ridge and furrow
appearance, the furrow corresponding to the apex of the wood sector
and the ridge to the primary ray.
The generic name Calamites was established for the casts, a great
many of which probably represent the Arthropitys type stem since
the latter were more abundant than Calamodendron or Arthroxy-
lon. The pith casts do not have characters that allow assignment
to the respective three “wood genera.”
The pith casts may reach a considerable size; Seward cites a
specimen from the Radstock Coal Measures in England that is 27
cm in width. Allowing for crushing this indicates a pith cavity of
perhaps 20 cm in diameter and would imply a stem with an over-all
diameter greatly in excess of the petrified ones mentioned above.

Roots

The anatomy of the roots (Fig. 9-4B) differs from that of the stems
in several ways. They lack the jointed organization, a solid and
well-preserved pith is often present even in quite large specimens,
and the protoxylary canals are absent. The cortical tissue is rarely
preserved, but in specimens in which it is still intact, large lacunae
are present.

Foliage

The calamitean stems bore foliage (Fig. 9-7) that is referable to

the genera Asterophyllites and Annularia. A useful monographic
study of the American species has been made recently by Abbott.
In Asterophyllites (there are four American species) the number
of leaves in a whorl varies from 4 to 40 with the different species
and they show a strong tendency to be cupped upward. With the
exception of the very small-leafed A. charaeformis (Sternberg)
Goeppert, in which they are only 2 to 3 mm long, the leaves are long
and essentially linear with a width-length ratio of 1:12 up to an
extreme of 1:100 in A. longifolius (Sternberg) Brongniart.
There are ten American species of Annularia; the number of
leaves per whorl varies from 8 to 32 and they tend to stand out at

Telegram @tubscstds
right angles to the stem, forming a perfect rosette in compression
 ARTHROPHYTA 267

A B
Fig. 9-7. Foliage of the calamites. A. Asterophyllites sp.; B. Annularia sp.; both from
the Upper Carboniferous of Missouri, about natural size. (B from Andrews, 1947.)

fossils. The individual leaves may be lanceolate but are more often
ovate to spatulate; in both genera they are uninerved.
Most of our knowledge of the anatomy of the leaves (Fig. 9-4C)
comes from a study made some years ago by Thomas on specimens
from the Halifax Hard Bed in the Lower Coal Measures of Lanca¬
shire; most of them are probably Asterophyllites leaves. The epi¬
dermis has a rather heavy cuticle and within this is a single row of

Telegram @tubscstds
268 STUDIES IN PALEOBOTANY

radially aligned cells which constituted the mesophyll; they are

separated by conspicuous air chambers as are the spongy mesophyll
cells of many modern plants.
The vascular bundle consisted of a single central strand of tra-
cheids surrounded by a narrow zone of thin-walled cells which
probably functioned as the phloem. Peripheral to this is a conspicu¬
ous layer of elongate cells which are partially filled with dark con¬
tents; this has been termed the melasmatic tissue and is considered
to be a food reservoir region. Directly above the bundle is a patch
of fiber cells which served to add rigidity to the leaves.

Cones

The diverse nature of the Carboniferous articulates is most force¬

fully illustrated by the variety in the morphology of their spore¬
bearing organs, and in many cases the preservation is far better than
the paleobotanist dares hope for.
The calamitean cones vary considerably in size, and the way in
which they were borne differs from Equisetum where we usually
find a single one terminating the shoot. In some of the calamites
the cones were borne individually at the nodes, others were arranged
in terminal groups or infructescences, and still others on specialized
branches. C. E. Weiss has given some excellent illustrations show¬
ing the general habit of the fertile shoots, Fig. 9-8 being taken from
his account; the serious student would do well to consult this
excellent work.
Calamostachys binneyana Carruthers from the European Coal
Measures is probably the most common and one of the best known
of the calamitean cones (Fig. 9-9A). It attained a length of 3.4 cm
and a maximum diameter of 7.5 mm, consisting of equidistant alter¬
nating verticils of bracts and sporangiophores. There are about 12
bracts in a whorl, which are fused to form a disc at the base, whereas
the upturned distal ends are separate. The fertile appendages are
arranged about six in a whorl and each one consists of a sporangio-
phore which terminates in a peltate or cruciate head bearing four
sporangia. To indulge in a fine point, but one that is quite apparent
in well-preserved specimens, the bracts in successive whorls alter¬
nate; it may be noted in the figure that one whorl shows the
upturned lamina, whereas in the one above or below, it appears to
be missing. All of the many specimens of this cone that have been
found in English coal balls indicate without doubt that it was

Telegram @tubscstds
homosporous. In 1909 Thomas reported a Calamostachys cone on
a twig with foliage of the Asterophyllites type.
ARTHROPHYTA 269

Fig. 9-8. Calamostachys lud-

wigi, a branch bearing nu¬
merous cones. (From C. E.
Weiss, 1884.)

Telegram @tubscstds
270 STUDIES IN PALEOBOTANY

Fig. 9-9. Diagrammatic longitudinal sections of, A. Calamostachys binneyana and

B. Palaeostachya andrewsii. (B from Baxter, modified.)

Calamostachys casheana Williamson, from the Coal Measures

of England, is quite similar to C. binneyana in its general organiza¬
tion but it is heterosporous. Some of the sporangia contain numer¬
ous small spores, whereas in others the number is much lower and
the spores three to four times as large in diameter. Figure 9-10 is
drawn from a specimen in the Williamson collection (British Mu¬
seum), the portion present being beautifully preserved. It is un¬
fortunate that this heterosporous species should be as rare as
C. binneyana is abundant; however, additional information has
recently turned up in the form of a spectacular cone, C. americana
Arnold, from the Pennsylvanian of Illinois. It is 4 cm in diameter
and at least 12 cm long. The axis bore whorls of bracts (40 to 45
each) which were united in the basal region to form a horizontally
expanded disc. Successive whorls are spaced about 4 mm apart and
midway between them is a whorl of sporangiophores; the latter are
about 1 cm long and the distal head bore four radially elongate spo¬
rangia. Aside from its size the most significant feature of the cone
lies in its spores; some sporangia contain microspores which range
from 85 to 114 ju, in diameter, whereas others contain megaspores
from 150 to 260 ju, in diameter. Some sporangia are reported to
contain megaspores at one end and microspores at the other; so

Telegram @tubscstds
far as I am aware such an association within a single sporangium
ARTHROPHYTA 271

has never been observed in any other plant. Although the speci¬
mens on which the original description is based are very well pre¬
served, the sporangium walls are quite delicate and not always
intact. It is therefore possible that the apparent occurrence of two
kinds of spores in a single sporangium may be the result of some
slight distortion and breaking of the tightly compressed sporangia.
Several interesting cones have been assigned to the genus Palaeo-
stachya which is distinguished by the alignment of the sporangio-
phores at an angle of about 45° to the axis and their attachment
immediately above the bracts. Whether Calamostachys and

Fig. 9-10. Oblique longitudinal section of Calamostachys casheana. (From a slide in

Telegram @tubscstds
the Williamson Collection, British Museum Natural History.)
272 STUDIES IN PALEOBOTANY

Palaeostachya should be regarded as two distinct genera, or whether

there are really more than two in the assemblage, is debatable.
P. andrewsii Baxter from the Des Moines series of Iowa (Fig.
9-9B) is a cone with 24 bracts in a whorl and these are free down
to their point of attachment to the axis; there are usually 12 spo-
rangiophores per whorl, each one terminating in a peltate head with
four sporangia. The spores are all the same size but quite large,
being 270 to 320 n in diameter. Although heterospory is not known
for certain in the genus, it seems probable that we are dealing with
a megasporangiate fragment of a heterosporous cone (the species is
based on a specimen from which both apical and basal parts are
missing) or possibly one in which microspores and megaspores were
produced in separate cones.
P. decacnema Delevoryas from the Des Moines series of Kansas
is a somewhat smaller cone with a diameter of about 8 mm and
spores that are 45 to 50 [i in diameter, whereas the sporangiophores
are attached slightly higher, relative to the associated whorl of
bracts.
A much discussed feature of Palaeostachya is the course followed
by the vascular strand from the cone axis to the sporangiophore.
In the British P. vera Seward the trace is described as passing up
to about the middle of the node and then descending through the
cortex before it diverges out into the sporangiophore. A similar
course has been observed in P. andrewsii, but in P. decacnema the
trace goes directly to the sporangiophore. It has been suggested
that the looped character of the strand, where it does occur in
Palaeostachya, indicates a “phyletic slide” downward of the spo¬
rangiophore from the position that it occupies in Calamostachys.
The cones of Mazostachys pendulata Kosanke (Fig. 9-11C) are
semipetrified fossils which were found in an ironstone concretion in
northern Illinois. They are 4 mm in diameter, about 2.6 cm long,
and the 12 bracts at each node are fused in the basal portion. Six
sporangiophores are attached just below the node and each bears
two pendant sporangia. The spores are about 60 n in diameter.
Annularia sphenophylloides foliage was found in the same concre¬
tion associated with the cones.
Two other cones may be mentioned briefly since they seem to
present some similarity to Mazostachys; these were described by
C. E. Weiss in the books cited in the References at the end of the
chapter. Cingularia typica Weiss (Fig. 9-1 IB) had strap-shaped
sporangiophores which are attached immediately beneath the whorl

Telegram @tubscstds
of bracts; each sporangiophore is divided at the distal end with two
 ARTHROPHYTA
273

Fig. 9-11. Representative portions of the cones of: A. Stachannularia tuberculata;

B. Cingularia typica; C. Mazostachys pendulata; Kallostachys scottii. (A, B from
C. E. Weiss, 1876; C from Kosanke; D from Brush and Barghoorn, 1955.)

sporangia borne on the under side of each bifurcation. The spo¬

rangia are nearly spherical and about 5 mm in diameter. In
Stachannularia tuberculata (Sternberg) Weiss the general organi¬
zation of the bracts and sporangiophores is similar to the above, but
here each sporangiophore bears but one pendant sporangium.
In contrast to the cones considered thus far, in Kallostachys
scottii Brush and Barghoorn (Fig. 9-1 ID) there is only one kind of
appendage which may appropriately be termed a sporophyll. The
cone is about 2 cm in diameter and was probably homosporous.

Telegram @tubscstds
274 STUDIES IN PALEOBOTANY

Each whorl consists of 12 units which are fused in their basal por¬
tion and the ascending distal lobes are bifurcated. In Fig. 9-11 two
whorls are included, one showing the external appearance and the
other a longitudinal section. On the under side of the basal part
of the sporophyll there are two radially elongate sporangia; since
these are borne side by side and closely appressed, only one appears
in longitudinal section.
The cone axis has a large central canal or cavity; exarch primary
xylem is quite well developed, occasional protoxylem canals can be
observed, and some secondary xylem is present. Cortical canals are
also present.
The whorled arrangement of the appendages as well as the pro-
toxylary and cortical canals point to the articulates, but the exten¬
sive development of exarch primary wood suggests lycopod anatomy.
The cone is perhaps best considered tentatively as representing a
distinct line of arthrophytes.

Equisetaceae

By comparison with the Carboniferous record the fossil history

of the articulates from early Mesozoic times to the present is quite
scanty. Numerous specimens, chiefly impressions or casts, of stems
have been referred to Equisetites; many suggest plants closely re¬
lated to the modern Equisetum, but others are too incomplete or
poorly preserved to allow a dependable generic comparison.
Stem impressions of Equisetites are distinguished from the cala-
mitean plants partly by the fusion of the leaves into a sheath, and
if the assumption that they are closely allied to Equisetum is cor¬
rect, they attained an appreciably greater size than the living
species. The stem casts of the Triassic Equisetites platyodon
Brongniart are about 6 cm in diameter and the stems of E. colum-
naris Brongniart from the Jurassic of Yorkshire reached a compara¬
ble size; this is insignificant by calamitean standards, yet appreciably
larger than any surviving Equisetums.
A few of the Equisetites species have been preserved with cones
intact. E. woodsi Jones and de Jersey from the Jurassic of Queens¬
land was a plant with stems about 7 mm in diameter and a leaf
sheath 11 mm long, of which the free teeth compose the distal 5 mm.
Associated cones, which are believed to have been borne on these
stems, are rather large, measuring 2.4 cm long and 1 cm in diame¬
ter. Several hundred sporangiophores were borne on the cone axis;
the head of the appendage was dome-shaped, 1.5 mm in diameter,

Telegram @tubscstds
ARTHROPHYTA 275

and three rows of 50 to 60 sporangia hang down from the under

side; this arrangement of the sporangia and the high number is quite
remarkable.
There are a few intriguing suggestions that plants closely akin to
Equisetum did live in the late Paleozoic. In his flora of Autun and
Epinac (France) Renault described a cone under the name of Bornia
radiata which has been variously referred to Asterocalamites and
Archaeocalamites, but its morphology is more interesting than its
taxonomic history. The cones are 5 mm in diameter and 13 to 15
mm long, but since they are only fragments the original length is
not known. No sterile appendages are present as in the calamite
cones, but in his description Renault notes that the fertile append¬
ages are occasionally interrupted by whorls of leaves; there are
eight to ten sporangiophores to a whorl and each apparently was
slightly forked at the distal end and bore four sporangia.
In the modern Equisetum the outer wall of the spore splits in a
distinctive fashion, resulting in four straplike appendages known as
elaters which apparently add buoyancy to the spores and allow them
to be more readily air-borne. The gametophytes of Equisetum,
which develop directly from the spores, are usually unisexual; thus,
since several spores usually become entangled, another asset of the
elaters may be the development of several gametophytes in close
proximity to one another. But whatever nature’s intended use of
the elaters may be they present a unique diagnostic character. It
is therefore of interest to note that elater-bearing spores (Elaterites
triferens Wilson) have been recorded from Des Moines age (Penn¬
sylvanian) rocks of Iowa. Each spore of Elaterites bears three
elaters. Somewhat comparable spores (Equisetosporites chinleana
Daugherty) have been found in the Upper Triassic of Arizona which
have two elaters.
There is, therefore, enough evidence to indicate that Equisetum
is a very ancient type reaching back into the early Mesozoic or per¬
haps earlier; the precise relationship of this modern relict to other
articulates is, however, still obscure.

SPHENOPHYLLALES

This is a distinct line of articulates which appears first in the late

Devonian and persisted to at least the early Triassic. The stems
and leaves are referred to the genus Sphenophyllum; the latter are
wedge-shaped and arranged in regular verticils, whereas the wood

Telegram @tubscstds
276 STUDIES IN PALEOBOTANY

of the stems presents unique features. Of particular interest is a

diverse assemblage of cones, generally referred to the genus Bow-
manites.
The sphenophylls were small plants probably forming a low-
growing herbaceous undergrowth in the Carboniferous forests. The
largest specimen that I am aware of is one reported by Seward, in
the collections of the Austrian Geological Survey; it consists of a
stem 85 cm long and 4 mm in diameter which bears a branch 61 cm
long. Thus, a meter in height is probably close to a maximum and
their slender stems were rarely over 1 cm in diameter.

Stem Anatomy

The wood of Sphenophyllum stems (Fig. 9-12) is unlike that

present in any other plant group. The center is occupied by a tri¬
angular, primary stele of tracheids only and with the small pro-
toxylem cells located at the apices. A considerable amount of
secondary wood was formed in the older stems, the tracheids oppo¬
site the protoxylary points being conspicuously smaller than those
in between.
A common Pennsylvanian species, S. plurifoliatum Williamson
and Scott, is distinguished by rays in the secondary wood which con¬
sist of groups of vertically aligned cells located between the trun¬
cated angles of the tracheids. These groups are connected, by
narrow horizontal cells, into a unified system. The tracheids of both
primary and secondary wood have closely aggregated, bordered pits.
S. insigne Williamson, a rather common British Lower Carbon¬
iferous species, has rays of the normal type, that is, they are uniform
throughout their radial extent, and the tracheids are scalariform.
This species also has protoxylary canals, but these have been
reported as occasionally occurring in S. plurifoliatum.
A new species {S. constrictum Phillips) has recently been described
from Indiana and Kansas coal balls in which the cortex and thick,
fleshy leaves (which are once forked) are composed of large, thin-
walled cells which lends a succulent aspect to the plant.
Sphenophyllum plants seem to have branched rather sparsely but
in various ways. In his study of numerous stems from Illinois and
Iowa, Baxter described a well-preserved specimen in which three
branches were found departing simultaneously from the angles of
the primary wood and apparently in between pairs of leaves rather
than in their axils; another specimen revealed an apparent dichot¬
omy of the stem.

Telegram @tubscstds
ARTHROPHYTA 277

Fig. 9-12. Sphenophyllum sp. A. A young stem in which secondary wood had just
started to develop, 15X. B. Part of a larger stem showing the distinctive central tri¬
angular primary wood and the abundant development of secondary wood, 10X.

Leaves

The leaves are arranged in whorls of six, eight, or nine and are
wedge-shaped in their over-all outline (Fig. 9-13). The whorls are
flattened in much the same fashion as with Annularia; thus, in life
they must have stood out at a 90° angle from the stem. In her study
of the American fossils Abbott recognized 18 species which vary a
good deal in size and the degree of dissection; brief notations on a
few of these suggest the range of leaf form in the genus:

S. emarginatum Brongniart. Regularly wedge-shaped, up to 16

mm long and 10 mm wide with a nearly entire (i.e., not dissected)
distal margin.
S. comutum Lesquereux. Up to 2 cm long; divided to about half
the length of the leaf into eight or more equal lobes. As in other
species a single vein enters the base of the leaf, dichotomizes sev¬
eral times, and a single strand penetrates each lobe.
S. tenerrimum Ettingshausen. Attain a length of 9 mm long and
may be dissected in the middle to the base of the leaf, thus dividing
it into two parts.
S. thoni Mahr. The distal margin is rounded and finely divided
(fimbriate).

Telegram @tubscstds
278 STUDIES IN PALEOBOTANY

Fig. 9-13. Sphenophyllum foliage, a

specimen from the Upper Carbonif¬
erous of Missouri, natural size.

The identification of species of Sphenophyllum foliage is actually

rather precarious due to the great variation in size, and degree of
dissection as well as over-all shape. This may occur within a single
specimen; Abbott figures a shoot of S. cuneifolium in which the
leaves are wedge-shaped with an entire margin on the upper parts
of the branches, whereas somewhat lower the leaves are partially
dissected and in the basal region they are divided to their base into
slender linear segments. As many as eight different names have,
in the past, been applied to the various leaf types which are now
regarded as falling within the range of this species!

Cones

One of the more significant recent advances in Carboniferous

paleobotany is the discovery of numerous cone types, some of which

Telegram @tubscstds
are remarkably well preserved, that have been attributed to the
ARTHROPHYTA 279

Sphenophyllales. There is, in fact, such an interesting array of

these fossils now known that it is difficult to decide what should be
omitted in a summary of this sort.
In a monographic account of many presumed sphenophyllalean
cones Hoskins and Cross have recognized 18 species in the genus
Bowmanites. Some three or four have been found in organic con¬
nection with stems bearing typical Sphenophyllum foliage, whereas
a few others have been found associated with Sphenophyllum twigs
in a way that leaves little doubt as to where they belong in the plant
kingdom. In most species of Bowmanites the spore-bearing organs
are arranged in distinct cones consisting of whorls of bracts that are
usually fused to form a basal disc; associated, immediately above
the bract whorl, are sporangiophores with terminally borne spo¬
rangia. The degree to which the sporangiophores are actually fused
with the bracts varies considerably and some authors have preferred
to designate the latter, sporophylls. In view of the diversity in
this genus it seems likely that several distinct genera will even¬
tually be recognized.
Bowmanites dawsoni (Williamson) Weiss (Fig. 9-14) is one of the
better known species. The cone is about 20 mm in diameter and

Fig. 9-14. Bowmanites dawsoni, slightly

oblique section of a specimen from Bel¬
gium, 4X.

Telegram @tubscstds
280 STUDIES IN PALEOBOTANY

composed of numerous whorls of 14 to 20 bracts each, fused to form

a shallow, funnel-shaped basal portion; the free distal limbs extend
upwards for a considerable distance. About twice as many slender,
terete stalks arise from the upper surface of the “funnel” near its
junction with the axis. Each stalk (sporangiophore) bears a single
sporangium in an anatropous position; that is, the stalk is reflexed
so that the sporangium is directed toward the cone axis. Although
there is apparently a single whorl of the sporangiophores associated
with each bract whorl, their lengths are so variable that, in a me¬
dian longitudinal section, the impression is gained of several rings
of sporangia. The cone axis has a triangular stele similar to that
of Sphenophyllum.
Bowmanites bifurcatus Andrews and Mamay (Fig. 9-15A) from
the Upper Pennsylvanian of Illinois is noteworthy for its very small
size. The cone is only 3 mm in diameter and the single specimen
known is approximately 1.5 cm long. Through this length there
were 17 whorls of appendages, one of which is shown in Fig. 9-15.
Each whorl consists of six bracts which are fused to form a basal
disc; the free segments ascend abruptly and each forks, resulting in
12 free tips. A sporangiophore arises from the upper surface of the
proximal part of each bract and bifurcates at the distal end, bear¬
ing two sporangia.
In spite of its minute dimensions it was possible to determine the
detailed vascular anatomy of the cone. The axis has three equidis¬
tant wood strands which may be compared with the apices of the
primary stele of a Sphenophyllum stem. At the node a branch
passes out horizontally from each of the three strands and immedi¬
ately divides into an upper and lower segment. The upper one
bifurcates horizontally and each of the two resultant strands sup¬
plies a sporangiophore; the lower segment does the same, and each
strand forks again to supply a single one to each of the terminal
divisions of the bract.
The occurrence of such tiny cones as this causes one to wonder
how many others have been missed in routine examinations of coal
ball petrifactions. The advantage of the peel technique (Chapter
17) is evident here where, in so small and complex a structure,
numerous transverse and longitudinal sections are required to
interpret its three-dimensional form.
Bowmanites fertilis (Scott) Hoskins and Cross is a remarkable
cone (Fig. 9-15B) from the Coal Measures of England and Belgium;
its organization is, in fact, so unique as to render usual morpholog-

Telegram @tubscstds
ARTHROPHYTA 281

ical terms somewhat inadequate. It attained a length of 6 cm and

a diameter of 2.5 cm. Six “appendage complexes” depart at a
node, each one consisting of an upper unit that will be called
a compound sporangiophore, and a lower pair of branches which
may be regarded tentatively as homologous with a bract in the

Fig. 9-15. A. Bowmanites bifurcatus, a

single verticil with three pairs of sporangia
removed to show organization of sporangio-
phores, 15X. B-D. Bowmanites fertilis;
B. two verticils of cone; C, D. terminal por¬
tions of two sporangiophores, each with a
pair of sporangia. (A from Andrews and
Mamay, 1951; B-D from Leclercq, 1936.)

Telegram @tubscstds
282 STUDIES IN PALEOBOTANY

previously described species. The compound sporangiophore

erupts into a dense cluster of about 16 branches, each of which
terminates in a pair of sporangia (Fig. 9-15C, D). The lower pair
of branches (“bracts”) extended downward and outward as long,
slender hornlike structures.
Bowmanites moorei Mamay is a small and much less complex
type. The total length is unknown, but it attained a diameter of
4 mm. There are three bracts or sporophylls at a node, but each
one divides to form a central and two lateral lobes. Two sporangi-
ophores are borne above each median lobe and these terminate in
an inverted sporangium; there are thus six sporangia per whorl.
Litostrobus iowensis Mamay from the Des Moines series of Iowa
is another cone of rather simple construction and quite small (3.5
mm in diameter). At a node there are 12 basally fused bracts and
half as many sporangia, each being borne erect on a stout pedicel;
the latter apparently are oriented directly above every other bract.
The vascular system is poorly preserved, but in the whorled
arrangement of the appendages, basally fused bracts and multiple
of threes, we have characters that point to the sphenophylls.
Sphenostrobus thompsonii Levittan and Barghoorn diverges in
its structure from the other cones described in several ways. It is
about 9 mm in diameter and consists of whorled appendages (16 at
each node) which are termed sporophylls rather than bracts since
the sporangia, also 16 in number, are sessile and attached to the
disc a short distance from the cone axis. The stele differs from
that of the other sphenophyll cones in being tetrarch.
In Sphenophyllum hauchecornei (Weiss) Remy from the West¬
phalian of Germany the close association of the sterile and fertile
appendages of the cone is even more pronounced. Although
described from compression fossils the preservation is good enough
to reveal many critical features. The cone is 1 cm in diameter and
in excess of 12 cm long. The 12 sporophylls at each node are
united in their basal portion; they extend out laterally for 2 or 3
mm, then turn up abruptly and each of the 12 segments divides
into two slender, tapering, hairy lobes. The latter are nearly 3 cm
long and thus overlap five or six of the nodes above. Two or three
sporangia were probably attached by short stalks to the basal
(disc) part of each sporophyll; in a longitudinal view the sporangia
appear in two verticils, one above the other. These cones are
believed to have been borne on stems with the Sphenophyllum
cuneifolium type of foliage.

Telegram @tubscstds
ARTHROPHYTA 283

SOME PROBLEMATICAL ARTICULATES

There are a few fossils attributed to the articulate group that

are of exceptional morphological interest, so much so as to deserve
real consideration in a general account such as this, yet one can do
little more than guess at their proper position in the Arthrophyta.
Eviostachya h(j>egi Stockmans is based on cones from the Upper
Devonian of Belgium which attain a length of 5.5 cm and a diam¬
eter of 0.8 cm although most of the specimens are somewhat
smaller. Each cone consisted of a peduncle about 2.5 mm long
with a single whorl of six bracts below the fertile region; above this
are as many as a dozen verticils of sporangiophores. At each
whorl there were three pairs of sporangiophores, a pair being
derived from each angle of the triangular stele of the cone axis.
Figure 9-15 illustrates one sporangiophore: the stalk is trisected at
its apex into three branches and each of these divides into two
laterals, and a shorter central one; each secondary branch termi¬
nates in a cluster of three sporangia, there being a total of 27 on
each sporangiophore. There are no associated sterile appendages
other than the single basal whorl. The spores range in size from
35 to 55 ju, the plant apparently being homosporous. Although

Fig. 9-16. Eviostachya htpegi, restoration

of a single sporangiophore, about 15X.
(From Leclercq, 1957.)

Telegram @tubscstds
284 STUDIES IN PALEOBOTANY

numerous specimens of Eviostachya have been found, most of the

critical details were worked out by Professor Leclercq from a single
specimen that was partially petrified with iron hydroxide. Since
the entire sporangiophore is only 2 to 3 mm long it is remarkable
that its organization could have been revealed in such precise detail.
In 1897 Scott described Cheirostrobus pettycurensis which I
believe still holds the record as the most complex of all pterido-
phytic sporangiate organs. It was derived from the well-known
basal Carboniferous deposits at Pettycur, Scotland, and although
only one specimen has ever been found it is well preserved and
there is no reason to doubt the accuracy of Scott’s restoration.
It is large as well as complicated (Fig. 9-17), measuring 3.5 cm in
diameter and consists of whorls of appendages that have been
called compound sporophylls. There are 12 such units in a whorl
and each divides within a few millimeters of the cone axis, into a
lower and upper segment. The lower segment divides horizontally
into three slender stalks (bracts) which become laminate toward
their distal end and form a conspicuous heel, whereas the ascend¬
ing limb bifurcates, resulting in a total of six free tips. The upper
segment likewise divides horizontally into three slender stalks
(sporangiophores), each of which is peltate at the distal end and
bears jfour very elongate sporangia. Each sporangium is 1 cm long
and 1 mm in diameter; the cone axis contains a 12-rayed stele.
A third plant may be included here, but it is perhaps well to
caution that its position in the Arthrophyta is somewhat less cer¬
tain than the preceding two. Prosseria grandis Read is a compres¬
sion fossil found at an Upper Devonian horizon in Yates County,
New York. The stem fragment that is preserved is 25 mm broad
and the leaves are whorled as in the articulates but otherwise
quite distinct; they are 6 mm wide, 33 cm long, and were probably
attached in groups of three.

Summary Comments

In all probability it will always be difficult to recognize the

earliest members of a great plant group; these are the plants in
which characters, that will later be diagnostic, are in the formative
stage and it is evident that the early Devonian and Silurian pteri-
dophytes formed a vast melting pot from which certain clearly
defined lines emerged in late Devonian times. Thus, one may be
reluctant to accept a plant such as Protohyenia as an early arti¬
culate. Yet, in view of the evidence afforded by Calamophyton

Telegram @tubscstds
ARTHROPHYTA
285

Fig. 9-17. Cheirostrobus pettycurensis, a complex articulate cone from the Lower
Carboniferous of Scotland, 3X. (From Scott, 1897.)

from the Middle Devonian and Eviostachya from the Upper

Devonian it is clear that the origins of the group must be found
ultimately in the early Devonian or in the Silurian.
Cheirostrobus is especially perplexing; one almost wishes that it
had been found in the Permian instead of at the base of the Car¬
boniferous! We may fall back once again on the explanation that
has been used several times, that this is a stray remnant from an
upland flora in which articulate evolution had already attained a
high degree of complexity.
The gap between any of the Middle Devonian articulates and
the dominant lines of the Carboniferous, the calamites and spheno-
phylls, is also a broad one.
The discovery in the past decade of so many new cone types in

Telegram @tubscstds
the Carboniferous is quite encouraging; they indicate clearly that
286 STUDIES IN PALEOBOTANY

the Carboniferous forests were by no means monotonous even

though the flowering plants were not present to add splashes of
color. Yet in spite of these discoveries evolutionary trends are by
no means crystal clear. Mamay has suggested a tendency toward
increasing simplicity in the evolution of sphenophyll cones; Cheiro-
strobus pettycurensis, the most complex of all pteridophytic fructi¬
fications, comes from the early Carboniferous; Bowmanites fertilis
is a rather complex cone from the English Lower Coal Measures;
Litostrobus, the simplest sphenophyll cone, is from the Des Moines
series; and Bowmanites bifurcatus occurs in the McLeansboro
formation. Thus he notes:
Although an absolute correlation between the geologic age and structural
complexity of the various other sphenophyllalean fructifications intermediate
between the extreme forms cannot be demonstrated, the wide stratigraphical
separation seems to suggest a general evolutionary trend toward reduction
rather than proliferation. Accordingly Litostrobus may best be interpreted
as a highly advanced and reduced fructification, rather than a primitive one.
(1954, p. 237)

This view seems to me acceptable at least as a working hypoth¬

esis, but of course we are left with the problem of how the older,
complex sporangiate organs reached that state!
The modern Equisetum is certainly closely akin to the calamites
yet we would like to know more about its exact relationships with that
group. Did the equisetums split off in Carboniferous times as a
semiherbaceous line and why are they the only articulates that
have been able to survive to the present?

REFERENCES

Abbott, Maxine L. 1958. The American species of Asterophyllites, Annularia and

Sphenophyllum. Bull. Amer. Paleont., 38: 289-390.
Ananiev, A. R. 1957. New Lower Devonian fossil plants from the southeast of west¬
ern Siberia. Akad. Nauk USSR (Botany), 42: 691-702.
Andrews, Henry N., Jr. 1947. Ancient plants and the world they lived in. Comstock
Pub. Co., Ithaca. 279 pp.
-. 1952. Some American petrified calamitean stems. Ann. Missouri Bot.
Card., 39: 189-218.
-and Mamay, S. H. 1951. A new American species of Bowmanites. Bot.
Gaz., 113: 158-165.
Arnold, Chester A. 1958. Petrified cones of the genus Calamostachys from the Car¬
boniferous of Illinois. Univ. Michigan, Contrib. Mus. Paleont., 14: 149-165.
Baxter, Robert W. 1950. Peltastrobus reedae: a new sphenopsid cone from the Penn¬
sylvanian of Indiana. Bot. Gaz., 112: 174-182.

Telegram @tubscstds
ARTHROPHYTA 287

. 1955. Palaeostachya andrewsii, a new species of calamitean cone from

the American Carboniferous. Amer. Journ. Bot., 42: 342-351.
Brush, Grace S., and Barghoorn, E. S. 1955. Kallostachys scotti: a new genus of
sphenopsid cones from the Carboniferous. Phytomorphology, 5: 346-356.
Daugherty, Lyman H., and Stagner, Howard R. 1941. The Upper Triassic flora of
Arizona with a discussion of its geologic occurrence. Carnegie Inst. Washington
Pub., 526: 1-108.
Delevoryas, Theodore. 1955. A Palaeostachya from the Pennsylvanian of Kansas.
Amer. Journ. Bot., 42: 481-488.
Hoskins, John H., and Cross, Aureal T. 1943. Monograph of the Paleozoic cone genus
Bowmanites (Sphenophyllales). Amer. Mid. Nat., 30: 113-163.
Jones, O. A., and de Jersey, N. J. 1947. Fertile Equisetales and other plants from the
Brighton beds. Univ. Queensland Papers, Dept. Geology 3 (4): 1-16.
Kosanke, Robert M. 1955. Mazostachys—a new calamite fructification. Illinois State
Geol. Surv., Rept. Investigations, No. 180: 1-22.
Lacey, William S. 1943. The sporangiophore of Calamostachys. New Phyt., 42: 1-4.
Leclercq, Suzanne. 1936. A propos de Sphenophyllum fertile Scott. Ann. Soc. Geol.
Belgique, 60: 170-172.
-. 1940. Contribution a l’etude de la flore du devonien de Belgique. Acad.
Roy. Belgique, ser. 2, 12: 1-65.
-. 1957. Etude d’une fructification de Sphenopside a structure conservee du
devonien superieur. Acad. Roy. Belgique, ser. 2, 14: 1-39.
-, and Andrews, H. N., Jr. 1960. Calamophyton bicephalum, a new species
from the Middle Devonian of Belgium. Ann. Missouri Bot. Gard. 47: 1-23.
Levittan, Edwin D., and Barghoorn, Elso S. 1948. Sphenostrobus thompsonii: a new
genus of the Sphenophyllales? Amer. Journ. Bot., 35: 350-358.
Mamay, Sergius H. 1954. A new sphenopsid cone from Iowa. Ann. Bot., n.s., 18:
229-239.
-. 1959. A new bowmanitean fructification from the Pennsylvanian of Kan¬
sas. Amer. Journ. Bot. 46: 530-536.
Phillips, Tom L. 1959. A new sphenophyUalean shoot system from the Pennsyl¬
vanian. Ann. Missouri Bot. Gard. 46: 1-17.
Read, Charles B. 1953. Prosseria grandis, a new genus and new species from the
Upper Devonian of New York. Washington Acad. Sci., 43: 13-16.
Reed, Fredda D. 1952. Arthroxylon, a redefined genus of Calamites. Ann. Missouri
Bot., Gard. 39: 173-187.
Remy, Winfried. 1955. Untersuchung von kohlig erhaltenen fertilen und sterilen
Sphenophyllen und Formen unsucherer systematischer Stellung. Abh. Deutschen
Akad. Wissen., 1: 5-40.
Renault, Bernard. 1893, 1896. Bassin Houiller et Permien d’Autun et d’Epinac.
Etudes des Gites Mineraux de la France. Paris. Atlas, 1893; Text, 1896.
Schultes, Richard E., and Dorf, Erling. 1938. A sphenopsid from the Lower Devonian
of Wyoming. Harvard Univ. Bot. Mus. Leaf., 7: 21-33.
Scott, Dunkinfield H. 1897. On Cheirostrobus, a new type of fossil cone from the
Lower Carboniferous strata (Calciferous Sandstone series). Phil. Trans. Roy. Soc.
London, 189: 1-34.
Seward, Albert C. 1898. Fossil Plants. Vol. I. Cambridge Univ. Press, pp. 1-452.
Thomas, H. Hamshaw. 1909. On a cone of Calamostachys binneyana (Carruthers)
attached to a leafy shoot. New Phyt., 8: 249-260.

Telegram @tubscstds
288 STUDIES IN PALEOBOTANY

-. 1911. On the leaves of Calamites (Calamocladus section). Phil. Trans.

Roy. Soc. London, 202B: 51-92.
Walton, J. 1949. On some Lower Carboniferous Equisetineae from the Clyde area.
Trans. Roy. Soc. Edinburgh, 61: 729-736.
Weiss, Christian E. 1876-1884. Beitrage zue fossilen Flora. Steinkohlen-Calamarien
mit besonderer Berucksichtigung iher Fructificationen. Abh. geol. Specialkarte
Preuss. Thiiring. Staaten. Berlin. I. (1876), pp. 1-149; II. (1884), pp. 1-204.
Wilson, Leonard R. 1943. Elater-bearing spores from the Pennsylvanian strata of
Iowa. Amer. Mid. Nat., 30: 518-523.

Telegram @tubscstds
 10
the CYCADOPHYTA

T he term Cycadophyta has been used to include three

groups of gymnospermous plants, the Pteridospermo-
phyta, the Cycadales, and the Bennettitales. The first was dealt
with in Chapter 5 where it was pointed out that it is a large and
varied assemblage of plants which in itself encompasses several
distinct lines. It is not my intention to discredit the concept that
the Cycadales and Bennettitales may have originated somewhere
in the seed-fern complex, but there is actually a very broad gap
between the pteridosperms and the other two orders, and we are
far from understanding just how it may have been bridged. A
consideration of the reproductive organs of the latter will make
this apparent.
The problem of classification is further complicated by the fact
that the seeds and microsporangiate organs of the Cycadales and
Bennettitales are quite different and it is now known that, in spite
of a gross similarity in the morphology of the foliage of the two,
the structure and distribution of the stomates are not closely com¬
parable. In his study of the Scoresby Sound fossils Harris, who
has probably contributed more to our knowledge of the cycado-
phytes than any other recent worker, concludes that the Cycadales
and Bennettitales are probably quite remote from one another. I
am inclined to feel that as evidence accumulates it will become ap¬
parent that this is the case and that we have been misled, largely
by a superficial resemblance of the foliage. To obviate any mis¬
understanding the following terminology will be used here: the
Cycadophyta are regarded tentatively as including two orders, the
Cycadales and Bennettitales. The word “cycadophyte” will be
applied to any plant in the entire assemblage. In the literature on
these plants it has been common practice to call the Cycadales
simply “cycads” and the Bennettitales “fossil cycads.” This has
resulted in a certain amount of confusion since the Cycadales are
represented in the fossil record as far back as mid-Mesozoic times.
Common names are useful if not unavoidable and the term “ben-

Telegram @tubscstds 289

290 STUDIES IN PALEOBOTANY

nettite,” rather than “fossil cycad,” will be used here in reference

to bennettitalean fossils.
Foliage that is probably referable to the Bennettitales makes its
appearance in Permian rocks and the group apparently became
extinct sometime during the Cretaceous. The Cycadales are
known from the Triassic to the present; nine genera, with about 90
species, now live in the tropical and subtropical parts of the Amer¬
icas, Asia, Australia, and South Africa. It is perhaps one of the
vagaries of preservation that we know less about the general habit
of such fossil Cycadales as have been found than we do about the
Bennettitales.
Since study material is not always as readily available as for
most of the extant groups described in previous chapters, it may be
helpful to consider briefly a living cycad.

Zamia Floridana—a Living Cycad

Zamia floridana grows in central and southern Florida and sev¬

eral other species are found in the West Indies and South America.
The stem of the mature plant (Fig. 10-1) is a rather massive and
irregularly branching structure and only the apical portion appears
above ground; the root system is sparsely branched but penetrates
to a considerable depth. Specimens are not infrequently encount¬
ered in the Florida woods which branch much more profusely than
the one illustrated; a cluster of a dozen or so leaves is found at the
tip of each branch, thus several crowns of leaves in close proximity
to each other usually indicate a single plant. The leaves are once
pinnate.
Zamia is dioecious, that is, a plant may bear seed cones or micro-
sporangiate ones but not both. The seed cones are stout barrel¬
shaped structures up to about 12 cm. long; the central axis bears a
dozen vertical rows of appendages, each of which (Fig. 10-lB) con¬
sists of a stalk and a hexagonal head from which two seeds are sus¬
pended toward the axis. The microsporangiate cones are essentially
the same in their general organization but are more slender and
elongate; on the under side of each appendage (microsporophyll)
there are 20 or 30 small spherical sporangia. (Fig. 10-lB).
In its cone and leaf structure Zamia is a representative cycad
although in many of the other species these organs may be con¬
siderable larger. The stem structure is somewhat less typical, for
in most of the living cycads it is an upright, sparsely branched,
columnar trunk covered by a rough armor of the persistent basal

Telegram @tubscstds
portions of the petioles. A maximum size is attained by certain
 CYCADOPHYTA 291

Fig. 10-1. A. A plant of Zamia floridana; one branch is shown bearing a seed cone,
about 'AX; B. a single seed-cone appendage (megasporophyll) with its two seeds;
C. a single microsporophyll from the underside showing numerous sporangia. (From

Telegram @tubscstds
Andrews, 1947.)
292 STUDIES IN PALEOBOTANY

species of Cycas, in the East Indies, in which the trunks reach a

height of 23 meters and a diameter of nearly 1 meter. Cycas is
represented by several species in the Asiatic-Australian region and
is commonly cultivated in the extreme southern parts of the
United States. It diverges notably from the other cycads in that
the seed-appendages (megasporophylls) are not aggregated into a
cone; some bear four or five seeds and may be pinnately divided,
thus presenting a slight resemblance to the leaves.

CYCADOPHYTE FOLIAGE

Foliage that is usually attributed to the cycadophytes is found

from the Permian to the present but is especially abundant in
Triassic and Jurassic rocks. The leaves vary from a few centi¬
meters long to well over 1 meter; for the most part they are un¬
branched and once pinnate, although some are entire or nearly so.
Beyond these generalities they present one of the most difficult
problems of identification in the whole field of paleobotany.
Dozens of genera and scores of species of leaves have been described
and the names of many have been changed several times in accord¬
ance with the views of different investigators. It was not until the
discovery was made that cuticular structures were more depend¬
able than the gross morphology that identifications began to take
on some real biological significance. Comprehensive studies of the
cuticle structure of these plants was initiated by Nathorst in the
early years of the century, were greatly advanced by Thomas and
Bancroft in 1913, and have been followed by several others in more
recent years, one of the most comprehensive and useful being
Harris’ discussion in Part 2 of his Scoresby Sound flora—a vade
mecum for so many paleobotanical topics.
Many of the cycadophytes seem to have had tough, leathery
foliage with a heavy cuticular coating as do the living cycads; con¬
sequently, preservation of the epidermal cell-wall structure is
occasionally quite good. With specimens in which these features
are not preserved it may be quite impossible to determine whether
the foliage belongs to a cycad or a bennettite, or whether it even
lies within the limits of these two.
A few examples of cycadophyte foliage are briefly considered
next with the intent of showing: (a) the difficulty of separating
cycads from bennettites on the basis of gross morphology of the
leaves; (b) the distinctive microscopic, epidermal details which

Telegram @tubscstds
seem to offer a more dependable way of delimiting the two.
CYCADOPHYTA 293

General morphology of foliage

A few examples of cycad and bennettite leaves, or representative

portions of them, are shown in the upper part of Fig. 10-2 (see also
Figs. 10-3A, 10-11, and 14-2B, C). It will be evident from this
small selection that it would be difficult to divide them into two
groups on the basis of gross form. There is actually much more
variety known than is indicated and the differences in size are
tremendous.

Cycadean (Cycadales) Stomatal Structure

The extant Zamia muricata is taken as an example. The accom¬

panying figures (10-2) show: a stomate in surface view (K); a sec¬
tion cut through the center and at right angles to the long axis (L);
and a section cut lengthwise through one of the guard cells (J). In
Fig. K the guard cells (g) are surrounded by a ring of somewhat
irregularly shaped subsidiary cells (s) and it has been shown that
these originate from independent mother cells and not from the
guard mother cell. The guard cells are somewhat sunken and in
the section cut parallel to the long axis (J) it is evident that the
central part (which is associated with the actual opening) is more
deeply depressed than the ends (poles) which turn up abruptly;
these appear at g' in Fig. K. The outer surface of the guard cells
is covered with a cuticle of quite even thickness.

Bennettitalean Stomatal Structure

In leaves of the Bennttitales there are two lateral subsidiary

cells associated with each stomate (Fig. 10-2G, H). These are
believed to have originated from the same mother cell that pro¬
duced the guard cells; presumably the mother cell divided to form
two, and each of these divided again in the same plane, the inner
pair of the four becoming the guard cells and the outer two the
subsidiary cells. The epidermal cells immediately adjacent to the
poles are not modified in any special way. A second distinctive
feature of the bennettitalean cuticle is the unique thickening of
the outer and dorsal walls of the guard cells (following the termin¬
ology of Harris, the term ventral wall is applied to the one facing
the aperture (stoma) and the dorsal wall is the one opposite to it).
The stomata in the Bennettitales are in general oriented trans¬
versely and the epidermal cell walls are conspicuously sinuous; in

Telegram @tubscstds
the Cycadales, and other pinnate-leaved fossil gymnosperms, the
Fig. 10-2. Cycadophyte leaf morphology and epidermis structure. CYCADALES:
D, E. portions of the leaves of Zamia intergrifolia and Z. furfuracea, respectively
(both extant). J, K, L. Zamia muricata (extant), all 350X; J. median longitudinal
section through guard cell; K. surface view; L. transverse section through middle of
stomate. F. Surface view of stomate and surrounding cells of Pseudoctenis spectabilis,
350X. BENNETTITALES: A. Nilssoniopteris vittata. B, G. Pterophyllum aster-
tense; B. leaf; G. stomate and subsidiary cells. C, H, I. Pterophyllum rosenkrantzi;
C. leaf; H. stoma and subsidiary cells, 900X; I. median transverse section through same.
s, subsidiary cells; g, guard cells; g', pole of guard cell. (All except A, D, E from
Harris, 1932.)

294
Telegram @tubscstds
CYCADOPHYTA 295

stomata tend to be irregularly or longitudinally oriented and the

epidermal cell walls are straight.
As might be expected there is a good deal of variation within the
two groups. For example, in the cycads the guard cells may be
depressed below the surface to varying degrees, whereas the sub¬
sidiary cells or others peripheral to them arch over the guard cells.
The latter are quite deeply sunken in the living Cycas revoluta.

THE CYCADALES

Although the Cycadales have existed from Triassic times, or pos¬

sibly earlier, well-preserved reproductive organs are scarce. One of
the most informative and interesting suites of fossils referable to
the order are leaves, seed organs, and microsporangiate cones that
have been described under the generic names Nilssonia, Beania,
and Androstrobus (Fig. 10-3). They are closely associated in the
mid-Jurassic Cayton Bay beds on the Yorkshire coast. The foliage,
Nilssonia compta Phillips, is one of the most abundant elements
in the flora (Fig. 10-3A); specimens attain a length of 40 cm and
consist of a strong rachis with a lamina that is divided into trun¬
cate segments which vary a good deal in size. Beania gracilis
Carruthers is a loose “cone” with an axis up to 10 cm long and
spirally borne appendages consisting of a stalk about 2 cm long
which expands at the distal extremity to form a broadly ovate,
recurved head with two seeds on its inner surface. The largest
seeds measure about 16 X 13 mm and are well enough preserved
so that the two layers of the integument can be recognized (Fig.
10-3C): an outer (possibly fibrous) zone and an inner stony one
which enclosed a vascularized nucellus. The male cone, Andro¬
strobus manis Harris, is about 5 cm long and 2 cm in diameter.
On the under side of each of the numerous, spirally disposed sporo-
phylls are several scores of cylindrical sporangia (Fig. 10-3E).
Androstrobus is closely comparable to the microsporangiate
cones of the living cycads, and if one were to stretch out the axis
of a Zamia seed cone the result would be something nearly identi¬
cal to Beania. Moreover, the stomatal structure of Androstrobus
and Beania, as well as the Nilssonia foliage, are sufficiently alike
as to leave little doubt that they represent organs of a single
species and that they are cycadean. Combined with their close
association at certain spots along the outcrop on the Cayton Bay

Telegram @tubscstds
beach, the case seems to be a good one.
296 STUDIES IN PALEOBOTANY

Fig. 10-3. Fossil Cycadales from Cay ton Bay, Yorkshire. A. Nilssonia compta.
B, C. Beania gracilis; B. infructescence or seed “cone,” reduced slightly; C. a single
megasporophyll with one seed, restored, in longitudinal section. D, E. Androstrobus
manis; D. the microsporangiate cone, about natural size; E. a single sporophyll.
(B-E from Harris, 1941.)

The mere association of fertile and vegetative parts of plants

that are suspected of belonging together has long been known to
constitute dubious evidence and has been the subject of much
criticism. In beds where large numbers of plants are found, many
of which are obviously unrelated, the evidence from association
may be worthless. The distribution of fossil plants at Cayon Bay
is, however, not an ordinary one, for distinct assemblages may be
encountered every few yards. On two visits to this memorable
locality I gained the impression that the association patterns of
the plants cannot be the result of indiscriminate deposition; others

Telegram @tubscstds
have arrived at the same conclusion, including Dr. H. H. Thomas
CYCADOPHYTA 297

whose several seasons of collecting led to the recognition of the

Caytoniales.

THE BENNETTITALES

Petrified cycadophyte trunks, most of which are probably refer¬

able to this order, have been found in many Mesozoic horizons and
are widely scattered geographically. They are unique as paleo-
botanical specimens and have long been prized by collectors. In
fact, a trunk of Cycadeoidea etrusca has the distinction of being
the oldest known fossil plant specimen collected by man. Accord¬
ing to Wieland it was “Placed with vases and other objects of
superstitious reverence on one of the supulchral chambers of the
ancient Necropolis at Marzabotto by the Etruscans more than
four thousand years ago” (1906, p. 12). It was rediscovered about
a century ago and given the binomial indicated above by Capellini
and Solms.
Many trunks have been found in Jurassic rocks of the Freezeout
Hills of Wyoming, smaller numbers have come from the Potomac
formation of Maryland, from other Mesozoic horizons in Kansas,
Colorado, Prince Edward Island, and several other North Amer¬
ican areas. Numerous specimens have also been found on the Isles
of Wight and Portland off the south coast of England and a few
have turned up in France, Belgium, the USSR, Poland, Germany,
and India.
The most prolific collections by far, however, have been made in
the Black Hills of South Dakota and are of early Cretaceous age.
The silicified trunks from this area were first brought to the atten¬
tion of botanists in 1893 and although many specimens are said to
have been damaged or destroyed through the depredations of
curiosity hunters, considerable numbers were acquired and pre¬
served through the efforts of T. H. MacBride of the State Univer¬
sity of Iowa, by Lester F. Ward and 0. C. Marsh, and particularly
G. R. Wieland whose vast collections are deposited at Yale
University.
For the most part these petrified trunks are heavy, bulky speci¬
mens, hard and quite difficult to prepare for study. Many of them
make attractive and unique museum specimens and there seems to
have been a general reluctance to cut them up sufficiently to allow
comprehensive study. The result is that few have actually been
investigated in their entirety.

Telegram @tubscstds
298 STUDIES IN PALEOBOTANY

Fig. 10-4. Cycadeoidea marshiana, shown from above. (From Wieland, 1906.)

The Black Hills area was rich in species as is evinced by the

variety in trunk shape and reproductive organs. The trunks may
be of upright columnar form, of conical shape, and under a meter
in height, or of a low growing and profusely branching habit (Fig.
10-4). These last are particularly striking and give the effect of
several pineapples aggregated into a dense bunch. It is estimated
that the columnar Cycadeoidea jenneyana may have attained a
height of several meters, but most of them were appreciably
shorter. Numerous illustrations of these unique plants are given
in Wieland’s American Fossil Cycads and in Ward’s Status of the
Mesozoic Floras.
Some of the silicified trunks have been found with partially
developed leaves still intact, so that one may readily visualize
them as they lived on the Cretaceous landscapes with a crown, or

Telegram @tubscstds
several crowns in the case of the branching species, of pinnate
CYCADOPHYTA 299

leaves presenting an appearance much like that of some modern

cycads. In the organization of their reproductive organs they
were, however, very different. As an introduction to these bizarre
and often well-preserved fossils, Cycadeoidea dacotensis will be
taken as an example (Fig. 10-5).

Note. A brief comment may be in order concerning the terminol¬

ogy that is used here and elsewhere. The term cone or strobilus
is certainly applied to far too many spore and seed-bearing organs
in which the comparative morphology is often obscure. The term
flower has been used rather generally for the reproductive organs
of gymnospermous plants as well as the angiosperms (Anthophyta);

Fig. 10-5. Cycadeoidea dacotensis. Semidiagrammatic sketch of a flower in longitu¬

dinal section; h, hairy bract; m, expanded microsporophyll (one at the left is shown in
immature condition); c, central axis with numerous seeds and interseminal scales.

Telegram @tubscstds
(From Wieland, 1906.)
300 STUDIES IN PALEOBOTANY

although it seems to me preferable to reserve this for the latter

group its usage is rather well established in bennettitalean litera¬
ture. One is torn between a choice of setting up an almost endless
series of terms to apply to spore, pollen, and seed organs in the
various groups or to hold to a bare minimum. The latter course is
perhaps preferable so long as the plant structures involved are
understood.

Unlike the living cycads in which the cones are borne at the apex
of the stem, they appear as rosettelike bodies in the cycadeoids,
scattered among the leaf bases; they are generally much smaller
than those of the cycads and in some species prodigious numbers
were formed—up to several hundred on a plant.
The flower itself presents some similarity to a magnolia in its
general organization, and, although some botanists have chosen to
see an evolutionary sequence here, a close comparison between the
two leaves a rather great gap. However, in the bennettites as in
the magnolias there is a basal aggregation of microsporophylls sur¬
mounted by a receptacle that may be flattened, dome-shaped or
conical, bearing on its outer surface seeds and interseminal scales.
The flower of C. dacotensis is, therefore, described as being
bisporangiate in that it includes both microsporangiate and seed
parts. It is 5 to 10 cm in diameter and 6 cm long, being borne on
a peduncle of comparable length. The upper two-thirds of the
peduncle bore a close spiral succession of 100 to 150 elongate hairy
bracts which enclosed the young reproductive organ.
There are hundreds of interseminal scales on the receptable of C.
dacotensis, with seeds distributed among them except in the basal
and apical regions. The seeds of the cycadeoids are generally
quite small, being little more than 1 mm long in this species. The
structure and relationships of the seeds and associated scales are
more clearly shown in Fig. 10-6 which is a median section through
the flower of C. wielandi; it differs from C. dacotensis in the short¬
ened receptacle and greatly elongate stalks of the seeds and inter¬
seminal scales.
The microsporangiate portion of C. dacotensis is composed of 19
or 20 once-pinnate appendages with their bases fused to form
a disc. In the restoration (Fig. 10-5) the one on the left is shown
in an immature condition and the one on the right fully expanded.
Two rows of complex synangia are arranged along each primary
branch. Each synangium is a flattened, tear-shaped organ attached
by a short, stout stalk; a single one is shown at a much higher

Telegram @tubscstds
enlargement in Fig. 10-7. The outer wall is a single layer of large,
CYCADOPHYTA
301

Fig. 10-6. Cycadeoidea wielandi. Longitudinal section through a flower. (From

Wieland, 1906.)

thick-walled cells; within this is a thin zone of parenchymatous

cells followed by a ring of sporangia extending around the periphery.
These dehisced longitudinally emptying the spores into the center
of the synangium which in turn opened at maturity.
At first glance the cycadeoids appear to have been exceptionally
prolific seed producers. The occurrence of several scores of strobili
on a trunk is not uncommon and, as an extreme case, some 500 to
600 were found on a single trunk of Cycadeoidea dartoni Wieland.

Telegram @tubscstds
Since the strobili are numerous when present at all and remnants
302 STUDIES IN PALEOBOTANY

Fig. 10-7. Cycadeoidea dacotensis. A. Longitudinal section through a synangium;

B. transverse section. (From Wieland, 1906.)

of older ones have not been observed among the leaf bases, Wieland
concluded that the plants probably were monocarpic, that is, they
fruited only once and died shortly thereafter. The evidence does
suggest that an individual plant fruited heavily and at long inter¬

Telegram @tubscstds
vals, but whether only once in its life is uncertain.
CYCADOPHYTA 303

It is possible to compute the age of the living cycads by dividing

the number of leaf bases on the trunk by the number of leaves
formed each year; allowing for some variation in the number put
out each year, and the fact that fewer were formed in the crown
in the earlier life of the plant, at least an approximate age may be
computed. In his book on the living cycads Chamberlain figures
a Mexican specimen of Dioon edule 5 feet in height that is approx¬
imately 1000 years old. Assuming that the growth of the cyca-
deoids went on at about the same rate as the modern cycads one
may estimate an age of several centuries for the larger trunks. It
is a little difficult to believe that they would have fruited but once
over so long a life span.
The morphology of the unique interseminal scales has perplexed
most serious students of the Bennettitales and in his study of the
Scoresby Sound fossils Harris encountered several fragmentary
strobili (Fig. 10-8) which offer a possible clue. Vardekloeftia
conica Harris differs from other bennettitalean seeds in the pre¬
sence of an additional integument or cupule, and in Bennetticarpus
crossospermus Harris the micropyle of the seed is surrounded by
a “micropylar plate” which may represent the distal portion of
such an outer integument. Harris points out that the micropylar
plate is closely comparable with the adjoining interseminal scales
and notes that this “supports the view . . . that the interseminal
scales are homologous with seeds and are in fact formed by the
diverted development of seed initials” (1932, p. 116).

Fig. 10-8. Reconstructions in longitudinal section of bennettitalean reproductive

organs from East Greenland; A. Vardekloeftia conica; B. Bennetticarpus crosso¬

Telegram @tubscstds
spermus. (From Harris, 1932.)
304 STUDIES IN PALEOBOTANY

We will return once again to the beach at Cayton Bay to con¬

sider a bennettite with reproductive organs that are basically
comparable with those of some of the Dakota fossils although the
general habit of the plant was quite different. The flowers, known
as Williamsoniella coronata Thomas (Fig. 10-9A), were probably
borne in the axils of leaves that have long been known under the
binomial Taeniopteris vittata, but more recently have been given
the revised name Nilssoniopteris vittata (Brongniart) Florin (Fig.
10-2C).
The flower, borne on a peduncle about 3.5 cm long, is bisporang-
iate; a ring of 12 or 14 separate microsporophylls is attached at the
base of a central column on which the seeds and interseminal
scales are arranged. The entire organism was enclosed by a peri¬
anth of hairy bracts like those of the cycadeoids considered
previously.
Each microsporophyll is shaped like an orange segment with
fingerlike branches in the central portion (facing the axis) which par¬
tially enclose two pairs of capsules. Each capsule is two-valved and
was probably constructed of several sporangia; this term is sug¬
gested by Harris (who has revised the original description of
Thomas) as preferable to synangium since it avoids the implication
that they evolved by a fusion of separate sporangia. The micro¬
sporophylls were separate from one another and probably were
shed following spore dispersal; they are frequently found isolated
in the rock.

Fig. 10-9. A. A flower of Williamsoniella coronata showing: the central, elongate col¬
umn bearing seeds (in solid black) and interseminal scales (with bulbous head); two
microsporophylls; the enclosing hairy bracts, about 3X. B. A flower of Sturiella

Telegram @tubscstds
langeri enlarged several times. (A from Harris, 1944; B from Krausel, 1948.)
CYCADOPHYTA 305

The gynoecial (seed) axis is only about 10 mm long, but bore some
300 minute ovules interspersed among 1200 interseminal scales,
the latter being characterized by a very slender stalk and a terminal
bulbous head. The abundance of the Nilssoniopteris leaves in the
Cay ton Bay beds implies that the plant was well represented
in this mid-Jurassic flora. They are entire, lanceolate to narrowly
ovate, and virtually all sizes may be found up to 20 cm long and
3 cm wide. Fragments of slender stems have been found associated
with the leaves. Thus in the organization of the vegetative organs
the plant was quite different from the Dakota cycadeoids with
their massive stems and pinnate foliage.
Some years ago Nathorst described a delicately branching ben-
nettite plant, Wielandiella angustifolia, from the Rhaetic of Scania,
in southern Sweden. In its general habit it was probably quite like
Williamsoniella and certainly was in contrast to the massive¬
stemmed cycadeoids. The stems (Fig. 10-13) for the most part do
not exceed 1.5 cm in diameter and divided by what is referred to
as a false dichotomy; apical growth was apparently terminated
by flower formation and two nearly equal branches developed as
shown in the restoration. Since it is not possible to determine the
exact mechanics of branching in a plant that has been dead for 150
million years, it is hardly advisable to distinguish dogmatically
between a dichotomy and a false dichotomy.
Leaves, about 8 cm long, classified as Anomozamites minor are
believed to have been borne on the stems. Such leaves are abun¬
dant in the same beds in which the stems are found, and there is
a close comparison between their winged petioles and the bracts
which enclosed the flower. The latter are not especially well
preserved; the structure of the microsporangiate organs is obscure
while the central ovulate axis bore numerous small seeds and
ovuliferous scales as in other bennettitalean flowers. The plant is
of particular interest for the slender, forking stems and the tend¬
ency of the leaves to be concentrated in the area of branching.
Another bennettitalean plant of apparently small and delicate
dimensions, Sturiella langeri Krausel, comes from the Triassic of
Lunz, Austria. It is known from an axis 4 cm long; slender side
branches terminate in a “cone” (Fig. 10-9B) that bears a superfi¬
cial resemblance to a miniature sunflower head. The distal end of
the branch flares out into a shallow cup, and arranged around the
outer edge are 25 to 30 ray like lobes; a sporangium (or possibly two
that are coalesced) is attached to the inner edge opposite the base of
each ray. The central part of the organ appears to have been

Telegram @tubscstds
306 STUDIES IN PALEOBOTANY

ovulate, possibly similar to that of Williamsoniella but much

shortened; there is, however, very little structure preserved.
The diversity of the Bennettitales is emphasized by the varied
types of microsporangiate organs that have been referred to the
group. Some of these almost certainly represent distinct male
organs, the corresponding female or seed axis having been borne on
another part of the plant or possibly a separate plant. Although
some of them are exceptionally interesting themselves, much
remains to be learned about the general habit of the plants and the
manner in which the reproductive organs were borne.
The genus Williamsonia includes several Jurassic species based
on male flowers. They are rather large structures with a cup¬
shaped base opening into numerous flaring lobes, with the sporang-
iate organs proper arranged along the upper (inner) surface.
Williamsonia spectabilis Nathorst (Figs. 10-10, 10-11) is a decid¬
edly striking fossil that measured about 9 cm across the greatest
diameter; the basal cup is 3 cm in diameter and expanded into
about a dozen segments which tended to bend out horizontally and
then turn abruptly upward, probably arching over the center of
the cup in life. Numerous slender branches are arranged on the
inner side of each sporophyll which in turn bear two rows of
synangia. The latter probably were similar in organization to the
synangia of C. dacotensis; they are divided into several chambers,
but spores have not actually been found in them.
Williamsonia whitbiensis Nathorst (Fig. 10-10, B, C) is of similar
size and shape although the stalk, a conspicuous feature of W.
spectabilis, is not present, and the synangia are borne in two rows
directly on the inner surface of the sporophylls and not on branches
arising from them. Another variant in bennettite microsporophyll
morphology has been described by Sitholey and Bose as W. santa-
lensis, which is known from incomplete specimens found in the
Jurassic Rajmahal series of Bihar, India. The authors caution that
their restoration (Fig. 10-10D) is tentative, but the several specimens
that they figure seem to substantiate it. The basal cup or disc is
perhaps somewhat more shallow than the two Yorkshire species
described previously, but the most novel feature of this Indian
species lies in the apparent forking of each microsporophyll into a
blunt terminal portion and a narrower side branch which bears
two rows of tapering fingerlike appendages. Each one includes two
rows of chambers which in all probability contained spores although
they were not preserved. It is thus not certain whether the unique

Telegram @tubscstds
appendages should be regarded as synangia or whether they
enclosed the synangia.
CYCADOPHYTA 307

Fig. 10-10. Some bennettitalean microsporangiate organs. A. A single microsporophyll

of Wonnacottia crispa. B, C. Williamsonia whitbiensis; B. restoration of the male
flower; C. a single sporophyll showing the two rows of synangia. D. Williamsonia
santalensis, reconstruction of a microsporophyll. E. A flower of Williamsonia specta-
bilis. (A from Harris, 1942; B, C from Nathorst, 1911; D from Sitholey and Bose,
1953; E from Thomas, 1913.)

In contrast to the more or less funnel-shaped organs considered

above, in which the microsporophylls are united basally, Wonna¬
cottia crispa Harris (Fig. 10-10A) is a very different type of apparent
spore-bearing organ. It was discovered several years ago in lower
Middle Jurassic deposits of Yorkshire and is essentially leaflike,

Telegram @tubscstds
being about 12 cm long and 1.5 cm broad with nearly opposite
segments (pinnules) which tend to taper outward slightly.
308 STUDIES IN PALEOBOTANY

Fig. 10-11. A fine specimen of Williamsonia spectabilis from Whitby, England, about
%X. (Swedish Natural History Museum.)

Numerous globose bodies containing pollen are partially sunken

in the under surface and the proximal segments are considerably
reduced, consisting of little more than the pollen-containing bodies.
Typical bennettitalean stomata clearly point to the general affini¬
ties of the fossil. Although originally described as a sporophyll,
and presumably a very primitive one, Professor Harris has informed
me that it is possible that this may actually be a leaf of Anomoza-
mites nilssoni that has been heavily attacked by a gall mite and
that the pollen has simply drifted into the pouches. It seems
unlikely to this writer that so many pollen would drift into a gall
cavity (as many as 100 have been found), but if it is in fact a leaf
and was borne immediately below a flower, this may be the case.
Sagenopteris leaves, according to Harris, are sometimes found
literally covered with Caytonanthus pollen grains.
A Williamsonia (W. sewardiana) from Jurassic rocks of the
Rajmahal Hills, India, affords the best evidence we have as to the
general habit of these plants. The restoration prepared by Profes¬
sor Sahni (Fig. 10-12) is based on fertile branches that he described
in 1932 and trunk specimens that had been recorded previously
under the name Bucklandia indica.
The branch shown on the left side of the trunk terminated in a

Telegram @tubscstds
female flower. The axis of the cone is dome-shaped and bore many
CYCADOPHYTA 309

interseminal scales with somewhat fewer seeds, the general organi¬

zation being similar to that of the cycadeoids. The difference in
habit is most pronounced, for the cone of W. sewardiana is borne
at the apex of a conspicuous branch instead of being sunken among
the leaf bases of the trunk. The seeds are also distinctive in that
they seem to have been composed of a short stalk, a greatly along-
ate nucellar region, and a distally extended integument called an

Fig. 10-12.A reconstruction of William-

soma sewardiana from the Rajmahal Hills,
India. (From Sahni, 1932.)

Telegram @tubscstds
310 STUDIES IN PALEOBOTANY

“apical funnel.” The cone as a whole was enclosed by long bracts

covered with a dense hairy ramentum above and scales below.
The microsporangiate organ is not known and there is no evidence
that one was ever attached at the base of the female flower. It may be
noticed that the fertile branch bears a somewhat smaller sterile
one; it seems probable that growth was terminated by the cone and
when it had shed its seeds the secondary branch assumed dominance.
The restoration is based on the known presence of columnar
trunk fragments from the same locality which bear round scars
that apparently represent the point of attachment of the branches;
some of these trunk specimens have been found with leaves of the
Ptilophyllum cutchense type.

Summary Comments

One might conclude from the distribution of the living cycads

and certain of their morphological features that they are of very
ancient lineage. Two genera enjoy an extensive geographical
range: Zamia, with 28 species, is found from Florida through the
West Indies and Mexico to northern South America and down the
west coast to Chile; Cycas stands as a sort of eastern counterpart,
its 16 species being distributed from Australia through the islands
to southern Japan, and the range swings west to include China,
India, and Madagascar. The two are as far apart in the organiza¬
tion of their seed-bearing organs as they are geographically, with
closely compacted cones in Zamia and leafy megasporophylls in
Cycas. The latter is often cited as the most primitive on the
basis of this character, but the wood anatomy does not correlate;
the tracheids are pitted in Cycas and scalariform in Zamia. The
explanation may lie in the underground habit of the Zamia stems
and they are perhaps best regarded as persistent juveniles, in this
character, as Arnold suggests. The staminate cones of these two
genera are, as in other living cycads, quite uniform, offering a sort
of “neutral” position as far as phylogeny is concerned. Of the
other genera, Macrozamia and Bowenia are confined to Australia;
Encephalartos and Stangeria are found only in South Africa;
Dioon and Ceratozamia occur in Mexico; Microcycas is limited to
western Cuba. This scattered distribution combined with the
fact that the nine genera are sufficiently distinct as to render
interrelationships obscure is indicative of considerable age.
A survey of the Mesozoic fossils leaves no doubts concerning the
diversity of the Cycadophyta. Two points in particular stand out

Telegram @tubscstds
CYCADOPHYTA 311

from studies of the foliage: there is a great deal of gross variation

in form suggesting considerable diversity; this is confirmed by
studies of the cuticles, which reveals two distinct types of stomatal
structure, and some leaves of cycadean aspect have been studied
which have a stomatal morphology that indicates they may not be
closely allied to either. As to specific and generic identifications
the situation is close to that of the ginkgophyte foliage; without
adequate cuticle material a name in many cases means little.
When one contrasts the trunk structure of Cycadeoidea marsh-
iana (Fig. 10-4) or C. gigantea (a stout columnar type from the Isle
of Portland, England) with the delicate shoot system of Wielan-
diella angustifolia (Fig. 10-13) it is evident that a broad gap sepa¬
rates them. Closely related plants may, however, be quite different
in certain of their vegetative organs and it is perhaps more
convincing to compare microsporangiate parts such as Wonna-
cottia crispa, Cycadeoidea dacotensis, and Williamsonia santa-
lensis; quite clearly these represent diverse branches of a great
race of plants concerning which we still know but little.
The similarity of Beania and Androstrobus to the seed and micro¬
sporangiate organs of a modern cycad such as Zamia suggests that

Fig. 10-13. Restoration of Wielandiella angustifolia from the Rhaetic of Scania,

Telegram @tubscstds
Sweden. (From Nathorst, 1909.)
312 STUDIES IN PALEOBOTANY

the Cycadales was a well-established group in the Jurassic. From

a somewhat earlier horizon (Rhaetic) in southern Sweden Florin
has described a cycad, Bjuvia simplex, with a stout columnar trunk
bearing a crown of pinnately veined leaves and a terminal cluster of
megasporophylls, each with two pairs of ovules. His restoration
suggests a plant very much like that of a modern Cycas. Bjuvia and
Beania would thus seem to indicate that the Cycas and Zamia
lines have been distinct for a very long time.
It seems more than likely that the cycadophytes existed as
a distinct group or cluster of groups by late Paleozoic times, but
precise details become obscure below the Triassic. Several cycado-
phyte-like leaves have been reported from the Upper Carboniferous.
In his Permian Shansi (China) flora Halle describes Dioonites
densinervis from pinnate frond fragments 4 cm wide and of
unknown length. Of interest also in this flora is an abundance of
Taeniopteris species; one of them, T. nystroemii, was a large leaf
attaining a breadth of 20 cm; these are possibly of cycadean
affinities.
The assumption made by many authors that the cycadophytes
had their origin in the pteridosperm complex seems to me to be
unwarranted. It is true that the stems have certain features
in common with the medullosan seed-ferns, but so far as I am aware
the leaves and more especially the reproductive organs of the two
are so different as to render such an origin for the cycadophytes
highly improbable. The view is therefore taken here that they
evolved independently as seed plants and very possibly along at
least two different lines from a very early stage, resulting in the
groups we refer to as cycads and bennettites.

REFERENCES

Arber, E. A. N. 1919. Remarks on the organization of the cones of Williamsonia

gigas (L. and H.). Ann. Bot., 33: 173-179.
Andrews, Henry N., Jr. 1947. Ancient plants and the world they lived in. Com¬
stock Pub. Co., Ithaca. 279 pp.
Arnold, Chester A. 1953. Origin and relationships of the cycads. Phytomorphology,
3: 51-65.
Chamberlain, Charles J. 1919. The living cycads. pp. 172. Univ. Chicago press.
Harris, T. M. 1932. The fossil flora of Scoresby Sound East Greenland. Part 3.
Caytoniales and Bennettitales. Meddelel. om Gronland, 85 (5): 1-130.
-. 1941. Cones of extinct Cycadales from the Jurassic rocks of Yorkshire.
Phil. Trans. Roy. Soc. London, 231B: 75-98.

Telegram @tubscstds
CYCADOPHYTA 313

-. 1942. Wonnacottia, a new Bennettitalean microsporophyll. Ann. Bot.,

6: 577-592.
-. 1944. A revision of Williamsoniella. Phil. Trans. Roy. Soc. London,
231B: 313-328.
Krausel, R. 1948. Sturiella langeri, nov. gen., nov. sp., eine Bennettitee aus der
Trias von Lunz (Nieder-Osterreich). Senckenbergiana, 29: 141-149.
Nathorst, A. G. 1909. Palaobotanische Mitteilungen. No. 8. Kungl. Svenska
Vetenskapsakad. Handl., 45 (4): 1-33.
-. 1911. Palaobotanische Mitteilungen. No. 9. Kungl. Svenska Veteskap-
sakad. Handl., 46 (4): 1-33.
Sahni, B. 1932. A petrified Williamsonia (W. sewardiana, sp. nov.) from the Rajma-
hal Hills, India. Mem. Geol. Surv. India, Paleont. Indica, n.s., 20 (3): 1-19.
Seward, A. C. 1900. The Jurassic flora. I. The Yorkshire coast. Catalogue Meso¬
zoic plants, Brit. Mus. Nat. Hist., pp. 1-341.
-. 1917. Fossil Plants. III. Cambridge Univ. press. Pp. 1-656.
Sitholey, R. V. and M. N. Bose. 1953. Williamsonia santalensis sp. nov.—a male
fructification from the Rajmahal series, with remarks on the structure of Onthean-
thus polyandra Ganju. The Palaeobotanist, 2: 29-39.
Stopes, Marie C. 1918. New Bennettitean cones from the British Cretaceous. Phil.
Trans. Roy. Soc. London, 208B: 389-440.
Thomas, H. H. 1913. The fossil flora of the Cleveland District. Quart. Journ. Geol.
Soc. London, 59: 223-251.
-. 1915. On Williamsoniella, a new type of Bennettitalean flower. Phil.
Trans. Roy. Soc. London, 207B: 113-148.
-and Nellie Bancroft. 1913. On the cuticle of some recent and fossil cyca-
dean fronds. Trans. Linnean Soc. London, ser. B, Botany, 8: 155-204.
Ward, Lester F. 1905. Status of the Mesozoic floras of the United States. U. S.
Geol. Survey Mon., 48, pts. I, II, pp. 1-616.
Wieland, G. R. 1906. American fossil cycads. Vol. I. Carnegie Institution Washing¬
ton Pub., 34: 1-295.
-. 1916. American fossil cycads, Vol. II. Carnegie Institution Washington
Pub., 34: 1-277.
-. 1920. Distribution and relationships of the Cycadeoids. Amer. Joum.
Bot., 7: 125-145.
-. 1934. Fossil cycads. with special reference to Raumeria reichenbachiana
Goeppert sp. of the Zwinger of Dresden. Palaeontographica, 79B: 85-130.

Telegram @tubscstds
 11
the CONDFEIROPHYTA
and

Introduction

As used here the Coniferophyta includes two subdivisions, the

Cordaitales and Coniferales. The former was a dominant arbores¬
cent group of Carboniferous seed plants and the Coniferales evolved
from them in the latter part of that period and became quite
diversified in early Permian times. Our knowledge of this sequence
of development has been revealed in recent years largely through
studies of the seed-bearing organs by the Swedish paleobotanist
Rudolf Florin and it stands as one of the great triumphs of pale¬
ontology. That the two orders are thus closely related seems to
me to be certain. There are, however, many gaps to be filled in
and it is expedient to retain the two ordinal names for the present.
In view of the very general application of the term cone in refer¬
ring to spore and seed-bearing organs of many groups of plants, it
may be noted that the use of the name conifer here and elsewhere
in this book will imply plants of the Coniferales. The group in¬
cludes many familiar evergreen trees such as pine, fir, spruce, red¬
wood, juniper, as well as a few that are deciduous such as bald
cypress and larch. There are in total some 550 extant species,
some of which belong to genera that are of very ancient vintage.
It seems to me safe to assert that the Coniferales, as a major group
of plants, exceeds all others in the combination of the length of
time that it has existed and its dominance over vast areas of the
earth’s surface. The oldest members of the order (from the Lower
Permian) flourished approximately 200 million years ago; their
evolution appears to have been rapid during the late Paleozoic and
early Mesozoic with relatively minor deviations since that time.
Until man began cutting away the coniferous forests for their choice
lumber they continued to compete favorably with other groups.

314
Telegram @tubscstds
CONIFEROPHYTA AND GINKGOPHYTA 315

The cordaite-conifer line appears to have originated independ¬

ently as seed plants; there is no reason to believe that they have
any common ancestors, at the seed plant level, with the pterido-
sperms, cycads, or bennettites.
The Ginkgophyta confronts us with many vexatious but intrigu¬
ing problems. Only one species survives today, the maidenhair
tree (Ginkgo biloba), and it is truly the dwindling remnant of a
race of plants that was quite diverse and, in places, a dominant
feature of the Mesozoic landscapes. The wood anatomy of Ginkgo
compares closely with that of the cordaites and conifers. Unfor¬
tunately, we know almost nothing about the stem structure of the
dozen and a half extinct genera of Mesozoic ginkgophytes, and it is
significant that the earliest presumed ginkgophyte in which repro¬
ductive organs are well preserved, the Lower Permian Trichopitys
heteromorpha, does not bear a close comparison with the contem¬
poraneous cordaites or conifers. In summarizing the relationships
of this fossil Florin has noted that
The Ginkgoinae, Cordaitinae, Coniferae, and Taxinae undoubtedly belong
to the same natural group of higher order . . . but they constitute parallel
evolutionary lines which probably were already separated from each other
in Upper Devonian or Lower Carboniferous times. At all events, a clear dif¬
ferentiation can be seen as far back as the available fossil records go. (1949,
pp.101-102)

I am inclined to go a little farther than this and proffer

the opinion that the ginkgophytes will ultimately be shown to have
evolved as a distinct and independent line of seed plants.

THE CONIFEROPHYTA

In order to demonstrate the relationships of the plants that are

usually classified in the Cordaitales and Coniferales it is not feas¬
ible to consider them under two separate headings. Thus, in the
pages that follow, certain aspects of the two, notably the seed
organs and wood structure, are described together although I have
tried to make clear the taxonomic status of the genera that are
dealt with.

The cordaitales, habit and vegetative structure

The cordaites formed one of the most imposing elements of the

Carboniferous and Permian forests. They were trees of monopo-
dial habit and attained a height of at least 100 feet; Seward has

Telegram @tubscstds
316 STUDIES IN PALEOBOTANY

recorded a trunk found near Newcastle, England, that was 72 feet

long and it was not complete. The branches bore spirally arranged
leaves (Fig. 11-1) that must have presented a close superficial
resemblance to those of an iris. They were elongate, strap-shaped
foliage organs which ranged, in different species, from 15 or 20 cm
to as much as a meter long and attained a maximum width of
about 15 cm. On a recent trip to the Minto coal field in south
central New Brunswick, Canada, I encountered fine specimens up
to 10 cm broad. These huge cordaite leaves along with fern,
pteridosperm, and articulate fossils bore vivid evidence of the con¬
trast with modern coniferous forests of that region.
The genus Cordaites was established for these foliar organs and
has also been applied to the stems; the latter, however, offer
special problems of nomenclature that will be considered separately.
The leaves were probably of a tough, leathery texture in life; in
some species the parallel veins alternate with I-shaped girders of
fibrous cells and in addition there may be a considerable develop¬
ment of fibrous tissue immediately within the epidermis. The
vascular bundle organization is essentially identical with that of a
cycad; the primary wood is predominantly centripetal (develops
toward the center) but is accompanied by a few centrifugal tracheids.
Of any extant trees the closest comparison in general habit seems
to be with certain species of Podocarpus and Araucaria; according
to Chamberlain the leaves of Podocarpus wallichianus may be 12.5
cm long and 3.5 cm broad, whereas the leaves of some species of
Araucaria are nearly as long and somewhat broader. However,
they fall short by a considerable margin of the larger leaved
species of Cordaites.
The stems that bore the Cordaites foliage have a transversely
chambered pith up to about 1.5 cm in diameter; it consisted of
discs of parenchyma alternating with lens-shaped gaps; this feature
is not confined to the cordaites, being present in some living plants
such as the walnuts. Beginning at the periphery of the pith
there is a transition in the primary wood from spiral, through
scalariform, to pitted tracheids; this may encompass a dozen cells.
The primary wood merges into the secondary wood uniformly so
that the distinction between primary and secondary wood is evi¬
dent only in a radial section. The pitting on the radial walls of the
secondary wood consists of quite regular vertical rows of closely
compacted hexagonal pits. The secondary wood is simple as far
as the number of cell types are concerned, there being only the ver¬
tically aligned tracheids and parenchyma cells of the rays; the

Telegram @tubscstds
CONIFEROPHYTA AND GINKGOPHYTA 317

latter are usually only one cell wide but may vary considerably in
height. The wood will be discussed further in a later section of this
chapter.

Reproductive organs of the cordaites and certain conifers

The genus Cordaianthus has been used for both pollen and
seed-bearing organs of the cordaites; in view of the close similarity
in the morphology of the two there seems to be justification for
this, but to the best of my knowledge well-preserved male and
female organs have not yet been correlated in a single species.
There is actually much less known about the pollen organs of both
cordaites and conifers than is known about the seed organs.
Some years ago Grand’Eury described cordaitean shoots with
leaves and male inflorescences attached. Figure 11-1 is taken from
his work. In addition to the leaves and a large vegetative bud
there is a male inflorescence apparently associated with each leaf.
Cordaitean inflorescences are common in coal-ball petrifactions of
Iowa and Kansas, but unfortunately well-preserved specimens are
of infrequent occurrence. Delevoryas has described specimens
from the Des Moines series of Kansas under the name Cordai¬
anthus concinnus consisting of a slender axis (Fig. 11-2A) 1 to 2 mm
in diameter which bore two rows of dwarf shoots that have been
referred to as cones or budlike bodies. It seems to me that the
phrase dwarf shoot is descriptive and morphologically accurate
and it will thus be used here for structures that are apparently
homologous. It may be noted that each shoot is borne in the axil
of a prominent bract that actually exceeds it in length.
Each dwarf shoot is about 6 mm long and consists of 25 to 40
closely imbricated scales. The latter are uniformly arranged on
the shoot axis in a spiral, although the distal fertile ones are some¬
what more slender than the others. There are usually six spor¬
angia, each 1 mm long, borne terminally on a fertile appendage;
the six sporangia are fused at the base.
Cordaianthus penjoni Renault which comes from the middle
Stephanian of France was of quite similar organization; the ap¬
pendages or scales of each dwarf shoot were more slender than
those of the American species and a higher number seem to have
been fertile. Florin, who has restudied them and added to Renault’s
original description, notes that all the scales are homologous as in
C. concinnus.
Before describing the seed-bearing organs a few words are ap¬
propriate concerning the history and significance of the investiga-

Telegram @tubscstds
318 STUDIES IN PALEOBOTANY

Telegram @tubscstds
Grand’Eury, 1877.)
CONIFEROPHYTA AND GINKGOPHYTA 319

Fig. 11-2. A. Cordaianthus concinnus; recon¬

struction of a portion of a male inflorescence
with two dwarf shoots; the bract is somewhat
longer than the shoot. B. Cordaianthus penjoni;
distal portion of a single microsporophyll with
six terminal sporangia, 25X. (A from Dele-
voryas, 1953; B from Florin, 1944.)

tions that have led to our present knowledge. The conifers have
long attracted the interest of layman and botanist; their economic
worth as a source of wood and pulp as well as their great esthetic
value have been widely appreciated. Among the most distinctive
features of many of them are their conspicuous seed-bearing cones,
and a great deal of effort has been expended in an attempt to
understand their morphology. Although there is considerable
diversity in the group, the seed cone of the Abietaceae (according
to the classification given by Chamberlain in his Gymnosperms,
Structure and Evolution) consists of a central axis with spirally
arranged “double appendages” which consist of a stiff, woody
ovuliferous scale bearing several seeds on its upper surface, and a
generally smaller and more delicate bract. The bract is quite
conspicuous in the cones of the fir (Abies) and larch (Larix) and is
readily detected with the naked eye when a cone is broken open
(Fig. 11-3); in the Douglas tree (Pseudotsuga taxifolia) the bract is
a conspicuous three-pronged appendage and is appreciably longer

Telegram @tubscstds
320 STUDIES IN PALEOBOTANY

than the ovuliferous scale which it subtends. The same organiza¬

tion is found in a pine cone although the bract is small and not as
readily discernible.
The question of the origin of this distinctive orientation of seeds,
ovuliferous scale, and bract has led to many studies on the com¬
parative morphology of the cone, and numerous theories have been
proposed which would require almost a page of print just to list.
Although there are still gaps in our knowledge, the elucidation of
the problem from a series of upper Paleozoic and lower Mesozoic
cordaites and conifers constitutes one of the great achievements in
evolutionary studies.

Seed organs of the cordaites

The female inflorescence of the cordaites has essentially the same

organization as the male one; it may reach a length of 30 cm and
it bears two rows of dwarf shoots, each in the axil of a bract. Two
main types are recognized:
Cordaianthus pseudofluitans Kidston is an older one (West¬
phalian) in which the dwarf shoots (Fig. 11-4) consist of numerous
scalelike appendages spirally arranged on the shoot axis. They are
all apparently homologous, but several of the distal ones are
greatly elongated, dichotomize several times, and bear two or more
terminal seeds that tend to be recurved toward the main axis of
the inflorescence.

Fig. 11-3. A. A modern fir cone, natural size; B. a single ovuliferous scale with the

Telegram @tubscstds
two-winged seeds on the upper (adaxial) surface; C. the under side of the scale showing
the bract (b).
CONIFEROPHYTA AND GINKGOPHYTA 321

Fig. 11-4. Cordaianthus pseudo-

fluitans. Part of an inflorescence
showing the axis and two rows of
dwarf shoots in several of which
the seeds are still intact, natural
size. (From Florin, 1944.)

In a geologically younger species, Cordaianthus zeilleri Renault

from the Stephanian, the general organization of the inflorescence
is similar, but each dwarf shoot bears not more than four fertile
appendages, and occasionally only one. They are unbranched and
produce but one terminal, erect ovule; the fertile appendages are
short and the ovule or ovules tend to be concealed among the
sterile appendages.

Seed organs of the conifers

Although the cordaite-conifer sequence in general seems to me

to be an established fact, it may be admitted that there is a con¬
spicuous break between the two; this is evident in both the foliage

Telegram @tubscstds
and the seed organs.
322 STUDIES IN PALEOBOTANY

Until rather recently much of the leafy twig material attributed

to the conifers, found in Upper Carboniferous and Lower Permian
horizons, was assigned to the genus Walchia. Although this is still
in use Florin has recognized, on the basis of cuticular structure and
seed cones, 14 species which are now placed in Lebachia (Fig.
11-5A) and one in Ernestiodendron (Fig. 11-5B); he notes that
there still remain 11 species of Walchia, most of which will ulti¬
mately prove referable to Lebachia. The leaves of Lebachia tend
to be more closely appressed to the stem axis than are those of
Ernestiodendron where they extend out at about 90° to the axis
and give the impression of having been quite stout and stiff. In
their general habit these plants were probably similar to the extant
Araucaria excelsa, the Norfolk Island pine. Several other genera,
known only from their foliage, are recognized from the Lower
Permian; in Carpentieria frondosa (Goeppert) Florin, for example,
the leaves are slender and deeply forked, but in their general
organization the leafy twigs appear “coniferous.”
The seed organs of the conifers are aggregated into distinct
cones in contrast to the open inflorescences of the cordaites. The
cone of Lebachia piniformis (Schlotheim) Florin is 6 or 7 cm long
and is composed of spirally arranged, closely imbricated bracts
with a forked tip. In the axil of each bract there is a dwarf shoot

Fig. 11-5. Foliage of two Lower Permian conifers: A. Lebachia piniformis; B. Ernes¬

Telegram @tubscstds
tiodendron filiciforme. Both about 5X. (From Florin, 1944.)
CONIFEROPHYTA AND GINKGOPHYTA 323

(Fig. 11-6) apparently homologous with the dwarf shoots in Cor-

daianthus; it consists in turn of several closely imbricated scales or
sterile appendages, with one bearing a terminal erect ovule. This
fertile appendage is usually so oriented that it is on the inner side
of the dwarf branch and not visible from the outside of the cone
as a whole.
It is apparent that a good deal of paleobotanical research is
needed to bridge the gap between the kind of foliage borne by the
cordaites and that of the Walchia type. The same holds true for
the female inflorescences; there are missing links between Cordai-
anthus with its dwarf shoots arranged on the central axis in two
rows and Lebachia in which the shoots are spirally arranged and
aggregated into a compact unit (cone); yet the homology is appar¬
ent. From Lebachia on we may observe a reduction (with appar¬
ently numerous evolutionary side lines) in the dwarf shoot to the
relatively simple ovuliferous scale of the modern Pinus and its
immediate relatives. For the most part the following descriptions
consider only the dwarf shoot.
Lebachia goeppertiana Florin (Lower Permian) has dwarf shoots
(Fig. 11-7B) with numerous sterile appendages and a single fertile
one, with an erect ovule, on the inner side. Florin points out that
there is a suggestion here of the shift to the flattening and bilateral
symmetry characteristic of the later conifers. Another Lower
Permian fossil, Ernestiodendron filiciforme (Schlotheim) Florin,
had cones about 2 cm long which were borne terminally on a next-
to-ultimate leafy branch. The bracts are forked as in Lebachia
and enclose dwarf shoots that are composed of several fertile ap¬
pendages only (Fig. 11-7D), each with an erect ovule. Walchia
(Ernestiodendron?) germanica Florin is very similar except that
the ovules are inverted (Fig. 11-7C). Under the name Walchio-
strobus sp. Florin has figured dwarf shoots that are somewhat
flattened. They include 20 to 30 sterile scales and 4 to 6 with ter¬
minal ovules; in one specimen the ovules were found erect and in
another they were inverted.
In the Upper Permian Pseudovoltzia liebeana (Geinitz) Florin
each dwarf shoot (Fig. 11-7G) is found as usual in the axil of a
bract. The axis of the shoot is quite rudimentary and there are
only five scales; three of which are larger than the other two, the
unit being distinctly flattened.
The ultimate in reduction of the dwarf shoot is found in Ullman-
nia bronnii Goeppert (Upper Permian); here the “shoot” is a disc¬
shaped structure (Fig. 11-7H) consisting of five wedge-shaped and

Telegram @tubscstds
324 STUDIES IN PALEOBOTANY

Fig. 11-6. Lebachia piniformis. A. Portion ot a leafy shoot with several ultimate
branchlets terminated by seed cones; B. a single dwarf shoot considerably enlarged,
showing the forked bract, the tips of which are missing; C. the dwarf shoot from the
inner (adaxial) side with one appendage bearing a terminal, erect seed, 10X. (From
Florin, 1944.)

Telegram @tubscstds
 CONIFEROPHYTA AND GINKGOPHYTA
325

Fig. 11 -7. Individual dwarf seed shoots of a cordaite and several conifers, all enlarged.
A. Cordaianthus pseudofluitans, 2X; B. Lebachia goeppertiana, 3.5X; C. Walchia
(Ernestiodendron?) germanica, 3X; D. Ernestiodendron fdiciforme, 2X; E. Walchio-
strobus sp., 3X; F. Walchiostrobus sp., 3X; G. Pseudovoltzia liebeana, 1.6X; H.
Ullmannia bronnii showing upper side (a) and under side (b), 1.4X. (From Florin,
1944.)

Telegram @tubscstds
326 STUDIES IN PALEOBOTANY

fused appendages. The fusion is so complete, however, that it

appears to be a single, nearly circular scale. One fertile appendage
is borne immediately above it with an inverted ovule.
Glyptolepis, an Upper Permian and Triassic genus, has shoots
consisting of five or more (depending on the species) sterile scales
which are even more strongly flattened than those of Pseudovoltzia
and there are only two fertile appendages. In the Triassic Voltzia
the sterile appendages are also strongly flattened and it differs
from previously noted genera in that the fertile appendages are
fused to the upper surface of the sterile ones. The dwarf shoot of
the Lower Jurassic Schizolepis consists of only three sterile ap¬
pendages; these are partially fused and three seed stalks or
modified fertile appendages are fused to their upper surface.
Finally, the Lower Jurassic Hirmeriella rhatoliassica Horhammer
appears to consist of a single sterile scale with two inverted ovules
borne on its upper surface. There is, however, some question as to
whether this is actually a simple (single) scale or composed of two
or three completely fused ones.
In summary, the fossil evidence indicates quite clearly that the
seed cones of the modern conifers, such as fir, larch, and pine, are
a much reduced inflorescence in which the ovuliferous scale and
seeds borne on its upper surface are the remnants of a radially sym¬
metrical dwarf shoot. As this evolutionary trend progressed,
numerous side lines branched out as is evident from the fact that
all the examples cited do not fit a perfect sequence.
The dwarf shoots, each in the axil of a bract, first appear aggre¬
gated into a distinct cone in the Lower Permian; these in turn
appear to have evolved from the Upper Carboniferous Cordai-
anthus in which the dwarf shoots are arranged in two rows along
a central axis. Racial development appears to have progressed
rapidly, for by Upper Permian times the “dwarf shoot” of Ullman-
nia is a simple disc-shaped structure not unlike the ovuliferous
scales of the modern abietinean conifers. Evolution in the Coni-
ferales since the Permian appears to have been a matter of minor
proliferations of certain lines and extinction of others. That some
of them have long been and still are a hardy lot is evident, but
their “morphological progress” essentially faded out some 150
million years ago.

Isolated seeds attributed to the Cordaitales

Numerous authors have figured and described Cordaianthus-

like inflorescences from compression specimens which show winged,

Telegram @tubscstds
CONIFEROPHYTA AND GINKGOPHYTA 327

bilaterally symmetrical seeds attached. This evidence combined

with the frequent association of such seed compressions with
cordaite foliage, and the occurrence of similarly shaped petrified
seeds with cordaitean stem and leaf remains in coal balls, is the
basis for the supposed relationship. Thus although one may safely
assign most petrified Carboniferous seeds that are radially sym¬
metrical in cross section to the pteridosperms and bilaterally sym¬
metrical ones to the cordaites, it must be regarded as a tentative
classification. The seeds described below are probably of cordaitean
affinities.
The genus Cardiocarpus, as defined by Seward in 1917, is based
on bilaterally symmetrical, petrified seeds from the Upper Car¬
boniferous which are quite common in Europe and in this country.
Cardiocarpus spinatus Graham is found in some abundance in coal
balls of Kansas and Iowa; the seed is about 15 mm long, and in a
median transverse section the diameters are approximately 15 X 10

Fig. 11-8. Cardiocarpus spinatus. A. Longitudinal section through the broad diam¬
eter of a seed, 4X; B. longitudinal section through narrow diameter of a specimen
containing a gametophyte, 5X; a, archegonium; n, nucellus; sc, sclerotesta; si, inner
sarcotesta; so, outer sarcotesta. (A photograph courtesy R. W. Baxter; B from

Telegram @tubscstds
Andrews and Felix, 1952.)
328 STUDIES IN PALEOBOTANY

mm. Thus, facing the broad side, the seeds are almost perfectly
circular, whereas from any other angle they appear broadly oval.
The integument is remarkable in its complexity and in Kansas
specimens recently described by Roth it is beautifully preserved
(Fig. 11-8). He recognizes five distinct tissue layers: A sarcotesta
composed of inner and outer zones; the cells of the outer layer are
large (about 165 /x in diameter) and thin-walled and are usually
filled with a brownish substance which gives them a rich amber
color; the inner sarcotesta differs from the outer only in the much
smaller size of the cells. The sclerotesta also consists of two zones,
the outermost one being conspicuous because of the dark-colored,
elongate, thick-walled cells and its tendency to proliferate outward
in the form of spines which occasionally extend to the epidermis.
In the seed that is illustrated they are not strongly developed, this
specimen having been selected for the well-preserved sarcotesta.
Another layer of thick-walled cells follows, the inner sclerotesta,
and within this is a tissue of thin-walled cells, the endotesta.
In mature seeds the nucellus appears as a thin, inconspicuous
band and in a few specimens rather well-preserved gametophytes
have been found (Fig. 11-8); the cells composing it are quite uni¬
form and thin-walled and at least two archegonia were formed at
the micropylar end. In his study of fossil seeds, published in
1881, Brongniart has described several with gametophytes pre¬
served; it is certainly one of the outstanding works of the nine¬
teenth century and should be consulted by any student of Paleozoic
plants.
The nucellus was free from the integument except at the base.
In European specimens of Cardiocarpus Seward describes the
vascular system as consisting of a central strand in the base of the
seed which gives off two branches that pass up through the inner
layer of the sarcotesta, and a second pair that penetrate the
peripheral tissue of the nucellus. So far as I know the nucellar
strands have not been observed in the American specimens.
The sclerotesta of these seeds was undoubtedly hard and highly
resistant to decay; in coal balls specimens are often found with
only this part remaining. There can be little doubt that most or
all of the Carboniferous seeds, known from impressions or compres¬
sions that are described as winged, are of this type; that is, the
tough sclerotesta was responsible for the impression of the central
body of the seed, whereas the fleshy sarcotesta formed the periph¬
eral outline or “wing”—since the sarcotesta was a uniform tissue

Telegram @tubscstds
around the entire seed the term wing is misleading. The number
of seeds of this type that have been described from impressions is
CONIFEROPHYTA AND GINKGOPHYTA 329

very considerable and suggests a greater diversity for the Cor-

daitales than is indicated by the leaves and wood.
Kamaraspermum leeanum Kern (Fig. 11-9) offers several unique
features and I would emphasize that it is only tentatively classified
as a cordaite seed. Several specimens were found in coal balls
from the Upper Carboniferous (Des Moines series) of Iowa; they
are approximately 11 mm long and in a median transverse section
(Fig. 11-9B) measure 11 by 3 mm. In its gross aspects two features
are especially striking: a basal cavity located below the nucellar
chamber; and a micropyle that consists of two distinct zones, a
massive one directly above the nucellus and a much narrower
elongate terminal part which had the form of a gently tapering,
flattened funnel.
The sequence of tissues composing the integument is unlike that
of other seeds in consisting of outer and inner sclerotic zones
separated by one of apparently more delicate cells. The inner
sclerotic layer may be seen to continue upward to form the inner
part of the proximal portion of the micropyle, and downward to
form a conspicuous lining for the basal chamber. It is thought
that this was an air chamber adding buoyancy which would have
aided dissemination by water. A seed with a similar basal cham¬
ber has been reported by Renault (Codonospermum olivaeforme)
from France, but it is a radially symmetrical seed and quite differ¬
ent in the structure of its integument from Kamaraspermum.

Some cordaitean and coniferous woods

Petrified logs and various sized fragments thereof that are attrib¬
utable to the cordaites, conifers, and ginkgo are frequently encoun¬
tered from the Devonian to the present. For the most part the
anatomy is rather stereotyped; many genera of extant conifers are
extremely difficult to identify on the basis of their secondary wood
alone, and although numerous studies have been devoted to this
type of fossil material it is my impression that they have not
entirely rewarded the effort involved. Since such wood specimens
are common and the paleobotanist is occasionally pressed for an
identification, a few comments seem appropriate. It is my intention
to mention a few plants in which distinctive characters are present
and to cite some of the problems presented by the remainder.
Most authorities have recognized three families under the
Cordaitales: Poroxyleae, Pityeae, Cordaiteae.
The genus Poroxylon (Poroxyleae) is known from slender stems

Telegram @tubscstds
which are rarely more than 2 cm in diameter, with large exarch
primary bundles around the periphery of the pith; the rays of the
330 STUDIES IN PALEOBOTANY

Fig. 11-9. Kamaraspermum leeanum. A. Longitudinal section through narrow diam¬

eter; B. median transverse section; be, basal chamber; i, integument; n, nucellus;
md, distal part of micropyle; mp, proximal part of micropyle. Both 13X. (From
Kern and Andrews, 1946.)

Telegram @tubscstds
CONIFEROPHYTA AND GINKGOPHYTA 331

secondary wood are quite high and several cells wide; hypodermal
sclerenchyma bands are present in the outer cortex. There is,
therefore, a resemblance to the stem of Lyginopteris, but the leaves
that are believed to have been borne on the Poroxylon stems are of
the Cordaites type and quite unlike those of the Carboniferous
pteridosperms. The viewpoint is expressed quite strongly else¬
where in this book that the coniferophytes and pteridosperms are
not at all closely related, but it seems only fair to mention this
plant which may possibly represent a link between the two great
groups.
The Pityeae includes several genera, some of them of great size,
such as Callixylon* in which the primary bundles are mesarch and
the secondary wood is for the most part of the coniferous type;
that is, the rays are usually not more than two cells broad, and the
tracheids have multiseriate-bordered pits on their radial walls.
The genus Callixylon is of unusual importance on several counts;
it includes several species distributed through the Upper Devonian
and they are found in numerous localities in the United States and
Europe. Most of our information on the American species has
come from the researches of C. A. Arnold who has described a
specimen from Indiana that is 3 feet broad at the base and tapers,
through a length of 9 feet, to about 18 inches at the upper end.
Trunks 5 feet in diameter have been reported from Oklahoma, a
size greatly exceeding that of any other Devonian plant. The
secondary wood is similar in its general organization to that of the
cordaites (compare Figs. 11-11A, B), consisting of tracheids and rays
only. In the specimen illustrated the rays are considerably larger
than those of the cordaitean wood. Particularly distinctive is the
grouped arrangement of the pits in the radial walls of the second¬
ary tracheids (Fig. 11-10). The foliage and reproductive organs
are not known and although it is a reasonable guess that Callixylon
should be regarded as of cordaitean affinities, and a seed plant, this
has not been proven.
In the Cordaiteae several genera have been recognized on the
basis of stem anatomy: Cordaites, in which the primary wood is
entirely centripetal, that is, the protoxylem is endarch, and
Mesoxylon in which some centripetal wood is present. These two
can be distinguished only with especially well-preserved specimens.
In view of the lack of information concerning the reproductive
organs of the Poroxyleae and Pityeae the above classification is of
course a tentative one.

Telegram @tubscstds
* See footnote on page 412.
332 STUDIES IN PALEOBOTANY

Fig. 11-10. Callixylon sp. A. Peripheral region of pith showing mesarch primary
bundle and some secondary wood, 60X; B. secondary tracheids in radial view showing
grouped arrangement of pits, about 150X.

Telegram @tubscstds
CONIFEROPHYTA AND GINKGOPHYTA 333

Of special concern in connection with the problem of wood iden¬

tification is the homogeneity of the secondary wood of plants in
the Pityeae and Cordaiteae. Fragmentary pieces of secondary
wood cannot be assigned with any confidence to a genus of these
two groups nor can it be separated from that of some coniferous
genera, notably Araucaria. (Callixylon with its grouped pits may
be regarded as an exception.) It is evident from the photographs
that there is little to differentiate between the cordaitean wood
(Upper Carboniferous) shown in Fig. 11-11A and the Cretaceous
araucarian in Fig. 11-llC. As a supplementary note it is of interest
to record the presence of a persistent leaf trace in Araucaria, that
is, the trace tissue continues to be formed by the cambium long
after the leaf has fallen off. They tend to become quite widely
separated in older branches.
Thus, in dealing with fragments of secondary wood of the
“cordaitean” type it has been general practice to use the generic
name Dadoxylon for Paleozoic specimens and Araucarioxylon for
woods of similar organization from the Mesozoic and Tertiary.
Many paleobotanists, including this writer, have been tempted to
describe new species of these nebulous genera. The result is that
there are a great many, but I am afraid that they contribute but
little to the sum total of our knowledge of the plant life of past
ages.
A few of the extant genera of the Coniferales possess features of
their secondary wood that are quite distinctive. For example,
Pinus (pine), Picea (spruce), Pseudotsuga (Douglas tree) and Larix
(larch) have conspicuous resin canals; the peripheral secretory cells
of those in pine are thin-walled whereas they are thick-walled in
the other three. In a few genera, including redwood, resin canals
are not normally present but may be produced as a result of
wounding. In Pseudotsuga and Taxus (yew) delicate tertiary
spiral thickenings characterize the tracheids. In Araucaria and
Agathis the pits in the radial walls of the tracheids tend to be
crowded and angular. However, the vast majority of living coni¬
fers present secondary wood in which generic distinctions are by
no means clear cut. Thus with much of the fossil coniferous wood
from Mesozoic and Tertiary horizons the names that have been
applied are little more than numbers which offer no clues to the
real identity of the plants concerned. For a detailed account of
coniferous woods the student is referred to Greguss’ recent
study.

Telegram @tubscstds
 STUDIES IN PALEOBOTANY
334

Telegram @tubscstds
CONIFEROPHYTA AND GINKGOPHYTA 335

El E2

Fig. 11-11. Transverse and tangential aspects of certain gymnosperm woods. Al, A2.
Dadoxylon sp., Upper Carboniferous; Bl, B2. Callixylon newberryi, Upper Devonian;
C1,C2. Araucarioxylon sp., Upper Cretaceous; Dl, D2. Pityoxylon sp. (probably a
pine), Eocene; El, E2. Ginkgo biloba (extant). (All 37X.)

THE GINKGOPHYTA

Probably no other plant has so effectively captured the imagina¬

tion of botanists and laymen, from the standpoint of its past his¬
tory, as has Ginkgo biloba. There are other extant plants with as
old or older fossil records but several factors seem to conspire to
give priority to ginkgo as the foremost “living fossil.” The tree is
unique in its appearance, possessing an abundance of characters

Telegram @tubscstds
that set it off sharply from any other; for the first 20 or 30 years
336 STUDIES IN PALEOBOTANY

it tends to do little other than grow straight up and the lower

branches gradually begin to spread out, a few of them becoming
quite massive in the course of a century. Slow in growth, it is a
shade tree to be planted for future generations, but even to a non¬
botanist its novel mode of growth in the early decades offers some
compensation. The fan-shaped leaves, unlike those of any other
plant, are borne singly along the terminal branches which may
shoot out a half meter in a single season; in their axils branch buds
are formed which for the most part grow very slowly, forming the
characteristic short shoots. Each one bears a dense cluster of
leaves at its apex (Fig. 11-12); the short shoots occasionally sprout
out to form new long shoots.
The plant is dioecious, that is, the short shoots of a tree may
produce a cluster of pollen-bearing catkins, whereas the seeds are

Fig 11-12. A branch of the extant Ginkgo biloba with three short shoots each bear¬
ing leaves and seeds, about VisX. (From Andrews, 1947.)

Telegram @tubscstds
CONIFEROPHYTA AND GINKGOPHYTA 337

found on a different tree (Fig. 11-12). The mature seeds are about
the size of a small apricot and the integument is composed of an
outer orange-colored fleshy portion and an inner, hard stony layer.
The fleshy coat is rich in butyric acid and when crushed emits an
odor that is unpleasant to some people. However, the seeds are
often produced in great abundance and tend to hang on for a
month or two after the leaves are shed; a female tree in this late
fall attire is a thing of great beauty and certainly outweighs the
disadvantage of the somewhat unsavory seeds when they have
fallen to the ground.
The ginkgo is known to have been cultivated for many centuries
in Chinese gardens and it is very probable that man’s interest in
the plant saved it from extinction—somewhat the reverse of the
usual fate of living things as human occupation has spread over
the earth. It has been reported as a native plant in a small area
in the Province of Chekiang in eastern China, but there is some
uncertainty as to whether or not such specimens are escapes from
gardens.
Western travelers first encountered the ginkgo in eastern China
and introduced it into Europe in the early eighteenth century; it
was brought to the United States a few decades later. It has since
been widely planted throughout the country and is slowly but
surely becoming one of our most valuable shade trees. One of
ginkgo’s great assets is the lack of any serious diseases, having ap¬
parently outlived any that may have bothered it in the past.
The fossil record of the ginkgophytes is an especially disconcert¬
ing one for two reasons:

Foliage has been found at many horizons and geographical points

in the northern hemisphere (less is known of southern records)
from the late Paleozoic to the present; some genera are quite dis¬
tinct but many of the fossils consist of leaves that fit into an
essentially continuous sequence beginning with deeply dissected
ones in the early Mesozoic to the nearly entire ones of a living
ginkgo. The difficulties of delimiting genera and species have per¬
plexed many a botanist.
In contrast to the abundance of foliage, reproductive organs are
almost nonexistent in the fossil record.

In view of the paucity of reproductive structures it is clearly not

possible to set up a satisfactory classification or to attempt to
establish evolutionary sequences other than on a tentative basis.

Telegram @tubscstds
338 STUDIES IN PALEOBOTANY

In order to deal with the evidence at hand it is proposed to con¬

sider fossil remains that are ginkgophytes or probable ginkgophytes
under the following headings:
Mesozoic and Tertiary fossils that seem closely related to the
extant Ginkgo biloba.
Mesozoic and Paleozoic fossils that are ginkgophytes but clearly
distinct at the generic level from Ginkgo.
Some Paleozoic fossils that are suggestive of being ancestral
ginkgophytes.

Mesozoic and Tertiary fossils probably closely related to G. biloba

It is well known that leaves on a living ginkgo may offer con¬

siderable variation in form, especially so with trees in the first
decade or two of growth. A mature plant displays much less
variation and although it is, therefore, reasonable to suppose that
the fossil record includes many species that are close to G. biloba
a clear-cut delimitation is not always possible. It seemed to me it
might prove most informative to consider approaches that certain
authorities have taken in their studies of these vexing fossils.
In dealing with fossils chiefly from western United States that
he regarded as “unmistakably assignable to Ginkgo” R. W. Brown
has suggested a classification of three species, as follows (see Fig.
11-13):

Ginkgo adiantoides (Unger) Heer. These are of reniform out¬

line, usually with a conspicuous apical notch; found most abun¬
dantly from the Paleocene onward.
Ginkgo lamariensis Ward. The outline is wedge-shaped and if
notched, only slightly so; predominantly of late Mesozoic age.
Ginkgo digitata (Brongniart) Heer. The general outline is
wedge-shaped, but the blade is deeply dissected; found in early
Mesozoic horizons.

Such a classification as applied to western United States is of

stratigraphical value provided one has reasonably abundant mate¬
rial from any one horizon. Brown emphasizes that this grouping
is one of expediency and that the three do not represent “biologic
species in the strict sense.” Perhaps the most obvious difficulty in
applying this grouping is that it is not easy, if indeed it is possible,
to distinguish in all cases between what is “unmistakably” a
Ginkgo and what is not. This is brought out in the next example.
Many students of the ginkgophytes have tended to classify fossil

Telegram @tubscstds
leaves that appear to be closely allied to Ginkgo biloba in the
CONIFEROPHYTA AND GINKGOPHYTA 339

C
Fig. 11-13. Ginkgophyte leaves from western United States; A. Ginkgo digitata;
B. Ginkgo lamariensis; C. Ginkgo adiantoides. (From Brown, 1943.)

genera Ginkgoites and Baiera. A remarkable assemblage of these

plants has been found in the Rhaetic-Jurassic rocks of East Green¬
land and I have chosen to deal with them in some detail, as con¬
stituting a representative and carefully studied collection, rather
than attempt a general survey of these two important genera.
Although it seems to be true in a general way that species of the
early to mid-Mesozoic, with deeply dissected leaves, gradually gave
way to ones with entire or nearly entire leaves, considerable
variety may be encountered within a narrow time range; this is
strikingly borne out by the Scoresby Sound collections.
Harris separates the two as follows;

Ginkgoites. A distinct petiole is present and the outline of the

lamina is semicircular.
Baiera. A distinct petiole is absent and the leaf is wedge-shaped.

From a detailed study of some 10 species of these two genera,

found in the Jurassic of Japan, Oishi cites the following characters
as diagnostic: the stomates occur on both upper and lower surfaces
but are much more numerous on the latter; guard cells are sunken
below the surface and the median slit (stoma) has no definite
orientation; the five to seven subsidiary cells form a circular group

Telegram @tubscstds
which arch over the guard cells. Oishi’s investigations agree with
340 STUDIES IN PALEOBOTANY

those of Harris in that the epidermal structures are indispensable

in delimiting species within the two genera, but the genera as such
cannot be separated in that way.
A few examples will indicate something of the diversity of these
plants in the Greenland flora.

Ginkgoites obovata (Nathorst) Seward. The lamina is entire or

nearly so and judging from the larger fragmentary specimens the
leaves attained a breadth of at least 16 cm.
Ginkgoites fimbriata Harris (Fig. 11-14A). The lamina is small,
about 1.5 cm broad and 2 cm wide, and is borne on a petiole about
4.5 cm long; quite distinctive are the delicate pointed teeth of the
terminal margin of the lobes. As to epidermal structure, there are
four to seven subsidiary cells surrounding the guard cells; the sur¬
face of the subsidiary cells is excessively thickened and forms an
irregular ring around the stomatal aperture.
Ginkgoites acosmia Harris (Fig. 14-2E). This is a dominant
species at certain horizons; the lamina generally tends to be deeply
divided into two halves and each one is subdivided into three lobes
which are in turn partially dissected.
Ginkgoites minuta (Nathorst) Harris (Fig. 11-14E). This is a
finely divided leaf of a type that has frequently been assigned to
Baiera; it would be impossible to distinguish it from that genus were
it not for the conspicuous petiole.
Ginkgoites hermelini (Hartz) Harris (Fig. 11-14F). Most of the
leaves tend to be divided into six lobes which are elongate-oval; the
stoma is surrounded by four to six subsidiary cells, each with
a papilla which points upward and does not project over the aper¬
ture in which the guard cells are sunken.
Baiera spectabilis Nathorst (Fig. 14-2D). This species seems to
have been very variable in size, degree of dissection, and width of
the segments; Harris figures one specimen that is 22 cm long.
Baiera boeggildiana Harris (Fig. 11-14C). The leaves are typi¬
cally 3 to 4 cm long with an apex that is divided into two rounded
lobes; the guard cells are located at the bottom of an oval pit and
the subsidiary cells bear papillae that project over the aperture.

Some other ginkgophytes of the Mesozoic and late Paleozoic

The most ancient fossil that may with some confidence be attrib¬
uted to the ginkgophyte line, and which offers some significant
clues concerning the evolution of the living species, is Trichopitys
heteromorpha from the Lower Permian of Lodeve in southern

Telegram @tubscstds
 CONIFEROPHYTA AND GINKGOPHYTA 341

Fig. 11-14. Ginkgophyte leaves from Scoresby Sound, East Greenland. A, B. Gink-
goites fLmbriata; A. leaf, 0.5X; B. stomate, 300X. C, D. Baiera boeggildiana; C. leaf,
IX; D. stomate, 400X. E. Ginkgoites minuta, %X. F, G. Ginkgoites hermelini;
F. leaf, 0.5X; G. stomate, 250X. H. Baiera leptophylla, 0.6X. (From Harris, 1935.)

France. Originally described by Saporta, our present under¬

standing of the plant is based on more recent investigations by
Florin; judging from photographs, his restoration (Fig. 11-15) offers
an accurate view of the fertile branches. They are known from
fragments about 8 mm in diameter which bore leaves that prob¬
ably were spirally arranged. The leaf is thought to have been more
or less terete and dichotomized several times. It is perhaps not
amiss to regard these foliar organs as representing a stage in gink¬
gophyte evolution comparable to that presented by Calamophyton
in the articulate series.
Seed-bearing branches are found in the axils of some of the

Telegram @tubscstds
leaves; these have been described by Florin as “sporangial trusses”
342 STUDIES IN PALEOBOTANY

Fig. 11-15. Trichopitys heteromorpha, a shoot fragment with leaves bearing seed-
branches in their axils, 0.7X. (From Florin, 1949.)

or “sporangiophoric complexes”; they probably represent a some¬

what “overtopped” dichotomous branch system with a single
inverted ovule borne at the end of each branch. Four to six ovules
per truss seems to have been most common, but as many as 20 have
been recorded.
The short shoot habit apparently had not evolved at this time
although it was established in some of the Mesozoic ginkgophytes.
Florin suggests that the reproductive complexes, which are present
in Trichopitys on long shoots (and this is occasionally so in Ginkgo),
were gradually transferred to the perennial short shoots so charac¬
teristic of the extant species. In the latter only one or two seeds

Telegram @tubscstds
usually mature on a single “complex,” but numerous instances are
CONIFEROPHYTA AND GINKGOPHYTA 343

on record in which several were initiated, forming a branch system

or truss very much like that of Trichopitys. In brief, the case for
regarding this fossil as an early member of the ginkgophyte line
seems to be a rather good one.
Sixteen or more genera of ginkgophytes have been recognized
from Mesozoic horizons ranging from the Triassic to the Upper
Cretaceous. One of the most interesting is Sphenobaiera, a genus
that extends from the Lower Permian to the Lower Cretaceous.
S. furcata (Heer) Florin (Fig. 11-16D) from the late Triassic near
Basle, Switzerland, bore foliage similar to that of Trichopitys, but
the leaves were arranged on short shoots as well as long shoots.
Microsporangiate organs have been found in organic connection
and appear to have been borne on short shoots; the central axis
divided to form branches which in turn bifurcated once or twice and
terminated in clusters of three to five sporangia.

Fig. 11-16. A, B. Ginkgoites lunzensis; A. short shoot with leaves; B. stomate. C, D.

Sphenobaiera furcata; C. long shoot with leaves; D. portion of short shoot with frag¬

Telegram @tubscstds
mentary microsporophyll. (From Krausel, 1943.)
344 STUDIES IN PALEOBOTANY

It is also of interest to find fossils in the Triassic that are

of quite advanced aspect; Ginkgoites lunzensis (Stur) Florin bore
leaves on short shoots that resemble those of some Jurassic species
and the epidermal structure tends to confirm this. Figure 11-16B
shows the stomates characteristically sunken and the papillae of
the subsidiary cells overhanging the aperture.
A considerable assemblage of ginkgophytes has been described
from high Arctic Jurassic rocks of Franz Joseph Land. The
leaves of Windwardia crookallii Florin are long, almost linear and
entire; they measure about 12 cm by 5 mm broad and have a
nearly truncate apex. The leaves of Stephanophyllum solmsi
(Seward) Florin are similar in form but have a rounded apex and
were larger, probably attaining a length of 30 cm and a breadth of
about 1 cm. The leaves of Arctobaiera flettii Florin are also of the
same general type, but are quite small, being about 6 cm long, 4
mm broad, and with a bluntly rounded apex; some are entire and
others are bifurcated for about one-third of their length. In all of
the above three the leaves are borne in clusters on short shoots.

Some Paleozoic fossils, possibly representing early ginkgophytes

Halle has described a species of Saportaea {S. nervosa) from the

Permian of Shansi, China (Fig. 11-17) in which the larger leaves
have petioles about 5 cm long and 4 mm broad; the blade attained a
breadth of at least 14 cm and was probably entire or nearly so in
life. The most distinctive feature of the leaf lies in the venation;
the vascular bundles of the petiole (probably two), instead of
dichotomizing uniformly as in a Ginkgo leaf, follow along the lateral
margins and produce secondary veins that take a nearly parallel
course toward the distal margin. This type of venation was

Fig. 11-17. Saportaea nerv¬

osa, a leaf restored. (From
specimens in the Swedish
Natural History Museum.)

Telegram @tubscstds
CONIFEROPHYTA AND GINKGOPHYTA 345

Fig. 11-18. Dichophyllum moorei, about 0.7X. (From Andrews, 1941.)

recorded previously for leaves from the Jurassic of Japan known

as Ginkgodium nathorsti Yokoyama; they are somewhat spoon¬
shaped, about 6.6 cm long and 2 cm broad, entire or lobed by a
median sinus and with a short petiole.
Ginkgodium was regarded by Seward in volume 4 of his Fossil
Plants as of uncertain affinities, but it is accepted by Harris as
ginkgoalean. Basically, both Ginkgodium and Saportaea seem to
me to be very close to the Ginkgo-Baiera type; the difference in
venation, although superficially rather striking in the former two,
is certainly not fundamental.
It is with some hesitation that the fossil considered next is
included in this section and it should be noted that it is by no means
certain that it is a ginkgophyte ancestor. Some years ago I spent
several days digging in an Upper Carboniferous shale bed in east¬
ern Kansas for specimens of Dichophyllum moorei Elias, a fossil that

Telegram @tubscstds
displays a particularly interesting branching pattern (Fig. 11-18).
346 STUDIES IN PALEOBOTANY

The available specimens show no distinct segregation of stems and

leaves; the main axis divides by more or less equal dichotomies
which ultimately terminate in slender, linear subdivisions. In some
cases they are rather densely clustered, which obscures the precise
details of the branching pattern. The plant was probably not a
fern, as a distinct cuticle is present although poorly preserved; the
only other evidence bearing on the affinities of the plant is the
association of seeds that are characterized by two prominent horn¬
like projections at the micropylar end.
It is only in the ferns that we have begun to obtain a reasonably
clear understanding of the evolution of the distinction between
stems and leaves; possibly Dichophyllum is an early ginkgophyte,
or more appropriately a “preginkgophyte” in which the characters
of the group are becoming established.
It certainly would be interesting to know what brought about
the decline of this unique assemblage of plants. Having flourished
quite profusely over a period of some 150 million years it would seem
they had been able to compete favorably with many other groups
and had become acclimatized to much climatic fluctuation. But
by Oligocene time only 2 out of a total of some 19 genera remain.
Although absent from western United States during the early
Tertiary, apparently due to the warm climate that prevailed,
Ginkgo reappears in the Oligocene, but the sole surviving species
disappeared at the close of the Miocene. It apparently held on a
little longer in Europe, having been recorded as occurring in late
Pliocene deposits.

REFERENCES

Literafure—Coniferophytes

Andrews, Henry N. Jr. and Charles J. Felix. 1952. The gametophyte of Cardiocarpus
spinatus Graham. Ann. Missouri Bot. Gard:, 39: 127-135.
Arnold, Chester A. 1930. The genus Callixylon from the Upper Devonian of central
and western New York. Papers Michigan Acad. Sci. Arts Letters, 11:1-50.
-. 1931. On Callixylon newberryi (Dawson) Elkins et Wieland. Univ. Michi¬
gan, Contrib. Mus. Paleont., 3: 207-232.
-. 1934. Callixylon whiteanum sp. nov., from the Woodford chert of Okla¬
homa. Bot. Gaz., 96: 180-185.
Beck, George F. 1945. Tertiary coniferous woods of western North America.
Northwest Science, 19: 67-69; 89-102.
Brongniart, Adolphe. 1881. Recherches sur les graines fossiles silifiees. Paris.

Telegram @tubscstds
Pp 34.
CONIFEROPHYTA AND GINKGOPHYTA 347

Chamberlain, Charles J. 1935. Gymnosperm Structure and Evolution. Univ. Chica¬

go press. 484 pp.
Delevoryas, Theodore. 1953. A new male cordaitean fructification from the Kansas
Carboniferous. Amer. Journ. Bot., 40: 144-150.
Florin, Rudolf. 1944. Die Koniferen des Oberkarbons und des Unteren Perms.
Paleontographica, 85B, sechstes heft: 365-456; siebentes heft: 457-654.
-. 1950. On female reproductive organs in the Cordaitinae. Acta Horti
Bergiani. 15: 111-134.
-. 1951. Evolution in Cordaites and Conifers. Acta Horti Bergiani. 15:
285-388.
Grand’Eury, Cyrille. 1877. Flore Carbonifere du Department de la Loire et du
centre de la France. Mem. Acad. Sci. Institut France, 24: 624 pp. Paris.
Greguss, Pal. 1955. Identification of living gymnosperms on the basis of xylotomy.
Budapest. 263 pp.
Jeffrey, Edward C. 1917. The Anatomy of Woody Plants. Univ. Chicago press.
478 pp.
Kern, Ellen M. and H. N. Andrews, Jr. 1946. Some petrified seeds from Iowa. Ann.
Missouri Bot. Gard., 33: 291-306.
Roth, Elmer A. 1955. The anatomy and modes of preservation of the genus Cardio-
carpus spinatus Graham. Univ. Kansas Sci. Bull., 37: 151-174.
Seward, Albert C. 1917, 1919. Fossil Plants, vol. Ill, 656 pp; vol. IV, 543 pp. Cam¬
bridge Univ. press.
Torrey, Ray E. 1923. The comparative anatomy and phylogeny of the Coniferales.
Part 3. Mesozoic and Tertiary coniferous woods. Mem. Boston Soc. Nat. Hist., 6:
41-106.

References—Ginkgophytes

Andrews, Henry N., Jr. 1941. Dichophyllum moorei and certain associated seeds.
Ann. Missouri Bot. Gard., 28: 375-384.
Brown, Roland W. 1943. Some prehistoric trees of the United States. Joum.
Forestry, 41: 861-868.
Dorf, Erling. 1958. The geological distribution of the Ginkgo family. Bull. Wagner
Free Inst. Sci., Philadelphia. 33(1): 1-10.
Florin, Rudolf. 1936. Die Fossilen Ginkgophyten von Franz-Joseph-Land. I. Spe-
zieller Teil. Palaeontographica, 81B: 71-173.
-. 1949. The morphology of Trichopitys heteromorpha Saporta, a seed
plant of Paleozoic age, and the evolution of the female flowers in the Ginkgoinae.
Acta Horti Bergiani, 15(5): 79-109.
Harris, Thomas M. 1935. The fossil flora of Scoresby Sound, East Greenland. Pt.
4. Ginkgoales, etc. Meddelelser om Gronland, 112 (1): 1-176.
Krausel, Richard. 1943. Die Ginkgophyten der Trias von Lunz in Nieder-Osterreich
und von neue Welt bei Basel. Paleontographica, 87B: 59-93.
Oishi, Saburo. 1933. A study on the cuticles of some Mesozoic gymnospermous
plants from China and Manchuria. Sci. Rept. Tohoku Imperial Univ., Sendai
(Geol. ser.), 12: 239-252.
Seward, A. C. 1938. The story of the Maindenhair tree. Sci. Progress, 32: 420-440.

Telegram @tubscstds
 1
some gymnospermous plants
of uncertain affinities

D t has been pointed out in previous chapters that the cor-

daite-conifer and the pteridosperm-cycadophyte lines in
all probability originated independently as seed plants; and within
the latter assemblage the morphological diversity is so great that
some paleobotanists regard it as composed of more than one divi¬
sion. I am strongly inclined toward the view that the seed has
evolved several times, for which further supporting evidence comes
from certain highly problematical fossils of gymnospermous affinities
that have been described in recent years. A few of these are con¬
sidered in the following pages, and although they cannot be assigned
with any degree of confidence to a recognized group of seed plants,
they are still important for several reasons. Some were numerically
significant elements in the landscapes of their day, all present intri¬
guing morphological features, and they indicate quite dramatically
the diversity of gymnospermous groups that is not revealed by the
extant flora.
It is safe to predict that many more such problematical fossils will
be excavated in the years to come; some may help to fill in notice¬
able gaps in the record as we know it and others will bring to light
previously unknown groups that flourished in the past and left no
close living relatives as a clue to their existence.
The plants included here are, therefore, a miscellaneous assem¬
blage and certainly not closely related to each other. Where a
reasonable guess can be made concerning their affinities it will be
offered, but it seems to me more scientific and more honest to admit
our ignorance rather than force fossils into some established taxo¬
nomic category that is highly questionable.
The order in which the several plants or plant groups are con¬

Telegram @tubscstds
sidered is of no significance.

348
GYMNOSPERMOUS PLANTS OF UNCERTAIN AFFINITIES 349

The Pentoxyleae

The name Pentoxyleae has been given to associated stems,

leaves, seed-bearing, and microsporangiate organs from a Jurassic
horizon in the Rajmahal Hills of India. They include a combination
of coniferous and cycadean characters, as well as others that are
wholly unique.
The stems, Pentoxylon sahnii Srivastava (Fig. 12-1A), reach sev¬
eral centimeters in diameter and contain a ring of five, or occasion¬
ally six, closely aggregated steles; each of these has a tangentially
elongated strand of primary wood enclosed in secondary wood that
is strongly endocentric in its development; that is, most of the

Fig. 12-1. The Pentoxyleae. A. Stelar system of stem showing endocentric develop¬
ment of secondary wood. B. Foliage-bearing branch. C. Pentoxylon sahni, female cone.

Telegram @tubscstds
Sahnia nipaniensis, male flower. (A-C from Sahni, 1948; D from Vishnu-Mittre, 1953.)
350 STUDIES IN PALEOBOTANY

cambial activity took place on the inner face of the steles. The
wood is typically coniferous, consisting of tracheids and small,
uniseriate rays; the tracheids have one or more rows of somewhat
crowded, circular-bordered pits on their radial walls. Well-defined
growth rings are present. Alternating with these steles are five
much smaller ones which consist chiefly of secondary wood.
Another genus of stems has been recognized (Nipanioxylon) in
which the number of steles is greater and the endocentric growth
is less strongly pronounced.
The stems bore branches 5 to 7.5 mm thick which are covered
with an armor of closely aggregated leaf cushions, and it is thought
that leaves, described under the name Nipaniophyllum raoi Sahni,
(Fig. 12-1B) were attached to these shoots; they attained a length
of 7 cm and were a little less than 1 cm broad. The vascular bun¬
dles compare quite closely in structure with those of a modem
cycad, whereas the stomata are of the bennettitalean type.
Two species of seed-bearing organs (Fig. 12-1C) have been de¬
scribed, Carnoconites compactum Srivastava and C. laxum Srivas-
tava; the peduncle of the infructescence divides into several branches
or pedicels, each of which terminates in a cone. The cone axis
contains a ring of five vascular strands and bore several closely
compacted seeds with a thick, fleshy integument and with the
micropyle directed out. No appendages are associated with the
seeds.
The microsporangiate organs or “male flowers” (Sahnia nipa-
niensis Vishnu-Mittre) were terminally borne on shoots (Fig.
12-ID) resembling those of Pentoxylon sahnii and consist of a ring
of filiform, spirally branched appendages which are fused at the
base to form a disc. Unilocular sporangia terminate short subdi¬
visions of the appendages and contain monocolpate or boat-shaped
pollen grains.
Our knowledge of the Pentoxyleae is the result of the work of
several Indian paleobotanists and in a summary account Sahni
made the following comment:
Some discoveries in science help, or appear to help, in the solution of the
old outstanding problems; others—and these are perhaps the most interest¬
ing—seem to create new difficulties in our path. My object here is to draw
attention to a recently recognized group of plants which defies classification
and presents a new problem in our understanding of the evolution of Gym-
nosperms. (1948, p. 47)

The supposition that these plant organs represent a single genus

Telegram @tubscstds
or closely related genera of plants seems well founded. Assuming
GYMNOSPERMOUS PLANTS OF UNCERTAIN AFFINITIES 351

this to be the case the Pentoxyleae are of particular interest for

their unique combination of characters; the wood is coniferous in
its detailed structure but is unlike any conifer in the gross organi¬
zation of the steles; the leaves show both cycadean and bennetti-
talean features, whereas the seed-bearing and microsporangiate
organs are peculiar unto themselves. Since the evidence as cited
in previous chapters indicates that there is no close relationship
between cycadophytes and coniferophytes, it seems most appropri¬
ate to regard the Pentoxyleae as a wholly distinct group of gymno-
sperms.

The Vojnovskyales

The greater part of the readily accessible knowledge of the later

Paleozoic floras comes from work that has been done in northwest¬
ern Europe during the past 150 years; to a lesser degree it is based
on American fossils and still less is known of the vast areas of the
USSR, China, and other parts of eastern Europe and Asia. Unfor¬
tunately, much of what has been published on the latter areas has
not been read extensively by western paleobotanists; but even a
hasty perusal of the Soviet literature dealing with Carboniferous
and Permian plants indicates the presence of many interesting
fossils and it is not surprising that wholly new types should be
encountered.
One of the most fascinating of these discoveries is Vojnovskya
paradoxa Neuburg, shown about natural size in Fig. 12-2. It was
probably shrubby or arborescent in habit, but the actual size of the
plant is not known. The leaves were fairly large fan-shaped organs
of the Nephropsis rhomboidea Neuburg type; scars where several
of these were attached are shown scattered over the surface of the
shoot. Several fertile branches or cones are preserved; these are
about 2.5 cm in diameter and include closely compacted, spirally
arranged microsporophylls, each of which bears two pairs of spor¬
angia. Scattered among them are megasporophylls described as
being similar to seeds known as Samaropsis; these are about 1 cm
long, bilaterally symmetrical, and with a notched tip.
Through the courtesy of Professor Neuburg I had the good for¬
tune to see this specimen on a recent visit to the Geological
Institute in Moscow. Additional information is particularly needed
to elaborate the details of the reproductive organs, but it immedi¬
ately gives one the impression of being totally unlike any other seed

Telegram @tubscstds
plant. The close association of microsporangiate organs and seeds,
352 STUDIES IN PALEOBOTANY

Fig. 12-2. Vojnovskya paradoxa; three branches with microsporangiate organs and
seeds. (From Neuburg, 1955.)

as well as the form of the leaves, are particularly distinctive and,

along with other unique details, segregate it from other Paleozoic
seed plants. One may see some resemblance to the bisexual cones
of the Bennettitales, but it is hardly a close one. The new order

Telegram @tubscstds
Vojnovskyales has been proposed by Neuburg for the plant.
GYMNOSPERMOUS PLANTS OF UNCERTAIN AFFINITIES 353

Czekanowskia and Associated Reproductive Organs

Czekanowskia is a rather well-known genus based on leaves that

are widely distributed in Jurassic rocks. Taking as an example C.
nathorsti Harris, from Greenland, we find that the leaves are
arranged in fascicles of about 15 on a short shoot. They reach a
length of 15 cm, dichotomize three or four times, the segments
being narrowly linear and only about 1 mm broad. They have
long been regarded by most paleobotanists as the foliage of a gink-
gophyte. Recently, however, Harris has described a unique seed¬
bearing organ, Leptostrobus, associated with the leaves. Although
not found in organic connection, the two occur together in Jurassic
rocks of East Greenland, in Yorkshire, and in eastern and south¬
western Siberia. Thus the evidence from association supports the
supposition that they may represent one and the same plant, and
there are also similarities in the cuticle structure.
Leptostrobus (Fig. 12-3) is an infructescence or loose cone with a
slender, unbranched axis; the appendages or “fruiting capsules” are
arranged at intervals of about 5 mm and consist of a pair of slightly
lobed valves up to 5 mm long and 5 mm wide with three to
five seeds borne on each inner surface. Harris notes that a single
one of these valves might be compared with the seed-bearing
structures of a number of other gymnospermous groups, such as a

Fig. 12-3. Leptostrobus longus. A. Restoration of a

part of the cone; B. a “capsule” in longitudinal sec¬
tion, megaspores in solid black. (From Harris, 1951.)

Telegram @tubscstds
354 STUDIES IN PALEOBOTANY

single Caytonia fruit, a Cycas megasporophyll, a Cupressus cone

scale, or a several-seeded cupule of a Lower Carboniferous pterido-
sperm, but the two-valved nature of the appendages of Leptostro-
bus places it in a morphological category of its own and seemingly
unrelated to any other group.

Glossopteris and Certain Allied Fossils

To do full justice to the literature that has developed around

Glossopteris and certain fossils that are considered as related to it
would require almost a separate volume. They have been the sub¬
ject of much discussion in very recent years because of problemati¬
cal reproductive organs that are found attached to the leaves (Fig.
12-4, 16-5B, C).
The name Glossopteris was first proposed by Brongniart in 1828
and it was considered for some time to be a fern. It is the key
plant of the so-called Glossopteris flora being distributed through
South Africa, India, Australia, and South America in rocks of late
Paleozoic and early Triassic age. The leaves are spathulate, ovate,
or linear-lanceolate and vary in length from 3 to 40 cm; in some
species a distinct petiole is present whereas in others it is lacking.
The side veins that pass out toward the margin from the conspic¬
uous midrib form an anastomosing network, and in some species
the leaves have been found arranged in whorls on slender stems.
Smaller specimens are quite similar in appearance to the leaflets
of Sagenopteris and on several occasions have been mistaken for
them.
Recent studies of the epidermal structure of several species leave
no doubt as to its being a seed plant and several types of unique
reproductive organs have been discovered attached to the leaves.
None of these is as well preserved as might be desired and they
have given rise to much speculation.
Mrs. Plumstead has described a considerable collection (Fig.
12-4) from the Lower Permian of Vereeniging in the Transvaal in
which the leaves bear reproductive organs thought to be seed-bear¬
ing cupules. Five such species are described under the name
Scutum. The structures in question are described as a round,
lanceolate or ovate, bilaterally symmetrical, cupules attached by a
pedicel to the midrib of the leaf; evidence has been offered to sup¬
port the view that the cupule was closed after fertilization, thus
allowing the possibility of angiosperm relationships.

Telegram @tubscstds
 GYMNOSPERMOUS PLANTS OF UNCERTAIN AFFINITIES 355

Fig. 12-4. Reproductive organs attached to or associated with Glossopteris leaves:

A. Scutum stowanum; B. Lanceolatus lerouxides; C. Scutum leslium; D. Scutum
dutoitides. (From Plumstead, 1952, 1956.)

Telegram @tubscstds
356 STUDIES IN PALEOBOTANY

What is referred to as the fertile half of an open cupule bears a

central head in which are embedded several small oval sacs which
range from 1 to 2 mm in size, with the entire head surrounded by
a wing. In a later study of these plants Mrs. Plumstead described
specimens which suggest that the fertile appendages may have
been bisexual, that is, with pollen organs as well as seeds. No
reasonable doubt exists that these are reproductive bodies, but
beyond that there is considerable divergence of opinion. It has
been suggested, for example, that the fructification is not a cupu-
late head but rather a strobilus, the supposed cupule actually con¬
sisting of laminate extensions of the closely aggregated “seeds.”
More recently H. H. Thomas has described sterile and fertile
leaves of the Glossopteris type from Natal; these are from a some¬
what younger Permian horizon than Scutum and because of the
distinctive nature of the reproductive bodies they have been given
the name Lidgettonia africana Thomas. The leaves, although
complete specimens are not known, range up to 15 cm long and 3
cm broad; the secondary veins are numerous, crowded, and anasto¬
mosing as is usual in Glossopteris foliage. Associated with these
are much smaller, so-called scale leaves, which bear several um¬
brella-shaped structures. The latter are borne in two rows, there
being a total of six or seven of the appendages, each consisting of a
stalk about 5 mm long and a peltate disc 5 mm in diameter. Both
isolated seeds and sporangia occur associated with the “fertile”
leaves and it seems very likely that they were borne by the peltate
appendages but had been shed prior to fossilization.
Gangamopteris is another characteristic leaf genus of the Glos¬
sopteris flora; it is similar in shape to Glossopteris and, indeed,
some authorities have questioned the validity of a generic distinc¬
tion. Gangamopteris leaves, however, lack a midrib and according
to Seward the venation, although reticulate, shows greater uni¬
formity in the size and shape of the meshes than does Glossopteris;
they are also somewhat larger, reaching a length of nearly 40 cm.
Specialized, apparently reproductive, appendages have recently
been found attached to the Gangamopteris leaf at its petiolar
region. These were originally described by Zeiller a half century
ago under the name Ottokaria, and the more recently discovered
specimens consist of a stalk several centimeters long which bore a
pair of disc-shaped appendages at its distal end; the margins of
these had a row of flat appendages, lending a superficial appear¬
ance to a sunflower head with its conspicuous ray florets. Repro¬
ductive bodies, possibly seeds, were borne on the inner face of one

Telegram @tubscstds
disc and the other may have served as a protective device.
GYMNOSPERMOUS PLANTS OF UNCERTAIN AFFINITIES 357

Although none of the presumed reproductive bodies found at¬

tached to Glossopteris and Gangamopteris leaves can be said to be
well preserved, they are sufficiently so as to indicate an assemblage
of plants that are not as closely related as the leaf structure might
imply. Glossopteris has been classified in some accounts as a
pteridosperm; Plumstead has suggested the new class Glosso-
pteridae to include plants that bore Ottokaria, Scutum, and related
fructifications. In the light of the cited facts, interesting but
tantalizing, one can only say that we are dealing with a diverse
group and one that is not closely related to any other.

Rhexoxylon

The genus Rhexoxylon is based on stems of rather complex

anatomy which attain a diameter of 25 cm. Three species are
known; R. africanum Bancroft, and R. tetrapteroides Walton from
the Triassic (Molteno beds) of South Africa, and R. priestleyi
(Seward) Walton from South Victoria Land, Antarctica. The fol¬
lowing notations are based on Walton’s description of R. tetraptero¬
ides (Fig. 12-5).
The largest specimen is a trunk fragment 10 feet long which
varies from 2 to 7 inches in diameter. A rather large pith occupies
the central part of the stem and consists of a parenchymatous
ground tissue which includes secretory ducts, sclerotic nests, and a

Telegram @tubscstds
Fig. 12-5. Rhexoxylon tetrapteroides. (From Walton, 1923.)
358 STUDIES IN PALEOBOTANY

few small irregularly scattered vascular strands. Around the out¬

side of the pith is a ring of strands or small steles which make up
what is known as the perimedullary system. Each strand usually
consists of two segments, an outer one that developed centrifugally
and an inner centripetal one; these are composed of only secondary
wood and are separated by a narrow gap which probably was occu¬
pied by parenchyma in life.
Outside the perimedullary steles is another ring of strands of
much greater radial extent; there is a small amount of primary
wood at their inner apex and the rest is of secondary origin.
In its detailed structure the wood, which is mostly secondary
throughout, is typically coniferous. The tracheids have one or two
rows of circular-bordered pits on their radial walls; the rays are
uniseriate and up to 15 cells high. Vascular strands have also been
observed running horizontally between the wood sectors; these
apparently originate from the perimedullary bundles and may
represent leaf traces.
The general organization of the wood in R. tetrapteroides is rem¬
iniscent of certain modern angiospermous lianas (vines) such as
Tetrapteris, Banisteria, and Wilbrandia. There is, however, no
reason to believe that the fossil was angiospermous, being appreci¬
ably larger than any of the modern lianas.
There is considerable variation in the three species. In R. priest-
leyi the outer ring of wood was quite compact, indeed almost as much
so as a coniferous stem, and the perimedullary system is not as
strongly developed as in R. tetrapteroides. The opposite tendency
is with R. africanum; in a specimen from Portuguese East Africa
the “pith” region (internal to the outer ring of steles) is 12 cm in
diameter and contains numerous bundles of the perimedullary
type, whereas the steles of the outer ring are widely separated.
There is thus some reason to suspect that the three may actually
represent growth stages of but one species.
When first discovered it was suggested that Rhexoxylon might
be related to the medullosas, but the detailed organization of the
stems is wholly different. It seems reasonable to suppose that
these are gymnospermous stems but no more can be said with
certainty.

Hydropteridangium

Few, if any, paleobotanical investigations have ever brought to

Telegram @tubscstds
light so many previously unknown plants as Harris’ account of the
GYMNOSPERMOUS PLANTS OF UNCERTAIN AFFINITIES 359

Rhaetic-Liassic deposits of East Greenland. Of the many problem¬

atical fossils that he has described Hydropteridangium marsilioides
Halle seems especially appropriate to this chapter. It is a micro-
sporangiate organ (Fig. 12-6A) with a main axis about 4 cm long
and 2 to 3 mm thick. The axis gives off lateral branches in all
planes and these in turn divide irregularly in a three-dimensional
pattern, ultimately terminating in capsules 3x2 mm. The cap¬
sule is a two-valved structure with about seven elongate sporangia
(probably embedded) on the inner face of each valve; the sporangia
contain numerous winged microspores.
On the basis of specimens found at Bjuf, Sweden, the plant was
originally likened to the water-fern Marsilia, but Harris points out

Fig. 12-6. A-D. Hydropteridangium marsilioides. A. Reconstruction of part of sporo-

phyll, about 2X; B. longitudinal section of capsule; C. transverse section of unopened
capsule; D. reconstruction of entire capsule. E. Ptilozamites nilssoni, about V2X.

Telegram @tubscstds
(From Harris, 1935.)
360 STUDIES IN PALEOBOTANY

that the branching pattern, a gymnospermous type of cuticle,

presence of only one kind of spore (Marsilia and the other water-
ferns are heterosporous), and the association of these organs with
the cycadlike foliage known as Ptilozamites nilssoni Nathorst (Fig.
12-6E), all render such a classification untenable.
The Ptilozamites fronds are unique in that the rachis dichoto¬
mizes twice, but otherwise they have the gross aspect of cycadophyte
foliage; the stomates are of the gymnospermous type but with
peculiarities unto themselves, and in general the cuticle structure
compares closely with that of Hydropteridangium.
Perhaps the most likely possibility, though admittedly little
more than a guess, is that Hydropteridangium-Ptilozamites repre¬
sent a unique group of gymnospermous plants derived from a com¬
mon ancester with the cycadophytes.

Nucellangium

It is perhaps verging on the indiscrete to include a discussion of

Nucellangium in a textbook. I have done so because it presents
morphological features that may be of some importance and be¬
cause they epitomize the more vexing types of problems encountered
in paleobotanical investigations.
Coal ball collections from an Upper Carboniferous horizon near
Des Moines, Iowa, contain abundant “seedlike” bodies of two dis¬
tinctly different kinds, yet there is reason to suppose that they
were borne by a single species of plant. They are abundantly
represented in the flora of that horizon; in a petrifaction 6 inches
in diameter a half dozen or more specimens are often revealed in a
saw cut and, because of the distinctive preservation of the epider¬
mal cells, they can sometimes be broken out of the coal balls intact.
For reasons that will be pointed out they will be called “normal”
and “proliferated” reproductive bodies in the following description.
The “normal seeds” (Fig. 12-7A) are about 12 mm long and broadly
ovate, the diameters in mid-section being approximately 6x9 mm.
They were attached at one end as indicated by a small circular
scar 0.5 mm in diameter, whereas the distal end tapers to a blunt
point. In cross section (Fig. 12-7B) the following sequence of tis¬
sues is encountered. Outermost is an epidermis of radially elongate,
thick-walled cells; it is probable that the original chemistry of the
walls is but little changed and the sharp physical and chemical
difference between them and the surrounding matrix is responsible

Telegram @tubscstds
for the ease with which they can be broken out intact. Next there
 GYMNOSPERMOUS PLANTS OF UNCERTAIN AFFINITIES 361

is a broad zone of nearly isodiametric, thin-walled cells which com¬

poses a major portion of the wall; then follows a prominent layer
of thicker walled cells that are at least twenty times as long as
they are broad, bearing numerous oval-shaped pits. The end walls
are transverse and do not appear to be truly tracheidal but seem

Fig. 12-7. Nucellangium glabrum; A. a “normal seed” showing point of attachment

at the base, 4X; B. median transverse section showing half of the organ; C. section

Telegram @tubscstds
through a “proliferated” organ, 8X. (From Andrews, 1949.)
362 STUDIES IN PALEOBOTANY

to be designed to serve as conducting elements. Finally there is an

innermost layer of thin-walled cells that is usually poorly preserved.
Within this tissue system is a distinct yellow band which is identified
as a megaspore membrane.
A small, vascular strand enters the fossil at the point of attach¬
ment and divides to produce two strands which extend up to the
distal end.
The so-called proliferated bodies (Fig. 12-7C) are slightly larger
than the normal ones, measuring about 10 X 13 mm, and they
possess an epidermis and outer parenchymatous layer that are
identical with the normal ones. The resemblance ends at this
point, for the parenchymatous tissue proliferates into numerous
fingerlike lobes which are directed in a general way toward the
center. It may be noted that the lobes are not preservation arti¬
facts as they have a clearly defined epidermis; moreover, the
peripheral portion of this parenchymatous tissue is vascularized
and a very slender strand extends out into each lobe.
The normal bodies may be interpreted as a complex sporangium,
with a single megaspore, or as a true seed. These possibilities will
be considered further in Chapter 13, but the former interpretation
may be tentatively accepted for two reasons: There is no evidence
of a micropyle at the distal end and it has not been possible to dis¬
tinguish between integument and nucellus.
The morphology of the proliferated bodies is even more vexing.
It seems virtually certain that these were borne on the same plant,
this conclusion being based on the close association of the two,
their similarity in size, and particularly the identity of the periph¬
eral tissues as described. It is therefore suggested that the
proliferations are an aposporous growth of the parenchymatous
tissue and that the body functioned as a kind of “gemma,” or in
somewhat simpler terms, that they represent a specialized asexual
reproductive structure.
There is little doubt that future investigations will turn up a
growing number of these “taxonomically difficult” plants adding
interest and complexity to our knowledge of the plant world.
Some of them will eventually be fitted into the evolutionary scheme
of things, whereas others may always remain enigmatic.

REFERENCES

Andrews, Henry N., Jr. 1949. Nucellangium, a new genus of fossil seeds previously

Telegram @tubscstds
assigned to Lepidocarpon. Ann. Missouri Bot. Gard., 36: 479-504.
GYMNOSPERMOUS PLANTS OF UNCERTAIN AFFINITIES 363

-and Mamay, S. H. 1955. Some recent advances in morphological palaeo-

botany. Phytomorphology, 5: 372-393.
Harris, Thomas M. 1951. The fructification of Czekanowskia and its allies. Phil.
Trans. Roy. Soc. London, 235B: 483-508.
-. 1935. The fossil flora of Scoresby Sound, East Greenland. Part 4. Gink-
goales, coniferales, lycopodiales and isolated fructifications. Meddelelser om Gron-
land, 112 (1): 1-176.
Neuburg, Maria F. 1955. New representatives of the Lower Permian Angara flora.
Doklady Acad. Sci. USSR, 102 (2): 613-616.
Pant, D. D. 1958. The structure of some leaves and fructifications of the Glosso-
pteris flora of Tanganyika. Bull. Brit. Mus. Nat. Hist. (Geology), 3: 127-175.
Plumstead, Edna P. 1952. Description of two new genera and six species of fructifi¬
cations borne on Glossopteris leaves. Trans. Geol. Soc. South Africa, 55: 281-328.
-. 1956. Bisexual fructifications on Glossopteris leaves from South Africa.
Palaeontographica, 100B: 1-25.
-. 1956. On Ottokaria, the fructification of Gangamopteris. Trans. Geol.
Soc. South Africa, 59: 211-236.
Sahni, Birbal. 1948. The Pentoxyleae: a new group of Jurassic gymnosperms from
the Rajmahal Hills of India. Bot. Gaz., 110: 47-80.
Sen, J. 1955. On some fructifications borne on Glossopteris leaves. Bot. Not.
(Lund), 108: 244-252.
-. 1955. A Glossopteris bearing sori-like structures. Nature, 176: 742-743.
Thomas, H. Hamshaw. 1958. Lidgettonia, a new type of fertile Glossopteris. Bull.
British Mus. Nat. Hist. (Geology), 3: 179-189.
Vishnu-Mittre. 1953. A male flower of the Pentoxyleae with remarks on the struc¬
ture of the female cones of the group. The Palaeobotanist, 2: 75-84.
Walton, John. 1923. On Rhexoxylon, Bancroft—a Triassic genus of plants exhibiting
a liana-type of vascular organization. Phil. Trans. Roy. Soc. London, 212B: 79-109.
-. 1956. Rhexoxylon and Dadoxylon from the Lower Shire region of Nyasa-
land and Portuguese East Africa. Colonial Geol. Min. Resources, London, 6:
159-168.

Telegram @tubscstds
heterospory and the
evolution of the seed

T here are several attractive problems centered around the

origin of the seed which seem worth considering in a sepa¬
rate chapter. For the purpose of discussion three questions may be
cited as being particularly pertinent, although it is neither possible
nor desirable to segregate them as wholly distinct from one another.

Are seeds of the gymnospermous groups strictly homologous;

that is, was the seed evolved once, with later modifications produc¬
ing the divergence in morphology that is now known, or has it
evolved several times?
What is the significance of the many heterosporous pteridophytic
groups in relation to seed evolution?
How early do seed groups appear in the fossil record?

Some exceptionally interesting discoveries have been made in

recent years which bear on these questions; they are not wholly
answered, but the accumulated evidence is impressive.
One point in particular should be emphasized—in all of the dis¬
cussion that follows, relative to seed evolution, it is the pterido-
sperm seed (or seeds that are presumed to be referable to that
group) that is chiefly alluded to. What bearing this may have on
the corresponding organ of the angiosperms is beyond the scope of
this account. Thus, in an attempt to answer the questions, at
least in part, the order of discussion will be as follows:

Heterospory in the pteridophytic groups

Earliest evidence of heterospory
Earliest seeds or supposed seeds
Seed evolution in the pteridosperms
Telomic concept

Telegram @tubscstds
Nucellar modification concept

364
HETEROSPORY AND THE EVOLUTION OF THE SEED 365

Heterospory in the Pteridophytic Groups

There seems to be general agreement that there has been an

evolutionary sequence from homospory to heterospory to the seed
habit. “General agreement” does not necessarily mean that we
are dealing with a truth, founded on clear-cut evidence, but in this
case there is a certain amount of support for such a viewpoint and
it is accepted as a reasonable working hypothesis. Homospory is
the rule with all of the earlier land plants; so far as I am aware
heterosporous plants do not appear in the fossil record until Upper
Devonian (or possibly upper Middle Devonian) times, and unques¬
tioned seeds have not been found below the Lower Carboniferous.
It is very possible that heterosporous plants may ultimately be
found in the mid-Devonian and that seeds will be identified in the
late Devonian, but even if such discoveries are made in due course,
the chronological order will remain unchanged.
It is perhaps not generally appreciated that heterospory is now
known in quite a number of highly divergent pteridophytic groups.
Most of them have been discussed in earlier chapters and they are
only listed below; a few cases that have not found an appropriate
place previously are described in more detail.
Arthrophyta. Heterospory is known in at least two Upper Car¬
boniferous species of Calamostachys: C. casheana (Fig. 9-10) and
C. americana. The large numbers of megaspores per megasporang¬
ium and relatively slight size difference (as compared, for example,
with certain lycopods) indicate that heterospory exists here in a
rudimentary stage.
Lycopodophyta. It does not seem amiss to say that heterospory
has run rampant in this group. Many petrified and compression
cone specimens have been recorded which reveal a general trend,
starting in the Lower Carboniferous, from large numbers of mega¬
spores to only one per sporangium, and heterospory is known in
both herbaceous and arborescent forms. Enclosure of the mega¬
sporangium by the sporophyll resulted in a seed in both herbaceous
lycopods (Miadesmia) and arborescent ones (Lepidocarpon).
In a study of 33 extant species of Selaginella, Duerden has
reported much variation in megaspore number. In five species the
numbers (of megaspores per megasporangium) were found to be in
excess of four; seven species showed numbers less than four,
whereas in a few both less and more than four were encountered.
A few examples will illustrate the variability: in S. willdenowii

Telegram @tubscstds
four cases were encountered in which a megasporangium produced
366 STUDIES IN PALEOBOTANY

only one megaspore, one case of 16 megaspores, two cases of 36

megaspores, and four cases of 42 megaspores; in S. lobbii the fol¬
lowing numbers were encountered: 8, 12, 14, 16, 20 and 24; in S.
inaequalifolia v. perelegans it was found that a megasporangium
usually contained two large spores and two small ones, and more
rarely one large and three small. In a summary comment Duer-
den notes:
The occurrence in Selaginella of megasporangia containing many compar¬
atively small spores suggests a condition possibly not far advanced beyond
the homosporous state, and on the other hand, the sporangia with fewer,
comparatively large spores, indicate an advance in the direction of the seed
habit.” (1929, p. 456)

Archaeopteris. Heterospory is reported in one species, A. lati-

folia, from the Upper Devonian.
Stauropteris. The Lower Carboniferous S. burntislandica (Fig.
13-5) presents a unique type of heterospory in which only two
megaspores matured. This is now known to have been a widely
distributed plant.
Enigmophyton. The cones attributed to this very curious fossil
(Fig. 2-11) from Spitsbergen were heterosporous; since they were
not found in organic connection with the leafy shoots some doubt
exists. The age is possibly as low as Middle Devonian.
Noeggerathia. Heterospory is clearly established here (Fig.
4-17) and the noeggerathias are certainly not closely related to any
other pteridophytic plants.
Protopitys. This case has not been described in previous chap¬
ters and is therefore given in some detail at this point. What may
represent an early stage in the evolution of heterospory has been
reported in Protopitys scotica Walton from the Lower Carboniferous
of Dunbartonshire, Scotland. Protopitys was originally established
on petrified stems from Silesia (P. buchiana Goeppert) in which
there is a characteristic elongate-oval pith with leaf traces depart¬
ing from the opposite ends, the leaf arrangement being distichous.
A good deal of secondary wood was formed resulting in stems up
to a foot in diameter. The tracheids have elongate, almost scalari-
form, pits, and the rays are uniseriate and but a few cells high. It
has enjoyed a varied taxonomic career, having been assigned
tentatively to the cordaites, pteridosperms, and Filicales by differ¬
ent authors.
Protopitys scotica is based on a small branch fragment 7 cm long
and 6 mm in diameter which presents the same general type of

Telegram @tubscstds
anatomy as P. buchiana, although only a small amount of second-
HETEROSPORY AND THE EVOLUTION OF THE SEED 367

ary wood is formed. The specimen bears two lateral branches

which subdivide by several unequal dichotomies and these are
terminated by sporangia. The sporangia are all about 3 mm long
and show no distinct annulus. The contained spores are 82 /x in
diameter, but in some they are 147 /x, and in a few others the
spores are more or less intermediate in size, measuring about 98 /x
in diameter.
Evidence from isolated spores. Chaloner has recently described
seven new species of megaspores from the Upper Devonian of
Ellesmere Island, Arctic Canada. Two of these are similar to cer¬
tain Carboniferous megaspores attributed to the lycopods, but the
assemblage displays great range in morphology. For example,
Ocksisporites is a triangular spore of 560 /x broad with a wide equa¬
torial flange which is scalloped into sharp teeth; Nikitinsporites,
which attains a diameter of 610 /x, is covered with exceptionally
stout spines 200 [x long and 30 /x in diameter. It is improbable
that such spores are lycopods, and being unlike those of any
previously known plant, they suggest a greater variety of hetero-
sporous plants than was formerly known to exist in the Upper
Devonian.
Nikitin has reported several distinct spore types from the
Voronezh region, some 500 kilometers north of the Sea of Azov; of
particular interest are associated megaspores and microspores
called Kryshtofovichia africani. The megasporangia are broadly
ellipsoidal, measuring 2 to 3 mm long and 1.7 to 2.5 mm thick.
The mature free megaspores are quite striking; the main body is
spherical with three long and broad valves extending outward,
forming a chamber called the androcamera. In addition the spores
bore about a dozen stout spines up to 0.7 mm long with a terminal
anchorlike hook. Microsporangia were found closely appressed to
the megasporangia. They are somewhat smaller and are described
as being occupied by a gelatinous, granulous material in which
clusters of microspores were embedded. The microspores were
found in some abundance attached to the megaspores and in the
androcamera. The nature of the spore-bearing fructification as a
whole is not known, but the unique form of the megaspores sug¬
gests a plant group that is distinct from any cited previously.
The diversity of heterosporous plants in the Lower Carboniferous
has been demonstrated by Dijkstra and others by discoveries of an
amazing variety of megaspores (Fig. 13-1). Although only a very
few localities have been investigated, 60 species have been described

Telegram @tubscstds
chiefly from Egypt and the Moscow basin. The spores vary in size
368 STUDIES IN PALEOBOTANY

Fig. 13-1. Some Lower Carboniferous megaspores: A. Cystosporites barbatus; B. Tri-

letes hamatus; C. Triletes agninus; D. Cystosporites strictus, abortive form; E. Triletes
furius; F. Cystosporites strictus, fertile form; G. Triletes acuminatus. (From Dijkstra
and Pierart, 1957.)

from about mm in diameter to almost 3 mm; some are smooth-

walled, others have minute warty protuberances, massive bulbous,
or long spiny ones, whereas still others display a distinct equatorial
fringe of branching appendages. Many of these spores are prob¬

Telegram @tubscstds
ably lycopods, but the affinities of some are uncertain.
HETEROSPORY AND THE EVOLUTION OF THE SEED 369

In view of the great diversity of the plants enumerated above, it

seems certain that heterospory has arisen independently many
times and in many different groups of plants. It also seems axio¬
matic that it has evolved from generally simpler homosporous
plants in the articulates and lycopods; of the other heterosporous
examples we have little or no information concerning their ancestry.

The Earliest Seeds or Seedlike Fossils

Although the presence of seeds in the Devonian has not been posi¬
tively demonstrated, it is possible that they did exist in the latter
part of that period or that they were actively evolving at that
time. The reasons for this supposition lie in the well developed
and, in some cases, rather elaborate seeds in the Lower Carbonif¬
erous, and the presence in the Devonian of fossils that are sugges¬
tive of being such and yet are not sufficiently well preserved to
settle the matter. Several of these intriguing structures have been
found in the Upper Devonian of Belgium:
Moresnetia zalesskyi Stockmans (Fig. 13-2A) is a rather slender
dichotomizing branch system, with the ultimate divisions termi¬
nated by an elongate, deeply lobed structure. The latter tends to
be rather strongly divided into two main lobes which are in turn
lobed less deeply.
Condrusia rumex Stockmans (Fig. 13-2C) is an ovoid body which
seems to have been partially enclosed by two narrow, spoon-shaped
structures attached at the base. Another species, C. minor Stock-
mans, is equally interesting and problematical and I am inclined to
believe should be placed in a distinct genus. It is an ovoid body
which was attached by a fairly stout stalk, very little of which
remains. At the other end there are two conspicuous hornlike
projections and a greatly elongate central spine.
Xenotheca bertrandi Stockmans (Fig. 13-2D, E) is a slender,
cupulelike organ with several lobes that tend to flare rather
slightly. This genus was originally described from the Devonian
of Devon, England.
It seems probable that all of these fossils were reproductive struc¬
tures, but beyond this one can only speculate. Condrusia rumex is
suggestive of a sporangium (nucellus?) partially enclosed by two
sterile lobes. C. minor is a most perplexing structure. It may be re¬
called, however, that a number of Carboniferous seeds are bicornute
in this fashion or have a lobed integument that might produce such

Telegram @tubscstds
an effect. It may be suggested that the long terminal “spine” is
370 STUDIES IN PALEOBOTANY

Fig. 13-2. Some problematical Upper

Devonian fossils from Belgium. A. Mor-
esnetia zalesskyi; B. Condrusia minor;
C. Condrusia rumex; D, E. Xenotheca
bertrandi. All about 3X. (Natural
History Museum, Brussels.)

the attenuate apex of the micropyle, although Dr. Stockmans has

interpreted this as the peduncle. As to Xenotheca, it seems most
likely that this was either a cupule which enclosed one or more
seeds or a microsporangiate organ. One of the specimens in the
collections of the Brussels Museum reveals a rather heavy carbona¬
ceous film in the basal portion that is suggestive of a seed.
Moresnetia may have been simply a foliar organ, but it would have
been a strange kind of leaf with such small “blades” borne out at
the extremeties of a slender dichotomous branch system.

Seed Evolution in the Pteridosperms—with Particular Reference

to the Nature of the Integument

There is admittedly a great gap in our knowledge between any

of the heterosporous organs considered above and the seeds of the
pteridosperms. Basically the pteridosperm seeds are regarded as
being composed of a nucellus, within which a megaspore develops,

Telegram @tubscstds
which is enclosed by an integument, and in some this is in turn
HETEROSPORY AND THE EVOLUTION OF THE SEED 371

partially surrounded by the cupule. Petrified seeds that have been

attributed to the pteridosperms may be separated into two groups
on the basis of the relationship between nucellus and integument;
in one group the two are said to be fused together and in the other
they are separate or nearly so. An understanding of these struc¬
tures, and their relationship to the cupule (or “outer integument”),
seems fundamental to a consideration of the evolution of the seed
and will be explored in the following pages. First, certain repre¬
sentative examples will be described briefly as a basis for discussion.

Seeds with nucellus and integument united

Lagenostoma lomaxi has been described in Chapter 5 (Figs. 5-2,

6-2A); its major features may be summarized briefly: the nucellus
and integument are free only at the micropylar end; the integu¬
ment is penetrated by a ring of nine vascular strands and is not
lobed; the cupule is divided for about two-thirds of its length into
several lobes which are also vascularized.
Sphaerostoma ovale (Williamson) Benson (Fig. 13-3B) from the
Lower Carboniferous of Pettycur, Scotland, has a very closely

Fig. 13-3. A. Tyliosperma orbiculatum, 12X; B. Sphaerostoma ovale, 2.5X; both in

Telegram @tubscstds
median longitudinal section; c, cupule; i, integument; n, nucellus.
372 STUDIES IN PALEOBOTANY

investing cupule. In the drawing I have taken the liberty to show

it spread out slightly in order to render it clearly recognizable from
the integument, the distal portion of which is lobed. The cupule
is vascularized and an inner ring of strands are described as pass¬
ing up through the integument; however, judging from the original
illustrations of Miss Benson it is not easy to be sure whether these
strands actually are running through the innermost part of the
integument or within the nucellus.
Tyliosperma orbiculatum (Fig. 13-3A) presents us with features
that are particularly cogent to the problem at hand. The distal
part of the integument is lobed and the cupule, which is more
deeply divided, is not as completely investing as is the case in
Sphaerostoma or Lagenostoma. A set of vascular strands is pres¬
ent in the outer part of the nucellus rather than the integument.
The epidermis extending over the apical portion of the nucellus is
continuous with the epidermis of the integument—to allow com¬
plete freedom of thought in making morphological interpretations
it may be more appropriate to note that there is a continuous
epidermis over the tissues that would ordinarily be interpreted as
integument and nucellus!

Seeds with integument and nucellus separate

The seeds included under this heading are ones that are classi¬
fied, in many previous accounts, as the Trigonocarpales. Since, as
noted in Chapter 5, there is some reason to believe that they were
borne on medullosan plants the seeds are accordingly dealt with
here as plants belonging to the Medullosaceae. The cupule (or a
structure recognized as such) is not present; the integument is
often massive, complex, and is free from the nucellus except in the
basalmost part of the seed. Numerous vascular strands rim
through the length of the integument and the nucellus is heavily
vascularized. Pachytesta illinoense (Fig. 5-13) and Stephano-
spermum elongatum may be regarded as representative; descrip¬
tions of both are given in Chapter 5.
There are two obvious problems of comparative morphology in¬
volved in an interpretation of the seeds considered above: What
are the nature and origin of the integument and cupule in the first
group? What are the nature and origin of the vascularized nucel¬
lus of the seeds of the second group (the medullosan seeds)?

The telomic concept

Telegram @tubscstds
As long ago as 1904 Miss Benson suggested that the pterido-
sperm seed may be interpreted as a synangium in which all but
HETEROSPORY AND THE EVOLUTION OF THE SEED 373

one sporangium has become sterilized. This implies that the “seed
state” was preceded by one in which a sporangium became closely
invested by an encircling ring of sporangia which lost their spore-
producing capacity and fused together to form the integument.
More recently this concept has been somewhat simplified by
Walton who proposes that the encircling elements (telomes) were
sterile, that is, ordinary branch tips. There is now a great deal of
evidence available in support of this theory.
Reference may be made first to the Silurian Hedeia corymbosa;
the restoration given in Fig. 2-5 is based on specimens, discovered
by Cookson, which demonstrated the three-dimensional nature of
the terminal branchlets of the plant. It is not implied that this is
a direct ancestor of the pteridospermous plants of the Carbonif¬
erous but merely as evidence of such a terminal clustering of sterile
and fertile telomes. It seems pertinent to divert our attention for
a moment to the pteridospermous microsporangia; Halle has sug¬
gested that the Whittleseya type synangium (see Chapter 5) is in
fact a ring of fertile telomes which fused to form a “cyclic fertile
syntelome.” Codonotheca, in which the sporangia are only par¬
tially fused, may be regarded as an earlier stage, whereas Aulaco-
theca, in which there is only a small spical opening, may be a more
advanced stage than Whittleseya. Thus in this sequence of
fructifications the pattern of evolution is as clear-cut as may be
expected from paleontological evidence.
Returning to the seeds, we find it significant to observe in Lag-
enostoma that the integument, although not lobed, is distinctly
chambered and a vascular strand penetrates each unit; in Tylio-
sperma the integument is partially lobed. An important link in
this line of evidence has been supplied by Long who has found Lower
Carboniferous seeds in which the integument is free from the nu-
cellus and divided into eight lobes for the greater part of its length.
I am indebted to Mr. Long for allowing me to include this especially
pertinent piece of information since the results of his investigations
were in press at the time my manuscript was completed. Several
stages in this proposed evolutionary sequence are shown in Fig.
13-4.
Evidence of a similar nature indicates a nearly identical evolu¬
tionary pattern for the cupule. In Tyliosperma it is deeply lobed;
in Lagenostoma and Calathospermum fusion has taken place
through the basal third of its length; and in Sphaerostoma there
is only slight lobing, if any, at the distal extremity.

Telegram @tubscstds
It seems appropriate to add a few correlating notes concerning
the origin of the integument in the cordaite seeds. The latter are
374 STUDIES IN PALEOBOTANY

Fig. 13-4. Diagrams showing suggested stages in the enclosure of a sporangium to

form nucellus and integument of a pteridosperm seed. A. Terminal branchlets
(telomes), based on a plant of the Rhynia type. B. Corymbose clustering of fertile
and sterile telomes with one sporangium assuming a central location; suggested by
Hedeia corymbosa. C. Hypothetical stage; one central enlarged sporangium is sur¬
rounded by telomes that are tending to web together. D. A seed with sporangium
(nucellus) enclosed by lobed integument. (Adapted in part from Walton, 1940.)

generally distinguished from pteridosperm seeds by their bilateral

symmetry, that is, they appear strongly flattened when observed
in cross section. It is not certain, however, that this character
holds true in all cases.

Telegram @tubscstds
Florin has observed that the fertile dwarf shoots of Cordaian-
thus williamsoni bear some appendages that are forked at the tip;
HETEROSPORY AND THE EVOLUTION OF THE SEED 375

a few of these have the primordium of a megasporangium, or what

may be interpreted as an aborted nucellus, between the two lobes.
It is therefore concluded that the integument of the flattened cor-
daitean seed originated by the forking of the megasporophyll and
the two resultant branches ultimately completely enclosed the
megasporangium.
It is evident here that the two lobes cannot be homologized with
sporangia. Furthermore, in view of the fact that the cordaitean
plant as a whole is so completely different from any pteridosperm,
however primitive, there can be no doubt that this apparent enclo¬
sure of a megasporangium by a ring of telomes (in the pterido-
sperms) or a pair of telomes (in the cordaites) took place
independently in the two groups.

The nucellar modification concept

There are several features of the pteridosperm seeds that are not
readily explained by the telomic concept. One is the problem of the
vascularized nucellus of the medullosan seeds; another is the fusion
of integument and nucellus in the seeds of the Lyginopteridaceae.
In a recent discussion on the evolution of the seed Walton has
pointed out that the so-called integument of the medullosan seeds
may really be equivalent to the cupule of Lagenostoma, Sphaero-
stoma, and Calathospermum (and Tyliosperma, which has played
an important role in our discussion, may be included); and in turn
the outer vascularized part of the nucellus may represent an (inner)
integument that is fused with a much reduced nucellus.
While admitting this as a valid explanation another possible
approach is offered which may be termed the “nucellar modifica¬
tion concept.” As a starting point reference may be made to the
remarkable heterosporous coenopterid fern, Stauropteris burntis-
landica, that was introduced in Chapter 3. The megasporangium
of this plant (Fig. 13-5) presents several exceptional features that
are pertinent to the present discussion:

A slender vascular strand runs through the sterile portion; so

far as I am aware this is the most extensive vascularization of any
fern or fernlike sporangium.
In a few specimens the tapering apex has been observed to ter¬
minate in a minute opening.
Much of the sporangium (the basal half or more) is sterile,
possibly serving as a food reservoir.

Telegram @tubscstds
Quite clearly this is no ordinary megasporangium, yet it is not
quite a seed. However, a seed might conceivably have evolved
376 STUDIES IN PALEOBOTANY

Fig. 13-5. Stauropteris burntislandica, a mega¬

sporangium in median longitudinal view showing
apical pore, two megaspores, and slender vascular
strand in center of lower half, about 80X.

from it along the lines indicated in Fig. 13-6. The first figure (A)
represents a diagrammatic section through the Stauropteris meg¬
asporangium showing the apical opening, the two megaspores
(enclosed in a cuticular envelope, the nature of which is problemat¬
ical), and the central vascular strand. Figures B and C are hypo¬
thetical steps involving:

The reduction of two megaspores to one, and increase in size of

the latter.
The “sinking” of the megaspore toward the basal part of
the sporangium.

Telegram @tubscstds
 HETEROSPORY AND THE EVOLUTION OF THE SEED 377

The displacment of the vascular strand and its division into two
or more branches which are forced to grow up around the mega¬
spore.

From this point (Fig. 13-6C) two possible lines of development

are postulated:
With only a slight modification of the sporangium wall and apex,
the vascularized nucellus of the medullosan type seed (Fig. D) is
reached. It might also be noted that this appears to be the kind
of organ that is represented by the “normal” Nucellangium seeds
described in the previous chapter.
With somewhat more extensive modification or specialization of
tissues in the peripheral wall of the megasporangium (C) a seed of

Fig. 1 3-6. Possible stages in the evolution of a pteridosperm seed based on an origin
from a vascularized megasporangium. A. Diagrammatic longitudinal view of Staurop-
teris burntislandica. B, C. Reduction to single megaspore, “sinking” of megaspore,
and branching of vascular strand. D. Seed of Pachytesta type with vascularized
nucellus; integument not shown. E. Seed in which megasporangium wall is differen¬
tiated into “integument” (vascularized) and “nucellus.” (Based in part on suggestions
by John Walton and Sergius H. Mamay.)

Telegram @tubscstds
378 STUDIES IN PALEOBOTANY

the lyginopterid type (with “fused” integument and nucellus) is

the end result (Fig. 13-6E).
In brief, this concept seems to offer a reasonable explanation for
the vascularization of the nucellus of the medullosan seeds, the
apparent “fusion” of integument and nucellus in the lyginopterid
seeds, and the particularly vexing matter of the continuous apical
epidermis in Tyliosperma, a feature that is difficult to account for
if the structures interpreted as nucellus and integument actually
developed as separate entities.
In presenting this concept I have drawn upon suggestions that
have been offered by Dr. S. H. Mamay and Professor John Walton.
It is not intended as a cure-all to explain the evolution of the pter-
idosperm seed; in fact, it is obvious that at best it only answers
some of the problems that confront us.

REFERENCES

Arnold, Chester A. 1935. On seedlike structures associated with Archaeopteris from

the Upper Devonian of northern Pennsylvania. Univ. Michigan, Contrib. Mus.
Paleont., 4: 283-286.
-. 1948. Paleozoic seeds. Bot. Rev., 14: 450-472.
Benson, Margaret J. 1904. Telangium scotti, a new species of Telangium (Calym-
matotheca) showing structure. Ann. Bot., 13: 161-177.
-. 1914. Sphaerostoma ovale (Conostoma ovale et intermedium, William¬
son), a Lower Carboniferous ovule from Pettycur, Fifeshire, Scotland. Trans. Roy.
Soc. Edinburgh, 50 (1): 1-15.
Chaloner, William G. 1958. Isolated megaspore tetrads of Stauropteris burntisland-
ica. Ann. Bot., 22: 197-204.
-. 1959. Devonian megaspores from Arctic Canada. Palaeontology, 1:
321-332.
Cookson, Isabel. 1949. Yeringian (Lower Devonian) plant remains from Lilydale,
Victoria, with notes on a collection from a new locality in the Siluro-Devonian
sequence. Mem. Nat. Mus., Melbourne, 16:117-130.
Dijkstra, S. J. 1956. Lower Carboniferous megaspores. Mededelingen Geol. Sticht-
ing (Heerlen), n.s., No. 10: 1-18.
-and Pierart, P. 1957. Lower Carboniferous megaspores from the Moscow
basin. Mededelingen Geol. Stichting (Heerlen), n.s., No. 11: 5-19.
Duerden, H. 1929. Variation in megaspore number in Selaginella. Ann. Bot., 43:
451-457.
Hoskins, John H. and Aureal T. Cross. 1946. Studies in the Trigonocarpales. Part
II. Taxonomic problems and a revision of the genus Pachytesta. Amer. Mid. Nat.,
36: 331-361.
Mamay, Sergius H. 1954. Two new plant genera of Pennsylvanian age from Kansas
coal balls. U. S. Geol. Survey Prof. Paper, 254-D: 81-95.
Martens, P. 1956. Un facteur evolutif neglige: le bee nucellaire de l’ovule. Rev. Gen.

Telegram @tubscstds
Bot., 63: 529.
HETEROSPORY AND THE EVOLUTION OF THE SEED 379

Nikitin, P. A. 1934. Fossil plants of the Petino horizon of the Devonian of the
Voronezh Region. I. Kryshtofovichia africani nov. gen. et sp. Bull. Acad. ScL,
USSR No. 7: 1079-1092.
Stewart, Wilson N. 1954. The structure and affinities of Pachytesta illinoense comb,
nov. Amer. Journ. Bot., 41: 500-508.
Stockmans, Frangois. 1948. Vegetaux du Devonien Superieur de la Belgique. Mem.
Mus. Roy. Hist. Nat. Belgique, Mem. 110: 1-85.
Thomas, H. Hamshaw. 1958. Fossil plants and evolution. Journ. Linn. Soc. London,
Zoology and Botany, 56: 123-135.
Walton, John. 1953. The evolution of the ovule in the pteridosperms. Adv. Sci., 10:
223-230. (British Assn. Adv. Sci., No. 38)
-. 1957. On Protopitys (Goppert): with a description of a fertile specimen
“Protopitys scotica” sp. nov. from the calciferous sandstone series of Dunbarton¬
shire. Trans. Roy. Soc. Edinburgh, 63: 333-339.

Telegram @tubscstds
 14
fossil plants of the
Arctic and Antarctic

Among the peculiar traits of mankind are his curiosity in

/^&a state of affairs other than what he finds here and
now, his tendency to discredit phenomena that fall beyond the
limitations of his own experiences, and his struggle to interpret the
events of long ago and far away that he cannot hope to experience
himself. A classic example of this is our interest in climates of the
past and particularly of the polar regions. Some of the most bitter
tales of man’s effort to extend his influence over the surface of the
globe come from his explorations beyond the Arctic circle, and
what we know of that territory has been won very largely in the
past century and a half. It is perhaps not surprising that, since
only men of high intelligence as well as tremendous physical endur¬
ance could penetrate the most northerly latitudes up until a few
years ago, they should have been intent on acquiring knowledge
from all possible sources. Thus many expeditions within the past
century have collected animal and plant remains where fossiliferous
horizons were exposed. I think it is safe to say that one of the
greatest surprises that the Arctic has yielded is the clear-cut
evidence of lands richly forested in past ages which now are almost
completely barren of plant life, or lie covered with snow and ice
most of the year.
It has seemed to me that the paleobotany of the Arctic is of
rather special interest and worth considering in a separate chapter,
but in so doing it is not possible to avoid a little repetition here
and there. It is clear, from the paleobotanical record as well as
other evidence, that the present rigorous climate of the northern
latitudes is very much the exception for that area during the past
three or four hundred million years. We have a good deal of
evidence from many geological horizons, but the difficulties and

Telegram @tubscstds
expense involved in Arctic travel have resulted in small collections,

380
FOSSIL PLANTS OF THE ARCTIC AND ANTARCTIC 381

and these have been acquired, for the most part, quite incidentally
to the particular expedition’s main objectives. Only in certain
parts of Greenland, Spitsbergen, and Alaska have paleobotanists
actually done the collecting and carried out studies that are com¬
prehensive. It is my impression, gained from the literature and
from studies of the magnificent collections in the Swedish Natural
History Museum, that we are only on the threshold of understand¬
ing the fossil floras of the Arctic; here is a vast expanse of the
earth’s surface awaiting paleobotanical investigation.
It has not been possible for the most part to pinpoint the age of
fossil plant collections from the far north as accurately as in the
more readily accessible parts of the earth, some having been identi¬
fied, for example, as simply Tertiary or Carboniferous.

Tertiary

Fossil plants of many ages have been found in Spitsbergen. An

older Tertiary flora has been described recently by Schloemer-Jager
from the Brogger Peninsula in which Sequoia langsdorfii, Meta¬
sequoia occidentalis, and Cercidophyllum arcticum (Katsura tree)
are the dominant elements. Two other conifers, Taxodium and
Taiwania schaeferi, were present in the flora and Ginkgo was well
represented. An aquatic monocot, Acorus brachystachys, is
reported, as well as the following dicots: Cercidophyllum crenatum,
Hamamelis clarus (witch-hazel), Acer spitzbergense (maple),
Aesculus longipedunculus, Planera ulmifolia (water-elm), doubtful
species of Alnus (alder), and Vitis (grape) plus a few indeterminate
dicot remains. Although some of the fossils in this flora are frag¬
mentary, it is noted that the branches of Metasequoia and Taxo¬
dium are quite well preserved and could not have been transported
any great distance. It is suggested that the plants indicate a tem¬
perate climate with fairly high rainfall.
Little has been done as yet with the microfossil floras of the
Arctic, but initial investigations indicate that they will ultimately
contribute a great deal. Spore studies of Paleocene-Eocene coals
of West Spitsbergen have revealed 52 different types, some of which
correlate with plants previously known from macrofossils and
others are new. Where Nathorst (1910) reported 6 pteridophytes,
Manum (1954) found 12 spore types, most of which are referred to
the Osmundaceae and Polypodiaceae, with one possible lycopod.
Numerous gymnosperms are represented by pollen, including ones

Telegram @tubscstds
that are tentatively identified as Abies and Sciadopitys, genera
382 STUDIES IN PALEOBOTANY

which are not known from macrofossils. The maple, water-lily,

heath, sweet gale, and palm families are evident from the fossil
pollen; the presence of palms in the Tertiary of Spitsbergen had
not been known from the macrofossil record.
In 1911 a very readable account of Arctic floras appeared from
the pen of A. G. Nathorst. In describing Tertiary plants from
Haron, near Waigattet, Greenland, he notes that:

In one of the deposits the fossil flora consisted almost exclusively of leaves
of the Maple {Acer), crowded like those which cover the ground in autumn,
and among these leaves large samaras, like those of A. otopteryx Gp., occur.
In another bed the tuff was formed of cinders and small lapilli, and the way
in which the vegetable fragments were embedded leads one to suppose that
the branches, leaves, and fruits of the trees were broken off by a shower of
cinders and lapilli. A medley of silicified branches of different sizes occurs,
and among them are the cones of the Spruce, the nuts of the Walnut {Jug-
lans), and the Hickory (Carya), with the leaves of Ginkgo, etc. (1911, p. 225)

Nearly a century ago Oswald Heer, in the course of his studies

of Arctic fossils, described a fairly extensive flora from Iceland; it
included Equisetum, several species of pine, alder, birch, hazelnut,
oak, elm, sycamore, grape, tulip tree (Liriodendron), maple, wal¬
nut, and several others. Many of Heer’s fossils are described from
fragmentary specimens and some of the identifications may be in
error. However, there are some quite well-preserved Tertiary
leaves from Iceland in the Swedish Natural History Museum, ap¬
parently collected at a later date, and it is evident that the Terti¬
ary flora of that island was more luxuriant than the present one.
Tertiary plants have been found in numerous localities in Alaska,
a summary account having been given by Hollick in 1936. He
recognized 327 species plus 42 that were identified only to the
genus. Many elements in the floras are suggestive of a temperate
climate. One hundred and twenty-six species (33%) are distributed
in 12 amentiferous genera: Populus (poplar), Salix (willow), Myrica,
Juglans (walnut), Betula (birch), Carpinus (hornbeam), Alnus
(alder), Fagus (beech), Castanea (chestnut), Quercus (oak), and
Ulmus (elm). A few of the dicots, however, indicate warmer con¬
ditions, as:

Piper (Piperaceae)
Engelhardtia (Juglandaceae)
Artocarpus, Ficus (Moraceae)
Grevillea (Proteaceae)

Telegram @tubscstds
Magnolia (Magnoliaceae)
Cinnamomum, Laurus, Persea (Lauraceae)
FOSSIL PLANTS OF THE ARCTIC AND ANTARCTIC 383

Sophora, Canavalia (Papilionaceae)

Semecarpus (Anacardiaceae)
Cissus (Vitaceae)
Elaeocarpus (Elaeocarpaceae)

Convincing evidence of a mild climate is found in the palm,

Flabellaria florissanti, from the Matanuska coal field in south cen¬
tral Alaska, at about 62° N. latitude. Other plants that seem to
point strongly to a milder climate than the present one are the
cycads Dioon and Ceratozamia, the conifers Taxodium, Sequoia,
and Metasequoia, and several ferns.
These Alaskan floras come from many localities scattered
through latitudes 57° to 65° and further study is needed to estab¬
lish their exact position in the Tertiary.

Cretaceous

An extensive fossil flora has been found in western Greenland

from Disco and Upernivik Islands and the Nugssuak Peninsula, an
area within 70° to 71° north. This was studied in detail by the
Swiss paleobotanist Heer many years ago and has been treated
more recently by Seward who made collections in the summer of
1921. One of the results of his trip was a delightful little book
entitled A Summer in Greenland, a short but informative account
of the history, geology, living flora, and something of the fossil
plants of this great ice-capped continent.
Two ferns are included in the Greenland Cretaceous flora,
Gleichenites and Laccopteris. Among the gymnospermous ele¬
ments are several cycadophytes, Pseudocycas and Ptilophyllum.
Ginkgophyte leaves are abundantly represented and of diverse form,
including specimens assigned to Ginkgoites, Baiera, and Phoeni-
copsis. Among the conifers are: Sequoia concinna, known from
foliage shoots and cones, which is rather similar to the living Cali¬
fornia bigtree; Pagiophyllum, a supposed araucarian relative;
abundant leaves of Sciadopitytes which compares most closely with
the modern umbrella pine (Sciadopitys) of Japan. Among the
dicots are:

Quercus (Fagaceae)
Artocarpus (Moraceae)
Menispermites (Menispermaceae)
Magnoliaephyllum (Magnoliaceae)

Telegram @tubscstds
Laurophyllum, Cinnamomoides (Platanaceae)
Dalbergites (Leguminosae)
384 STUDIES IN PALEOBOTANY

The identification of many of the conifers and dicots in this flora

is admittedly questionable, but in this respect it is in no way un¬
usual as a Cretaceous flora. Regardless of the accuracy of the
names the assemblage presents striking contrasts with the extant
flora of Greenland. The Cretaceous flora was more diverse, includ¬
ing trees of respectable size, and it is of a distinctly warmer climatic
relation. The presence of an Artocarpus leaf and fruit (Fig. 14-1),
which seem very similar to modern breadfruit (this is an Indo-
Malayan genus), has served as a focal point for discussion of the
Greenland climate. Seward notes:
Nathorst’s discovery in the Cretaceous beds at Igdlokunguak (latitude 70°
on Disco Island) of splendid leaves and pieces of the inflorescences exhibit¬
ing a remarkably close resemblance to the foliage and fertile shoots of
Artocarpus incisa Forst. affords one of the most impressive illustrations of
the contrast between Cretaceous and recent Arctic vegetation. (1926, p. 115)

The specimen referred to is preserved in the Natural History

Museum in Stockholm and is undoubtedly the most fascinating
plant that has ever come out of the Arctic and one of the most
discussed of all fossils.
It seems certain that the Cretaceous flora could not have lived
in a climate that approached the rigor of the present one in Green¬
land, yet many of the plants probably could have withstood some
frost. The ginkgophytes were abundant here as in other Arctic
Mesozoic floras—a short diversion from our main theme is inserted
here for whatever bearing it may have on the general problem of
the growing conditions of past Arctic floras. For a tree that appar¬
ently came close to extinction Ginkgo biloba does remarkably well
under a variety of situations in the United States. There are
specimens in the Missouri Botanical Garden in St. Louis that have
thrived under the violent contrasts that that city can produce, fluc¬
tuating from a minimum of —22° in winter to 110° F in summer.
There are fine specimens on the campus of the University of Massa¬
chusetts at Amherst (latitude 42°) where the winter minimum is
— 26°. However, in the Bergianska Tradgarden in Stockholm there
is one tree which the Director, Dr. Rudolf Florin, has informed me is
at least 40 or 50 years old, but it is small (under 20 feet tall), of
stunted, atypical growth, and has never produced reproductive
structures; it is clearly an unhappy ginkgo tree. The winter mini¬
mum in Stockholm, which lies at latitude 59° is —22° F. Ginkgo
does not survive outdoors in Leningrad (latitude 60°) where the
winter minimum is —35° F. It would seem as though some factor

Telegram @tubscstds
other than the cold is responsible for its failure in Stockholm and
Fig. 14-1. Leaf (A) and fruit (B) of Artocarpus dicksonii from the Cretaceous of
Greenland, about 0.5X. C. An Artocarpus growing on the island of Viti Levu, Fiji.

Telegram @tubscstds
(A, B Swedish Natural History Museum; C photograph by O. H. Selling.)

385
386 STUDIES IN PALEOBOTANY

one cannot help wondering, in view of the great difference in lati¬

tude between that city and Amherst, whether light relations are
responsible.
The dicots in the Greenland flora for the most part suggest a
temperate climate, with Artocarpus and one or two others present¬
ing notable exceptions. Seward concluded that the Cretaceous
climate there was comparable with that of southern Europe today.
In his studies of the Cretaceous floras of Alaska Hollick recog¬
nized two distinct assemblages, one from the Yukon River region
(latitude about 63 to 65°), considered as early Upper Cretaceous,
and a second from the Alaska Peninsula. The Yukon River local¬
ities are rich in Ginkgo leaves and the cycadophyte Nilssonia; the
dicots include:

Populus (Salicaceae)
Quercus (Fagaceae)
Macclintockia (Proteaceae)
Aristolochia (Aristolochiaceae)
Castalites, Nymphaeites, Paleonuphar (Nymphaeaceae)
Menispermites (Menispermaceae)
Magnolia (Magnoliaceae)
Benzoin, Laurus, Daphnogene, Cinnamomum (Lauraceae)
Platanus, Credneria (Platanaceae)
Cotinus (Anacardiaceae)
Sapindus (Sapindaceae)
Vitis (Vitaceae)
Sterculia (Sterculiaceae)
Juglans (Juglandaceae)

A preliminary report has been given by Dorf on a newly dis¬

covered late Cretaceous flora from the Knob Lake district of
Labrador (latitude 55°). There are 36 species including 3 conifers
and 27 angiosperms which are believed to have grown under humid,
warm temperate conditions.

Jurassic Plants of Spitsbergen and King Karl’s Land

Nathorst has described several florules from Spitsbergen (about

78.5° N latitude) of Jurassic age. The plants are predominantly
leafy shoots and cones of conifers, cycadlike foliage, and fragments
of fern fronds and ginkgophytes. Some are sufficiently scrappy as
to suggest transport over some distance, but others such as Elatides,

Telegram @tubscstds
and a few nearly perfect cycadophyte leaves, indicate a nearby
origin.
FOSSIL PLANTS OF THE ARCTIC AND ANTARCTIC 387

That the vegetation of Spitsbergen in Jurassic times was quite

different from that of today is rather vividly brought out by
Nathorst:
One of the coal seams at Cape Boheman furnishes a great abundance of
Podozamites and Pityophyllum; sometimes the surface of the schists is as
completely covered with the leaves of Ginkgo digitata, as the soil beneath
a living Ginkgo tree may be in autumn. Since branches and seed of the
same plant are also associated, it is natural to suppose that a Ginkgo forest
occurred not far away from this spot. (1911, p. 221)

Late Jurassic silicified woods from King Karl’s land (latitude

78°) have been reported up to 80 cm in diameter and showing 210
annual rings. In some of them the rings are more accentuated
than in woods found at essentially the same horizon in continental
Europe, which lead Nathorst to conclude that they grew in
the region in which they were found and had not been transported
from the south by marine currents. Five or six genera of coniferous
woods have been described by Gothan from the Jurassic of Spits¬
bergen; the relationships of the woods with modern conifers is
somewhat uncertain, but they at least reveal some diversity in the
flora.

The Rhaeto-Liassic Flora of East Greenland

In a series of monumental works Harris (1926,1931-1937) has given

us the most informative account of an Arctic flora, as well as a very
extensive and diverse one, that has been recorded to date. This is
the Rhaeto-Liassic flora of Scoresby Sound (Fig. 14-2) on the coast
of East Greenland at 71° N latitude. The plants come from two
zones or horizons, the Lepidopteris zone which is Rhaetic and is
regarded as transitional between the Triassic and Jurassic; and the
Thaumatopteris zone which is considered to be lowermost Jurassic.
Since it is the objective of this chapter to reveal the general flor-
istics of the Arctic in the major periods rather than attempt a
detailed stratigraphic analysis, the Scoresby Sound plants will be
dealt with as a unit. Some have been described in other chapters
and space here allows little more than a listing of the better known
elements which display a striking contrast with the present vege¬
tation of Greenland.
The liverworts are represented by both thalloid (Thallites) and
leafy (Hepaticites) types. Several species of Equisetites have been
recorded and one of these (E. muensteri) is described as being very
close to certain living species. This is perhaps not surprising in

Telegram @tubscstds
view of the fact that the horsetails grow in Greenland today, but
388 STUDIES IN PALEOBOTANY

Fig. 14-2. Fossil plants from the Mesozoic of Scoresby Sound, East Greenland.
A. Stenomischus athrous, a male coniferous cone, possibly related to the extant Cun-
ninghamia. B. Anomozamites amdrupiana. C. A. nitida; this bennettitalean genus
is represented by several species. D. Baiera spectabilis. E. Ginkgoites acosmia.
F. Neocalamites hoerensis, an articulate with especially long slender leaves. G. Todites
recurvatus; both sterile and fertile fronds of several species of this osmundaceous genus
are present. H. Storgaardia spectabilis; coniferous foliage twig possibly related to
the podocarps. A is 6X, all others 14X. (From Harris, 1931-35.)

Telegram @tubscstds
FOSSIL PLANTS OF THE ARCTIC AND ANTARCTIC 389

we encounter a contrast in the fossil Neocalamites with stems up

to 10 cm in diameter and whorls of leaves which reach a length of
12 cm (Fig. 14-2F).
Several families of ferns are represented by abundant sterile and
fertile foliage. There are five species of Todites (Osmundaceae)
which bore dimorphic fronds apparently quite like certain living
species. The Marattiaceae, a modern tropical family and one that
we have traced back to the Carboniferous, is well represented.
Rhinipteris, from Scoresby Sound, is compared with the Carbonif¬
erous Ptychocarpus and Asterotheca. Marattiopsis, which is very
close if not identical with the living Marattia, is represented by
fertile specimens with synangia which reach a length of 7 mm and
include up to 17 sporangia on each side. Spore masses have been
isolated intact and measure 900 p, by 350 ju, and like the modem
Marattia consists of several hundreds of spores.
Three other families of ferns are present, the Gleicheniaceae,
Matoniaceae, and Dipteridaceae; since these are distinctive in their
frond morphology and sporangial structure and well-preserved
fertile specimens are present, there seems to be no question as to
the identity of these fossils.
The variety of leaves assigned to the Ginkgoales is an especially
distinctive feature of the flora; although they probably did not all
live at one time in that area one gains the impression that it would
have been an ideal place to have studied ginkgophyte evolution.
There are several species of Ginkgoites which range from the
slightly lobed G. obovata (Nathorst) Seward to species with deeply
dissected leaves such as G. acosmia Harris, G. minuta (Nathorst)
Harris, and G. hermelini (Hartz) Harris. There are also several
species of Baiera as well as Hartzia and Torellia.
The coniferous elements of the Greenland flora were fairly num¬
erous but they cannot be compared closely with modern ones.
Storgaardia (Fig. 14-2H) is based on foliage shoots, the leaves
being oppositely arranged, 10 cm long and 5 to 10 mm wide. The
difficulty of classifying the Greenland conifers is typically illus¬
trated here. Harris notes:

This striking fossil (Storgaardia spectabilis Harris) recalls Podocarpus

subgenus Eu-Podocarpus in the form of its leaves, but subgenus Nageia in
their insertion and the genus Taxus in the general appearance of its cuticle,
although the stomata differ in detail in being monocyclic. (1935, p. 59)

Elatocladus is a genus with foliage superficially similar to Taxus

and is represented by 16 species. Podozamites is a genus of broad-

Telegram @tubscstds
390 STUDIES IN PALEOBOTANY

leafed conifers and Harris notes that it is probably not a natural

one and may include species “with purely Cycad affinities.” Other
coniferous genera are Araucarites and several new ones erected for
reproductive organs whose precise affinity is not known.
The Bennettitales is represented by 19 species of leaves and 12
of reproductive organs. Included in the former are Anomozamites,
Taeniozamites, Pterophyllum, and Otozamites. Finally, the
Scoresby Sound beds include several other fossils of particular
interest which are not referable to the above mentioned groups
and are considered elsewhere in this book: the Caytoniales, Lepi-
dopteris, Nilssonia, Hydropteridangium, and others.
The Scoresby Sound flora is rich in its variety of plants and is a
classic example of the wealth of information that can be obtained
by the application of the proper techniques. As to the floristic
contrast of this part of the Arctic between early Jurassic times and
the present the following comment of Seward’s seems especially apt:
It is an impressive experience to stand on a flower-sprinkled Greenland
heath and to see on a piece of shale split from the face of a ravine fragments
of fronds closely allied to those of a Gleichenia which one had previously
seen in very different circumstances growing on a tangled bank of ferns on
the edge of a Malayan forest. The interval of millions of years separating
the living from the dead is for the moment forgotten. In place of the bleak,
treeless hills and the heath-covered slopes one sees a luxuriant undergrowth
of ferns against a background of conifers and broad-leaved trees.

Carboniferous

Rather extensive Carboniferous floras have been found in Spits¬

bergen and northeast Greenland at points approximately 79° and
81° N respectively; the floras of both regions include lycopods,
calamites, and sphenopterid foliage that is probably pteridosperm-
ous. Lepidodendron stems, from Spitsbergen, are described by
Nathorst up to 40 cm in diameter and Stigmaria was observed
with the rootlets radiating out through the sediments in which
they were entombed; it is thus evident that the trees grew in the
immediate vicinity and they were of respectable size. Silicified
cordaitean (Dadoxylon) wood which lacks annual rings is also
reported from Spitsbergen; this is admittedly an isolated bit of
evidence which one would wish to confirm before arriving at any
general conclusions, but if the long winter Arctic night prevailed
at that time a pronounced indication of seasonal growth would be
expected.

Telegram @tubscstds
FOSSIL PLANTS OF THE ARCTIC AND ANTARCTIC 391

Lower Carboniferous collections from Traill Island, on the east

coast of Greenland (latitude 72°), include Catamites and Lepido-
dendron stems.

Devonian

One of the most interesting Upper Devonian floras on record is

from Bear Island, lying at approximately 74° N latitude between
Spitsbergen and the North Cape of Norway. The early ferns were
represented by magnificent fronds of Archaeopteris roemeriana
(Goeppert) and A. fimbriata Nathorst, the latter being distin¬
guished by delicately dissected pinnules which must have been
magnificent in life. Another fern of exceptional interest is
Cephalotheca mirabilis Nathorst, a genus that is closely related
to, or identical with, Rhacophyton (see Chapter 3); massive
clusters of sporangia were borne in pairs at the base of the frond
and the general habit of the plant was probably quite like that of
Rhacophyton. A fragmentary specimen of Stigmaria which meas¬
ures about 15 cm in diameter is indicative of lycopods of arbores¬
cent dimensions; Lepidodendron stems are included in the flora
and on some specimens isolated megaspores up to 2 mm in diam¬
eter are abundant. Spores of this size suggest lycopods of an
advanced type. The diversity of Devonian heterosporous plants,
including lycopods, is also indicated from Chaloner’s study of
megaspores from Ockse Bay, on the southwest coast of Ellesmere
Island; these are mentioned in further detail in Chapter 13.
Most unusual of all the Bear Island plants is Pseudobornia
ursina, an articulate that Nathorst regarded as constituting a dis¬
tinct order, the Pseudoborniales. The shoots are about 0.5 cm in
diameter and bear whorls of leaves, probably four at each node.
Each leaf is several times dichotomized, the ultimate divisions
being delicately fimbriate to the point of resembling small feathers.
Associated articulate stems about 9 cm in diameter are believed to
have borne the leafy shoots.
Professor Hdeg has given us the most complete picture of Arctic
Devonian vegetation for any single area in his study of several
fossil floras from Spitsbergen. Some of the more interesting plants
have been described in previous chapters—the problematical
Enigmophyton superbum, the protopterid Svalbardia polymorpha,
and the articulate Hyenia vogtii with its distinctive branch
morphology. The lycopods are represented by stems of Bergeria

Telegram @tubscstds
mimerensis H0eg up to 10 cm in diameter; they are partially
392 STUDIES IN PALEOBOTANY

decorticated, making comparison with other genera rather difficult.

Slender axes that are not more than 1 cm in diameter and char¬
acterized by a small, circular leaf scar located at the upper part of
the slender, vertically elongate cushions are given the name Proto-
lepidodendropsis pulchra. Both of these lycopods are from Mid¬
dle Devonian or basal Upper Devonian horizons. Actinopodium
nathorstii is a small, leafless axis with a star-shaped stele consist¬
ing of a central region of mixed parenchyma and multiseriate
pitted tracheids surrounded by a narrow zone of radially aligned,
apparently secondary tracheids. Unfortunately not enough of this
plant is available to afford a very sound basis for speculation, but
it may possibly represent an advanced psilophyte.

Summary Comments on the Arctic Climate

The evidence from fossil plants alone leaves no doubt that the
frigid climate of the Arctic, which has prevailed through the
Pleistocene to the present time, is very much the exception rather
than the rule. The records from the Devonian on indicate condi¬
tions quite in contrast to what is known today. There have prob¬
ably been significant fluctuations during the past few hundred
millions of years and it is not implied that tropical forests ever
forged their way into extreme northern latitudes. Yet, even though
the present inhospitable climate may be temporary in terms of
geologic time, it is with us and we are anxious to know not only
what caused the change but what the future may hold in store.
Are we now living in an interglacial period of the Pleistocene or are
we on our way back to the generally warmer climates of pre-Pleis-
tocene times?
There is a voluminous literature on the causes of climatic change
in which one may find a sufficient range of explanations to suit al¬
most any reader; a few have been especially popular such as
changes in the position of the poles, drifting continents, and shift¬
ing ocean currents, but none thus far has met with anything
approaching general agreement. At the risk of adding speculation
on speculation it has seemed to me most appropriate to mention
some of the approaches that have been undertaken by recent
investigators.
Types of evidence are coming from sources that only a few years
ago were quite unknown; the use of oxygen isotopes is one of these
methods. It has been shown by H. C. Urey that when water

Telegram @tubscstds
evaporates oxygen 16, 17, and 18 are lost at different rates.
FOSSIL PLANTS OF THE ARCTIC AND ANTARCTIC 393

Slightly higher proportions of the common oxygen 16 are carried

off in water vapor and as a result water becomes slightly enriched
in the rare heavy isotopes 17 and 18. The ratios of the two will
vary slightly in carbonates such as limestone, coral, or limy skele¬
tons of other aquatic animals, partly according to the temperature
of the water. Studies of bottom-dwelling foraminifera taken from
deep sea drill cores in the Pacific indicate that the bottom temper¬
atures, resulting from cooler surface waters of the Antarctic sink¬
ing and flowing northward, have dropped about 16° F in the past
32 million years.
A new and somewhat controversial approach to an understanding
of fossil climates comes from paleomagnetic studies. Recent investi¬
gations of the magnetization of rock formations indicate that some
sedimentary deposits may retain their direction of magnetization
over long periods of geologic time and the direction perpetuates
that of the earth’s field at the time they were formed. For
example, certain Carboniferous strata of Derbyshire and Yorkshire,
England, are said to be magnetized with a low angle of dip which
suggests an original magnetization in a field fairly close to the
equatorial region. Paleomagnetic measurements in western United
States also indicate a closer proximity to the equator in the late
Paleozoic. In both western United States and in Great Britain
studies of windblown sand deposits correlate with the paleomag¬
netic evidence in suggesting that these areas were within the belt
of the northern trade winds, assuming that the present width of
this belt is typical. Thus the evidence from this direction sup¬
ports the view that the axis of rotation of the earth, also referred
to as polar wanderings, may have undergone significant changes in
the past.
I have mentioned elsewhere the possibility of physiologic changes
in the plants themselves. This is summed up rather well in the
following comment by Axelrod:

Cain (1944, p. 75) has pointed out that breadfruit (Artocarpus), which is
now strictly tropical, and which has been recorded at higher latitudes, may
have been represented in Cretaceous and Early Tertiary floras by one or
more temperate to warm temperate species as in the case of certain genera
today, such as persimmon (Diospyros), avocado (Persea) and magnolia
{Magnolia). We may add in this connection that a number of genera which
are frequently regarded as temperate, such as alder (Alnus), maple {Acer),
birch {Betula), haw {Crataegus), katsura {Cercidophyllum), sweet gum
{Liquidambar) and sycamore {Platanus), were widely represented in the
warm temperate to subtropical zone during the Cretaceous and Early

Telegram @tubscstds
Tertiary. Pertinently, many of the widely distributed families in the north-
394 STUDIES IN PALEOBOTANY

ern hemisphere which are often considered as temperate, including the

Aceraceae, Betulaceae, Fagaceae, Rosaceae, Ulmaceae and others, extend
into the tropics. The most primitive living species of maple, poplar, birch,
oak, cherry and hackberry are largely evergreen types, and are found in
warm temperate to tropical regions. (1952, p. 51-52)

Some Fossil Plants of the Antarctic

As compared with the Arctic we have as yet very little informa¬

tion on the past floras of the South Polar region. During the
course of the Swedish Southpolar Expedition of 1901 to 1903
a rather large collection of Jurassic plants was made at Hope Bay
which is located near the tip of the Palmer Peninsula at latitude 63°.
Included in the flora are: Equisetites stems up to 3 cm broad; well-
preserved fertile and sterile fronds of Todites, an osmundaceous
fern that is widely distributed through the Jurassic; and several
species of Cladophlebis and Coniopteis; cycadophyte foliage is the
most conspicuous element of the flora with large specimens of
Nilssonia taeniopteroides Halle, Pseudoctenis, Zamites, Otozam-
ites, and Ptilophyllum; judging from the long fragments pre-
seved the leaves of Nilssonia reached a length of about 1 meter and
those of Pseudoctenis are at least 14 cm broad; a few conifers are
included but no angiosperms or lycopods.
Several coniferous and dicot woods have been found in lower
Tertiary deposits on Seymour and Snow Islands, also located on
Palmer peninsula, at 64° S latitude.
A Tertiary leaf flora from Seymour Island includes Lauriphyl-
lum, Drimys, Iliciphyllum, Lomatia, Knightia, Fagus, Nothofagus,
Myrica, several ferns, and two species of Taeniopteris. The preser¬
vation of these plants is for the most part rather poor and many
of the identifications perhaps questionable. A preliminary study
by Cranwell of spores and pollen found in a rock sample from
Seymour Island offers the hope that this approach will greatly
enhance our knowledge of Antarctic floras. The age of the sample
is somewhat uncertain and may include mixed Cretaceous and
mid-Tertiary fossils. The dominant elements are conifers, includ¬
ing Araucaria, Agathis and Podocarpus, and several southern
beeches (Nothofagus). The fern families Cyatheaceae and Schiz-
aeaceae are represented and pollen of the following flowering
plant families are tentatively identified: Cruciferae, Myrtaceae,

Telegram @tubscstds
Proteaceae, Loranthaceae, Oenotheraceae, and possibly the
Winteraceae.
FOSSIL PLANTS OF THE ARCTIC AND ANTARCTIC 395

It seems evident from the little that is known at present that the
Antarctic, like the Arctic, has enjoyed a much milder climate in
the past.

REFERENCES

Cranwell, Lucy M. 1959. Fossil pollen from Seymour Island, Antarctica. Nature,
184: 1782-1785.
Dorf, Erling. 1959. Cretaceous flora from beds associated with iron-ore deposits in
the Labrador trough, (abstract). Geol. Soc. Amer. 1959 program of meetings,
p. 33A.
Dusen, Carl H. 1908. Uber die Tertiare flora der Seymour-Insel. Wissensch.
Schivedischen Sudpolar-Exped., 1901-1903. vol. Ill, lief 3: 1-27.
Emiliani, Cesare. 1958. Ancient temperatures. Scientific Monthly, 198 (2): 54-63.
Gothan, Walther. 1908. Die Fossilen Holzer von der Seymour und Snow Hill Insel.
Wissensch. Schwedischen Sudpolar-Exped., 1901-1903. vol. Ill, lief 8:1-31.
-. 1908. Die fossilen Holzer von K5nig Karls Land. Kungl. Svenska
Vetensk. Akad. Handl., 42 (10): 1-44.
-. 1910. Die fossilen Holzreste von Spitzbergen. Kungl. Svenska Vetensk.
Akad. Handl., 45 (8): 1-56.
Halle, Thore G. 1913. The Mesozoic flora of Graham Land. Wissensch. Schwedi¬
schen Sudpolar-Exped., 1901-1003. Vol. Ill, lief 14: 1-123.
-. 1931. Younger Palaezoic plants from East Greenland collected by the
Danish expeditions 1929 and 1930. Meddel. orn Gronland, 85 (No. 1): 1-26.
Harris, Thomas M. 1926. The Rhaetic floras of Scoresby Sound, East Greenland.
Meddel. om Gronland, 68: 45-147.
-. 1931-37. The Fossil flora of Scoresby Sound, East Greenland. Meddel.
om Gronland. Pt. 1, vol. 85, No. 2; pt. 2, vol. 85, No. 3; pt. 3, vol. 85, No. 1; pt. 4,
vol. 112, No. 2; pt. 5, vol. 112, No. 3.
Heer, Oswald. 1868. Die fossile Flora der Polarlander, enhaltend die in Nordgron-
land, auf der Melville-Insel, im Banksland, am Mackenzie, in Island und in Spetz-
bergen entdeckten fossilen Pflanzen. pp. 1-192. Zurich.
H0eg, Ove Arbo. 1942. The Downtonian and Devonian flora of Spitsbergen. Norges
Svalbard-ogIshavs-Undersokelser, 83: 1-228.
Hollick, Arthur. 1930. The Upper Cretaceous floras of Alaska. U. S. Geol. Survey
Prof. Paper, 159: 1-116.
-. 1936. The Tertiary floras of Alaska. U. S. Geol. Survey Prof. Paper,
182: 1-185.
Manum, Svein. 1954. Pollen og sporer I Tertiaere Kull fra Vestspitsbergen. Blyttia
(Oslo), 12: 1-9.
Nathorst, Alfred G. 1890. Ueber die reste eines Brotfrucktbaums, Artocarpus dick-
soni n. sp., aus den Cenomanen Kreideablagerungen Gronlands. Kongl. Svenska
Vetensk. Akad. Handl., 241 1-10.
-. 1894. Zur Palaozoischen flora der Arctischen Zone. Kongl. Svensk. Vet.
Akad. Handl., 26, No. 4.
-. 1897. Zur Fossilen Flora der Polarlander. Kongl. Svenska Vets. Akad.
Handl., 30 (1): 1-77.

Telegram @tubscstds
-. 1911. Contribution to the Carboniferous flora of Northeastern Greenland.
Meddel. on Gronland, 43: 339-346.
396 STUDIES IN PALEOBOTANY

-. 1911. On the value of the fossil floras of the Arctic regions as evidence of
geological climates. Geol. Mag., 48, 8: 217-225.
-. 1907. t)ber Trias und Jurapflanzen von der Insel Kotelny. Mem. Akad.
Imp. Sci. St. Petersbourg, ser. Ill, vol. 21, No. 2.
Opdyke, N. D. and S. K. Runcorn. 1959. Palaeomagnetism and ancient wind direc¬
tions. Endeavour, 18: 26-34.
Runcorn, S. K. 1955. The permanent magnetization of rocks. Endeavour 14:
152-159.
Schloemer-Jager, Anna. 1958. Altertiare Pflanzen aus Flozen der Brogger-Halbinsel
Spitsbergens. Palaeontographica, 104B: 39-103.
Seward, Albert C. 1922. A Summer in Greenland. Cambridge Univ. Press. Pp.
1-100.
Seward, A. C. 1925. Arctic vegetation past and present. Journ. Roy. Horticultural
Soc., 50:1-18.
-. 1926. The Cretaceous plant-bearing rocks of western Greenland. Phil.
Trans. Roy. Soc. London, 215B: 57-175.

Telegram @tubscstds
 1
HIPATOPIHYTA and
SPYOPHIYTA
— the bryophytic plants

T he mosses and liverworts are ancient groups and their

position in the plant kingdom is controversial; they are
all small by comparison with the vascular plants although very
common and widely distributed over the surface of the earth.
Quite a few favor moist shady places, many are epiphytes on large
plants in both temperate regions and the tropics, and others are
able to thrive in the most exacting habitats. Although they are
not often preserved as fossils we have a few records that are
of exceptional interest.
In order that the use of terms may be clear a brief consideration
of the classification of the plants under consideration seems essen¬
tial. It is, however, recommended that the student who is not
familiar with representative mosses and liverworts should consult
one of the many good elementary botany texts, virtually all of
which devote a chapter to these plants. For the purpose of
the following discussion it will be helpful to become familiar with
the essential features of such plants as Marchantia, Porella, Antho-
ceros, Polytrichum, and Sphagnum. In all of them the sexual phase
is the dominant one in the life cycle, the opposite being the case
in the pteridophytic groups and seed plants. The first three genera
cited are liverworts (Hepatophyta) although they differ from one
another in no small degree. In Marchantia the main body of the
plant is a flat, ribbon-shaped, dichotomously branching structure
with a complex chambered internal anatomy. Porella is one of the
many leafy liverworts and superficially resembles certain mosses,
although the lack of a midrib in the leaf and other important
characters distinguish it from the group. Anthoceros (the horned

Telegram @tubscstds
liverwort) consists of a thin, irregularly dichotomizing thallus

397
398 STUDIES IN PALEOBOTANY

which lacks the chambered organization of Marchantia, but it has

a sporophyte that is larger and more conspicuous than that of
other liverworts; it is, moreover, photosynthetic and thus partially
independent and it possesses a basal meristem which allows con¬
tinued growth for several months. Polytrichum (hair-cap moss) is
a very common moss with upright, radially symmetrical leafy
stems with conspicuous sporophytes that develop at the top of the
female plant. Sphagnum is the peat moss, a distinctive feature of
the northern bog vegetation. Perhaps its most striking feature lies
in the leaves which are composed of two types of cells, small pho¬
tosynthetic ones which give the plant its pale green color, and
large empty ones that serve as minute water storage tanks and
which render the plant valuable in gardening as a conditioner for
heavy soils.
Living species and genera of mosses and liverworts are often
delimited by structural details of microscopic magnitude. Conse¬
quently it is rarely possible with fossils to compare them with
modern species, genera, or even families. Thus we need not be
concerned with a detailed classification although a few notations
oh the major categories may be in order. The mosses and liver¬
worts have long been treated as two subgroups under the major
division Bryophyta; the differences between the two are, however,
great and it seems to me that the classification given by Bold in his
recent book Morphology of Plants is more realistic in regarding
them as two distinct divisions. Since it is, nevertheless, convenient
to have an inclusive designation for the entire assemblage, the term
bryophytes or bryophytic plants will be employed here in that way.
A few other comments on the classification and evolution of these
plants will be more appropriate at the close of the chapter.

Standard classification Classification of Bold

BRYOPHYTA HEPATOPHYTA (liverworts)
Hepaticae (liverworts) BRYOPHYTA (mosses)
Musci (mosses)

Fossil Hepatophyta

The oldest authentic liverwort is Hepaticites kidstoni Walton

from the Yorkian series of the Upper Carboniferous of Shropshire,
England. Although its exact affinities are not known it is remark¬

Telegram @tubscstds
ably similar to some of the modern leafy liverworts and indicates
that the group had already enjoyed a long period of evolution. It
HEPATOPHYTA AND BRYOPHYTA 399

was a small plant, the specimen shown (Fig. 15-1A) being but
a few millimeters long. The cellular structure is quite well pre¬
served; in addition to two lateral rows of large conspicuous leaves
there are two alternating series of small semicircular flaps on the
under (or upper?) surface.
A thalloid liverwort, Hepaticites willsi Walton, is also known from
the Upper Carboniferous (Staffordian series of Staffordshire). It
is a dichotomously branching plant averaging about 1 mm in
breadth. In another thalloid type, Hepaticites langi Walton from
the Upper Carboniferous of Shropshire, the rhizoids were found to
be preserved.
The fossils described above indicate that both thalloid and leafy
types existed in the Upper Carboniferous, but little more can be
said about their affinities within the Hepatophyta.
The naming of fossil liverworts or presumed liverworts has
tended to be somewhat haphazard; owing in part to considerable
variation in the quality of preservation, a completely acceptable
system is probably not possible, but the following use of terms
seems reasonably satisfactory:
Thallites Walton is employed for thalloid plant remains which
may be liverworts but which cannot positively be distinguished
from certain other groups of plants.

Telegram @tubscstds
Fig. 15-1. A. Hepaticites kidstoni; Upper Carboniferous, Shropshire, 30X. B. Hepa¬
ticites willsi; Upper Carboniferous, Staffordshire, 16X. (From Walton, 1925.)
400 STUDIES IN PALEOBOTANY

Hepaticites Walton is employed for fossils that can with some

assurance be determined as liverworts and in which characters are
present that enable one to distinguish them from algae, pteri-
dophytes, and other groups of plants, yet their structure is not
sufficiently complete to warrant the use of a distinctive generic
name or one that implies close affinity with living genera.
Marchantites Brongniart, according to Harris “is restricted to
hepatics agreeing with M. sezannensis Brongniart, the type species,
in which marchantiaceous air chambers, ventral scales and repro¬
ductive organs have been demonstrated.” (1942, p. 394)
There are many reports of thalloid liverworts from Mesozoic and
Tertiary horizons, but for the most part they only inform us that
these plants existed, probably in about the same form that we find
them today; the preservation is such that one can usually do little
more than assign them to the noncommittal Hepaticites, although
a few offer more precise information.
A rather rich liverwort assemblage has been described by Miss
Lundblad from the Rhaetic-Liassic coal mines of Skromberga,
Sweden, which includes Ricciopsis fiorinii, a small dichotomously
branching thallus about 1 to 3 mm wide which developed a rosette¬
like growth 2.5 cm in diameter. It is considered to be closely
related to certain modern members of the genus Riccia.
Marchantites hallei Lundblad from the Lower Cretaceous of
Patagonia shows air pores and two rows of ventral scales plus
numerous rhizoids, but the internal structure is not known.
Harris has described several species of Hepaticites from Scoresby
Sound, Greenland: H. rosenkrantzi is a thalloid plant up to 2 cm
long and 3 mm wide; no midrib is evident but round or oval cups
1.5 to 3 mm in diameter were borne on the surface which may
represent gemma cups such as are found in Marchantia.
A few fossil liverworts are well enough preserved to allow a some¬
what more precise identification. Metzgeriites glebosus (Harris)
Steere, also from Scoresby Sound, is a dichotomously forking thal¬
loid plant with a lamina that is divided into irregular lobes, and
numerous unicellular unbranched rhizoids are attached to the
underside of the midrib. It has been assigned this generic name
because of undoubted affinities with the anacrogynous Jungerman-
niales.

Fossil Bryophyta

Telegram @tubscstds
Fossil mosses were described by Renault and Zeiller in 1888
from the Upper Carboniferous of Commentry in central France
HEPATOPHYTA AND BRYOPHYTA 401

under the name Muscites polytrichaceus. Professor Walton has

reexamined the specimens and concludes that they have been cor¬
rectly identified. It is, however, of considerable interest to note the
discovery of beautifully preserved mosses in Upper and Lower
Permian horizons in the Kuznetsk basin and in Vorkouta, respect¬
ively, of the USSR. Professor Neuburg has described some
remarkably well-preserved plants (Fig. 15-2) representing several
species; the preservation is such that perfect leaf compressions can
be removed from the shale and mounted on a slide for study
in much the same way that a modern moss leaf would be handled.
They show the characteristic leaf structure of the Bryales, a sub¬
division of the Bryophyta; the central vein is conspicuous and in
at least one species there seems to be a tendency toward the
development of side veins. During a visit to her laboratory in
October of 1958 Mrs. Neuburg also showed me small rosette-shaped
fossils composing a branching moss plant, or a portion of one,
in which the leaves have two distinct cell types which are strongly
suggestive of the Sphagnales (peat mosses). The spore-bearing

Fig. 15-2. Intia vermicularis, a moss from the Permian of the USSR. A. a nearly

Telegram @tubscstds
complete leaf; B. a portion more highly enlarged.
Maria F. Neuburg.)
(Photographs courtesy of
402 STUDIES IN PALEOBOTANY

organs of these Permian mosses have not been found, so that more
precise comparisons with living forms are not possible. Their
importance lies in the fact that they are well preserved, they are
unquestionably mosses, and sufficiently similar to modern ones in
their vegetative organization as to suggest no major changes in
moss evolution since that time.
Quite a few Tertiary mosses have been described but, as with the
liverworts, close comparisons with modern genera are usually not
possible.
Polytrichites spokanensis Britton from the Upper Miocene of
Washington is an upright or slightly decumbent plant with stems
up to 2 cm long; it is considered to bear a close resemblance to
Polytrichum.
The genus Plagiopodopsis has been proposed for acrocarpous fos¬
sil mosses (with upright stems and terminal sporophyte). Mac-
Ginitie has figured a fine specimen of P. cockerelliae (Britton and
Hollick) Steere from the Florissant beds in Colorado; sporophytes
are present but, judging from the presence of the calyptra, they were
immature at the time of fossilization.
Palaeohypnum has been established by Steere for fossils that
are referable to the Bryales Pleurocarpi—mosses with creeping
stems and laterally attached sporophytes. Several species have
been reported from various Tertiary horizons of western United
States.
Although most fossil specimens are sterile (the gametophyte
only is preserved), a moss theca has been reported from Oligocene
brown coal in Victoria and given the name Muscites yallournen-
sis Clifford and Cookson. It is quite small, being 0.8 mm long and
0.2 mm in diameter, and only the epidermis of the theca and its
operculum were preserved.

Some Problematical Bryophytic Plants

Naiadita lanceolata Buckman

There are a few fossil plants which in themselves are expressive

of the whole history of paleobotanical investigation—confused
nomenclature, unique preservation, occasional abundance of speci¬
mens, mistaken identity, and the wealth of information that may
be wrested from the rocks with the proper techniques and a com¬
petent investigator. Of such landmarks in the science Naiadita
deserves a place of some prominence; the information given here

Telegram @tubscstds
is taken from a comprehensive study by T. M. Harris.
HEPATOPHYTA AND BRYOPHYTA 403

Naiadita comes from an English Rhaetic outcrop extending

from Somerset east of the Mendip Hills to Worcestershire and
Warwickshire. Although specimens are fragmentary they occur in
abundance, as many as 33 leaves, leaf fragments, and other parts
being found per square centimeter of rock surface, and this is said
to be typical over a stretch of about 90 miles of the outcrop.
It was a smallish plant (Fig. 15-3) consisting of stems usually
unbranched and rarely over 2 cm high which bore lanceolate,
spirahy arranged leaves, mostly under 5 mm long. Large numbers
of gemmae (small, specialized vegetative reproductive structures)
are found among the leaf and stem remains; they are oval in
shape and range from 500 X 300 \x to 200 X 120 ju and a few have
been found in terminal gemmae cups. Archegonia have been
found in some abundance on the stems and their position relative
to other organs is problematical; they do not occupy a position in
the leaf axils and it has been suggested that they may take the place
of a leaf. Some of the archegonia are enclosed in a “perianth” of
several leafy lobes. Following fertilization (the antheridia have not
been found) a pedicel developed resulting in a short branch which
carried the developing sporophyte at its tip. The sporophyte had
a short bulbous foot and a globose sporangium varying from 0.3 to
1.2 mm wide and with 100 to 400 spores which are lens-shaped
with a characteristic equatorial flange.
In its general habit, notably the arrangement of the leaves,
Naiadita is typically mosslike, but the lack of a midrib in the
leaves suggests a liverwort. There are other liverwort characters
such as unicellular rhizoids, an archegonium with a one-layered
venter and lack of tissue differentiation in the stem. The general
form of the sporophyte is comparable with that organ in the
Marchantiaceae (a family of Hepatophyta), but elater cells are not
found in Naiadita.

Sporogonites exuberans Halle

In 1916 Halle described a fossil from the Lower Devonian of

Norway which consisted of an oval-shaped sporangium borne at
the end of a slender unbranched stalk at least 4 cm long; the
remainder of the plant was not found. The sporangium was suffi¬
ciently well preserved to reveal a multilayered wall with a dome¬
shaped mass of spores within; it was assumed that this indicated
the presence of a columella but lack of preservation in the central
portion left the matter in doubt. Columellate sporangia are found

Telegram @tubscstds
in the modern Sphagnum (peat moss) and in Horneophyton.
404 STUDIES IN PALEOBOTANY

Fig. 15-3. Naiadita lanceolata; A. leafy shoot, 6X; B. gemma cup, 55X; C, a single
gemma, 50X; D. archegonium, 200X; E. plant with a mature sporophyte, 8X. (From
Harris, 1938.)

When Sporogonites was first discovered, Halle favored a bryo-

phytic alliance but when the Rhynie plants were made known a
comparison with Horneophyton seemed more likely and in most
accounts dealing with early land plants Sporogonites is listed
under the psilophytes.
A short time ago this writer had occasion to examine abundant
collections of S. exuberans in the Brussels Museum which have
been collected by Dr. Stockmans from the Lower Devonian of
Belgium. A notable feature of the Belgian specimens is the con¬
sistent parallel alignment of the long sporangiophores; this is true

Telegram @tubscstds
for literally hundreds of specimens and it did not seem likely that
HEPATOPHYTA AND BRYOPHYTA 405

it could be the result of chance deposition. A careful survey of

the collection revealed several attached to an irregularly shaped
carbonaceous film which is interpreted as a thalloid structure (Fig.
15-4). In one specimen this “thallus” measured about 15 X 5 cm
and was clearly only a fragment of the whole. No evidence of
vascular tissue was detected in this structure nor has vascular
tissue ever been found in the stalks. The plant is therefore tenta¬
tively considered to be a bryophyte.

Relationships of the Bryophytes to Other Plant Groups

It is necessary to reflect back for a moment to the classification

of the plants considered in this chapter. In dividing them into two
distinct divisions Professor Bold noted that:

When one reviews the structure and reproduction of the members of these
several groups, the diversities among them appear more striking than the
resemblances. (1957, p. 312)

Fig. 15-4. A tentative restoration of Sporogonites exuberans based on specimens in

Telegram @tubscstds
the Natural History Museum, Brussels. (From Andrews, 1960.)
406 STUDIES IN PALEOBOTANY

Probably one of the primary reasons for retaining the mosses

and liverworts within one group has been the general similarity in
their life cycle; that is, both have a dominant sexual phase and a
much smaller sporophyte one that is either partially or wholly
dependent on the former. This is quite the opposite of what is
found in the pteridophytic groups. This in itself does not neces¬
sarily imply close relationship; some of the algae, for example, dis¬
play essentially the same type of life cycle. It is precarious to
depend on such comparisons, but I hazard the suggestion that the
mosses and liverworts, so different in their gametophytic structure,
are no more closely related than the various groups of pterido-
phytes are to each other.
The fossil record reveals both thalloid and leafy liverworts of
essentially modern aspect in the Upper Carboniferous; mosses are
also known from that age although not well preserved. However,
the Permian specimens seem quite modern in their general organ¬
ization. It seems reasonable to assume that the origins of the two
groups will be found at a much lower horizon. All of the bryo-
phytic plants are small and there has been very little systematic
searching for fossil remains as yet.
Plants like Naiadita complicate the evolutionary picture in that
they suggest a greater diversity among the bryophytes than is in¬
dicated by the modern flora; it may well be that many lines will
ultimately be found in the Paleozoic.
We do not know for sure where Sporogonites fits into the scheme
of things, but enough information is available to suggest that it
may represent a group of plants equal in status with the Hepatophyta
and Bryophyta. The size and shape of the structure that is inter¬
preted as thallus and the densely borne, unbranched sporangiate
stalks are distinctive characters.
The bryophytes are of more than passing interest, for botanists
have long toyed with the notion that some of them may have been
the progenitors of vascular plants; Anthoceros has been pointed to
as a plant in which the sporophyte shows a tendency towards
dominance over the gametophyte. Others have chosen to regard
the group as reduced, and as a third possibility they may be wholly
independent of any relationship with the pteridophytes.

REFERENCES

Telegram @tubscstds
Andrews, Henry N. Jr. 1960. Notes on Belgian specimens of sporogonites. The
Palaeobotanist, 7: 85-89.
HEPATOPHYTA AND BRYOPHYTA 407

Bold, Harold C. 1957. Morphology of Plants. Harper and Bros. 669 pp.
Clifford, H. T. and Isabel C. Cookson. 1953. Musettes yallournensis, a fossil moss
capsule from Yallourn, Victoria. The Bryologist, 56: 53-55.
Halle, Thore G. 1916. A fossil sporogonium from the Lower Devonian of Roragen in
Norway. Bot. Notiser, Lund (1916): 79-81.
-. 1936. Notes on the Devonian genus Sporogonites. Svensk. Bot. Tidskrift,
30: 613-623.
Harris, Thomas M. 1938. The British Rhaetic Flora. Brit. Mus. Nat. Hist. pp.
1-84.
-. 1942. On two species of hepaticae of the Yorkshire Jurassic flora. Ann.
Mag. Nat. Hist., ser. 11, 9: 393-401.
Lundblad, Britta. 1954. Contributions to the geological history of the hepaticae.
Svensk. Bot. Tidsk., 48: 381-417.
-. 1955. Contributions to the geological history of the hepaticae II. Bot.
Notiser, 108: 22-39.
MacGinitie, Harry D. 1953. Fossil plants of the Florissant beds, Colorado. Carne¬
gie Inst. Washington Pub., 599.
Neuburg, Maria F. 1955. Bryophytes from Permian sediments of the USSR. Dokl.
Akad. Nauk. SSSR, 107 (2): 321-324.
Renault, Bernard and Zeiller, Rene. 1888-1890. Etudes sur le terrain houiller de
Commentry. Flore fossile. Pts. 1, 2. Soc. Industrie Min. St. Etienne Bull.
Steere, William C. 1946. Cenozoic and Mesozoic bryophyta of North America. Amer.
Mid. Nat., 36: 257-380.
Walton, John. 1925. Carboniferous Bryophyta. I. Hepaticae. Ann. Bot., 39: 563-
572.
-. 1928. Carboniferous Bryophyta. II. Hepaticae and musci. Ann. Bot.,
42: 707-716.
-. 1949. A thalloid plant showing evidence of growth, etc. Trans. Geol.
Soc. Glasgow, 21: 278-280.

Telegram @tubscstds
 16
some Paleozoic and
Mesozoic floras

D n this chapter floristic surveys are made of certain ages

and areas that have been omitted thus far or touched on
only incidentally. The coverage is by no means complete, the ap¬
proach being similar to that used in presenting the Tertiary angio-
sperm floras; that is, plant assemblages of particular interest have
been selected from Paleozoic and Mesozoic horizons with the intent
of revealing their distinctive features as well as the problems that
remain unsolved. In large part the floras dealt with are from sec¬
tions of the world that have not been emphasized in previous
chapters and from which we may expect to learn a great deal in
the future.
In certain of the older works on fossil plants, as well as some
recent ones, there is a tendency to regard floras of certain ages as
world-wide in extent. As our knowledge of them has increased,
through more extensive collecting and improved techniques for
extracting information, the floras from various parts of the world
for a particular period, say the Jurassic or Permian, show marked
differences in composition. It is my impression that future study
will continue to emphasize the diversity of floras in different
regions for any one point in time. In the following pages a few
vistas are opened which reveal a little of what has gone on in the
past, but to attempt a generalization in the nature of world-wide
floristics would be precarious if not quite misleading.
A tremendous amount of floristic information is gathered to¬
gether in Seward’s Plant Life through the Ages and, although
much has been added to our fund of knowledge since its date of
publication, it will continue to be a valuable reference. It is par¬
ticularly recommended to students who have already acquired
some understanding of fossil plants.

Telegram @tubscstds
408
SOME PALEOZOIC AND MESOZOIC FLORAS 409

The Distribution of Vascular Plants in the

Lower and Middle Devonian

The statement has been made by several competent students of

early land vascular plants that the floras of the Lower and Middle
Devonian were essentially uniform throughout the world; it seems
to me that this viewpoint has been somewhat overemphasized.
In the first place the localities from which these early evidences of
plant life on the land have been obtained are somewhat short of
being world-wide in coverage. Second, in view of the relative sim¬
plicity of many of the early Devonian plants and the fact that they
are often incomplete or poorly preserved, it is my feeling that their
interrelationships may not be as close as has been supposed.
The greater part of our knowledge of the earlier Devonian floras
has come from studies of fossils found in northwestern Europe,
eastern North America, Australia, and more recently, Siberia.
The earlier chapters of this book have drawn chiefly from these
geographical sources only because they have yielded the most
significant plant remains. The more important ones found in
other parts of the world may be summarized briefly.
A Middle or Lower Devonian flora has been reported recently
by Teichert and Schopf from the canyon of the Salt River in cen¬
tral Arizona. Initial studies reveal plants similar to Cooksonia,
Rhynia, and Hicklingia as the more common elements; it may be
noted that Hicklingia edwardi, as originally described by Kidston
and Lang from the Middle Old Red Sandstone of Scotland, was a
densely tufted, dichotomously forking, leafless plant with terminal
sporangia and is possibly quite close to Rhynia. Other plants of
the Salt River flora suggest Hedeia and Dawsonites; isolated spores
have been found and although poorly preserved it is interesting to
note that some of them are regarded as being the spores of
heterosporous plants. The only other early Devonian flora from
the western states is the one from Beartooth Butte, Wyoming,
which has been mentioned previously in connection with Psilophy-
ton wyomingense and Bucheria ovata. It also includes Hostimella
and Broggeria strobiliformis Dorf, the latter being a problematical
strobiluslike body bearing some resemblance to an Equisetum cone.
The type species, B. noruegica Nathorst, was found in the lower
Middle Devonian of Norway and appears to be a unique cone of
some sort, but nothing more precise can be said about it.

Telegram @tubscstds
410 STUDIES IN PALEOBOTANY

Sahni has reported very fragmentary small, naked, forking

branches and one spiny specimen from Spiti province in the northeast
Himalayas; although the preservation is not good, these offer
evidence of early vascular plants. From the Lower or Middle
Devonian of South Africa H0eg has reported Dutoitia pulchra,
consisting of small dichotomous fragments with terminal, flat-
topped sporangia about 3 mm broad; no spores have been found in
them. Kryshtofovich has recorded Psilophyton, Sporogonites
exuberans, and Dawsonites arcuatus from Turkestan in Central
Asia.
Hsii has described several plants of particular interest from
rocks of Lower or Middle Devonian age of central Yunnan in south¬
western China. Plants tentatively referred to Taeniocrada dubia
are preserved in the form of thin, flattened incrustations up to 1
mm thick and he suggests that it was not a ribbon-shaped axis;
the tracheids of the central strand have small, round, uniseriate
bordered pits with elliptic pores. Other fossils are tentatively
assigned to Asteroxylon and Protolepidodendron; additional lyco-
pods are present which indicate a horizon somewhat higher than
the Lower Devonian.
Halle discovered some rather poorly preserved plants from the
Devonian (presumably Lower) in the Falkland Islands. These in¬
clude slender, leafless, dichotomously forking shoots, some with
globose bodies at the tip which may be sporangia; there are also
doubtful fragments which may represent the terminal spike of a
Zosterophyllum. Some problematical plants have been described
by Termier from central Morocco; they are of lower Middle
Devonian age and suggest species of Psilophyton, Asteroxylon, and
Aneurophyton.
So far as I am aware the records cited above are the chief ones
that we have for the earliest land vascular plants aside from the
North American, northwest European, and Siberian and Australian
localities. It should be noted that very few are positively known
to be Lower Devonian and the preservation for the most part is
such that identification to the genus is precarious. The conclusion
is unavoidable that our knowledge of the earliest land vascular
plants is wholly inadequate to allow any generalization on a world¬
wide basis.
By Middle Devonian times a rather diverse assemblage of plants
was established on the land; they were generally larger and more
complex than those from the Silurian and Lower Devonian. One

Telegram @tubscstds
of the most significant insights that we have into the landscapes
SOME PALEOZOIC AND MESOZOIC FLORAS 411

of this age comes from Professor H0eg’s studies of the Spitzbergen

flora; a few of the plants have been mentioned previously and some
others may be introduced here. It has not been possible to deter¬
mine the exact stratigraphic position of all of the Spitsbergen
fossils; however, those cited below fall within the range of the
upper part of the Lower Devonian and the lower part of the
Upper Devonian.
Psilophyton arcticum H0eg seems to have borne a fairly close
resemblance to P. princeps and P. wyomingense; the stem frag¬
ments attain a breadth of 5 mm and larger specimens branch by
unequal dichotomies; thus the plant probably had an upright habit
similar to that of Asteroxylon. The spines measure 5 to 6 mm
long with some as much as 8 mm. Psilodendrion spinulosum
H0eg is another probable psilophyte; the largest stems reach a
diameter of 1 cm and the main branches are arranged oppositely
or nearly so. These primary order branches are monopodial,
whereas those of higher orders are dichotomous. Spines are
present but rather sparsely distributed and somewhat under 2 mm
long. Svalbardia polymorpha has been described previously as a
plant that is seemingly transitional between the psilophytes and
protopterid ferns. Lycopod stem casts described under the name
Bergeria mimerensis H0eg range up to 10 cm in diameter and it is
quite clear that they represent small trees. Other stem impres¬
sions have been found, as much as 15.5 cm broad, which date to
the lower Middle Devonian or older; although unidentifiable they
present drammatic evidence of trees of respectable size in the early
part of the Devonian. It might be added that the specimens show
distinct longitudinal striations, with ridges several millimeters
apart, which suggest cortical fibers; it is thus certain that they do
not represent the presumed alga, Prototaxites. A few petrified
plants are present in the Spitsbergen flora, one of the most unusual
being Actinopodium nathorstii H0eg. It is a stem about 16 mm in
diameter containing a seven-rayed stele with a central mixed pith
of parenchyma and tracheids; the protoxylem is probably mesarch,
and of special interest are the radially aligned, presumably second¬
ary, tracheids with multiseriate pits in their radial walls. In the
mesarch protoxylem, fairly well-developed pith, and apparent
secondary wood there is a suggestion of pteridosperm affinities.
Adding to the variety of this mid-Devonian flora is Enigmophyton
superbum and a few other problematical plants.
Several years ago Professor Suzanne Leclercq discovered a
remarkable Middle Devonian plant horizon near the village of Goe

Telegram @tubscstds
412 STUDIES IN PALEOBOTANY

in eastern Belgium, the specimens of Calamophyton bicephalum

described in Chapter 9 having been obtained from there. The
preservation is generally excellent; the flora includes several dis¬
tinctly new plants and it may be expected to shed a great deal of
light on others that are inadequately known at present. It is safe
to assert that the Goe deposit, when fully studied, will yield a
more diverse assemblage of plants than has yet been found in the
Middle Devonian and it will reveal vascular plants of greater com¬
plexity than was formerly supposed existed at that time. I am in¬
debted to Professor Leclercq for allowing me to insert this comment
on the Goe plants since it will be some few years before they are
fully studied.
Upper Devonian floras are known from eastern Canada and the
United States, Ireland, Great Britain, Germany, the USSR, Bear
Island, Spitsbergen, and Australia. Among the widely distributed
elements of them are several species each of Archaeopteris and
Callixylon;* many others occur in one or more localities and it is
evident that the vegetation of this age was a highly diversified one.
To present, briefly, representative views of Upper Devonian land¬
scapes two areas will be considered, Belgium and the east central
United States.

The Upper Devonian Plants of Belgium

As a result of the investigations of Stockmans, Leclercq, and

others the late Devonian rocks of Belgium have yielded a consider¬
able number of interesting plants. Archaeopteris, Aneurophyton,
and Rhacophyton have been mentioned in Chapter 3. Six species
of Sphenopteris and one of Diplotmema have been described;
among the fernlike fronds Sphenocyclopteridium begicum Stock-
mans is particularly distinctive, the fronds being at least twice pin¬
nate and with pinnules that are nearly circular in outline. In
Chapter 13 note was made of several fossils that are suggestive of
being seeds or cupules, such as Moresnetia, Condrusia, and Xeno-
theca; the association of sphenopterid foliage adds to the probabil¬
ity that pteridosperms, or plants close to that level of organization,
were already in existence. The high level that the articulates had
attained is evinced by the Eviostachya h</>egi cones with their com¬
plex sporagiophores. Stem impressions nearly 10 cm in diameter

* As this was going to press the important discovery has been announced (Beck,
1960) of a pyritized stem with Callixylon wood structure bearing several fragments

Telegram @tubscstds
of Archaeopteris fronds.
SOME PALEOZOIC AND MESOZOIC FLORAS 413

have been found and, while not identifiable, they indicate here, as
elsewhere in the Devonian, plants of arborescent dimensions.

Upper Devonian Plants from East Central United States

In view of what is already known and what may be expected

within the next few decades the fossil floras of the Black Shales of
the east central states, especially central Kentucky, are beginning
to assume a niche of great importance. There is some disagree¬
ment concerning the exact stratigraphic position of the plant hori¬
zons, whether uppermost Devonian or basal Mississippian, the
latter being most probable for the majority of the 28 or more
genera discovered thus far. Unlike most Upper Devonian floras
the Kentucky one is based entirely on petrified remains; it is of
special interest by virtue of the variety of plant form that is rep¬
resented, and by the fact that many of the fossils are quite unlike
any discovered elsewhere.
Among the more unusual fossils are several that are best
regarded as problematical pteridophytic plants.
Stenokoleos simplex (Read and Campbell) is known from stems
about 3.5 mm in diameter with a more or less cruciform vascular
strand that occupies a considerable portion of the stem (Fig. 16-1).
There is an apparent resemblance to Asteroxylon, but the two

Telegram @tubscstds
Fig. 16-1. Stenokoleos simplex, a transverse view showing the conspicuous stele and
two traces (t) departing at the left, about 25X. (From Beck, 1960.)
414 STUDIES IN PALEOBOTANY

differ fundamentally. In Stenokoleos the stele occupies a larger

portion of the stem, it lacks the middle trabecular cortex of
Asteroxylon, possesses numerous (about 14) mesarch protoxylem
points, but most striking is the organization of the appendages.
Rather large triangular-shaped traces are given off in pairs from
opposite arms of the stele; the diameter of these appendages some¬
what exceeds that of the main axis at the point of departure, but
they rapidly taper to about 2 mm in diameter. The plant is
regarded by Beck as possibly referable to the Zygopteridaceae, but
this is admittedly uncertain.
Guycampbellia microphylla (Read and Campbell) is another
actinostelic type in which the stems are known up to 5 mm in
diameter. It apparently bore closely appressed microphyllous
leaves 6 to 10 mm long which were twice forked. There is some
resemblance to Protolepidodendron in the form of the foliage but
the stele, with deeply embedded mesarch protoxylem, is certainly
not typically lycopodiaceous.
Polyxylon elegans (Read and Campbell) is known from stems 5
to 8 mm in diameter in which the stelar system consists of a ring
of U-shaped strands; the tracheids are large and scalariform.
Nothing is known of the appendages of this interesting fossil.
Pietzschia polyupsilon (Read and Campbell) is based on stems
of about 25 mm in diameter with a single peripheral ring of about
54 radially aligned primary strands, some of which appear Y-shaped
as a result of frequent branching and anastomosing of the meri-
steles. It has been suggested that Pietzschia, which is also known
from the Upper Devonian of Saxony, is related to Cladoxylon.
Siderella scotti (Read) is especially perplexing. The stem is a
little under 1 cm in diameter and the star-shaped stele, with 8 to
10 rays or lobes, gives off vascular strands of two types: there are
small ones that depart in whorls and each trace divides once or
twice in its outward course; the others are relatively large, H-
shaped “zygopterid” strands and these depart in opposite pairs.
This distinctive vascular nodal anatomy has led to the supposition
that the plant may be allied to both the articulates and the zygo¬
pterid ferns. It seems most unlikely, however, that there could
be “intermediate” forms between two such divergent groups; it
may be recalled that whorled appendages are not exclusive to the
articulate group. Siderella has received only a cursory study as of
this writing, but it seems very possible that it may represent a
group of plants quite distinct from any known at present.

Telegram @tubscstds
Small lycopod stems have been attributed to Lepidodendron,
SOME PALEOZOIC AND MESOZOIC FLORAS 415

but these are for the most part rather poorly preserved. A cone,
Lepidostrobus fischeri, described some years ago by Scott and
Jeffrey, stands in contrast to the small and scanty vegetative
remains. It is known from a fragment only 8 cm long but is suffi¬
ciently intact to indicate an over-all diameter of 4 cm; thus it com¬
pares favorably with the large cones of the arborescent Upper
Carboniferous lycopods. The opinion has also been expressed that
the bisporangiate cone Lepidostrobus noei (mentioned in Chapter
8) was derived from a horizon close to that from which the other
Black Shale plants originated.
Certain stems and petioles have been assigned to several species
of the Calamopityeae, a group of presumed pteridosperms. Diich-
nia kentuckiensis Read may be taken as an example; it is a stem
with an angular pith about 7 mm broad and although predomi¬
nantly parenchymatous there are tracheids or clusters of tracheids
scattered through it. Excentrically mesarch primary vascular
strands are present in the angles of the pith which is enclosed by
a broad zone of secondary wood; the latter is composed of multi-
seriate pitted tracheids and rays which vary considerably in width,
but many are eight cells broad and thus lend a conspicuous pattern
in transverse sections. The outer cortex consists of alternating
bands of parenchymatous and sclerotic (apparently fibrous) cells.
The petiole vascular supply originates from branches of two pri¬
mary strands; as these enter the cortex they tend to elongate and
divide to form a semicircle of bundles in the base of the petiole.
The genus Calamopitys, which is known from several species
both in this country and Europe, is basically similar to Diichnia;
perhaps the most notable difference lies in the development of the
petiole vascular supply from one peripheral primary bundle of the
stem rather than two.
Callixylon is represented by numerous small fragments and occa¬
sional logs in excess of 1 foot in diameter.
In summary, the 45 or more species now known from the Black
Shales are an especially diverse aggregation of plants; the Calamo¬
pityeae are probably seed plants and several pteridophytic groups
are included, some of which are very likely new. Not a few are
especially problematical, such as Steloxylon and the Foerstia-Pro-
tosalvinia complex, and investigations which are now in progress
will undoubtedly bring others to light. Although the petrifactions
are not especially abundant it seems to me that this is a most
promising source of information concerning a very critical age in

Telegram @tubscstds
the evolution of pteridophytic and early seed-plant groups.
416 STUDIES IN PALEOBOTANY

Asian and Southern Floras of the Late Paleozoic

and Early Mesozoic

According to our present knowledge the floras of the late Car¬

boniferous to the early Triassic fall into four groups as follows:

1. The Euramerican flora extends from the Ural Mountains

through western Europe and takes in the United States and
Canada to eastern Kansas, with western outliers in Colorado and
northern New Mexico. Fossils found in this general area form a
large part of the subject material considered in Chapters 3 to 5
and 8 to 9.
2. The Angara (also known as the Siberian or Kuznetzk) flora
occupies the territory known as Angaraland which is bounded by
the Urals in the west and extends east through Siberia to the
Pacific, and from the Arctic ocean south to outer Mongolia.
3. The Cathaysia (or Gigantopteris) flora extends from Shansi,
China and Korea south through Sumatra, New Guinea, and with
what appear to be outliers in Colorado, Oklahoma, and Texas. It
may be noted on Fig. 16-2 that this overlaps a corner of the Angara
province.
4. The Glossopteris flora occupied an area referred to as Gond-
wanaland which includes India, the central and southern part of
Africa, the southern half of South America, Australia, and
Antarctica.

It is convenient but not quite correct to apply the phrase “the

flora” to the fossil assemblage to any of these areas. We are deal¬
ing with large geographic areas and long time spans in each one;
the Euramerican and Angara provinces are best known and it is
well established that the plant assemblages vary considerably, both
geographically and stratigraphically. We must of necessity deal
with them here in rather general terms.
One of the major problems involved in understanding these
floras is the matter of their distribution, this being especially true
of the Euramerican and Glossopteris provinces. The most favored
explanations have been the former presence of land bridges or
continental drift. The latter has been pronounced both impossible
and indispensable; a rather tremendous literature has been devel¬
oped by its proponents and opponents and since it is still a very
live issue in connection with the distribution of the Glossopteris
plants a brief consideration of it seems appropriate. For further

Telegram @tubscstds
SOME PALEOZOIC AND MESOZOIC FLORAS 417

information the student is referred to Chaloner’s well-formulated

summary of the problem cited at the end of the chapter.
In 1915 Alfred Wegener published his detailed treatise of the
proposal that the continents were once aggregated together into a
single huge land mass; this concept was later modified by A. L.
DuToit who proposed two primitive land masses, a northern one

Telegram @tubscstds
418 STUDIES IN PALEOBOTANY

including what we now call North America, Europe, and Asia, and
a southern one composed of South America, Africa, Antarctica,
Australia, and India. The separation of this great southern land
mass is thought, by many supporters of the idea, to have started
early in Tertiary times.
Some of the points which have been presented in support of con¬
tinental drift are the general similarity in outline of the west coast
of Africa and the east coast of South America and a close compar¬
ison in composition and grain size of the rocks along the respective
opposing coasts. There is also agreement in a line of ancient
mountain chain systems along the east side of North America and
continuing through parts of Britain and northern Scandinavia; a
comparable structural line is found extending through the tip of
the Cape and the opposite coast of South America.
Next, there is evidence of widespread glaciation in late Paleozoic
times in southern Australia, extreme southeastern South America,
the southern tip of Africa and in India. The latter area is espe¬
cially perplexing since the direction of ice movement seems to have
been north from a point near the equator. The great difference in
both latitude and longitude of these presently remote areas is dif¬
ficult to explain other than on the basis of a former close associa¬
tion; correlated with this is a presumed change in the position of
the poles, a point that is anathematic to many physical scientists.
A considerable fund of biological evidence has been marshalled
in defense of the idea from distribution patterns of both modem
and fossil plants. Several genera and families of flowering plants
seem to afford cogent support: the Proteaceae is distributed
through most of South America, the southern half of Africa, Aus¬
tralia, India, and Madagascar; the Restionaceae is found in the
extreme southern tips of South America and Africa, in Madagas¬
car, Malaya, and Australia; the genus Symphonia (Guttiferae) is
distributed through the northern half of South America and in
west central Africa. An interesting case of an animal is found in
the small Triassic reptile, Mesosaurus, which is known only from
the east coast of South America and along the African coast on
the opposite side of the ocean; having been a small animal that is
thought to have lived in a delta swamp environment it does not
seem possible that it could have crossed the south Atlantic with
its present breadth. Most impressive of all the positive evidence
is the distribution of the Glossopteris flora, which is presented in
some detail below.

Telegram @tubscstds
SOME PALEOZOIC AND MESOZOIC FLORAS 419

Proponents of former land bridges have frequently voiced their

case with enthusiasm and not entirely without evidence that com¬
mands some consideration. The mid-Atlantic Ridge presents an
interesting case in point; one of the most intriguing bits of evidence
that has turned up recently supporting the belief that islands, per¬
haps of considerable magnitude, once existed comes from Kolbe’s
report of abundant fresh-water diatoms as well as the distinctive
silica “casts” of the epidermal cells of grasses and sedges, found in
deep sea cores taken about 1000 kilometers off the west coast of
Africa between the equator and approximately 10° N latitude. It
is possible that the diatoms may have been transported by sea or air
currents from the African mainland, but Kolbe regards it as much
more likely that the cores, in which the diatoms are most abun¬
dant, came from a lake bed of an island of the mid-Atlantic Ridge.

The Angara or Kuznetzk Flora

The Angara flora comes chiefly from the Kuznetzk and Minus-
sinsk depressions in central Asia within the Altai-Sayany mountain
regions. In the Kuznetzk basin the succession of sediments bearing
the fossil plants begins with the Lower Carboniferous, but it is not
until the Lower Permian that the flora, with all of its unique ele¬
ments, reaches its full development. There is a strong lycopod
component in the Lower Carboniferous and this gives way to what
has been called a cordaite flora, although many other groups are
present; there is, in fact, a total of some 600 species according to
a recent summary by Neuburg (1958), although full descriptions of
all of them have not been given.
The following summary refers chiefly to late Carboniferous and
Permian plants of the floral succession. The cordaites are repre¬
sented by the leaf genus Noeggerathiopsis, 17 species having been
recognized; in general organization they are comparable with the
leaves of Cordaites and may be as large, but in most species they
are smaller, some being only a few centimeters long, and they
vary from lanceolate to spatulate in outline.
Fernlike fronds referable to Sphenopteris, Pecopteris, Callipteris,
and Neuropteris are represented by several species each, whereas
there are many other genera that indicate plants of a rather dif¬
ferent nature from those found in the Euramerican upper Paleo¬
zoic, such as Gondwanidium, Tschirkoviella, Angaropteriduim,
and Angaridium (Fig. 16-3); reproductive organs are known for

Telegram @tubscstds
420 STUDIES IN PALEOBOTANY

Fig. 16-3. Some plants of the Angara flora. A. Ginkgophyllum vsevolodi; B. Angari-
dium mongolicum; C. Gondwanidium petiolatum; D. G. sibiricum; E. Tschirkoviella
sibirica; F. Angaropteridium cardiopteroides. (All from Neuburg, 1948.)

very few and future studies will almost certainly reveal many new
ferns and pteridosperms. The articulates are represented by
Sphenophyllum, Annularia, Phyllotheca (with exceptionally long
slender leaves), and Paracalamites robustus with pith casts at

Telegram @tubscstds
least 8 cm in diameter. Among a few apparent ginkgophytes
SOME PALEOZOIC AND MESOZOIC FLORAS 421

Ginkgophyllum vsevolodi Zalessky is suggestive of Dichophyllum

moorei, a Kansas fossil from the top of the Carboniferous men¬
tioned in Chapter 11.
Several true mosses (Bryales) as well as early peat mosses (Pro-
tosphagnales) have been found which are considered to indicate a
trend toward a more temperate climate in Angaraland.
Conifers are very scarce in contrast to their relative abundance
in the late Paleozoic floras of the Euramerican province. Finally,
as suggestive of the unique elements that will probably continue to
turn up in the Angara flora, reference may be made to Vojnovskya
paradoxa which was described in Chapter 12.

The Cathaysia Flora

Much of our knowledge of this flora comes from Halle’s studies

of uppermost Carboniferous and Lower Permian plants from the
province of Shansi in east central China. His original collections,
made during several months of field work in 1917, were lost when
the ship carrying them to Stockholm went down in a typhoon;
later collections formed the basis of his classic work (1927). In
view of the many interesting plants that it contains one cannot
help wondering what was lost and what may turn up in the future.
The following plants are representative of the flora:
Annularia seems to have been quite abundant and Halle notes
that specimens of A. stellata agree closely with European ones. A
distinct articulate genus, Lobatannularia, is characterized by
markedly excentric whorls of leaves; that is, some are notably
smaller than others (Fig. 16-4D); it also differs from Annularia in
that there are four branches at a node. Several species of Spheno-
phyllum are present; the leaves of S. thoni Mahr are quite large,
being nearly 5 cm long and the vascular system is fan-shaped with
vein endings distributed along the sides as well as the distal edge.
In some specimens of S. rotundatum Halle the leaves are almost
round.
The Gleicheniaceae was represented by Oligocarpia gothanii
Halle which apparently was a small plant. Halle figures a speci¬
men consisting of several leaves not much more than 10 cm
long attached to what is probably a rhizome; it is suggested that
it was a creeping plant occupying mud flats or was possibly float¬
ing. A closely allied genus, Chansitheca, is distinguished by
elongate sori and more sporangia per sorus (as many as 16 in C.

Telegram @tubscstds
kidstoni Halle) than in Oligocarpia.
A species of Cladophlebis is included in the flora; this is a com-
422 STUDIES IN PALEOBOTANY

Fig. 16-4. Some plants of the Shansi flora. A. Taeniopteris latecostata; B. Plagio-
zamites oblongifolius; C. Tingia carbonica; D. Lobatannularia ensifolius; E. Norinia
cucullata; F. Protoblechnum wongii; G. Gigantopteris nicotianaefolia. (Drawn from
specimens in the Swedish Natural History Museum and from Halle, 1927.)

mon Mesozoic genus and is usually regarded as representing the

Osmundaceae.
The presence of the pteridosperms is implied by quite a number
of the genera and there can be no doubt that there were many
members of the group in the flora. A few have been found with
seeds attached, namely, Pecopteris wongi, Alethopteris norinii,
and Emplectopteris triangularis (see Chapter 5).

Telegram @tubscstds
SOME PALEOZOIC AND MESOZOIC FLORAS 423

The form-genus Taeniopteris is an abundant element of the flora,

but Halle himself recognized the uncertainty of clearly defined limits
for the 11 species that he named. Taeniopteris is widely distributed
and ranges from the uppermost Carboniferous to basal Cretaceous; it
may be recalled that one species was cited in Chapter 10 as the
foliage of Williamsoniella coronata. The entire, lanceolate, leaves
have a strong midrib and side veins that depart at an angle of 70
to 90°; these usually divide but once immediately after leaving the
midvein and thus appear parallel. In his study of the Scoresby
Sound plants (Chapter 7) Harris proposed the segregation of cer¬
tain species into three natural groups: those that are known to be
the fronds of marattiaceous ferns; those with bennettitalean type
stomata (as Taeniozamites)\ those with cycadean type stomata (as
Doratophyllum); and finally, species that are otherwise indeter¬
minate are retained in Taeniopteris. Some of the Shansi taenio-
pterids attained a respectable size; there are fragmentary speci¬
mens of T. nystroemii Halle in the Swedish Natural History
Museum that are 20 cm broad; very likely they appeared similar
to a fair-sized banana leaf in life.
The “feature element” of this flora is Gigantopteris nicotianae-
folia Schenk, and lends the frequently used name “Gigantopteris
flora.” The fronds were large although the total size and form are
not fully known. They were at least once pinnate, 30 cm or more
broad, and the main rachis, which was 13 mm in diameter, bore
opposite, slightly overlapping pinnae. The latter attained a length
of 15 cm and were dentate to almost entire; a conspicuous midvein
produced nearly parallel laterals that terminate in the teeth apices,
and the finer venation in between these is netted. Towards the
apex of the frond the pinnae merge into each other forming a mas¬
sive terminal portion. Halle established another species, G.
lagrelii, which he regarded as intermediate between G. nicotianae-
folia and Emplectopteris triangularis. In a later study he
described tendrillike structures which have the same general
branching pattern as the normal leaves of G. nicotianaefolia, indi¬
cating the possibility of a vinelike habit for the plant.
A few of the other plants of the flora may be mentioned briefly.
Cordaites is represented by two species and in one the leaves at¬
tain a breadth of 45 mm. In a more recent study Sen and Bose
have described a petrified stem, Cordaites sahnii, from the Upper
Carboniferous or Lower Permian of central Shansi; it is a fragment
with a radial dimension of about 4 cm and shows no evidence of

Telegram @tubscstds
424 STUDIES IN PALEOBOTANY

growth rings. Saportaea nervosa has been mentioned in the chap¬

ter on ginkgophytes as a possible early member of that group;
there are also two species of Baiera recorded and one of these, B.
tenuistriata, is based on exceptionally large “leaves” in which the
segments attain a breadth of 1.5 cm. There is a fragmentary
specimen in the Museum in Stockholm that is 15 cm long and ob¬
viously only a small portion of the leaf; this is a most interesting
fossil but must be regarded as a doubtful ginkgophyte. There are
several species of the problematical genus Tingia; as an example,
T. carbonica (Schenk) Halle consists of an axis 3 to 8 mm thick
which bears a strong resemblance to a cycadophyte frond, but in
addition to two lateral rows of “leaflets” there are two more on the
under side although they are somewhat smaller.
In summary the following features of the flora seem most signi¬
ficant. Nearly 50% of the species are attributed to the ferns or
pteridosperms. It seems safe to assume that the latter were
abundantly represented and included some unique forms; of special
interest are the three species with seeds attached to the foliage.
The Equisetales and Sphenophyllales are characterized by aniso-
phyllous species, that is, ones in which leaves of varying size are
found in the same whorl. The lycopods compose a very minor
element of the flora, and the cordaites, although fairly abundant
in specimens, are represented by only two species. The extreme
scarcity of the conifers is also noteworthy, especially the absence
of Walchia which is a conspicuous feature of Lower Permian
floras in the Euramerican province. Several plants are attributed to
the ginkgophytes and although of considerable interest it seems to
me their classification must be regarded as tentative.

The Glossopteris Flora

The literature dealing with the Glossopteris flora is now exten¬

sive and quite scattered and having accumulated over a period of
about one century it is encumbered by some error and no small
amount of divergence of opinion. Krishnan has recently given a
useful stratigraphical analysis of the entire Gondwana era and it
includes charts for the beds in several parts of India as well as cor¬
relations for South Africa, Australia, Brazil, and Argentina. In
India the Gondwana era is regarded as beginning with the Talchir
tillite in the Upper Carboniferous and continuing into the middle
Cretaceous. As might be expected this range includes a great

Telegram @tubscstds
SOME PALEOZOIC AND MESOZOIC FLORAS 425

variety of plant forms. Glossopteris and its characteristic asso¬

ciates appear in the Upper Carboniferous and extend into the early
part of the Triassic but are rare thereafter; this constitutes the so-
called Lower Gondwana flora with which we shall be concerned in
the following pages. Although this extends through southern
South America, the Falkland Islands, Antarctica, South Africa,
Australia, and India, the discussion will deal largely with the flora
as it occurs in India. The geology of the Gondwana system is best
known in that country and especially active studies of the flora
have been conducted there in recent years.
Glossopteris and Gangamopteris are the most notable elements
of the flora; some 40 species of the former have been described and
about 18 of Gangamopteris. Recent studies of the epidermal
structure confirm the fact that numerous species existed, but the
number may not be quite as high as the figures cited. In both
genera the leaves are net-veined, Gangamopteris being separated
by the supposed lack of a distinct midrib. A third leaf type may
be appropriately introduced here, Palaeovittaria kurtzi Feistman-
tel, which is of the same general shape but is usually described as
having dichotomous, open (not netted) venation. However, accord¬
ing to recent studies the three cannot be sharply delimited by
gross macroscopic characters or cuticular structure. Vertebraria
indica Royle is a name that was given to axes 2 or 3 cm in diam¬
eter, with characteristic radially aligned wood sectors, that are
thought by some workers to have been the rhizome of Glossopteris.
More recently Glossopteris leaves have been found attached in
whorls to slender axes little more than 1 mm broad. The repro¬
ductive organs found attached to Glossopteris and Gangamopteris
leaves have been mentioned in Chapter 12. Mrs. Plumstead has
summarized the evidence bearing on the habit of the glossopterids
and suggests that they may have been large deciduous trees with
the foliage arranged on both long and short shoots.
Noeggerathiopsis, although present, apparently was not a domi¬
nant element as in Angaraland; several species of cordaitean wood
(Dadoxylon) have been found in India and other parts of Gond-
wanaland, most of which display distinct annual rings. There are
only three genera of conifers known, Buriardia, Paranocladus, and
Walkomiella; W. australis (Feistmantel) Florin is regarded as hav¬
ing been a large forest tree probably similar in general appearance
to the modern Araucaria cunninghamii which is found in Queens¬
land and New South Wales. In another species, W. indica Surange

Telegram @tubscstds
426 STUDIES IN PALEOBOTANY

Fig. 16-5. Some plants of the Glossopteris flora. A. Schizoneura gondwanensis;

B. Glossopteris sp.; C. a portion of the leaf enlarged to show venation; D. Sphenophyl-
lum speciosum; E. Phyllotheca etheridgei. (A from Feistmantel, 1881; D from a speci¬
men in the British Natural History Museum; E from Saksena, 1954.)

and Singh, small seed-bearing shoots have been identified which

compare with those of Lebachia; very possibly they were arranged
in a cone like the dwarf-shoots of the latter genus.
There are a few distinctive articulate types. Schizoneura gond¬
wanensis Feistmantel has been described as having leaf sheaths
that are usually divided into two parts which are ovate-lanceolate
with a rounded tip; occasionally the sheath (or paired leaves) may
be dissected into several segments each. The genus also has been
reported from several widely scattered Triassic-Rhaetic localities
in Europe and China. Srivastava has described a cone from the
Raniganj coalfield of Bengal which is believed to have been borne
on £>. gondwanensis; it is 30 mm long and bore appendages that

Telegram @tubscstds
seem to have been similar to those of Equisetum, there being no
SOME PALEOZOIC AND MESOZOIC FLORAS 427

evidence of associated bracts. Recent restorations of two species

of Phyllotheca show them as plants with foliage sheaths that were
deeply dissected in the younger parts of the branchlets and as they
matured the sheath developed an umbrellalike form. Sphenophyl-
lum was fairly abundant as an undergrowth plant.
The lycopods are not common in the Glossopteris flora and the
few specimens available are poorly preserved; as a consequence
there has been considerable disagreement over their identification.
Lycopodiopsis derbyi Renault is known from petrified stems in
which the primary xylem cylinder is discontinuous, that is, dis¬
sected; otherwise it is said to be typically lepidodendraceous.
Lepidodendron and Sigillaria have both been reported, but in a
recent study of southern hemisphere lycopods Edwards concludes
that the Glossopteris flora does not include any species that are
common to both southern and northern floras, and that there is no
satisfactory evidence to indicate that any genera are common to
both.
A few other plants of the flora may be cited briefly: there are
several species of sterile Sphenopteris, the affinities of which are not
known. A fertile pecopterid frond fragment from the Raniganj
coalfield of India has been tentatively assigned to the marattiace-
ous genus Ptychocarpus; the discovery of a rather well-preserved
Psaronius stem in Brazil many years ago tends to confirm the
presence of the Marattiaceae in the flora. Fronds originally
described under the name Neuropteridium appear to be quite
closely related to some of the Angara species of Gondwanidium
and have been transferred by certain authorities to that genus.
Belemnopteris woodmasoniana Feistmantel is a distinctive arrow-
shaped frond with net venation which has been found only in the
Upper Permian of India. Surange has recently described a fossil
from the Raniganj coalfield consisting of a cylindrical axis about
1 cm broad and 5 cm long which was borne on a slender stalk 1
mm in diameter; numerous sessile, exannulate sporangia, each con¬
taining many spores, were borne on, or slightly embedded in, the
axis; it is regarded as being a fern or pteridosperm but is a unique
spore-bearing organ.
Relatively little has been done as yet with the palynology of the
Lower Gondwanas, but in two recent studies more than 40 distinc¬
tive microspore types were found in coal from Bihar and 20 species
of megaspores have been identified. The megaspores are classified
in the genus Triletes which is regarded as being lycopodiaceous,

Telegram @tubscstds
but it is by no means certain that all the species are correctly as-
428 STUDIES IN PALEOBOTANY

signed to that group. There is, however, a strong suggestion that

the lycopods were more abundant than is indicated by the macro¬
fossil record.
The evidence available at present indicates that the Glossopteris
assemblage was not nearly as rich a one as is found in the northern
floras. In view of the widespread glaciation, as evidenced by exten¬
sive tillites or boulder beds, there has been considerable discussion
over the matter of the Gondwana climate. Gangamopteris and
Phyllotheca have been found as low as the Dwyka series (Upper
Carboniferous of South Africa) and spores have been found in some
abundance in the Dwyka and in the Bacchus Marsh beds (Upper
Carboniferous of Victoria). It thus seems evident that the flora
existed during times of extensive glaciation and the conclusion that
it was a cool temperate one seems reasonable.
The effect of a glacier on the vegetation of the immediate vicin¬
ity depends in part at least on the size of the ice sheet and
the latitude. Tree ferns grow within a mile of the terminal face
of the Franz Josef Glacier in New Zealand, a dramatic photo of
this contrast being given in Seward’s Plant Life through the Ages.
On the other side of the world, glaciers in western Norway come
down very nearly to sea level and exist in close proximity to
forests of fir, birch, and alder, as well as fine apple orchards. One
of the most memorable sights I have ever encountered for geolog¬
ical and botanical interest, as well as sheer beauty, is the view
across Sogne Fjord in western Norway. Looking north from the
highland above the town of Vik one can see part of the great
Jostedals Glacier, covering some 1076 square kilometers, and in
autumn the apple harvest will be in full swing in the many small
farms bordering the fjord—an impressive contrast to the nearby
icefield.

Triassic Floras

The Middle Triassic Ipswich flora of Queensland

The early Mesozoic floras of the southern hemisphere are fre¬

quently referred to as Thinnfeldia floras because of the numerous
species of this widely distributed genus. The fronds are once or
twice pinnate with a stout rachis that may display a single dichot¬
omy and the pinnules vary, in different species, from ovate to nar¬
rowly elongate. The presence of a rather heavy cuticle has led
most authors to regard them as pteridosperms, but in view of the

Telegram @tubscstds
morphological variation it is possible that the natural relationships
SOME PALEOZOIC AND MESOZOIC FLORAS 429

may be correspondingly diverse. In the instance of T. feistmanteli

Johnson the pinnules are more or less rhombic and attached along
their entire base; exannulate sporangia are grouped three or four
to a sorus, and there are two or three rows of the latter on each
pinnule. In another species, T. lancifolia (Morris), in which the
pinnules are variable in shape but more or less lanceolate, sporan¬
gia have also been reported and seem to be nearly identical with
those of T. feistmanteli.
Stenopteris is another difficult genus that has been considered as
closely related to Thinnfeldia and is reported from northern as
well as the southern floras. The rachis may or may not be dichot¬
omized and the lamina of the pinnules is almost nonexistent; the
strongly developed cuticle is regarded as indicating gymnosperm
affinities. Of particular importance in this problematical assem¬
blage is a series of specimens (Fig. 16-6C) reported by Jones and de
Jersey from the Ipswich coal measures that reveal an evolutionary
sequence of increasing complexity; it is rarely possible to clearly
demonstrate such trends through a known stratigraphic sequence.
All of the forms that occur in the Ipswich flora were formerly as¬
signed the binomial S. elongata (Carruthers) Seward but this no
longer seems tenable. Specimens taken from the base of the series
bear “pinnules” in which essentially no lamina is present (Fig.
16-6C1); at a higher horizon some lobing appears in the base of
the pinnules and these are supplied with branch veins (Fig. 16-6C3);
still higher this lobing becomes constant throughout the length of
the pinnule and finally the lobes in turn begin to proliferate (Fig.
16-6C6). The transition indicated in the figures takes place
through approximately 3000 feet of sediment, and it is also inter¬
esting to note that the simpler forms do not immediately disap¬
pear but with the advent of the more complex ones they do show
a marked decrease.
Plants that can be positively identified as true ferns are not
abundant in the Ipswich flora. Three or four species of Cladoph-
lebis have been recorded, but the few fertile specimens are not well
preserved. Several species of Dictyophyllum are known and these
are assigned to the Dipteridaceae.
In one of several contributions on the Ipswich flora Walkom
(1917) described a considerable variety of ginkgophyte foliage
under the names Ginkgo and Baiera. Most of the leaves are quite
deeply dissected and petiolate, thus falling within the range of
Ginkgoites as defined by Harris (see Chapter 11). It is a little

Telegram @tubscstds
disturbing, however, to find Ginkgo antarctica Saporta at this low
430 STUDIES IN PALEOBOTANY

Fig. 16-6. Some plants from the Triassic Ipswich flora of Queensland. A. Thinnfeldia
talbragarensis; B. Yabeiella brackebushiana; C. a series showing evolution in Stenop-
teris leaves; 1, 2, S. elongata; 3-5, S. spinifolia; 6, S. tripinnata. (All from Jones and
de Jersey, 1947.)

Telegram @tubscstds
SOME PALEOZOIC AND MESOZOIC FLORAS 431

horizon; it is a species in which the leaf is entire or nearly so and

closely resembles the foliar organs of the modern G. biloba. The
variety of leaf types figured by Walkom suggests that the group
was quite diversified and very likely had its origin not later than
the Permian.
The flora includes a few cycadophytes as well as some leaf types
that are of questionable affinities. The Bennettitales is repre¬
sented by very fragmentary remains of Williamsonia reproductive
organs, and leaves of Pterophyllum. The Cycadales seem to have
been slightly more abundant with Ctenis, Nilssonia, and Dorato-
phyllum present, the last being a taeniopterid leaf with the cyca-
dean type stomata. Another interesting and problematical leaf
genus, Yabeiella, is represented by several species; the leaves are
entire, lanceolate to oval-lanceolate, and up to 15 cm long. The
venation is quite distinctive; two lateral veins depart near the base
of the blade and assume a near-marginal position; the numerous
other minor laterals occasionally anastomose and terminate in
these marginals.

The Triassic Flora of Arizona

In conclusion it seems of interest to compare the Ipswich flora

with one on the other side of the globe, namely, the Upper Triassic
Chinle flora of Arizona. The two were not exactly contempor¬
aneous, the latter being somewhat younger, but the differences in
composition are quite striking.
The Chinle flora is known from western Texas to Nevada al¬
though the most productive localities are in Arizona. The exten¬
sive accumulations of silicified logs composing the “petrified forest”
of Arizona have long been famous; however, a more significant
contribution to our understanding of the plants of that area was
made by Daugherty a few years ago following the discovery of im¬
pression fossils from several localities. The logs are colorful and
impressive but are often poorly preserved and comprise only a very
few species.
The cycadophytes and certain ferns are the dominant elements;
the Osmundaceae is represented by a petrified stem, Osmundites
walkeri Daugherty, which is of quite modern aspect; another stem
has been referred to the family, Chinlea campii Daugherty, which
is rather poorly preserved but appears to have scalariform tracheids
scattered through the central portion and is possibly intermediate

Telegram @tubscstds
432 STUDIES IN PALEOBOTANY

between Thamnopteris and Osmundites. Leaf impressions of

Todites and Cladophlebis are also referred to this family.
Some fine frond specimens of Phlebopteris smithii testify to the
presence of the Matoniaceae and a leaf specimen, regarded as suf¬
ficiently distinct to merit the name Apachea arizonica Daugherty,
is assigned to the Dipteridaceae.
There are several cycadophytes in the Arizona flora, but the
absence of preserved cuticular structure renders their identification
unsatisfactory. Otozamites powelli (Fontaine) Berry is a typical
once pinnate cycadophyte frond with oblong pinnules having a
truncated tip. Leaves of Macrotaeniopteris are described (M.
magnifolia (Rogers) Schimper) which are simple and up to 17 cm
broad and a meter long; poorly preserved fructifications were ob¬
served which may be fern sori but this is uncertain. Whatever the
plant’s affinities were it was abundant; although recorded at only
one locality it occurs as a mass of leaves more than a foot thick.
Macrotaeniopteris is associated with Equisetum-like stems and
largely on this account the plant is regarded as having grown in
low, marshy places.
A few additional notations may be inserted here concerning the
articulates; their record consists of very fragmentary specimens
but several are worthy of note: Neocalamites virginiensis (Fon¬
taine) Berry is known from stem casts that exceed 10 cm in diam¬
eter; these are small by comparison with the Carboniferous
calamites but considerably larger than our modern equisetums.
The stems of Equisetites bradyi Daugherty attain a diameter of
only 14 mm, but have leaf sheathes that appear very similar to
those of a modern Equisetum. The distinctive spores of Equiseto-
sporites chinleana, with their two elaters, have been mentioned in
Chapter 9.
Many of the petrified logs are those of a conifer, Araucarioxylon
arizonicum Knowlton; specimens have been found up to 7 feet in
diameter and 120 feet long and they may, therefore, be properly
referred to as a dominating as well as a dominant element of the
flora! Growth rings are faintly discernible. Schilderia adamanica
Daugherty is a wood with unique ray organization; it is thought
to have been a small tree since the diameter of the trunks does not
exceed 30 cm shortly above the fluted base. The wood consists of
tracheids, with some parenchyma cells toward the end of each
faintly defined growth ring, and small uniseriate rays; it is thus
essentially coniferous with the exception of rather regularly spaced
multiseriate rays. These are 0.3 mm broad and 10 mm high and

Telegram @tubscstds
SOME PALEOZOIC AND MESOZOIC FLORAS 433

form a very conspicuous feature; they are composed of irregularly

shaped parenchyma cells as well as some tracheids that bend into
them. The affinities of this curious wood are not known.
It is precarious to speculate on the climatic conditions under
which these plants grew. In all probability several habitats are
represented including a low, swampy one as well as an upland
environment. The ferns of the Matoniaceae and Dipteridaceae
imply a warm climate, but it seems to me that the cycadophytes
or, more precisely, the cycadlike foliage, is not classified precisely
enough to give any dependable clues. Growth rings of several
species are sufficiently well defined to indicate some seasonal
fluctuation.

REFERENCES

Arber, E. A. Newell. 1905. Catalogue of the fossil plants of the Glossopteris flora in
the Department of Geology, British Museum (Natural History), London. 255 pp.
Beck, Charles B. 1960. Studies of New Albany shale plants. I. Stenokoleos simplex
comb. nov. Amer. Journ. Bot., 47: 115-124.
--—. 1960. Connection between Archaeopteris and Callixylon. Science, 131:
1524-1525.
Chaloner, William G. 1959. Continental Drift. New Biology (Penguin Books), No.
29: 7-30.
Coleman, A. P. 1926. Ice Ages Recent and Ancient. Macmillan and Co., London.
296 pp.
Daugherty, Lyman H., and Howard R. Stagner. 1941. The Upper Triassic flora of
Arizona. Carnegie Instit. Washington Pub., 526, 108 pp.
DuToit, A. L. 1937. Our Wandering Continents, an Hypothesis of Continental Drift¬
ing. Oliver and Boyd, London. 366 pp.
Florin, Rudolf. 1940. On Walkomia n. gen., a genus of Upper Palaeozoic conifers
from Gondwanaland. Kungl. Svenska Vetenskapsakad. Handl., ser 3, 18 (5): 1-23.
Frenguelli, Joaquain. 1952. The Lower Gondwana in Argentina. The Palaeobota-
nist, 1: 183-188.
Halle, Thore G. 1911. On the geological structure and history of the Falkland
Islands. Bull. Geol. Inst. Univ. Uppsala, 11: 115-229.
-. 1927. Palaeozoic plants from central Shansi. Geol. Survey China,
Palaeontologia Sinica, ser. a, 2 (1): 1-316.
-. 1929. On the habit of Gigantopteris. Geol. Foren. Stockholm Forhandl.,
51: 236-242.
-. 1937. On the relation between the late Palaeozoic floras of eastern and
northern Asia. Deux. Congres. Strat. Carbonifere Heerlen, 1: 237-245.
H0eg, Ove A. 1930. A psilophyte in South Africa. Konelige Norske Videnskabers
Selskab Forhand., 3 (24): 92-94.
-. 1937. Plant fossils and paleogeographical problems. Deux. Congres
Strat. Carbonifere, Heerlen, 1: 291-311.

Telegram @tubscstds
434 STUDIES IN PALEOBOTANY

Holden, Ruth. 1917. On the anatomy of two Palaeozoic stems from India. Ann.
Bot., 31: 315-326.
Hoskins, John H., and Aureal T. Cross. 1951. The structure and classification of four
plants from the New Albany shale. Amer. Mid. Naturalist, 46: 684-716.
-. 1952. The petrifaction flora of the Devonian Mississippian Black Shale.
The Palaeobotanist, 1: 215-238.
Hsii, J. 1947. Plant fragments from Devonian beds in Central Yunnan, China. Joum.
Indian Bot. Soc., (Iyengar Commem. Vol., 1946): 339-360.
Jones, O. A. and N. J. de Jersey. 1947. The flora of the Ipswich coal measures-
morphology and floral succession. Univ. Queensland Papers, Dept. Geol., 3: 1-88.
Just, Theodor. 1952. Fossil floras of the southern hemisphere and their phytogeo-
graphical significance. Bull. Amer. Mus. Nat. Hist., 99: 189-203.
Kidston, Robert and William H. Lang. 1923. Notes on fossil plants from the Old
Red Sandstone of Scotland. I. Hicklingia edwardi K. and L. Trans. Roy. Soc.
Edinburgh, 53: 405-407.
Kolbe, R. W. 1957. Fresh-water diatoms from Atlantic deep-sea sediments. Science,
126: 1053-1056.
Krishnan, M. S. 1954. History of the Gondwana era in relation to the distribution
and development of flora. Seward Memorial lecture, Sahni Institute of Palaeobot.,
Lucknow, pp. 1-15.
Mendes, Josue C. 1952. The Gondwana formations of southern Brazil: some of their
stratigraphic problems, with emphasis on the fossil flora. The Palaeobotanist, 1:
335-345.
Neuburg, Maria F. 1948. Upper Paleozoic Flora of the Kuznetsk Basin. Paleontol¬
ogy of the USSR, Acad. Nauk USSR, vol. 12, pt. 3, fasc. 2, pp. 1-342. Moscow.
-. 1958. Present state of the question on the origin, stratigraphic signifi¬
cance and age of Paleozoic floras of Angaraland. (Rept. for 4th Internat. Congr.
Carbonif. Stratigr. and Geol., Heerlen, 1958), USSR Acad. Sci., Geol. Institute,
Moscow, pp. 1-27.
Plumstead, Edna P. 1958. The habit of growth of Glossopteridae. Trans. Geol. Soc.
South Africa, 61: 81-94.
Rao, H. S. 1940. On the anatomy of Lycopodiopsis derbyi Renault with remarks on
the southern Palaeozoic lycopods. Proc. Indian Acad. Sci., 11B: 197-217.
Read, Charles B. 1936. The flora of the New Albany shale, Part 1. Diichnia kentuck-
iensis, a new representative of the Calamopityeae. U. S. Geol. Survey Prof. Paper,
185-H: 149-155.
-and Guy Campbell. 1939. Preliminary account of the New Albany shale
flora. Amer. Mid. Naturalist, 21: 435-453.
Sahni, Birbal. 1953. (Posthumous work edited by T. M. Harris). Note on some pos¬
sible psilophyte remains from Spiti, north-west Himalayas. The Palaeobotanist,
2: 1-3.
Saksena, Shivdayal. 1954. Reconstruction of the vegetative branches of Phyllotheca
etheridgei Arber and P. sahni Saksena. The Palaeobotanist, 3: 51-53.
Scott, Dukinfield H., and Edward C. Jeffrey. 1914. On fossil plants, showing struc¬
ture, from the base of the Waverley shale of Kentucky. Phil Trans. Roy. Soc.
London, 205B: 315-373.
Seward, Albert C. 1931. Plant Life through the Ages. Cambridge Univ. press.
601 pp.
-, and John Walton. 1923. On a collection of fossil plants from the Falk¬
land Islands. Quart. Journ. Geol. Soc. London, 79: 313-332.

Telegram @tubscstds
SOME PALEOZOIC AND MESOZOIC FLORAS 435

Srivastava, P. N. 1954. Studies in the Glossopteris flora of India: 1. Some new fossil
plants from the Lower Gondwanas of the Raniganj coalfield, India. Th'e Palaeo-
botanist, 3: 70-78. (And several other contributions under this title in the same
journal.)
Surange, K. R., and Prem Singh. 1953. The female dwarf shoot of Walkomiella
indica—a conifer from the Lower Gondwana of India. The Palaeobotanist, 2: 5-8.
-, and P. N. Srivastava. 1953. Megaspores from the west Bokaro coalfield
(Lower Gondwanas) of Bihar. The Palaeobotanist, 2: 9-17.
-, P. N. Srivastava, and Prem Singh. 1953. Microfossil analysis of some
Lower Gondwana coal seams of west Bokaro, Bihar. Bull. Nat. Institi. Sci. India,
2: 111-127.
Teichert, Cart, and James M. Schopf. 1958. A Middle or Lower Devonian psilophyte
flora from central Arizona and its paleogeographic significance. Journ. Geol., 66:
208-217.
Termier, Henri, and Genevieve Termier. 1950. La Flore Eifelienne de Dechra Ait
Abdallah (Moroc central). Bull. Soc. Geol. France, ser. 5, 20: 197-224.
Thomas, H. Hamshaw. 1952. A Glossopteris with whorled leaves. The Palaeobota¬
nist, 1: 435-438.
Walkom, A. B. 1915-1917. Mesozoic floras of Queensland. Queensland Geol. Survey.
Pub. 252, pt. 1, pp. 1-50, 1915; Pub. 257, pt. 1 cont., pp. 1-66, 1917; Pub. 259, pt. 1
concl., pp. 1-48, 1917.
Walton, John, and Jessie A. R. Wilson. 1932. On the structure of Vertebraria. Proc.
Roy. Soc. Edinburgh, 52: 200-207.
Wegener, Alfred. 1922. The Origin of Continents and Oceans. E. P. Dutton and Co.,
New York, 3rd ed. 212 pp.

Telegram @tubscstds
 17
an introduction to
palynology
by Charles J. Felix

T he term “palynology” was introduced by Hyde and Wil¬

liams in 1944 to formally include all work with pollen and
spores. It encompasses a wide variety of investigations in pure
and applied botany, both modern and fossil. Palynology is a
research tool in such diverse fields as paleobotany, systematic
botany, plant geography, archaeology, geochronology, allergies,
melittopalynology, and petroleum exploration. Because of this
heterogeneity, comprehensive reference works are lacking and only
a few segments are adequately covered by textbooks. However,
palynology is unique in possessing a near complete bibliography,
which has been published regularly since 1927.
Although palynology is commonly regarded as a new botanical
science, it is more applicable to say that it is only now coming into
its own. Actually palynological history is associated with the
development of the microscope, and nearly 300 years ago both
Nehemiah Grew and Marcello Malpighi observed and described
pollen. There have been many botanists who have contributed to
our knowledge of plant microfossils, and Wodehouse gives an excel¬
lent historical review, which surveys pollen work from its earliest
days to the modern era. Carl Hugh Fischer is widely accepted as the
founder of pollen morphology from the modern viewpoint, although
only his late nineteenth century work dealt with pollen. Lennart
von Post presented the first modern percentage pollen analyses in
1916 and brought realization of the true potentialities of the
method. Erdtman presented von Post’s methods outside of the
Scandinavian countries in 1921. Erdtman’s subsequent contribu¬
tions have been numerous, and today he stands as a major figure
in world palynology.

Telegram @tubscstds
436
Fossil spore and pollen research has grown considerably in the
INTRODUCTION TO PALYNOLOGY 437

last decade. The journal Micropaleontology gave names and ad¬

dresses for 90 North American paly nolo gists in its 1958 directory.
Palynologie Bibliographie in 1959 listed the same data for 516
palynological personnel on a world-wide basis. There are certainly
several thousand scientific workers with direct or indirect interests
in palynology. Two international journals are devoted solely to
palynology, and numerous publications regularly present plant
microfossil articles. The first international palynological confer¬
ence met at Stockholm in 1950, and in 1954 palynology was repre¬
sented for the first time with its own section at the Eighth
International Congress of Botany in Paris. In Europe and America
palynological meetings with restricted membership now meet
regularly. The first American national pollen conference met in
1953 and has continued informally, with the fifth national confer¬
ence held in 1958.
lH is perhaps the successful utilization of plant microfossils in
stratigraphic work of petroleum geology that has given palynology
its greatest impetus. However, plant spores have been a proven
correlation tool in coal investigations for several decades both in
North America and Europe. Nearly every segment of the plant
kingdom has contributed the microscopic reproductive dissimules,
termed pollen or spores, to palynology. This comprises a number
of different sources, but only a small number of major groups
produce most of the plant microfossils, and palynology is largely
concerned with those of the vascular plants. Pollen is produced
by members of the angiosperms and gymnosperms, the former
commonly referred to as the flowering plants and the latter as non¬
flowering. Spores are contributed by a number of groups such as
the ferns, fern allies, bryophytes, fungi, and algae. The ferns are
by far the major source, with fern allies a minor contributor and
the remainder to a lesser degree.
A frequently overlooked phase of palynology is the role of the
pollen specialist in certain allergies. Many people are sensitive or
susceptible to pollen of certain plants, and such an allergy is popu¬
larly called hayfever. To induce hayfever the offending pollen
must be buoyant and thus easily transported; it must be toxic to
allergic people. Most hayfever is caused by wind-transported
pollen although all such plants do not cause hayfever. For illus¬
tration, the cat-tails and many of the conifers produce large
amounts of pollen, but with the exception of a few conifer species,
these do not cause hayfever. A few insect-pollinated plants, such as
the goldenrod, have pollen toxic to some, but they require intimate

Telegram @tubscstds
handling to produce an allergy and are regarded as minor hayfever
438 STUDIES IN PALEOBOTANY

causes. Among the better known such plants responsible for wide¬
spread hayfever are the grasses and ragweeds. Allergy pollen is of
such importance that most major cities now make atmospheric
pollen counts during hayfever seasons by exposing slides covered
with an adhesive, and many newspapers publish daily pollen
counts for the interest of hayfever sufferers. Little is known of
the role of fern spores in allergies, but it is suspected that they
may be a contributing factor.
There should be no objection to including melittology in palyn-
ology, since knowledge of pollen is recognized as an essential part
of honey investigations. The authenticity of honey sources can be
verified by examination of the included pollen of honey. Pollen is
indispensable as a food material for bees and all pollen does not
contain the same food value. Protein content varies from 7 to 11%
in pine, to 35% in date palm pollen, and average fat content ranges
from 1% in birch to 17.5% in black walnut. Bee-collected pollen is
known to be an excellent source of vitamins B and C, as well as
some vitamins D and E. However, vitamins A and K are not
known to be present in pollen. On occasions when there is a
scarcity of pollen in nature, the beekeeper finds a knowledge of
pollen to be valuable in preparation of pollen substitutes for bees.
Although the scope of palynology is botanical, the uniqueness of
floras through the ages has made it necessary to approach pollen
from a different aspect from spores. From the advent of pollen¬
bearing plants the problems of morphology, nomenclature, and
techniques in pollen research have been different from those of
spore studies. Cryptogam spores and gymnosperm pollen differ
from angiosperm pollen both in chemical composition and cell wall
morphology, and parallelism of structure is still ill-defined. Pre¬
sumed angiosperm pollen has been recorded from the Jurassic of
Scotland and Sweden; it became abundant throughout the world
in the lower Cretaceous and dominant in the upper Cretaceous.
Gymnospermous plants have an ancient lineage, with some of them
extending back possibly into the Devonian. Some primitive
gymnosperms have spores which show considerable organizational
advance beyond the cryptogamic microspores, yet lack some signi¬
ficant features of modern pollen. For these some have proposed to
use the term “prepollen”; however, the existence of pollen tubes
in any Paleozoic plants has never been demonstrated. There are
references to spores from the Cambrian, but the most authentic
published assignments of vascular plant spores places their earliest
known occurrence in the Silurian.

Telegram @tubscstds
INTRODUCTION TO PALYNOLOGY 439

Although it is a general practice to use the terms “pollen” and

“spore” as synonyms, literature abounds with discussions on the
distinction between spores and pollen and there are divergent
views on the homology of the two. In spore studies particularly,
there is disagreement on usage of the terms “microspore” and
“megaspore.” In the botanical sense, the microspore is a reproduc¬
tive body which germinates to form the male gametophyte, whereas
the megaspore germinates to form a female gametophyte. A pol¬
len grain can be said to be the germinated microspore of the seed
plants. The megaspore of the seed plants gives rise to the embryo
sac containing a reduced female gametophyte. In a strict sense
the microspore is determined solely by the presence of the male
gametophyte. In isosporous plants the spores develop directly
into both male and female gametophytes. Since isospores of homo-
sporous plants are very similar to microspores of heterosporous
plants in size and form, discrimination between them is difficult in
living plants and often impossible in fossils. It is true that micro¬
spores are frequently small, but all small spores are not micro¬
spores. In several instances the microspore is larger than the
megaspore, especially in the gymnosperms where in the majority
of instances the so-called microspore is the larger. Therefore the
term “microspore” refers to fundamentally functional differences
entirely aside from size distinctions, for it is possible to be unable
to differentiate in the fossil state as to whether a spore is iso-,
micro-, or megaspore. Among suggested terminology by palynolo-
gists is the term “miospore” for all fossil spores less than 200 /x in
diameter. The term “polospore” has been introduced into palyn-
ology to include pollen and/or spores. Also suggested is 200 /x as
the arbitrary size limit, with megaspores exceeding 200 ju, and
those below this size comprising microspores, isospores, and pre¬
pollen. Such new terminology or artificial limits are of question¬
able value since botanically spores and pollen are well defined on
functional characters. Additional terminology will not aid in
recognition in the isolated fossil state unless functions can be
inferentially established, and only detailed botanical research can
do this.

POLLEN MORPHOLOGY

The development of the pollen tube in the angio- and gymno¬

sperms provides the best means of differentiation between pollen

Telegram @tubscstds
440 STUDIES IN PALEOBOTANY

and spores, but a general descriptive terminology for pollen is still

lacking. Probably the most detailed system for describing pollen
is that of Erdtman, but its great number of morphological char¬
acteristics renders it more acceptable for the specialist, and the
beginner would do well to utilize a less detailed classification. In
general, the pollen grain consists of three concentric layers. The
inner layer, consisting of the living cell or protoplast, disappears
quickly if pollination is not achieved. The middle layer is the
intine, which is a cellulosic coat and easily destroyed. The third
and outer layer is the exine and is a typically complex structure
of extraordinary durableness. It is the exine which survives ages
of decay processes and lengthy chemical treatment that is rela¬
tively unchanged; it may be said that its remarkable durability is
the basis of palynology. Its resistance to destructive forces is due
to its unusual chemical composition. The exine of pollen, as well
as the walls of fungal spores, contains a highly polymerized, cyclic
alcohol termed “sporopollenin” by Frey Wyssling. It is related to
suberine and cutin but is more resistant than either. The quantity
of sporopollenin differs specifically, and resistance to decay
evidently changes in accordance with the quantity of the chemical.
The exine is as variable morphologically as in chemical composition.
It seems to be completely absent in aquatic species, and thus their
remains are seldom preserved. The exine is very simple in some
species, such as Larix, where it is a thin, homogeneous sheet.
Typically the exine is complex structurally, with at least two
layers (Fig. 17-1). An inner layer, the endexine, is a continuous,
homogeneous membrane. The outer layer, the ektexine, is com¬
posed of numerous small elements whose development and distri¬
bution produce great variability in structure. Two principal types

TECTATE INTECTATE

Fig. 17-1. Principal exine types.

Telegram @tubscstds
INTRODUCTION TO PALYNOLOGY 441

of wall structure are recognized because of ektexine variability. If

ektexine elements are free and isolated or form an open pattern in
such manner as not to place a continuous cover over the endexine,
the type is called intectate. If the ektexine forms a continuous
coat outside the endexine such as by a fusion of ektexinous ele¬
ments, it is labeled as tectate. These two types are not absolutely
distinct, and several transitional wall types exist. There are
numerous exceptions to the ektexine division, as in some species
(Zostera) both exine layers appear rudimentary or are absent, and
in Juncus a very thin exine may correspond to the endexine.
The casual use of the term “pollen morphology” by palynologists
obviously has a restricted meaning since it refers only to the
morphology of the exine. Most investigations are concerned with
fossil pollen that has lost intine and protoplasmic contents or
recent pollen in similar conditions due to acetylation or other
chemical means. The majority of diagnostic features are found in
the exine, and presence of other layers obscure these features.
Consequently, untreated grains are not suitable for morphological
analysis of the exine.

Apertures

Most pollen grains possess openings or thin areas of the exine

through which the pollen tube emerges at germination. Two general
types of such apertures exist, and the arrangement and number of
these on the grain produce a variety of pollen types. A few grains
possess no visible aperture, and the exine is continuous over the
entire surface (Populus). When apertures are present, the number
varies from one to as high as 100. The terminology of apertures
is still unsettled, but generally as purely morphologic definitions
they are designated as furrows and pores. However, the nomen¬
clature of Erdtman is more elaborate and creates some difficulty
in application. Furrows are somewhat boat-shaped depressions in
the exine, the ektexine being much reduced but with the endexine
less affected. The openings are also the harmonegathi of Wode-
house and exist as flexible parts of the exine to accommodate changes
in volume with varying water content. The typical exit is a rather
isodiametric pore. In some instances it is closed by an operculum,
and it is often surrounded by an annular area (annulus). Unlike
pores, furrows do not completely penetrate the exine, and con¬
sequently if pores are lacking, the pollen tube must force its growth
through the covering membrane of the exine at germination. Ar-

Telegram @tubscstds
442 STUDIES IN PALEOBOTANY

rangement of apertures varies, with one-furrowed grains frequent

in monocotyledonous (Fig. 17-3, 8) and gymnospermous plants.
Three furrows and/or pores are common in dicotyledonous plants.
They are often arranged equidistantly from each other and merid¬
ionally along the grain’s equator. If more than three openings are
present (Fig. 17-3, 6), they may be arranged as above or evenly
distributed over the entire grain surface. The number and distri¬
butional pattern of apertures are generally easily observed. They
provide valuable diagnostic characters, and several practical classi¬
fication systems have been devised on aperture types. The num¬
bers and disposition of apertures promise to be useful in classification,
as major taxonomic units are generally uniform in this respect, and
the use of aperture characters would afford a natural classification.

Size

Size is important since structural differences are sometimes in¬

adequate for distinguishing species, and size becomes a reliable
criterion. In Picea, for instance, such measurements have aided in
species identification. Pollen grains of angiosperms range from
about 5 to over 200 /x in diameter, although extreme sizes are rare
and many such records are open to question. Most grains seem to
fall in a range of 25 to 100 [x in living angiosperms. There are a
few authentic instances of pollen about 5 ju in greatest diameter;
the Boraginaceae, Piperaceae, Crypteroniaceae, and Cunoniaceae
are families having species in the lowest size ranges. There are
also recorded instances of grains in excess of 200 /x, with species in
the Dipsacaceae, Nyctaginaceae, and Oenotheraceae having occa¬
sional grains of such extraordinary dimensions. Cranwell has ob¬
served that most of her monocotyledonous pollen studied lies be¬
tween 15 and 80 jix, with very small grains being rare. The pollen of
Zostera, an aquatic monocot, is among the most unusual in the plant
world with regard to size. The grains resemble a membranous tube
and exceed 2500 jix in length, although only 3 to 4 /x wide. There often
appears to be a direct relation between grain size and number per
anther. The very large pollen of Mirabilis occurs 32 grains per
loculus, and the minute grains of Borago probably exceed 50,000.
It also appears that, as a rule, the largest grains are produced by
ephemeral flowers lasting only a day. Strict rules for size measure¬
ments are not standardized, but it should always be clear in
descriptions which layers are included, whether over-all dimensions
are given, or if processes are given separately.

Telegram @tubscstds
INTRODUCTION TO PALYNOLOGY 443

GYMNOSPERMS

As in the angiosperms, the pollen tube usually serves as a fertili¬

zation agent in the gymnosperms. However, in certain members
of the Cycadales it is simply a nutritive haustorium. A general
description of gymnosperm pollen is impossible since there is great
diversity among the orders. Among gymnosperms the pollen of
the Coniferales is perhaps most familiar and most frequently en¬
countered in sediments. The most conspicuous character is the
presence of air sacs or bladders in genera of the Abietineae (Fig.
17-3, 1, 2, 3, 5) and Podocarpineae, although both tribes contain
genera without bladders. When such appendages are present, the
Abietineae have two, whereas the Podocarpineae may have one to
six. The saccate grains in recent gymnosperms, as exemplified by
the genus Pinus, are generally bilateral with proximal distal sur¬
faces that are distinctly different (heteropolar). The aperture is
distal between the bladders and is rather poorly defined. It is
often referred to as a “sulcus,” but is seldom more than a thin area
merging into the adjacent exine.

SPORE MORPHOLOGY

The main types of spore configurations are bilateral and radial.

Spores are seldom completely spherical due to close proximity in
the tetrad. Depending upon tetrad arrangement, divergence from
a sphere is toward an elongate shape (Fig. 17-2, 2), produc¬
ing bilateral spores, or toward a tetrahedral shape with three
contact faces and radial symmetry (Fig. 17-2, 3-16). That part of
the spore nearest the tetrad’s center and in contact with adjacent
members is the “proximal surface” and the opposite free surface is
the “distal surface.” The safest and most universal distinction be¬
tween the two fundamental spore types is the tetrad scar. It is
usually well defined and readily discernible as “monolete” (Fig.
17-2, 2) or “trilete” (Fig. 17-2, 3-16). There are a few instances
in extant and fossil spores where specimens are considered to be
inaperturate (alete). Spore differentiation based on tetrad char¬
acters refers almost entirely to microspores or isospores. The
megaspore is generally identified as such upon size features or an
awareness of its botanical relationship. Even spore studies have
their specialists and there is an increasing trend toward specializa-

Telegram @tubscstds
444 STUDIES IN PALEOBOTANY

Telegram @tubscstds
INTRODUCTION TO PALYNOLOGY 445

tion in either microspores or megaspores. This is in part because

megaspores are not produced by all vascular plants, and true
microspores are present in vastly greater numbers than mega¬
spores. Microspores and homosporous spores are much more suit¬
able for the prevalent mass treatment methods than are the larger
megaspores, which are subject to damage in drilling.
The exine in spores is generally structureless. Some stratifica¬
tion may be visible, but the elaborate divisions seen in pollen walls
is lacking in spores. The exine does have, however, a variety of
sculpturing which aids in identification. Living pteridophytes
have a considerable size range with extremes of 15 to 125 p for
microspores and isospores. Harris found 67% of the New Zealand
fern spores to be in a medium size class of 25 to 50 p.

POLLEN AND SPORE DISSEMINATION

The method of pollination is important to the palynologist since

it provides a clue to the dispersal of pollen. This is valuable in
interpreting fossil floral assemblages, especially in oil geology
where environment plays a significant role in oil accumulation. To
attain correct interpretations, the investigator must be cognizant
of a number of facts: (1) The amount of pollen produced by speci¬
fic plants. (2) The chief mechanism of transport. (3) The possible
distance from its growth site that pollen is likely to be carried. In

Fig. 17-2. Schematic drawings of important spore and pollen classes.

1-12. Microspores.
1. Aletes. 2. Laevigatosporites. 3. Granulatisporites. 4. Punctatisporites. 5. Tri-
quitrites. 6. Densosporites. 7. Raistrickia. 8. Reinschospora. 9. Reticulatisporites.
10. Cirratriradites. 11. Lycospora. 12. Calamospora.
13-16. Prepollens.
13. Florinites. 14. Endosporites. 15. Alatisporites. 16. Illinites.
17-22. Gymnospermous pollen classes.
17. Inaperturate (also in 23-29, 30-57). 18. Monocolpate (also in 23-29). 19-20.
Polyplicate. 21-22. Vesiculate.
23-29. Monocotyledonous pollen.
23. Monocolpate. 24. Monoporate. 25-27. Operculate. 28-29. Trichotomocolpate.
30-57. Dicotyledonous pollen (polar and equatorial view).
30. Diporate. 31. Triporate. 32-33. Stephanoporate. 34-36. Periporate. 37. Dicol-
pate. 38. Tricolpate. 39-40. Stephanocolpate. 41-43. Pericolpate. 44. Dicolporate.
45. Tricolporate. 46-47. Stephanocolporate. 48-50. Pericolporate. 51. Syncolpate.
52. Heterocolpate. 53-54. Zonorate. 55. Dyad. 56. Tetrad. 57. Polyad. (From
Kuyl, Muller and Waterbolk, 1955.)

Telegram @tubscstds
446 STUDIES IN PALEOBOTANY

general, pollination mechanisms accountable for grain dispersal

may be grouped as: Anemophilous (wind); Entomophilous (insect);
Hydrophilous (water); and Ornithophilous (bird). In addition
there are reported instances of Chiropteriphily (bat) and Malaco-
phily (snail and slug pollination).
There are fundamental structural differences in pollen grains
which afford clues as to their mode of distribution. In general,
hydrophilous pollen grains lack an exine, whereas anemophilous
grains are small and smooth-walled. Greater variation is noted in
entomophilous pollen where the grains are generally adhesive, and
the ektexine bears prominent protuberances and pits of a wide
variety of forms. Most of the gymnosperms are widely distributed
by wind, and catkin-bearing angiosperms such as Quercus, Corylus,
and Betula lend themselves to wind dispersal of pollen. Some
extraordinary distances of wind transport are recorded, with Pinus
pollen recovered in the Arctic some 60 miles from the nearest conif¬
erous forest, and Erdtman has cited examples of Picea and Pinus
pollen being wind transported for a distance of over 600 miles.
One of the most precise figures is Faegri and Iversen’s of 30 to 60
miles as the natural limits of pollen dispersal. In all probability
the extreme distance carried is much overrated, and movement
from a source area by wind of more than 100 miles is an extreme
case. Nevertheless, the researcher must remain aware of long dis¬
tance pollen movement, especially when the question of species
migration arises. It will be difficult to ascertain whether a low
percentage of a pollen species is due to a small local occurrence or
to long distance transport. It must be realized that with increas¬
ing distance from the forest site that the importance of long dis¬
tance species increases. The spores of ferns and mosses are dis¬
seminated by wind and water almost exclusively, but there are few
data available on comparative distances of transport.
Anemophilous species contribute vast amounts of the pollen
deposited, and the quantity of pollen produced by some plants is
enormous. Such pollination is a very inefficient process since vast
numbers of pollen grains are involved, although an infinitesimal
proportion complete their role in the life cycle. It has been esti¬
mated that a single anther of Canabis sativa produces 70,000
grains. Pollen productivity is usually highest among wind-polli¬
nated species. For instance in Linum catharticum, which is
entomophilous, the number of grains for anther is only about 100.
Certainly, the forest trees, which contribute so greatly to pollen
spectra, produce huge quantities of pollen. Dissemination from a

Telegram @tubscstds
INTRODUCTION TO PALYNOLOGY 447

mature Pinus, Picea, or Quercus may amount to hundreds of mil¬

lions of grains yearly.
Because of differences in habitat and plant habit, few Filicales
may be expected to have as wide spore dispersal as anemophilous
tree pollen in post-Paleozoic deposits. Spore production is varied
in extant ferns and often abundant, but it does not approximate
the prodigious numbers of the pollen producers. Production
ranges from thousands of spores in each Eusporangiate sporangium
to the single spore of the Hydropteridea megasporangium. Ophio-
glossum with about 15,000 spores per sporangium is probably most
prolific, and some mature Dryopteris plants have been estimated
to produce a half million spores in a single season. The Filicales
are certainly productive and ubiquitous enough to warrant atten¬
tion in statistical investigations. There is apparently no relation¬
ship in the ferns between spore size and number per sporangium.
The quillwort, Isoetes, is another prolific spore producer; it has
been estimated that as many as one million microspores are borne
in a sporangium.

POLLEN IDENTIFICATION

A basic prerequisite in palynology is correct identification of the

microfossils. This is a considerable task in pollen work in view of
the vast number of pollen producers, and the evaluation of palyn-
ological data must be cautious. Although it may be said that
pollen grains in related genera are usually more or less of the same
type, striking differences exist. In the Caryophyllaceae the grains
for the most part have many pores (cribellate), but Spergula,
Pteranthus, and a few other genera have tricolpate grains. There
are also numerous instances of pollen grains of similar types occur¬
ring in plants not at all related, such as in the genera Salix (Salicaceae)
and Adoxa (Adoxaceae). Most genera do show a marked consist¬
ency of pollen form, but there are exceptions where more than one
pollen type will occur in certain genera. Tulipa of the Liliaceae, a
genus with constant floral characteristics and seemingly natural,
bears monocolpate pollen in some species and tricolpate pollen in
others. It is unique in being the only monocot known to have the
latter type of grain. The genus Crocus has pollen either nonaper-
turate or with several parallel, ringlike furrows. Some other genera
with comparable differences are Symplocos, Anemone, and Morina.
A step further is the production of more than one pollen type by

Telegram @tubscstds
448 STUDIES IN PALEOBOTANY

a single species. The dimorphic and heterostylous flowers of

Primula veris have a decided difference in pollen from long and
short styled flowers, the grains of the latter being smaller in size
along with differences in shape and in aperture numbers. In¬
stances of pollen dimorphism are also known in the Plumbagin-
aceae and Rubiaceae. A classic example of heterostyly is Lythrum
salicaria of the Lythraceae. Three kinds of flowers, based on style
and stamen length, are present in the species. Not only is there a
size difference in pollen from each but the longest stamens produce
green pollen, the other two lengths having yellow pollen. It can
be said that the generic identification of pollen is usually possible,
but specific identification with any degree of certainty is less easily
accomplished. However, there are many families in which char¬
acteristic features may be sufficiently constant for diagnosis. The
Cyperaceae have tetrahedral, psilate grains with a prominent basal
pore and three slitlike lateral pores as the common aperture type.
The Betulaceae have characteristic thickened streaks of exine,
known as arci, which form sweeping curves from pore to pore.
Compound grains due to failure of the spore tetrads to separate
are found in families and genera of both the monocots and dicots.
The Ericaceae are notable because the pollen remains permanently
in tetrads, a peculiarity shared with the Empetraceae, Typhaceae,
(Fig. 17-3, 11) and Juncaceae among others. The permanent cohe¬
sion of pollen into groups larger than tetrads (polyads) is con¬
spicuous in the Mimosoideae (Fig. 17-3, 9) of the Leguminosae.
Even greater variation is present among the Orchidaceae where
isolated grains, tetrads, and polyads occur in different genera.
Most of the notable anomalies occur in the dicotyledonous
plants. Monocotyledonous pollen is generally recognizable, but
the narrow range of diagnostic characters often render generic and
family differentiation difficult. There are two basic monocot grain
types: (1) essentially spheroidal without a recognizable aperture
and (2) with bilateral symmetry and characterized by a single fur¬
row (Fig. 17-3, 8) or pore (Fig. 17-3, 13). There are a few variants
such as the many-pored Alisma and Hypoxis with two apertures.

COLLECTION AND STUDY TECHNIQUES

Reference Collection

Despite the increase in number of publications in recent years,

Telegram @tubscstds
studies of Tertiary and Quaternary pollen suffer from a lack of
 INTRODUCTION TO PALYNOLOGY 449

Fig. 17-3. Representative extant gymnosperm and angiosperm pollen.

1-5. Gymnosperm pollen.
1. Abies grandis. 2. Picea sitchensis. 3. Cedrus atlantica. Tsuga heterophylla.
5. Pinus strobus.
6-13. Angiosperm pollen.
6. Hibiscus althea. 7. Ochroma limonensis. 8. Iris kaempferi. 9. Pithecolobium
oblongum. 10. Fremontia californica. 11. Typha latifolia. 12. Cissus ampelopsis.

Telegram @tubscstds
13. Hordeum vulgare.
450 STUDIES IN PALEOBOTANY

published illustrations for use in identifications. The necessity for

determining botanic affinities of unknown entities has led to the
establishment of reference collections of modern pollen and spores.
This is one of the most valuable tools available, for regardless of
the quality of illustrations, no drawing or photograph can show
satisfactorily the minute diagnostic features that can be discerned
in reference material. Illustrations at best can only convey sug¬
gestions as to the identification of pollen grains and spores and are
not intended to replace pollen and spore preparations. Since the
majority of Tertiary and Quaternary plants belong to extant
genera or families, a carefully planned reference collection is indis¬
pensable. In enabling correct assignment of pollen and spores, it
can help prevent creation of artificial genera for living ones. It is
almost essential that the investigator make accurate botanical
identifications if accurate interpretations are to be made of flora
evolution or ancient climates, and the reference slides can provide
the necessary comparative material.
There are over 200,000 species of living vascular plants. There¬
fore, the basic reference collection must be prepared with the
specific problem in mind. Obviously a near complete collection
would be a vast undertaking, and the collection of the Palynolog-
ical Laboratory of the Swedish Natural Science Research Council
in Solna, Sweden, is probably the largest in existence today. It
comprises about 25,000 slides representing more than 20,000
species of plants. The most valuable collection is one in which
specimens are referable to a specific sheet in an established herbar¬
ium. This would provide the most accurately identified material,
and in the event of discrepancy, the exact specimen could be
rechecked. A serious objection to this source of collecting is the
danger of mutilating valuable taxonomic material, but conscien¬
tious collecting can alleviate this. It is quite possible that, as
palynology grows and demands for reference specimens increase,
herbaria sources for standards may be exhausted, and the palyno-
logical laboratories may conceivably have to build herbaria or
contribute to existing ones.
By examination of floral material with a hand lens the capable
botanist can usually determine whether pollen is present. Flower
buds collected just before anthesis are the most reliable since little
of the pollen will be shed. There are objections to this on the
premise that the pollen may be immature, but the possibility is so
remote as to be negligible. The fertility of gymnosperm cones and
cryptogam sporangia can also be determined with a lens. Fern

Telegram @tubscstds
 INTRODUCTION TO PALYNOLOGY 451

Fig. 17-4. Representative Paleozoic spore genera.

1. Reinschospora. 2. Convolutispora. 3. Raistrickia. 4. Proprisporites. 5. Knoxi-
sporites. 6. Grandispora. 7. Densosporites. 8. Dictyotriletes. 9. Rotaspora. 10. Tri¬
partite s. 11. Mooreisporites. 12. Schulzospora. 13. Calamospora. 14. Laevigatos-
porites. 15. Punctatisporites. 16. Callisporites.

Telegram @tubscstds
452 STUDIES IN PALEOBOTANY

spores may create a somewhat different situation than pollen since

several palynologists have experienced difficulty in obtaining satis¬
factory spore preparations from herbarium sheets. Evidently
shedding is so efficient in the Filicales as to disperse most mature
spores and, unlike pollen grains, fern spores appear to undergo a
final developmental stage of external morphology prior to normal
shedding and mature spores are often rare. Specimens collected
are best stored in small vials or envelopes and all herbaria data
should be recorded.
Various published preparation methods exist, but most are basi¬
cally the acetylation procedure of Erdtman, frequently with modi¬
fications. It involves acetylation of the carbohydrate fraction of
plant microfossils, leaving only the spore or pollen exine with its
diagnostic features. This method involves treatment with nine
parts acetic anhydride to one part concentrated sulfuric acid. The
general practice is to divide the sample, submitting one fraction to
a bleaching procedure and staining the unbleached fraction. This
provides specimens useful for viewing wall sculpture when stained,
and permits examination of internal wall structures in bleached
specimens. The acetylation method has its chief disadvantage in
that pollen grains are overexpanded. Numerous investigations
have confirmed that recent and fossil grains expand to different
degrees, and therefore acetylation is not reliable in statistical
studies of size in recent pollen or in comparison of recent and fossil
specimens. It is advisable to use untreated specimens when size
determination is critical, even though permanent preparations may
not be retained.
Proper mounting media are of inestimable importance, for many
early preparations have deteriorated to uselessness due to inade¬
quate mountants. Unfortunate, too, has been the practice of
lactic acid mounts of pollen from herbarium sheets. This method
results in such expansion of grains as to nullify their usefulness in
morphological studies, and its clearing action is so rigorous that
grains are virtually destroyed. Two main schools of thought exist
on media. Absolute permanency in a medium is probably favored
by most Paleozoic spore investigators. However, a viscous medium
is considered desirable by many post-Paleozoic pollen workers. A
viscous preservative permits orienting the grains for examination,
and pollen is not generally subjected to the deformation by move¬
ment that occurs in moving the many winged and flanged forms of
the Paleozoic. Factors to be considered in selecting a medium

Telegram @tubscstds
are: (1) refractive index; (2) viscosity; (3) permanency; and
(4) crystalization.
 INTRODUCTION TO PALYNOLOGY 453

Among better known palynological mounting media are Canada

balsam, diaphane, and glycerin jelly. The first two require dehy¬
dration of the specimens before mounting. Glycerin is water-
miscible and has its supporters among those desiring a viscous
medium. A number of synthetic resins are marketed, but for
proven durability balsam and diaphane remain the most reliable.
A quick drying plastic medium used by many palynologists is
Clearcol, and it has a good history of permanency. It is advisable
to seal cover slips, especially glycerin mounts, to reduce the pos¬
sibility of drying. Even though there are records of unsealed slides
several decades old, the addition of a ringing cement is good insur¬
ance. If slides are permitted to dry in an inverted position,
specimens settle on or near the cover glass and greatly facilitate
microscopic examination.

Systematics

As with many phases of paleontology, nomenclatural problems

are an important aspect of palynology. In Paleozoic investiga¬
tions most spores have been treated as species of organ genera
independent of any natural phylogenetic system. Their true
botanic relationships can be determined only by associating them
with fructifications of known affinity. Thus artificial systems with
genera defined by arbitrary morphological features suffice quite
well until true relationships of spores are discovered, and species
can then be transferred to the appropriate natural genera. Some
Paleozoic students have completely abandoned normal botanic
practices and use systems of letters and numbers in classification.
However, for eventual intelligent comparison of results and the
incorporation of spores into a natural classification system, stand¬
ard nomenclatural practices are most desirable.
In the Quaternary, spores and pollen are largely assignable to
living genera, and modern nomenclature, often specific epithets,
can be used. Tertiary palynological taxonomy has problems addi¬
tional to those of the Paleozoic and Quaternary, and there is
always the possibility that new artificial genera will be created
without knowledge of extant genera. It is also possible that
Tertiary pollen assigned to a living genus may instead belong to a
related, extinct genus with near indistinguishable pollen. There
is a wide disagreement among palynologists regarding taxonomy,
and this has resulted in several nomenclatural systems. Some are
compromises between the use of artificial systems of letters and

Telegram @tubscstds
symbols, and normal paleontologic taxonomic systems.
454 STUDIES IN PALEOBOTANY

Among constructive developments have been the use of artificial

organ genera where the microfossil’s relationship is questionable,
but showing the botanic relationship wherever it is known. Also
there is a growing insistence that new taxonomic units be amply
described, well illustrated, and that a holotype be designated.

Microscopy

Good optical equipment is most essential in palynology. Plant

microfossils are not only so small as to require high magnifications,
but the time that must be devoted to microscopy entails consider¬
able eyestrain. A binocular microscope is recommended rather
than a monocular type to ease eye fatigue and also because stero-
scopic relations of grain characters are more easily viewed. Work¬
ing magnifications of about 100 to 1200 are generally considered most
desirable. The more difficult specimens require about 1000 X lens
with a very high numerical aperture. Apochromatic and fluorite
lenses have a wide preference among palynologists, and good
synthetic fluospar objectives are now available. All of these lenses
should be used in conjunction with wide-field compensating eye¬
pieces; and very necessary for advanced pollen-analytic research is
an immersion objective of high quality.
Electron optics is only beginning to be introduced into pollen mor¬
phology. Pollen is not the optimum subject for such study, but elec¬
tron microscopy has already provided important information on the
fine structure of exines. The high magnifications of the electron
microscope promise to shed new light upon structural differences
between wall layers that presently lie beyond the utmost powers
of the light microscope. Phase-contrast microscopy also has its
adherents, but it has wrought little material change in palynology
investigations. Pollen is not well suited for phase-contrast obser¬
vations, and little has been accomplished by these techniques
beyond affording clearer resolution of some wall features.

Sampling Methods

The collection of the sample is one of the most important aspects

of palynology, for the greatest care must be taken to prevent the
introduction of foreign matter. Contamination is a constant
source of error in palynology from the moment of collection of the
specimen until the final slide preparation. Sources of impurity are
so numerous that cleanliness and vigilance must be considered in-

Telegram @tubscstds
INTRODUCTION TO PALYNOLOGY 455

dispensable. The very atmosphere at the collecting site and in the

laboratory are hazards. Improper collecting techniques, soiled
laboratory equipment, and carelessness all contribute to the danger
of contamination. Cleanliness in the palynological laboratory must
be equal to that practiced in bacteriology if absolute confidence in
final analysis is realized.
Both surface and subsurface material is used. The quantity
varies with problems and investigators, but about 20 grams appears
to be a reliable amount. In collecting surface samples, column sam¬
ples from mine and outcrop exposures can be equally good, if uniform
segments are taken. In coal sampling the seam is cleared of debris
with hammer and divided into benches. Partings and bands often
provide natural benches. A continuous sample, providing com¬
plete vertical representation, is made by sampling perpendicular to
the bedding plane. Tools should be cleaned carefully after each
collection, and the outcrop is collected from bottom to top to
minimize contamination from sifting debris.
Subsurface samples are obtained most accurately from well
cores, but the majority of wells drilled employ the rotary drilling
technique in which a hollow drill stem terminated by a drill bit is
rotated into the subsurface. A stream of drill mud, circulated
down the hollow stem and back up the hole to a pit, facilitates
drilling by lubricating the drill bit and by removing rock chips
from the hole. The chips are collected at regular intervals and are
the most common type of sample. They are also least satisfactory
because of contamination due to caving from the hole walls. The
lack of care taken by well crews in obtaining the cuttings is often
a principal source of error. However, when properly collected,
rotary cuttings have proved valuable in correlation studies. De¬
spite these disadvantages, well cuttings must be utilized since most
wells are drilled by this method, and pollen and spores are espe¬
cially valued as most megafossils are severely damaged in drilling.
Most dependable, but costly, are “standard cores” obtained by
receiving a cored section in a hollow core barrel attached to the
lower end of the drill stem. Other types of core are “wire line
core” and “side wall core.” Although expensive, core samples are
most reliable in the establishment of standards.

Sample Preparation

Preparation techniques vary in palynology since fossils differ in

size, chemical composition, and in lithologies, and preparation meth-

Telegram @tubscstds
456 STUDIES IN PALEOBOTANY

ods are about as numerous as rock types. Paleozoic spore studies

have necessitated techniques different from those in pollen research.
Particularly in coal, where spore preservation is generally excellent,
treatment is quite specialized. A well-known preparation is the
maceration method first described by Franz Schulze in 1855 when
he treated coal with potassium chlorate and nitric acid to render
microscopic structure clearer. Consisting of partial oxidation of
coal and dispersal of the humic matter, the Schulze’s method is
still widely used today with various modifications. This solution
oxidizes and partially dissolves the humic matrix. After being
washed free of acid, the residue is treated with a basic solution
such as potassium hydroxide or ammonium hydroxide until all
humic material is thoroughly dispersed. No fixed procedure can
be recommended for various coals. Weathered coal generally
macerates more readily than fresh coal and the procedure best
suited for each type of sample must be determined by trial.
Hydrochloric acid is the most common reagent for dissolution of
carbonates, whereas hydrofluoric acid eliminates silicates. After
maceration of the sedimentary matrix, complete separation is often
attained by differential flotation with heavy liquids, with the
organic remains floating and the mineral fraction sinking. These
procedures have many variations, depending upon the matrix. A
new development in sample processing has been the use of the
ultrasonic generator. Ultrasonics are useful in the disaggregation
of some rock matrices. However, the principal utilization has been
the removal of debris and foreign particles from the plant micro¬
fossils following the initial breakdown by chemical methods.

APPLICATION OF PALYNOLOGY

So vast is the scope of botany and the varied utilization of data

that palynology has developed its specialized divisions in regard to
time. Pleistocene palynology has encompassed extensive study of
post-glacial vegetation, climatology, and paleoecology. Paleozoic
research has concerned itself largely with the Pennsylvanian coals
and their correlation. Mesozoic palynology has devoted most
attention to the Cretaceous and early pollen floras, but the Juras¬
sic spore assemblages are beginning to receive attention. The
Tertiary includes such diverse flora and climatic changes that
several specialized divisions have developed.

Telegram @tubscstds
INTRODUCTION TO PALYNOLOGY 457

Every age therefore has its specifically different goals which are
necessitated by the changing spectra throughout the geologic
column due to plant evolution and the migration of vegetational
units. The wind-transported entities are of interest to the geolo¬
gist as they are not, as a rule, restricted to particular facies. In
their dispersal by wind the pollen and spores are so well mixed as
to settle in a depositional basin and leave a fairly representative
picture of the regional vegetation. This picture may be distorted
in several ways: (1) The microfossils are differentially resistant to
decay not only in fossilization but also in withstanding chemical
preparation treatment; (2) Local overrepresentation must be con¬
sidered since an exceptionally abundant pollen producer in the
proximity of the depositional area can distort the interpretation of
a region’s vegetational history; (3) Long distance transport is
usually a minor factor, but it must be considered, as minor acces¬
sory pollen is often important and the possibility of its transport
can affect a spectrum.
Plant microfossils are more likely to be found in unweathered
sediments of reducing environments. These consist of marine,
brackish and fresh-water shales, limestones, bituminous coals, lig¬
nite, and peat. A high organic content, usually characterized by a
dark color, is often an indication of plant microfossils being pres¬
ent. Sandstones are generally barren since the coarse grain size
permits entrance of oxygen and ground water to a destructive
extent. Pollen and spores are independent of basin environments,
which sets palynology apart from traditional paleontology. Most
microfossils used for stratigraphic work (foraminifera, ostracodes,
fusulinids) are products of their depositional basins and thus are
controlled by ecologic conditions of the environment. Plant micro¬
fossils occur in marine, brackish, and continental sediments. The
varied environments of plants, and the unique pollen transport
mechanisms, serve to release pollen from the usual environmental
restrictions of most organisms.
The defining of ancient shore lines bears special significance in
oil accumulation, and palynology has proven a valuable aid to the
geologist in determining strand lines. As the pollen rain is pro¬
gressively less in the seaward direction, sedimentary environments
with pollen assemblages are limited to near shore marine and
lacustrine waters. In modern marine sediments pollen and spores
are frequently associated with diatoms, microforaminifera, dino-
flagellates, hystrichospherids, and various oceanic plankton. It is

Telegram @tubscstds
458 STUDIES IN PALEOBOTANY

possible to determine distance and direction of ancient shore lines

by the kinds and quantities of microfossils, as pollen and spores
will decrease in density in a seaward direction, with a correspond¬
ing increase in marine forms.
It is in the Paleozoic that palynology has been utilized most
effectively in correlation work. Especially in the Pennsylvanian,
spore analyses of coal seams are an integral part of coal correla¬
tions, and coal beds may often be identified on the basis of relative
abundance of certain spore genera. Figure 17-5 illustrates the
remarkable uniformity of the microflora composition of the Indiana
Coal V over a regional area. Such a seam traced laterally usually
remains similar over a coal field, whereas an adjacent seam usually
has a different histogram due to the different proportions between
the constituent spore genera.
In Quaternary pollen analysis correlations are based on changes
in relative abundance of carefully selected species or groups of
pollen grains. These are mainly pollen grains of forest trees.
Shifts in the relative abundance of these forms reflect important
vegetational changes due to altered climatic conditions. Plant
microfossils have proven of value in determining changes in climate
associated with changes in the extent of the Pleistocene ice. Most
plant genera of the Mesozoic and Paleozoic are extinct. Thus
ecological evaluation of such assemblages is dependent on correla¬
tive evidence from associated fossils and the physical character of
the deposit.
Although much of the work that has been done in palynology to
date has been of a practical nature, in connection with stratigraphic
problems, fossil spores and pollen give promise of adding greatly
to our understanding of the past distribution of floras. In all
probability floristic studies of the future will combine the evidence
from microfossil and macrofossil remains, thus adding to the accu¬
racy of our knowledge of the floras. Pollen will be particularly
helpful in cases where they offer distinctive diagnostic characters
and with plants that are not often preserved in macrofossil form.
The Gnetales, a unique assemblage of gymnospermous plants,
offers a case in point. Three living genera, the interrelationships
of which are not clear, are included in the group: Gnetum, with 30
species, is found in moist tropical regions; Welwitschia is a mono-
typic genus found in desert areas of southwest Africa; Ephedra
includes about 42 species that are widely distributed through arid
tropical and temperate regions. None of the habitats is favorable
for deposition.

Telegram @tubscstds
 lN30d3d J.N30d3d lN30d3d
o o o o o o o
10 in CO eg rM O

n ii I i i

(From
S31IN0SHM
S3iiyim0iyi
vdOdsoznnHOS
S31lddOHOS
S3iiyodsu\nnoii3d

Fig. 17-5. Histograms showing relative abundance of spore genera in eight samples of Indiana Coal V.
vyodsoHoswi3y
vixomisivy
s3iiyodsnviDNnd
VdOdSOOAT
S31ld0dS01V0IA3\n

S31INIT1I
z -on aavaNVis J

S3iiyodsn\nnNvyo
S31INIdOU
s3iiyodsoaN3
LlI S3imodsosN3a
LU
5 S3iiavyiyivyyio
ZD vyodsowvnvo
<
2 vyodsoiavo
1111111 1 I I I I I I L.l 1 I I I I S3iiyodsu\nv
L

I I I I I I I S31IN0SHM
S3iiyim0iyi
vyodsonnHOs
S31ldd0H0S
s3iiyodsn\nnoii3y
Vd0dS0H0SNI3d
VIMOId-LSIVd
S3ildOdSllV10Nnd
VdOdSOOAl
S31ld0dS01VDIA3\n
S31INITII

S3iidOdsn\nnNVdo
S31INId03d
Guennel, 1952.)

S31ldOdSOaN3
S31ldOdSOSN3a
hsvsvm

S3iiavdidivyyi3
vdodsowvnvo
vyodsoiavo
S3iid0dsu\nv
o
CO
J.N30d3d ±N30d3d lN30d3d ±N30d3d

Telegram @tubscstds 459

460 STUDIES IN PALEOBOTANY

The macrofossil record of the Gnetales is extremely scanty.

However, the pollen grains are readily recognized as Ephedra and
Welwitschia have distinctive longitudinal ridges; the former is
polycolpate and Welwitschia monocolpate. Gnetum does not
possess any visible germinal opening. Pollen attributed to Ephedra
has been found at early Tertiary horizons in Tasmania, Victoria,
and South Australia, in the Eocene Green River shales of Colorado
and in the Paleocene of Hanover, Germany, and at Cretaceous
horizons in Iraq, Venezuela, and Long Island in New York. Wel¬
witschia pollen has been reported from Russia and Australia, as
well as the United States. The most ancient records to date for
both are from the Middle Permian of Oklahoma; L. R. Wilson has
reported pollen (Ephedripites sp. and Vittatina sp.) which is
believed to be that of early ancestral forms of Ephedra and
Welwitschia.

REFERENCES

Arnold, Chester A. 1950. Megaspores from the Michigan Coal Basin. Michigan
Univ., Contr. Mus. Paleontology, vol. 8, No. 5, pp. 59-111.
Bhardwaj, D. C. 1957a. The palynological investigations of the Saar coals. Palaeon-
tographica, vol. 101, pt. B, pp. 73-125.
-. 19576. The spore flora of Velener Schichten (Lower Westphalian D) in
the Ruhr coal measures. Palaeontographica, vol. 102, pt. B, pp. 110-138.
Cookson, Isabel C. 1956. Pollen grains of the Ephedra type in Australian Tertiary
deposits. Nature, vol. 177, pp. 47-48.
Couper, R. A. 1958. British Mesozoic microspores and pollen grains. A systematic
and stratigraphic study. Palaeontographica, vol. 103, pt. B, pp. 75-179.
Cranwell, Lucy M. 1953. New Zealand Pollen Studies. The monocotyledons. Bull.
Auckland Institute and Museum, No. 3, 91 pp.
Dijkstra, S. J. 1946. Eine monographische Bearbitung der karbonischer Megasporen.
Meded. Geol. Stichting., Ser. C-III-1. 101 pp.
-. 1956. Lower Carboniferous Megaspores. Meded. Geol. Stichting., n. s.
No. 10, pp. 5-18.
Erdtman, Gunnar. 1952. Pollen Morphology and Plant Taxonomy (An Introduction
to Palynology. I). The Chronica Botanica Co., Waltham. 539 pp.
-. 1954. An Introduction to Pollen Analysis. The Chronica Botanica Co.,
Waltham. 239 pp.
-. 1957. Pollen and Spore Morphology/Plant Taxonomy {An Introduction
to Palynology. II). The Ronald Press Co., New York. 151 pp.
Faegri, Knut. 1956. Recent trends in palynology. Bot. Rev., vol. 22, No. 9, pp.
639-664.
-, and Iversen, J. 1950. Textbook of Modern Pollen Analysis. E. Munks-
gaard, Copenhagen. 168 pp.

Telegram @tubscstds
INTRODUCTION TO PALYNOLOGY 461

Guennel, G. K. 1952. Fossil spores of the Alleghenian coals in Indiana. Indiana

Geol. Surv. Rpt. of Progress, No. 4. 40 pp.
Harris, William F. 1955. A Manual of the spores of New Zealand Pteridophyta.
New Zealand Dept. Sci. and Indust. Research, Bull., No. 116, 186 pp.
Hyde, H. A., and Adams, K. F. 1958. An Atlas of Airborne Pollen Grains. Mac¬
Millan and Co. Ltd., London. 112 pp.
Ikuse, M. 1956. Pollen Grains of Japan. Hirokawa Publ. Co., Tokyo. 303 pp.
Kosanke, Robert M. 1950. Pennsylvanian spores of Illinois and their use in correla¬
tion. III. Geol. Surv. Bull., No. 74. 128 pp.
Kuyl, 0. S., Mueller, J., and Waterbolk, H. T. 1955. The application of palynology to
oil geology with reference to western Venezuela. Geologie en Mijnbouw, No. 3, n. s.,
vol. 17, pp. 49-75.
Leopold, E. B., and Scott, R. A. 1958. Pollen and spores and their use in geology.
Smithsonian Institution Rpt. for 1957. Pp. 303-323.
Morgan, J. L. 1955. Spores of McAlester coal. Oklahoma Geol. Surv., Circ. 36.
52 pp.
Ogden, E. C. 1957. Survey of airborne pollen and fungus spores of New York State.
New York State Mus. and Sci. Ser. Bull., 356. 62 pp.
-, and Lewis, D. M. 1960. Airborne pollen and fungus spores of New York
State. New York State Mus. and Sci. Ser. Bull., 378. 104 pp.
Potonie, Robert. 1956. Synopsis der Gattungen der Sporae dispersae. Beihefte z.
Geol. Jahrbuch, vol. 23. 103 pp.
-. 1958. Synopsis der Gattungen der Sporae dispersae. Beihefte z. Geol.
Jahrbuch, vol. 31. 114 pp.
-and Kremp, G. 1954. Die Gattungen der palaozoischen Sporae dispersae
und ihre Stratigraphie. Geol. Jahrbuch, vol. 69, pp. 111-194.
Schopf, James M., Wilson, Leonard R., and Bentall, R. 1944. An annotated synopsis
of Paleozoic fossil spores and the definition of generic groups. III. Geol. Surv. Rpt.
Inv., No. 91. 72 pp.
Selling, Olaf H. 1946. Studies in Hawaiian pollen statistics. I. Bishop Mus. Spec.
Publ., 37. 87 pp.
-. 1947. Studies in Hawaiian pollen statistics. II. Bishop Mus. Spec. Publ.,
38. 430 pp.
_. 1948. Studies in Hawaiian pollen statistics. III. Bishop Mus. Spec. Publ.,
39. 154 pp.
Traverse, Alfred. 1955. Pollen analysis of the Brandon lignite of Vermont. Bur.
Mines Rpt. Inv., 5151. 107 pp.
Van Der Hammen, Thomas. 1956. A palynological systematic nomenclature. Boletin
Geol., vol. 4, No. 2-3, Bogota, Colombia. Pp. 63-101.
Wenner, C. G. 1947. Pollen diagrams from Labrador. A contribution to the quater¬
nary geology of Newfoundland-Labrador, with comparisons between North America
and Europe. Geografiska Annaler. 241 pp.
Wilson, Leonard R. 1959. Geological History of the Gnetales. Oklahoma Geol.
Notes, vol. 19, No. 2, pp. 35-40.
Wodehouse, Roger P. 1933. Tertiary Pollen II. The oil shales of the Eocene Green
River formation. Bull. Torrey Bot. Club, vol. 60, pp. 479-524.
__ 1935. Pollen Grains. Their Structure, Identification and Significance in
Science and Medicine. McGraw Hill Book Co., New York. 574 pp.

Telegram @tubscstds
462 STUDIES IN PALEOBOTANY

Periodicals Devoted to Palynology

Catalog of Fossil Spores and Pollen. College of Mineral Industries, The Pennsylvania
State University.
Grana Paly nologica: An International Journal of Palynology. Almqvist and Wiksell,
Stockholm.
Palynologie Bibliographic. Service d’information Geologique du Bureau de Recher-
ches Geologiques, Geophysiques et Minieres, Paris.
Pollen et Spores. Museum National d’Histoire Naturelle, Laboratoire de Palynologie,
Paris.

Telegram @tubscstds
 18
techniques for
studying fossil plants;
some basic references

T he increase in knowledge that may be made in any field

of science is in large part limited by the tools that are
available to work with; the tools need not always be elaborate but
they must be adequate. Great advances have been made in paleo¬
botany in recent decades partly because new ways and means have
been found to extract information from fossil plants; certain older
methods have been improved and entirely new ones have been
developed.
This chapter is not, however, intended as a laboratory manual
nor as a survey of all the techniques that are in use at present.
Some of the basic methods are outlined, but the student is strongly
urged to consult the original accounts or better still visit labora¬
tories in which various types of paleobotanical research are being
conducted. Even though space were available and the writer’s
knowledge adequate to the task it would not be possible to com¬
pile a paleobotanical techniques manual that would lead one to
the best preparation methods for all types of preservation. The
latter are sufficiently variable so that a good deal of ingenuity
must frequently be brought into play to adapt standard techniques
to the material at hand. The advances of recent years should
leave no doubt that techniques will continue to improve and our
knowledge should expand correspondingly in the future. The
techniques used in certain highly specialized branches of paleo¬
botany such as palynology and coal studies are not included here.
Since most fossil plant materials fall in the compression or petri¬
faction categories it seems most expedient to consider them
separately.

Telegram @tubscstds 463

464 STUDIES IN PALEOBOTANY

PETRIFACTIONS

Ground Thin Sections

This classical method of preparing petrifactions for study has

been employed for about 150 years and is still useful. The major
items of equipment needed are a diamond impregnated saw and
one or more grinding laps. There are many different saws and
laps available which vary considerably in price; the only feasible
approach for a beginner who is setting up a laboratory is to visit
several in which such equipment is in use and acquire some under¬
standing of what seems most suitable. General purpose saws and
laps are adequate for much paleobotanical work; however, if a
great deal of thin section preparation is to be done special equip¬
ment is available. The preparation of a ground thin section is
essentially as follows:

1. The specimen is cut to a convenient size; this will, of course,

vary with the botanical nature of the material. For example, spec¬
imens of petrified coniferous or dicot wood about 1 cm square are
usually adequate.
2. The surface to be studied is smoothed with #400 or #600
Carborundum on a grinding lap or a sheet of plate glass.
3. The smoothed surface is then affixed to a glass slide. This is
most effectively accomplished on a hot plate by warming both
specimen and slide, applying melted resin to a slide, then placing
the specimen, smooth side down in the resin, pressing and rotating
gently to eliminate any air bubbles; then allow the mounted
specimen to cool until the resin hardens. This will take only a few
minutes.
Synthetic resins prepared especially for the purpose should be
used as they will harden to the proper viscosity. Paleobotanists
often require larger slides than the small ones used in petrographic
work; these can be cut, at very little expense, from ordinary double
weight window glass and are safe and satisfactory if the edges are
ground slightly. If one is doing much thin sectioning, it is well to
have a variety of sizes on hand.
4. Fasten the specimen (not the slide) to the jaws of the saw and
cut off as close to the glass as possible.
5. The slice of the petrifaction thus affixed to the slide is then
ground on a revolving lap. It is usually safe to use #100 Carbo¬
rundum until light begins to appear through the specimen; then

Telegram @tubscstds
TECHNIQUES AND BASIC REFERENCES 465

wash thoroughly and transfer to another lap or a piece of plate

glass, and finish with #400 Carborundum or finer. This procedure
is an art and requires some practice before one is able to prepare a
uniformly thin section. A rather simple grinding instrument has
been described by Croft (1950) which facilitates the preparation of
uniformly thin sections. Petrifactions differ tremendously in hard¬
ness and correspondingly in the time required for grinding; in the
last stages, with the fine abrasive, the slide should be washed and
examined frequently until the desired thinness is reached.
6. When sufficiently thin the slide should be washed, the excess
resin cut away, and the remainder washed off with a soft cloth and
the resin solvent. A cover slip may then be mounted over the sec¬
tion using a liquid synthetic resin mounting medium.
There are several disadvantages to this method, which may be
partially overcome with certain minor innovations. The finished
section is usually very fragile, but can be much less so by the addition
of this step: following (2) the smoothed surface may be etched very
lightly and a film of peel solution poured on and allowed to dry
(see Peel Method for details). The method continues without
change, the only difference being that there is now a very thin cel¬
lulose film between the specimen and the slide. Two advantages
accrue: the specimen does not tend to chip away as readily in the
late stages of grinding, and the film serves as a backing for the sec¬
tion so that it can be removed from the slide if necessary by soak¬
ing in a dish of the resin solvent.
With extremely soft or porous petrifactions it may be helpful to
allow the specimen to stand for a few minutes in melted resin
before affixing to the slide.
If material is scarce or for other reasons it is desirable to cut
sections closer together and thinner, this may be accomplished
with a fine copper wire; the wire is drawn back and forth over the
specimen using a sludge of Carborundum to do the cutting. The
chief disadvantages of the ground section technique is that it is
wasteful of material and at best the resultant sections are several
millimeters apart.

The Peel Method

This is applicable to well-preserved petrifactions, that is, ones in

which a considerable amount of organic material is still present.
It has several advantages over the ground section method: it is
rapid; essentially serial sections may be prepared; the size of sec-

Telegram @tubscstds
466 STUDIES IN PALEOBOTANY

tions is limited only by the cutting and grinding equipment, and

the resultant sections (“peels”) are durable and easily stored.
A. The (original) liquid technique
1. The surface to be sectioned is smoothed with #400 Carbo¬
rundum. Prior cutting and rough grinding are as given above.
2. The smooth surface is etched in dilute acid. In most of the
petrifactions one encounters, the minerals are silica or corbonates.
With carbonates HC1 is used, while HF1 is employed with silica.
Great care must be taken with the use of hydrofluoric acid and the
beginning student should consult a person who is experienced in its
use.
A convenient procedure, with carbonate petrifactions, is to
sprinkle a few small pieces of siliceous gravel (Vs to lA inch in diam¬
eter) over the bottom of a shallow glass tray and flood with HC1;
the specimen is then placed with the smooth side down in the acid.
A minimum quantity of acid should be used so that the sides of
the specimen are attacked as little as possible; the gravel simply
prevents the surface to be etched from contacting the bottom of
the tray.
In the author’s laboratory we maintain a constant solution of
acid (about 2 or 3%) and regulate the etching by time, which will
vary from one to five minutes. The etching dissolves away the
mineral matter leaving the cell walls standing in relief; two or
three trial peels should be made to determine the best thickness of
the finished section.
3. The specimen is washed in a gentle stream of water and
allowed to dry. The etched surface is very fragile and should not
be touched.
4. Place the specimen in a tray of sand or gravel so that the
etched surface is level. A small spirit level with a wire loop at¬
tached as a handle is satisfactory and, if allowed to contact the
surface gently, will cause no damage. If many peels are to be
made from one surface of a specimen it may save a little time to
embed it in plaster so that the desired surface is level when
the specimen is placed on a flat surface, thus eliminating leveling
each time.
5. The peel solution is then poured over the dry, etched surface.
A considerable variety of formulas have been used, the one given by
Darrah being highly satisfactory:

Parlodion, 28 grams xylol, 10 cc

butyl acetate, 250 cc castor oil, 3 cc

Telegram @tubscstds
amyl alcohol, 30 cc ether, 3 cc
TECHNIQUES AND BASIC REFERENCES 467

The solution must cover the surface evenly and a little practice
is required to obtain a uniform peel. It may be desirable to add a
few drops of butyl acetate first to prevent entrapping small air bub¬
bles; if the resultant peel is too thick, less solution should be used
and gently spread with a small piece of paper or the solution may be
diluted with butyl acetate. It is then allowed to dry for 12 to 20
hours in a reasonably dust free place.
6. The peel is removed by starting an edge with a scalpel or
razor and then carefully pulling it off. The result is an actual thin
section of whatever plant materials were exposed and not merely
an impression.
The peels may be studied directly with reflected light or mounted
under a cover slip in the usual way if higher magnifications are
needed. When mounting in balsam or a synthetic resin the rough
side should be moistened with a drop or two of solvent to prevent
entrapping air bubbles.
7. Before the next peel is made the surface should be smoothed
very lightly with the fine abrasive; otherwise the specimen will
etch differentially, ultimately forming a rough surface.
Each worker will devise his own modifications of the technique.
One innovation that has frequently proven useful may be worth
noting: it is possible to prepare sections in two planes, such as the
transverse and longitudinal faces of a stem, and mount them as a
single unit for study. The two surfaces are smoothed and etched
as usual; one surface is then “poured” and allowed to set; the block
is then turned and the second surface treated; after hardening, the
peels from the two surfaces may be removed carefully and will
remain attached at the adjoining edge.
B. The cellulose sheet technique
Recently the method as outlined above has been modified using
cellulose acetate sheets instead of a solution, as follows:
1. Steps 1 through 4 as above.
2. The etched surface is flooded with acetone and a sheet of
cellulose acetate 0.003 inch thick is quickly and gently rolled down
onto the surface. If the film is bowed slightly, and started at one
edge, air bubbles will be eliminated.
3. The film may be removed in about V2 hour.
This method is highly satisfactory and has two distinct advan¬
tages over the use of a solution: absolute uniformity of the result¬
ant peels is assured and a dozen or more may be made in a day.
The peel method has largely supplanted the older grinding
technique although the latter is by no means obsolete. It often

Telegram @tubscstds
produces better thin sections with certain materials, particularly
468 STUDIES IN PALEOBOTANY

where the preservation is of poorer quality. A paleobotanist work¬

ing with petrified plants must become adept at the ground thin
section method.
In his Methods of Palaeobotany Lacey has recorded a dramatic
instance of the advantage of the peel thin section method, although
one that we may hope will never be repeated. In 1940 certain
slides of the articulate cone Calamostachys, preserved in the
British Museum of Natural History, were partially destroyed by a
bomb blast. Upon searching through the debris W. N. Croft was
able to retrieve seven of the original eleven sections and, although
the slides and cover glasses were shattered, the peel sections, which
remained intact, were remounted and were essentially as good
as ever.

Microtome Techniques for Sectioning Petrifactions

To the best of my knowledge very little has been done with this
approach to handling petrifactions; it is mentioned as a technique
that is possibly worth more attention than it has received. It is
applicable with fossil woods in which much of the cell wall struc¬
ture is intact:
1. Prepare cubes about 1 cm square.
2. These are allowed to stand in hydrofluoric acid (the usual
commercial 52% should be diluted with equal parts of water for
safety in handling) for a few days. The specimen can be tested
with a needle to determine when the mineral has been largely dis¬
solved away.
3. Wash in running water overnight.
4. Embed in celloidin using the usual technique for modern plant
materials. Standard technique books such as Johansen, 1940, or
Sass, 1940, should be consulted for details of this embedding
procedure.

Treatment of Opaque Petrifactions

Some petrified plant materials are opaque, notably where sulfides

of iron are the mineral agent, and the grinding or peeling methods
are of little use. A good deal of information may be extracted from
such fossils by studying polished surfaces under reflected light.
Examples of what may be obtained with the use of modern tech¬
niques of this sort are found in Leclercq’s study of Rhacophyton

Telegram @tubscstds
TECHNIQUES AND BASIC REFERENCES 469

and Beck’s investigation of Tetraxylopteris (see Chapter 3). In

both cases these were compression fossils but fragmentary portions
were found here and there with vascular tissue preserved. In the
former study a plastic embedding medium was used and in the
latter a resin; these are outlined briefly:

Balsam coating method (Beck, 1955)

A thin slice of the petrifaction is placed in a solution of three parts

toluene and one part Canada balsam or a synthetic resin and boiled
gently on a hot plate until most of the toluene has evaporated.
The boiling should be carried far enough so that the resin, when
cooled, will be firm but not excessively brittle. The section is then
removed, excess resin scraped off, and the two surfaces ground with
#600 Carborundum followed by chromium oxide; grinding should
be kept to a minimum since there is little actual penetration of the
petrifaction. The section is then washed, dipped in xylol or
toluene, and mounted under a cover slip; both surfaces may then
be examined under reflected light.

Plastic embedding method (Leclercq and Discry, 1950;

Leclercq and Noel, 1953)

Since the use of plastic as an embedding medium is now widely

applied in biology, it does not seem necessary to give a detailed
description of this method.
As carried out in Professor Leclercq’s laboratory the technique
involves the following steps:
Small porcelain trays (as used in paraffin embedding) or boxes
lined with aluminum foil may be used as a container. A small
portion of the plastic is poured in and allowed to set; the specimen
is then placed in the container and covered with the plastic which
is allowed to harden overnight. It is then cut according to the
surface desired and polished with a fine abrasive.

COMPRESSIONS

Well-preserved compression fossils may reveal an astonishing

amount of information, and specimens of a type that were once
examined only superficially or even discarded can now be made to
yield critical data. The fragment of a Carboniferous fertile frond
(Senftenbergia), shown in Figs. 1-2A-C, is representative. The

Telegram @tubscstds
470 STUDIES IN PALEOBOTANY

study of compression fossils can for the most part be conducted in

three rather different ways, depending on the nature of the plant
materials: by direct excavation, maceration, or transfer.

Excavation Technique

When the plant remains ramify rather freely through the rock
matrix and consequently are not confined to a narrow bedding
plane, there is no method presently available for removing the fos¬
sil intact from the matrix (as in the transfer methods described
below). A great deal may be learned, however, by judicious exca¬
vation with mircrovibratory machines. If such equipment is not
available or if one is working with minute structures as is often the
case, a small hammer and steel needles may be employed. The
specimen may be held on a small sand bag and the excavation car¬
ried out under a binocular dissecting microscope. This is slow,
tedious work but may be repaid with highly productive results
that can be obtained in no other way; the structure of the minute
leaves and sporangiophores of Calamophyton (see Chapter 10)
were worked out using this method.

Transfer Techniques

When the compression is preserved in an essentially flat bedding

plane and is too delicate (or lacking a tough cuticle) to allow
removal by maceration (see below) one of the transfer methods
may be used.

The Ashby cellulose film transfer method (Lang, 1926)

1. The plant compression and immediately adjoining rock sur¬

face is coated with the peel solution described above; two coats are
usually necessary to produce a tough film.
2. When the film is dry, the rock matrix is cut away from the
back side as much as possible without damaging the compression.
3. The specimen is then placed in a suitable container and
covered with dilute hydrofluoric acid (about 25%). The acid will
attack the rock matrix but will not affect the cellulose film or plant
compression.
The time required to completely dissolve away the rock matrix
will vary with the size of the specimen and the nature of the min¬
erals. The dissolving action may be speeded by changing the acid

Telegram @tubscstds
TECHNIQUES AND BASIC REFERENCES 471

each day and by carefully scraping away the disintegrating rock.

The last of it may be removed under water with a small, soft
artist’s brush.
This procedure is, as noted, particularly helpful with fertile
frond fragments. The sporangia are made more accessible for
study and small portions of the transferred compression may be
treated in a small glass dish with a few spoonfuls of nitric acid
containing about 5% potassium chlorate in solution. This is a
powerful oxidizing agent and will soon disintegrate all except
epidermal cuticles and spores. It is necessary to watch the macera¬
tion process and stop it at the desired point by diluting the solu¬
tion with water.

The Balsam transfer method (Walton, 1923)

1. The compression specimen, or a representative portion of it,

is trimmed to a convenient size and as much as possible of the rock
matrix in back is cut away.
2. The specimen is then pressed, compression side down, into a
mass of melted balsam on a glass slide.
3. The entire preparation is then heavily coated with wax by
dipping it into melted paraffin; when hard, the wax is scraped
away from the back of the rock (the side opposite the compression).
4. The preparation is then immersed in a bath of hydrofluoric
acid. The paraffin serves to protect the glass slide and allow the
acid to act only on the rock. The process may be aided and the
final remnants of rock cleaned away as indicated above.
Of these two transfer techniques the balsam one is perhaps more
satisfactory with highly fragile material.

The Abbott transfer method

A modification of the film transfer has been developed by Abbott

(1950, 1951) which allows the removal of large compressions. The
compression is given a rather heavy coat of fingernail polish fol¬
lowed by sheet acetate or a nail polish-varnish mixture. The film
is then pulled off bringing the compression with it. Compressions
up to 4 by 6 inches may be removed in this way. For research
work such large preparations are usually not necessary, but the
method produces excellent display specimens and is especially
valuable for preserving a compression that is found in a rock that
is not durable.

Telegram @tubscstds
472 STUDIES IN PALEOBOTANY

Bulk macerations

When compression fragments are abundant in a shale or are

tough enough so that they may be removed without the use of a
transfer medium such as cellulose or balsam, a bulk maceration
may be employed.

Harris’ (1926) method

1. The shale specimen is immersed for several days in a solution

of strong nitric acid containing about 5% potassium hydroxide. It
is suggested that the beginner start with small specimens of per¬
haps 1 or 2 square inches and only as much of this rather violent
oxidizing agent as is needed to cover it.
2. The acid is then removed with numerous changes of water.
3. The specimen is next placed in dilute sodium hydroxide; the
matrix breaks down into a fine mud and the plant materials, which
are now reduced for the most part to cuticularized parts, may be
screened off.
4. The cuticular fragments are finally cleaned with 25% hydro¬
fluoric acid.

Hydrofluoric acid method

Although it is perhaps a matter of personal preference it seems

to me that it is sometimes advantageous to remove the plant
materials intact before subjecting them to the oxidation process.
1. The fossiliferous rock sample is placed in about 25% hydro¬
fluoric acid; several hours to a day or two may be required to break
down the mineral matrix.
2. The sludge containing the plant fragments should be thor¬
oughly washed with water to remove the acid.
3. The larger plant fragments may then be removed and the
smaller ones separated by screening. They may then be treated
individually with nitric acid and potassium chlorate followed by
sodium hydroxide as outlined above. The time required for this
treatment varies from an hour or less to several days.

REFERENCES

Abbott, Maxine L. 1950. A paleobotanical transfer method. Journ. Paleont., 24:

619-621.
Abbott, Ralph E., and M. L. Abbott. 1952. A simple paleobotanical transfer tech¬

Telegram @tubscstds
nique. Ohio. Journ. Sci., 52: 258-260.
TECHNIQUES AND BASIC REFERENCES 473

Beck, Charles B. 1955. A technique for obtaining polished surfaces of sections of

pyritized plant fossils. Bull. Torrey Bot. Club., 82: 286-291.
Croft, W. N. 1950. A parallel grinding instrument for the investigation of fossils by
serial sections. Journ. Paleont., 24: 693-698.
Darrah, William C. 1936. The peel method in paleobotany. Harvard Univ. Bot.
Mus. Leaflets, 4: 69-83.
Graham, Roy. 1933. Preparation of paleobotanical sections by the peel method.
Stain Technology, 8: 65-68.
Harris, Thomas M. 1926. Notes on a new method for the investigation of fossil
plants. New Phyt., 25: 58-60.
Johansen, Donald A. 1940. Plant Microtechnique. McGraw-Hill Book Co., New
York. 523 pp.
Joy, K. W., A. J. Willis, and W. S. Lacey. 1956. A rapid cellulose peel technique in
palaeobotany. Ann. Bot., n. s., 20: 635-637.
Lacey, William S. 1953. Methods in palaeobotany. The North Western Naturalist,
Arbroath, Wales, 24: 234-249.
Lang, W. H. 1926. A cellulose-film transfer method in the study of fossil plants.
Ann. Bot., 40: 710-711.
Leclercq, Suzanne, and M. Discry. 1950. De l’utilization du plastique en paleontolo-
gie vegetale. Ann. Soc. Geol. Belgique, 73: 151-155.
-, and R. Noel. 1953. Plastic—a suitable embedding substance for petro¬
graphic study of coal and fossil plants. Phytomorphology, 3: 222-223.
Sass, John E. 1940. Elements of Botanical Microtechnique. McGraw-Hill Book Co.,
New York. 222 pp.
Selling, Olof H. 1944. Studies on Calamitean cone compressions by means of serial
sections. Svenska Bot. Tidskrift, 38: 295-330.
Walton, John. 1923. On a new method of investigating fossil plant impressions or
incrustations. Ann. Bot., 37: 379-390.
-. 1928. A method of preparing sections of fossil plants contained in coal
balls or in other types of petrifaction. Nature, 122: 571.
-. 1930. Improvements in the peel method. Nature 125: 413.
-. 1952. Notes on the preparation and permanence of peel sections.
Compte Rendu, 3ieme Congr. Strat. Geol. Carbonifere, Heerlen, 1951. Pp. 651-653.

Some Basic References

The problem of “keeping up with the literature” is a potential source of despair in

almost any major branch of science. As to paleobotany it is hardly possible for an
individual to be thoroughly familiar with all that is being done in the various branches,
but an acquaintance with the standard reference works and literature reports allows
one to keep up with what seem to be the more important works within his own area
of interest.
In the foregoing chapters some of the basic as well as more recent references are
listed and the bibliographies given in them will in turn lead the reader to a more
detailed survey. Certain other useful works are cited below.

Standard Textbooks

Arnold, Chester A. 1947. An Introduction to Paleobotany. McGraw-Hill Book Co.,

New York. 433 pp.
Darrah, William C. 1960. Principles of Paleobotany. 2nd ed. Ronald Press Co.,

Telegram @tubscstds
New York. 256 pp.
474 STUDIES IN PALEOBOTANY

Emberger, Louis. 1954. Les plantes fossiles dan leurs rapports avec les vegetaux
vivants. Masson et Cie., Paris. 492 pp.
Hirmer, Max. 1927. Handbuch der Palaobotanik. R. Oldenbourg, Munich. 708 pp.
This is an encyclopaedic treatment of the cryptogamic groups, copiously illustrated
and with extensive bibliographies. It is volume I of what was apparently intended
as a two-volume work.
Jongmans, Willem J. 1949. Het Wisselend Aspect van het Bos in de Oudere Geol-
ogische Formaties. Issued as a part of “Hout in alle tijden,” edited by W. Boerhave
Beekman. Deventer, Netherlands. 164 pp. This is essentially a compilation of
individual plant as well as general landscape restorations and it stands alone in this
way as a valuable and remarkble volume.
Krausel, Richard. 1950. Versunkene Floren. Waldemar Kramer, Frankfurt a. M.
152 pp.
Krystofovich, A. N. 1957. Paleobotany. Leningrad. 650 pp.
Magdefrau, Karl. 1956. Palaobiologie der Pflanzen. 3rd ed. Gustav Fischer. Jena.
443 pp.
Seward, Albert C. Fossil Plants. 4 vols. Cambridge Univ. press. In addition to
making many original contributions Sir Albert Seward was the greatest compiler
that paleobotany has had. Although this work is out of date in many aspects, it
contains a wealth of information and is still indispensable. The four volumes ap¬
peared as follows: I. 1898; II. 1910; III. 1917; IV. 1919.
Scott, Dukinfield H. Studies in Fossil Botany, vol. I. 1920; vol. II. 1923.
Walton, John. 1953. An Introduction of the Study of Fossil Plants. 2nd ed. A & C,
Black Ltd., London. 201 pp.

Journals Devoted Exclusively to Paleobotany

The Palaeobotanist. Birbal Sahni Institute of Palaeobotany, Lucknow, India.

Palaeontographica (Abt. B). Stuttgart, Germany.

Catalogues and Literature Reports

Andrews, Henry N., Jr. 1955. Index of generic names of fossil plants, 1820-1950.
U. S. Geol. Survey Bull., 1013. 262 pp. This is taken from the United States
Geological Survey’s Compendium Index of Paleobotany, a card file that maintains a
citation of the original source of publication of all species of fossil plants except the
diatoms, spores, and pollen.
Knowlton, Frank H. 1919. A catalogue of the Mesozoic and Cenozoic plants of
North America. U. S. Geol. Survey Bull., 696. 815 pp.
Kremp, G. O. W., et al. 1957-. Catalogue of Fossil Spores and Pollen. Palynologi-
cal laboratories, The Pennsylvania State University. Several volumes have
appeared each year, starting in 1957.
LaMotte, Robert S. 1944. Supplement to Catalogue of Mesozoic and Cenozoic plants
of North America 1919-1937. U. S. Geol. Survey Bull., 924, 330 pp.
-. 1952. Catalogue of the Cenozoic plants of North America through 1950.
Geol. Soc. Amer. Mem., 51, 381 pp.
World Report on Palaeobotany. Edited by E. Boureau and collaborators. Prior to
1956 several regional literature reports were issued more or less regularly. These
included Reports for Europe, Britain, India, and the United States. In 1956 the
first World Report appeared as an official organ of the International Paleobotanical
organization. The second number appeared in 1958.

Telegram @tubscstds
TECHNIQUES AND BASIC REFERENCES 475

Studies in the History of Polaeobotany

Edwards, Wilfred N. 1931. Guide to an exhibition illustrating the early history of

paleontology. Brit. Mus. Nat. Hist., 68 pp.
Gordon, William T. 1934. Plant life and philosophy of geology. Pan-Amer. Geolo¬
gist, 62: 161-186, 329-346.
Ward, Lester F. 1884. Sketch of paleobotany. U. S. Geol. Survey. Fifth Ann. Rept.,
pp. 363-452. Washington.

Telegram @tubscstds
 .

'

Telegram @tubscstds
 NDIX

Abbott, M. L., 277 Angiosperms, earliest, 169-173

Abbott transfer method, 471 evolution in the Arctic, 180
Acrangiophyllum pendulata, 121 floras of Upper Cretaceous, 184
Actinopodium nathorstii, 392, 411 from Lower Cretaceous of Canada, 181
Actinostele, 24 from Lower Cretaceous of USSR, 181
Adiantites, 90, 91 modes of preservation, 167
Alabama, 19, 121 primitive, 173
Alaska, Upper Cretaceous flora, 197 problems of identification, 188
Aldanophyton antiquissimum, 47 woods from Aptian of England, 181
Alder, 196, 199 Ankyropteris glabra, nodal anatomy, 77
Alethopteris, 90, 91 hendricksi, 78
grandifolia, with seed, 154 Annual rings in Jurassic Arctic woods, 387
norinii, with seed, 153 Annularia, 267
Algae, 16 sphenophylloides, 272
Alloiopteris, 90, 91 Anomozamites amdrupiana, 388
Alnus, carpinoides, 205 minor, 305
smithiana, 209 nitida, 388
Amber, 16 Antarctica, 357, 394
Amboy flora of New Jersey, 184 Apotropteris minuta, 73
Amino acids, in animal fossils, 15 Araucarioxylon, 334
Anachoropteridaceae, 72 arizonicum, 432
Anachoropteris clavata, 74, 75 Archaeopteris fimbriata, 391
Androstrobus manis, 295, 296 hibernica, 63
Anemia adiantifolia, 104 latifolia, 63
fremonti, 104, 105, 109, 117 roemeriana, 391
mexicana, 103 Archaeosigillaria primaeva, 251
poolensis, 104 Archangiopteris, 94
sporangium, 103 Arctic plants, 55, 180, 380
Aneurophyton germanicum, 68 Arctobaiera flettii, 344
Angara flora, 419 Arcto-Tertiary flora, 205
Angaridium mongolicum, 420 Aril, 183
Angaropteridium cardiopteroides, 420 Arizona, 14, 110, 275
Angiopteris, 94 Triassic flora, 431

Telegram @tubscstds 477

478 INDEX

Arthropitys, 262 Bold, H. C., 21, 405

Arthroxylon williamsonii, 264 Bornia radiata, 275
Artocarpus dicksonii, 385 Bothrodendron kiltorkense, 240
incisa, 384 mundum, 241
Ash, 202, 204, 208 Bothrostrobus bailyanus, 240
Ashby transfer method, 470 mundus, 241
Asia Paleozoic floras, distribution of, 417 Botrychioxylon paradoxum, 79
Asterophyllites, 263, 267 Botrychium lunaria, 78
charaeformis, 266 Botryopteridaceae, 69
longifolius, 266 Botryopteris antiqua, 72
Asterotheca meriani, 102 elliptica, 72
parallela, 97 forensis, 72
Asteroxylon mackiei, 36, 37, 39 globosa, 70, 71
Aulacotheca elongata, 149, 151 ramosa, 72
Australia, 41, 80, 274 trisecta, 70, 71, 85, 86
Austria, 102 Bower, F. 0., 99
Austroclepsis australis, 80 Bowerbank, J. S., 188
Axelrod, D. I., 196, 393 Boweria schlatzlarensis, 121, 122
Axillary branching in Ankyropteris, 77 Bowmanites bifurcatus, 280, 281
Azolla intertrappea, 16, 17 dauosoni, 279
fertilis, 280
Baiera, 339 moorei, 282
boeggildiana, 340, 341 Bracken fern, 204
leptophylla, 341 Bradley, W. H., 17
spectabilis, 340 Branching, axillary, in Ankyropteris, 77,
Bailey, I. W., 179 78
Balsam transfer method, 471 in Hyenia, 260
Banana, fossil, 18 Brandon, Vermont, lignite, 15
Baragwanathia longifolia, 215 Brazil, 115, 424, 427
Barghoorn, E. S., 15 Breadfruit, 180, 200
Baxter, R. W., 79, 224 in Cretaceous of Greenland, 384
Beania gracilis, 295, 296 Bridge Creek flora, 204
Bear Island, 240, 391 Brown, R. W., 20, 199
Beech, 204 Bucheria ovata, 44, 45
Belgium, 43, 54, 216, 258, 280, 283, 369, Bucklandia indica, 308
412
Bell, W. A., 181 Calamitaceae, 261
Bembridge flora, 194 Catamites, cones, 268-274
Bennetticarpus crossospermus, 303 cruciatus, 265
Bennettitales, 297 foliage, 266-268
microsporangiate organs, 306 pith casts, large size of, 266
stomatal structure, 293, 294 roots, 266
Bergeria mimerensis, 411 secondary wood types, 262-264
Berry, E. W., 8 twig, 263
Bertrand, P., 82 Calamodendron, 263
Birch, 195, 198, 199 Calamophyton bicephalum, 258-260
Birth wort, 198 Calamostachys americana, 270
Biscalitheca musata, 81 binneyana, 268, 270
Bjuvia simplex, 312 casheana, 270, 271

Telegram @tubscstds
Bochenski, T. A., 245 ludwigi, 269
INDEX 479

Calathospermum scoticum, 158 Coniferales, foliage of early fossils, 322

Callistophyton poroxyloides, 137, 138 wood, 333
Callixylon, 332, 334 Continental drift, 418
size of trunks, 331 Cooksonia, 42
Cambium, 24 hemispherica, 41
Cambrian plant records, 47 pertoni, 41
Carbon-14 method, 4 Coprolites, 16, 19, 20
Cardiocarpus spinatus, 327 Cordaianthus concinnus, 317
Camoconites compactum, 350 penjoni, 317
laxum, 350 pseudofluitans, 320, 321, 325
Carpel, conduplicate, 174, 175 zeilleri, 321
evolution of, 178 Cordaitales, isolated seeds, 326-329
Carpentieria frondosa, 322 leaves, 316
Cathaysia flora, 421 seed organs, 318-321
Cat-tail, 203 size attained, 316
Caulopteris, 100 wood, 329
Caytonanthus arberi, 178 Cordaites sahnii, 423
kochi, 177 Com, supposed fossil ear, 20
Caytonia nathorsti, 177 Comer, E. J. H., 182
sewardi, 176 Corycium enigmaticum, 49
thomasi, 178 Corystospermaceae, 159
Caytoniales, 176 Croft, W. N., 465
Centrarch, 27 Crossotheca, 136
Chalk Bluffs flora, 200 Cupulate disc, 164
Chaloner, W. G., 76 Cupule, 129, 157
Chandler, M. E. J., 189, 191, 193 Cyathotrachus altissimus, 95
Chaney, R. W., 196, 210 Cycadales, distribution of living genera,
Cheirostrobus petty curensis, 284, 285 310
Chemnitz, 74 fossil, 295
Chile, 104 stomatal structure, 293, 294
China, 153, 216, 410 Cycadeoidea dacotensis, 299, 300, 302
Chinlea campii, 431 dartoni, large number of strobili, 301
Christensenia, 94 etrusca, 297
Cingularia typica, 272, 273 jenneyana, 298
Cinnamon, 181, 185 marshiana, 298
Cladoxylon radiatum, 82, 83 wielandi, 301
scoparium, 82 Cypress, 198, 203
taeniatum, 82, 83 Czekanowskia nathorsti, 353
Classification, general, 21
Clathropteris meniscoides, 111 Dadoxylon, 334
Climatic change, use of oxygen isotopes, Dakota sandstone flora, 184
392 Dating of rocks, 3
Coal balls, chart of horizons, 126 Dawson, J. W., 38, 51
where found, 14 Dawsonites, 42
Coenopteridales, distinctive features of, 69 arcuatus, 38
Colorado, 170, 185, 222 Degeneriaceae, 174
Colpodexylon trifurcatum, 218 Degeneria vitiensis, 174
Compression fossils, 9, 10 Delevoryas, T., 75, 144, 148
Condrusia minor, 369, 370 Dendrites, 19, 20
Devonian floras, distribution of, 409-415

Telegram @tubscstds
rumex, 369, 370
480 INDEX

Diatoms, fresh water forms from Atlantic Falkland Islands, 410

drill cores, 419 Ferns, classification, 93
Dichophyllum moorei, 345 primitive, 61
Dicots, vesselless, 174 Fig, 181, 184, 185, 197, 198
Dicroidium, 160, 161 Fir, 196, 203
Dictyostele, 24 seed cone of, 320
Diichnia kentuckiensis, 415 Fischer, C. H., 436
Dijkstra, S. J., 367 Flagellates, in Cretaceous flint, 18
Dinichthys terelli, 15 Florin, R., 312, 315, 317, 341, 342
Dioon edule, 303 Florissant, Colorado, 203
Diplotmema, 90, 91 Flower, use of term, 299
adiantoides, 140 Foerstia, 52, 53
Dipteridaceae, 111 Form-genera of leaves, 89
Dipteris conjugata, 111, 112 Fossil, definition, 4
lobbiana, 111 Fossil forest, of Arizona, 14
Discinites delectus, 124 of Florissant, Colorado, 12
major, 124 of Yellowstone Park, 11
Dissemination of spores and pollen, 445 France, 146, 157, 170
Dogwood, 195 Frond, definition, 60
Dolerotheca formosa, 149 Frontier formation, fossils of, 104, 197
Dorf, E., 13, 202 Fruitland formation, 197
Dorsiventral stem, 72 Furcula granulifer, 170
Drepanophycus spinaeformis, 216
Durian theory, 182 Gametophytes, of Biscalitheca, 81
Dutoitia pulchra, 410 of Cardiocarpus, 328
of Lagenostoma, 133
of Lepidocarpon, 232
Earth, age of, 3 of Mazocarpon, 244
Elaterites triferens, 275 Gangamopteris, 356
Elaters of articulate spores, 275 Gaspe, 38, 51, 216
Ellesmere Island, 62 Geisel Valley fossils, 18
Elm, 203, 204 Geologic time chart, 6
Emplectopteris triangularis, with seed, Germanophyton psygmophylloides, 54
153, 154 Germany, 18, 42, 54, 68, 82, 144, 216, 217,
Encrustation fossils, 15 282
Endarch, 27 Gigantopteris nicotianaefolia, 422, 423
England, 79, 96, 104, 114, 116, 122, 130, Gilboa, New York, 66
141, 151, 188-194, 241, 266, 268, 280, Ginkgo adiantoides, 338, 339
369 biloba, general morphology, 336
Enigmophyton superbum, 54, 55 present distribution, 384
Eoangiopteris andrewsii, 96, 98 wood, 335
Eospermatopteris, 66, 67 digitata, 338, 339
Equisetites bradyi, 432 lamariensis, 338, 339
platyodon, 274 Ginkgodium nathorsti, 345
woodsi, 274 Ginkgoites, 339
Equisetosporites chinleana, 275 acosmia, 340, 388
Emestiodendron filiciforme, 322, 323, 325 fimbriata, 340, 341
Escherichia, 16 hermelini, 340, 341
Etapteris lacattei, 81 lunzensis, 343

Telegram @tubscstds
Eviostachya hoegi, 283 minuta, 340, 341
Exarch, 27 obovata, 340
 INDEX
481

Ginkgophyllum usevolodi, 420 Heterospory, evidence from isolated

Ginkgophyta, 335-346 spores, 367
Glaciers of New Zealand and Norway, 428 in Lepidostrobus, 229-231
Gleichenia, distribution of, 108 in Noeggerathiales, 124
longissima, 103 in Protopitys scotica, 366
sporangium, 103 in Selaginella, 214, 365, 366
Gleicheniaceae, 107 in Selaginellites, 247
Gleichenites, 383 in Sigillariostrobus, 245
coloradensis, 109 in Stauropteris, 75
rostafiniskii, 109 summary, 365
Glossopteris flora, 354, 424 Hicklingia edwardi, 409
Gnetales, fossil pollen, 460 Hickory, 198, 199, 200
Gnetopsis elliptica, 157, 158 Hirmeriella rhatoliassica, 326
Goldenbergia glomerata, 151 Hpeg, 0. A., 61, 260
Golden rain tree, 199 Holly, 195
Gondwanidium petiolatum, 420 Homoxylon aviassi, 176
sibiricum, 420 neocaledonicum, 176
Gordon, W. T., 156 Hordle flora, 193
Goshen flora, 199 Hornbeam, 195, 196, 199, 204
Gosslingia breconensis, 45, 46 Horneophyton lignieri, 33, 36
Grammatopteris baldaufi, 74, 116 Hungary, 114
Grape, 195, 198 Hydropteridangium marsilioides, 359
Greenland, 163, 170, 176, 180, 247, 382, Hyenia elegans, 258, 261
383, 391 vogtii, 258
Green River flora, 199
Ground thin sections, 464, 465
Iceland, 382
Gunflint iron formation, Ontario, 48
Idaho, 117
Guppy, H. B., 192
Illinois, 74, 99, 140, 219, 244, 251, 262,
Guycampbellia microphylla, 414
265, 280
Gymnosperm, use of term, 22
Impression fossils, 8, 9
India, 110, 309, 424-426
Hackberry, 203, 204
Integument, origin in cordaites, 374
Halonia, 234
origin in pteridosperms, 370-378
Harris, T. M., 289, 292, 303, 308, 339, 387 Interseminal scales of bennettites, 303
Hausmannia dentata, 112 Intia vermicularis, 401
nariwaensis, 112 Iowa, 70, 96, 138, 142, 275, 282
Hawthorn, 203, 204 Ireland, 64, 240
Hazel, 195 Isoetes horridus, 252
Hedeia corymbosa, 41, 42 serratus, 252
Helminthostachys, 78 Ivy, 198
Hemitelia crenulata, 120
Hemlock, 196
Japan, 106, 111, 339
Hepaticites kidstoni, 398
Jongmans, W. J., 150
langi, 399
Juniper, 208
rosenkrantzi, 400
willsi, 399
Heterangium grievii, 140 Kcillostachys scottii, 273
Heterospory, in Archaeopteris, 63, 64 Kamaraspermum leeanum, 329, 330
in Calamostachys, 270 Kansas, 70, 140, 152, 184, 247, 272
in Devonian plants of Ellesmere Island, Katsura tree, 200

Telegram @tubscstds
391 Kentucky, 76, 84, 231, 413
482 INDEX

Kidston, R., 32 Litostrobus iowensis, 282

King Karl’s Land, 387 Lobatannularia ensifolius, 422
Kirtland formation, 197 Lonchopteris, 90, 91
Klukia exilis, 106 bricei, 85, 86
Kolbe, R. W., 419 London Clay flora, 188
Krausel, R., 217, 218, 304 Lycopod cones, evolution of, 231
Kroscienko flora, 195 growth, 241
Kryshtofovichia africani, 367 herbaceous types, 245-248
leaves, anatomy, 226
Laccopteris, 383 size variation, 227
Lagenostoma lomaxi, 133, 175 Lycopodites, 245
ovoides, 133 Lycopodium, 214
Lance flora, 185 Lyginopteridaceae, 130
Land bridges, 419 Lyginopteris oldhamia, 130-137
Lang, W. H., 32 Lygodium palmatum, 103
Laurel, 180 pumilum, 104
Lava flows of Columbia Plateau, 205 skottsbergii, 104
Leaf evolution, 84 sporangium, 103
Leather fern, 194
Lebachia goeppertiana, 323 Maceration technique, 472
piniformis, 322, 324 MacGinitie, H. D., 203, 304
Lepidocarpon magnificum, 232 Macrotaeniopteris magnifolia, 432
Lepidodendron acutum, 227 Madro-Tertiary flora, 205
in Arctic floras, 390 Magnetite, 3
dichotomum, 227 Magnolia, 185
johnsonii, 222 Mamay, S. H., 286, 378
lycopodioides, 227 Maple, 198, 199, 202, 208
obovatum, 227 Marattia, 94
scleroticum, 219-226 Marattiaceae, 94, 389
Lepidophloios pachydermatikos, 220 Marattiopsis hoerensis, 102
scoticus, 234-235 macrocarpa, 102
with cones attached, 233 Marchantites hallei, 400
Lepidophylloides, 225 sezannensis, 400
Lepidophyllum, 225 Mariopteris, 91
Lepidopteris natalensis, 163 Mascall flora, 207
ottonis, 163 Matoniaceae, 109
Lepidostrobus braidwoodensis, 231 Matonidium americanum, 110
diversus, 228-231 indicum, 110
fischeri, 415 Mazocarpon oedipternum, 244
foliaceus, 231 Mazostachys pendulata, 272, 273
noei, 231, 415 Medicine Bow flora, 197
size of cones, 231 Medullosa endocentrica, 144, 145
Leptopteris wilkensiana, 113 evolution of stelar system, 146-148
Leptostrobus longus, 353 gigas, 146
Levicaulis arranensis, 224 grandis, 145
Libby, W. F„ 4 heterostelica, 144
Lidgettonia africana, 356 leuckarti, 145, 146
Ligule, 230, 241 leuckarti v. lignosa, 146
Linden, 199, 204 noei, 145, 147, 151

Telegram @tubscstds
Linopteris, 90, 91 primaeva, 144
INDEX 483

Medullosa solmsi, 145 Nipaniophyllum raoi, 350

stellata v. typica, 144-146 Noeggerathiales, 124
thompsoni, 142, 143 Noeggerathiostrobus bohemicus, 124, 125
Medullosaceae, 142 Norinia cucullata, 422
Megaphyll, 60 Norway, 216
Megaphyton, 100 Nova Scotia, 243
Megatheca thomasi, 157 Nucellangium glabrum, 360-362
Mercenaria mercenaria, 15 Nystroemia pectiniformis, 154, 155
Mesarch, 27
Mesoxylon, 331 Oak, 180, 195, 197, 198, 208
Metasequoia occidentalis, 206 Ocean currents and transport of plant
Metaxylem, 26 materials, 192
Metzgeriites glebosus, 400 Ohio, 15
Miadesmia membranacea, 251 Oldest fossil plants, 48, 49
Michigan, 124 Oligocarpia brongniarti, 109
Microorganisms, in coprolites, 16 capitata, 108
Microphyll, 60 Ontario fossils, 48
Miospore, 439 Ophiobolus utinensis, 18
Mock orange, 204 Oregon grape, 185
Montana, 120 Osmunda cinnamomea, 113
Moonseed, 180 claytoniana, 113
Moresnetia zalesskyi, 369, 370 regalis, 103, 113
Moseley, H. N., 191 Osmundites braziliensis, 114
Muscites polytrichaceus, 401 chandleri, 114
yallournensis, 402 dunlopi, 114, 115
kolbei, 114, 115
Naiadita lanceolata, 402-404 spetsbergensis, 115
Natal, South Africa, 163 walkeri, 431
Nathorst, A. G., 382, 387 Ostrya oregoniana, 206
Nathorstia alata, 102 Ottokaria, 356
Nathorstiana arborea, 253 Ovuliferous scale of conifers, evolution of,
Naumova, S. N., 47 317-326
Nematothallus pseudovasculosa, 50 Oxygen isotopes, 392, 393
Nemejc, F., 231
Neocalamites hoerensis, 388 Pachytesta illinoense, 151, 152
Nephropsis rhomboidea, 351 Pachytheca vera, 150
Neuburg, M. F., 351, 401 Palaeostachya andrewsii, 272
Neuropteris, 91 decacnema, 272
heterophylla, with seed, 154 vera, 272
schlehani, 150 Palaeovittaria kurtzi, 425
tenuifolia, with seed, 154 Paleomagnetism, 393
New Brunswick, 316 Palm, earliest record, 170
New Caledonia, 113, 176 Tertiary of Alaska, 383
New Jersey, 15 Paper coal, 251
New York, 66, 68, 83, 251, 284 Parichnos, 228
New Zealand, 115 Patagonia, 102
Nikitin, P. A., 367 Pecopteris, 91
Nilssonia compta, 295, 296 pluckenetii, with seed, 155
Nilssoniopteris vittata, 304 Pectinophyton bipectinatum, 45

Telegram @tubscstds
Nipa burtini, 189 Peel method, 465-467
484 INDEX

Peltaspermaceae, 163 Protein in animal fossils, 15

Pennsylvania, 64 Protobarinophyton obrutschevii, 45
Pentoxyleae, 349 Protoblechnum wongii, 422
Pentoxylon sahnii, 349 Protohyenia janovii, 260, 261
Periderm of lycopods, 224 Protolepidodendron scharyanum, 218
Petiole, definition, 60 Protopitys buchiana, 366
Petrifaction fossils, 11, 12 scotica, 366
Pettycur, Scotland, 140, 284 Protopteridales, 61
Phlebopteris smithii, 110 Protosalvinia, 52, 53
utensis, 110 Protostele, 24
Phyllites, 170 Prototaxites southworthii, 51, 52
Phyllotheca etheridgei, 426 Protoxylem, 26
Physostoma elegans, 141 Psaronius blicklei, 99
Pietzschia polyupsilon, 84, 414 Pseudobornia ursina, 391
Pilophorosperma granulatum, 160, 161 Pseudofossils, 19
Pine, 196, 199, 203, 208 Pseudovoltzia liebeana, 323, 325
Pinna, definition, 60 Psilodendrion spinulosum, 411
Pinnule, definition, 60 Psilophytales, as defined by Kidston and
Pitting transition in primary wood, 27 Lang, 41
Pityoxylon, 335 Psilophyton arcticum, 411
Plagiopodopsis cockerelliae, 402 princeps, 38
Plagiozamites oblongifolius, 422 robustius, 43
Planation, 85 ivyomingensis, 40
Pleuromeia sternbergi, 248, 249 Psilotum, 25, 26, 35, 36
Plumstead, E. P., 354 Psygmophyllum gilkineti, 54
Poland, 195 Pteridophyte, use of term, 22
Pollen, identification of, 447 Pteridosperms, characteristics of, 129
morphology, 439-442 Pteruchus africanus, 160, 161
protein content, 438 Ptilozamites nilssoni, 359
sampling methods, 454 Pur year, Tennessee, 8, 9
size of, 442
treatment for study, 452 Queensland, 46
Polospore, 439 Quercus hannibali, 209
Polycyclic stele, 25
Polystelic stems of medullosas, 142 Rachiopteris aspera, 132
Polytrichites spokanensis, 402 Radforth, N. W., 97, 106
Polyxylon elegans, 414 Rajmahal Hills, India, 102
Pont-de-Gail flora, 194 Reduced, as distinct from primitive plant,
Poplar, 181, 198 36
Populus alexanderi, 209 Renault, B., 72
pliotremuloides, 209 Renaultia gracilis, 121, 122
Portugal, 180 Reuverian flora, 194
Potoniea, 151 Rhacophyton zygopteroides, 64, 65
Prefems, 59 Rhacopteris, 91, 92
Preservation, of leaves, seeds, fruits, 191 Rhexoxylon africanum, 357
of organic compounds, 15 priestleyi, 357
of plants as fossils, 7 tetrapteroides, 357
Primary tissues, 23 Rhodea, 91, 92
wood, maturation of, 27 Rhodesia, fossil algae, 49

Telegram @tubscstds
Propalmophyllum laisinum, 170 Rkynia gwynne-vaughani, 32, 33
Prosseria grandis, 284 major, 34
INDEX 485

Ricciopsis florinii, 400 Seymour Island, 394

Root mantle, in Osmundites, 115 Shadbush, 203
in Psaronius, 100 Shansi, China, 421
Siberia, 45, 47, 181, 215, 260
Sabal palm, 185, 208 Siderella scotti, 414
Sagenopteris nilssoniana, 178 Sigillaria coriacea, 243
phillipsi, 177, 178 mammillaris, 243
Sahni, B., 16, 79 Sigillariostrobus, 245, 246
Sahnia nipaniensis, 350 Silurian, tracheid fragments of, 48
Salt River canyon flora, 409 plants of Australia, 41
Sanmiguelia lewisi, 170, 172 Smith, A. C., 174
Saportaea nervosa, 344 Snigirevskaya, N. S., 225
Sassendorfites benkerti, 170 Snowball, 195, 196
Schilderia adamanica, 432 Soapberry, 199
Schizaeaceae, 102 Sorus, use of term, 96
Schizoneura gondwanensis, 426 South Africa, 3, 160, 357, 410
Schizopodium davidi, 39, 46 South Dakota, 297
Schopf, J. M., 76 Spencerites insignis, 241, 249
Schopfiastrum decussatum, 138, 139 Sphaerostoma ovale, 371
Scleropteris illinoiensis, 79 Sphenobaiera furcata, 343
Scolecopteris incisifolia, 96, 98 Sphenophyllum, cones, 278-282
iowensis, 96, 97 constrictum, 276
Scotland, 32, 132, 140, 151, 156, 224, 236, cornutum, 277
237, 247 cuneifolium, 278
Scott, D. H., 79, 232 emarginatum, 277
Secondary wood, of Cladoxylon, 82 insigne, 276
internal, in Lyginopteris, 131 hauchecornei, 282
primitive nature of, in Schizopodium, leaf types, 277
46 plurifoliatum, 276
strong development of, in Medullosa speciosum, 426
gigas, 146 stem anatomy, 276, 277
in Zygopteris, 79 tenerrimum, 277
Seedlike fossils of the Devonian, 369, 370 thoni, 277
Seeds, attached to fernlike foliage, 150 Sphenopteridium, 91, 92
of Cardiocarpus with gametophyte, 328 Sphenopteris, 91, 92
of Cordaitales, 326-329 alata, with seed, 153
evolution of, 372-378 baumleri, dictyoxylon cortex, 135
of lycopods, 232, 252 hbeninghausi, 132
of pteridosperms, 140 tenuis, with seed, 153
with air chamber, 329 Sphenostrobus thompsonii, 282
Selaginella amesiana, 247 Spider in amber, 16
hallei, 247 Spitsbergen, 54, 61, 115, 258, 381, 386
Selaginellites canonbiensis, 247 Sporangial aggregate, in Biscalitheca, 81
crassicinctus, 247 in Botryopteris, 70
polaris, 247 Sporangiostrobus, 231
Senftenbergia, 10 Sporangium evolution in Schizaeaceae,
ophiodermatica, 106, 107 106
plumosa, 106, 107 Spore analysis of coal seams, 458
sturi, 106, 107 Spore morphology, 443
Sequoiadendron chaneyi, 209 Sporocarps, 53

Telegram @tubscstds
Seward, A. C., 108, 180, 384, 390 Sporogonites exuberans, 403, 405
486 INDEX

Sporophyll evolution in lycopods, articu¬ Telome concept, of leaf, 84

lates, ferns, 123 of integument, 372-375
Spruce, 195, 196, 203 Temperature, importance of minimum,
Stachannularia tuberculata, 273 204
Staff tree, 185 Tempskya knowltoni, 120
Staurastrum, 16 wesselii, 117-120
Stauropteris burntislandica, 74-77, 376 Tempskyaceae, 116
oldhamia, 77 Tertiary floras of Europe, 188
Stele, definition, 24 Tetrastichia bupatides, 156
evolution, in Medullosa, 148 Tetraxylopteris schmidtii, 68
in Osmundaceae, 116 Thamnopteris schlechtendalii, 114, 115
types, 24, 25 Thinnfeldia feistmanteli, 429
Steloxylon irvingense, 84 floras, 428
sanctae-crucis, 84 lancifolia, 429
Stenokoleos simplex, 413 talbragarensis, 430
Stenomischus athrous, 388 Thomas, H. H., 173, 176, 356
Stenopteris densifolia, 160, 161 “Thunder-eggs,” 19, 20
elongata, 429, 430 Time, measurement of, 2
spinifolia, 430 Tingia carbonica, 422
tripinnata, 430 Todea barbara, 113
Stephanophyllum solmsi, 344 Todites recurvatus, 388
Stephanospermum elongatum, 151 Transfer techniques, 470, 471
steward, 152 Trichopitys heteromorpha, 340-343
Sternberg, C. H., 184 Triletes, 252
Stewart, W. N., 144 Trimerophyton robustius, 43, 45
Stigmaria fi-coides, 237-239 Tschirkoviella sibirica, 420
Storgaardia spectabilis, 388 Tubicaulis scandens, 73
Stopes, M., 181 solenites, 73
Sturiella langeri, 304, 305 stewartii, 73
Stylites andicola, 253 Tulip tree, 196
gemmifera, 249, 253 Tyliosperma orbiculatum, 140, 371
Sugambrophyton pilgeri, 217
Sumac, 199 Ullmannia bronnii, 323, 325
Surange, K. R., 75 Ulodendron, 234
Svalbardia polymorpha, 61, 62 Umkomasia macleani, 160, 161
Swamy, B. G. L., 179 Urals, 115
Sweden, 173, 305, 359 Uranium, use of in time measurements, 2
Sweet gale, 180
Sweetgum, 197, 199 Vancouver Island, 186
Sycamore, 180, 185, 198, 204, 208 Vermejo flora, 197
Syngenesis, 87 Vessel elements, 174
Virgin’s bower, 195
Taeniocrada decheniana, 42 Vojnovskyales, 351
langi, 42, 43 Vojnovskya paradoxa, 351, 352
Taeniopteris latecostata, 422 Volcanic action, 11
nystroemii, large size, 423
vittata, 304 Walchia germanica, 325
Takhtajan, A., 182 Walchiostrobus, 325
Tehachapi flora, 208 Wales, 46, 153, 216

Telegram @tubscstds
Telangium affine, 156 Walnut, 185, 198, 199, 204
INDEX 487

Walton, John, 378 Wood in New Jersey clays, 15

Wegener, A., 417 Wyoming, 40, 104 109, 252
Weiser flora, 208
Weiss, C. E., 268, 272 Xenocladia medullosina, 83
Whittleseya, attached to Neuropteris foli¬ Xenotheca bertrandi, 369, 370
age, 150
elegans, 150 Yabeiella brackebushiana, 430
media, 150 Yellowstone Park, age of fossil flora, 202
Wieland, G. R., 12, 297 profile of fossil forests, 13
Wielandiella angustifolia, 305, 311 Yorkshire, England, 8, 106, 170, 176, 295,
Williamson, W. C., 243, 264 393
Williamsonia santalensis, 307
sewardiana, 308 Zamia floridana, 290, 291
spectabilis, 306, 307 muricata, 293
whitbiensis, 306, 307 Zelkova hesperia, 206
Williamsoniella coronata, 304 Zimmermann, W., 85, 122
Willow, 180, 197, 199 Zosterophyllum australianum, 44
Wilson, L. R., 460 fertile, 44
Windwardia crookallii, 344 rhenanum, 44
Wingnut, 195, 196 Zuberia zuberi, 161, 162
Wonnacottia crispa, 307 Zygopteris illinoiensis, 79
Wood anatomy, 23-28, 329-335 primaria, 79

Telegram @tubscstds
 -

Telegram @tubscstds
Telegram @tubscstds
- •

Telegram @tubscstds
Telegram @tubscstds
Telegram @tubscstds
Telegram @tubscstds
 Date Due
QCT 2 b‘St
KYl'O m
MY ? 4 ’R6
JAlg '67
MY 2 0 '6f s

JA 31 *69
FE 28 ‘69
-q l l’S9
OC 2 5 '69
no i rm
AP 16 7CL
MY 12 TO
JE l 1 ’87

Demco 293-5

Telegram @tubscstds
 MARYGROUE COLLEGE LIBRARY
Studies in paleobotany
561 An2

3 11E7 □□□3DT1L a

Telegram @tubscstds
Telegram @tubscstds