I

Digitized by the Internet Archive

in 2007 with funding from
Microsoft Corporation

http://www.archive.org/details/earlylatinverseOOIinduoft
EARLY LATIN VERSE
 OXFORD UNIVERSITY PRESS
London Edinburgh Glasgow Copenhagen
New York Toronto Melbourne Cape Town
Bombay Calcutta Madras Shanghai
HUMPHREY MILFORD
Publisher to the University
EARLY LATIN VERSE

BY

W. M. LINDSAY, F.B.A.
MEMBER OF THE DANISH ACADEMY

PROFESSOR OF HUMANITY AT THE UNIVERSITY OF ST. ANDREWS

OXFORD
AT THE CLARENDON PRESS

1922
NOV 21 1958

Printed in England
 TO THE MEMORY

OF

FRANZ SKUTSCH
A TRUE SCHOLAR

WHO WOULD GLADLY LEARN AND GLADLY TEACH

 PREFACE
The removal of the lava and cinders from Pompeii discovered
to us a town of the Early Empire. We saw the very room in
which a Roman had lived. Would that we could hear him speak !
We know his language — the significance of each word, the sound
of each letter. But words and letters are the dry bones of
a language. It is the tone of utterance that breathes life into
them. And that is what this volume claims to discover— Plautus',
Terence's (and presumably Cicero's) intonation of the sentence.
To disclose it the rubbish of half a century had to be cleared
away. Klotz's large book on Early Latin Verse gathered up all
the wisdom and — alas! — much of the folly of Ritschl's time.
And no more egregious folly than the * metrical ' theory of the
Brevis Brevians, that Plautus scanned ' apud me ' when the
metrical ictus happened to fall on ap-, ' ad Illos ' when the
metrical ictus happened not to fall on ad. Clear that rubbish
away, and you see that ' apud me ', ■ ad Illos ' go with emphasis
on the pronoun, while in 'apud me', 'ad illos' the pronoun is
a sentence-enclitic. This was pointed out many years ago in a
magazine-article (indeed Ritschl had given a hint of the same
kind). But on Klotz's layer of rubbish (itself embedded on
C. F. Muller's Early Latin Prosody, 1869) the volumes of the
Teubner edition have been superimposed, keeping it firmly in its
place. And so another /xiya Pl&\lov has to be inflicted on the
reader. Satan must cast out Satan.
The word ' rubbish ' is not too strong. What censure can be
too strong on people who imagine that the Infinitive of venio
could be put at the beginning of a trochaic line and (with ve-
1 under the ictus ') be pronounced ' venire ' (True. 504) ; who fail
to see that the difference between Plautus' prosody and Virgil's
viii PREFACE

must be (in the main) the same as that between any Comedy and
any Epic, namely that the lines of Comedy echo conversational
utterance, ' I'll ' for ' I will ', \ he's ' and ' we're ' for ' he is ' (or ' he
has ') and ' we are ' ?
This book will justify its length and its excessive care over
seemingly insignificant details * if it achieves any of these three
things :
(i) if it gathers up every ray of light in Plautus' pages on
spoken Latin and illuminates that dark background from which
emerge into literary Augustan Latin isolated specimens like male,
calefacio (and calfacio), nescioquis, ac, dein, sorsum, siquidem;
Virgil's viden, modo (Adverb), uno eodemque, abjete ; Horace's
puta, palus, mentiS, dixero, pituita, vietus, vindemjator ; the
monosyllabic cui(u)s of the poem to Messala in the Catalepta.
(2,) if it prepares the way for an adequate presentation of
Plautus' lines by editors. Editors nowadays present a Trochaic
Septenarius of the Poenulus (1207) in this fashion :
Nos fore invito domino nostro diebus paucis liberas.
All or most of their readers find this quantity of each syllable :
Nos for(e) invito domino nostro diebus paucis liberas,
and complain that Plautus' ' running ' line does not run ; in fact,
goes lame more than once. But what Plautus meant was :
Nos for(e) in|vito |domin5 |nostrd || diebuJ | paucis j llbe|ras.
That echoed the spoken sentence.

(3) if it vindicates Plautus' artistic skill. Klotz's unlucky

* metrical ' theory of the Brevis Brevians blinded him to the
scrupulous care shewn in the division of e. g. an Anapaest between
words in dialogue-metres. The Teubner editors' shortcomings in
the editing of Cantica are well known. If the Fayoum has more
to teach us about the New Comedy, delay might improve the
concluding paragraphs of chapter IV. But the reappearance of
Leo's Forschungen, with almost all the sins of the first edition
1 Readers may avoid these by skipping the small-print pages.
 PREFACE ix

unshriven, and some extraordinary articles in English and Irish

journals show the danger of delay.
Go then, large book. If you restore order and harmony among
lovers of Plautus who seem at present to follow each his own
path, the labour which has been spent on you will not have been
spent in vain.
W. M. LINDSAY.

St. Andrews,
St. Andrew's Day> 191 9.
 CONTENTS
PAGE
I. INTRODUCTORY : THE SATURNIAN METRE . i
II. PLAUTUS AND MENANDER . . . . .11

III. EARLY LATIN PROSODY : HIATUS . . .113

IV. EARLY LATIN METRES 260
APPENDIX.
A. Regard for Accent 317
B. Diaeresis 323
C. Hiatus 331
D. Final Monosyllable 339
E. Bibliography 359
INDEXES.
A. General Index t . .367
B. Index of Lines whose form is discussed . . .370
 INTRODUCTORY. THE SATURNIAN METRE

i. It is only Dramatic Verse which has been so well preserved that

we can detect the Early Latin usage. For early Epic Verse we have
quite inadequate material ; for other early verse, hardly any. A poem
like Ennius' Hedyphagetica would differ in technique from his Annals
much as Horace's Satires and Epistles differ from the Aeneid ; but we
have only a few lines of it. They are enough to whet our curiosity with
such an opening of a Dactylic Hexameter as meidnurum (Anapaest for
Dactyl) and such a close as glaucumque apud Cumas (Brevis Brevians
Law : apud Cumas, like calefio). Not until the time of Lucilius (c. 130 j
b. c, a generation later than Terence) do we get even as much material!
for investigating this homely type of Dactylic Hexameter as we have for
the other type, the Epic of Ennius. And by that time it must have been
influenced by the Epic type far more than at its origin.

2. The Hedyphagetica Fragment. The fragment of Ennius' adaptation of

Archestratus' poem, a manual of gastronomy for cooks and their masters, is
preserved in the famous eleventh-century Monte Cassino MS. of Apuleius
(who cites the lines in his Apologia, ch. 39 paucos versus memini, eos dicam).
In spite of the weakness of the lines* tradition, their interest justifies their
insertion here. Athenaeus has preserved a little of the Greek original.
Omnibus ut Clipea praestat mustela marina !
Mures sunt Aeni ; aspra ostrea plurima Abydi.
[mus] Mitylenae est pecten, caradrumque apud Ambraciaet finis.
Brundisii sargus bonust ; hunc magnus si erlt sume.
(Or si erit sum(e), with Synaphea.)
5 Apriculum piscem scito primum esse Tarenti.
Surrenti a(utem) elopem fac emas, glaucumque apud Cumas.
Notes :— I. Clipea 'from Clipea', Abl. of origin, a. Hiatus at pause. For the
syncopated Adjective cf. Virg. Aen. 2, 379 Improvisum aspris veluti qui sentibus
anguem. 2-3. Archestratus' lines were :
Tovs fivs Alvos 6X« ntyaKow, oarpeta 8' "AjSuSos.
. . . tovs 51 Krivas 17 MiTvk-qvrj'
nKeicrrovs 5' 'AfifipaKia iraptx*1'
3. Perhaps Ambraciai ended the line and the next line (not quoted in full by
Apuleius) began with Finis. Another suggestion is Ambraci' finis. Another,
Ambraciensis. 4. Brundisii. This is the correct Locative ending. Cf. Sunii (Ter.
Eun. 519). 6. The second word is doubtful. Tu is another suggestion.
2348 B
2 INTRODUCTORY

Quid scarum praeterii, cerebrum Iovi' paene supremi

(Nestoris ad patriam hie capitur magnusque bonusque),
Melanurum, turdum, merulamque umbramque marinam ?
10 Polypu' Corcyrae, calvaria pinguia (a)carnae,
Purpura, muriculi, mures, dulces quoque echini.
As the style of Cato's manual for farmers differs from the style of his
speeches, so would this unpretending verse show the diction of everyday life
(e. g. 6 fac emas ' be sure you buy ') and the pronunciation too (e. g. 2 aspra,
6 apud Cumas) ; while artificial (or archaic) ' literary ' diction and pronuncia-
tion (like our ' wind ' pronounced with the vowel of ' kind ') would be reserved
for the Annals.

3. The fragments of Ennius' Annals, however inadequate, are

numerous enough for a rough comparison of his treatment of Epic with
Virgil's. The difference between the first Roman imitation of this
Greek quantitative metre and the final form that it takes in Virgil's
hands is really not very great (IV 37). It is points of Prosody which
most obtrude themselves. Some syllables were pronounced differently
in Ennius' time : e. g. the last syllable of clamor had a long vowel (III 15)
and a final s after a short vowel was faintly uttered. Hence a line like
(Ann. 531):
Clamor ad caelum volvendu' per aethera vagit,
a line which at once strikes us as remote from Virgil. Now a language
is always in a state of unrest. Pronunciation changes from century to
century. This utterance of s did not go out of fashion till Cicero's
middle age and even that elegant poet, Catullus, has it in a poem which,
though printed at the end of our editions, was probably one of his first
pieces (116, 8):
At fixus nostris tu dabi' supplicium.
Virgil was merely a few years too late for it. And in the next few
years the short pronunciation of final o, that characteristic of Silver Age
poetry (along with pyrrhic cut for cut, etc.), crept into fashion. Silver Age
poetry differs from Virgil's, if Virgil's differs from the early Republican.
One does not find in these points of Prosody any great difficulty in
reading the fragments of Ennius and Lucilius.
4. Fragments of Republican Poetry. Since we shall have to cite
Ennius, Lucilius, and those Dramatists whose plays have not been preserved,
we had better assure ourselves of the worth of the tradition of their fragments.
There is a great difference between the citations which Nonius Marcellus makes
from his own copies of these authors and those which he merely takes over
from other grammarians. The former were accurate (unless Nonius' copy
was at fault), the latter often most inaccurate. Servius, like many of these
 THE SATURNIAN METRE 3

grammarians, is a very careless quoter. Not so Festus (or rather Verrius

Flaccus) ; but Paulus' Epitome of Festus cuts down many quotations most
absurdly. Festus usually cites (as we do in this book) a complete line (not
sentence). Editors of Fragments have too often ignored these points, and
some of the editions appeared before Onions' collation of the Nonius MSS.
was published in the small Teubner text of the Compendiosa Doctrina. So
that, although we make reference of each fragment to the standard editions,
our version of it wilt sometimes differ from what is printed there. These
standard editions are : Ribbeck's Tragicorum Romanorum Fragmenta ;
Comicorum Romanorum Fragmenta (small Teubner texts, 1897 ; 1898) ;
Vahlen's Ennianae Poesis Reliquiae (Teubner, 1903); Marx' C. Lucilii Car-
minum Reliquiae (Teubner, 1904). (Accius' non-dramatic fragments we/
refer to pp. 303 sqq. of L. Mueller's edition of Lucilius, Teubn. 1872.)
Mere fragments like these cannot provide good material for our investigation.
A consecutive passage of seventy lines would be better than seventy times
seven fragments. We must content ourselves with a very limited use of
them ; and the argument ' ex silentio ' must not be used by our readers.
If we do not cite examples of this or that Plautine usage from lost Tragedies
or Comedies, it is often because, although a fragment may be printed by one
editor as a complete line, others will make it a mere portion of a line, or
even the end of one line and the beginning of the next. And it is usually the
presence of a peculiar scansion (the very thing which would make the line
worth citing) which causes these editorial distortions. Evidence like this,
which can be so easily challenged, we are forced to ignore. Plautus must
be our piece de resistance ; then follows— longo sed proximus intervallo— the
unedited Terence ; the rest can claim only an occasional mention. The rude
verse of inscriptions (not to mention * sortes ') had better be left alone (see
Amer. Journ. Phil. 14, 147).

5. Why is it that one feels Plautus' verse to be something so remote

from the Virgilian type ? It is because we have not Dramatic poetry of
Virgil's time to compare with it. We have indeed fragments of Laberius
(time of Julius Caesar), but they are few and insufficient. And they are
not familiar as Virgil is familiar. We rarely read them, and, when we
do, we think of them as the imitation of an antique type, a late survival
of an obsolete technique. Quite unjustly. We might as well think of
Virgil's hexameter as a mere resuscitation of Ennius.
Since it has so happened that Plautus' (and Terence's) Comedies are
the only fully preserved verse of early Republican times, and that
Augustan Epic is fully preserved, but not Augustan Comedy, we most
unfairly compare Plautus' verse with Virgil's. We should compare
Plautus with Laberius1 or Pomponius (of Sulla's time), and reserve
1 Laberius, like Plautus, tried to make his lines echo talk, the slurred pronuncia-
tions, not the literary forms of words. Is that what he means by frag. 55 (versorum,
non numerorum numero studuimus) ?
B 2
4 INTRODUCTORY
Ennius for comparison with Virgil. If we did that, we should find that
the difference between Early Republican and Classical Comedy (or
Mime) was no greater than between Early Republican and Classical
Epic. Further, that the two types of poetry had developed in much the
same way. The pioneers in the imitation of these Greek quantitative
types followed their Greek models too closely j subsequent generations
of poets pruned away one roughness after another and adapted the metre
more and more to the requirements of the Latin language. But
Our remedies oft in ourselves do lie
Which we ascribe to heaven.

The greatest obstacle to our appreciation of Plautine verse is self-

imposed. We do not print a Comedy of Shakespeare in the same way
as the Paradise Lost. We spell ' i' faith ' in Shakespeare but ' in faith '
in Milton, ' we'll ' in Shakespeare but ' we will ' in Milton ; even
a Comedy of Aristophanes or Menander is printed in our editions
according to the colloquial pronunciation, e. g. eywSa (not eyw 018a),
Trova-Tif ovfios, kolvto?, and so on. But our editions of Plautus and
Terence present each word in the full form of deliberate utterance, with
very few exceptions (such as viden, bonust). We print eius, whether the
word is to be pronounced as a Trochee or as a monosyllable (eis), meos
whether an Iambus or a monosyllable (mebs). We give no hint when
the second syllable of voluptatem is to be read as a long, and when as
a short syllable. All this must be changed if Plautus and Terence are
to find appreciative readers.
6. The Text of Plautus and Terence. The MSS. of Terence have not
yet been all collated ; at least, collations have not been published. And for
a critical edition there is as yet nothing better than Umpfenbach's (shall we
say?) pre -scientific volume, a product of the time of Ritschl, which is far
from presenting to us even the rival readings of the (at least two) ancient
editions. Much less can it enable us to answer the all-important question :
« How far has the text been left as Terence wrote it ? How far has it been
modernized or otherwise altered * by Rhetoric Professors, to suit the require-
ments of the Rhetoric Class?' This uncertainty should not be exaggerated.
Nor need it be much of an obstacle, since Plautus, as the older and more
voluminous writer, must take the leading place in our investigation. Until
better advised, we may think of the Bembine codex (A) as representative of
the purer text. And with A we must include those old corrections in the
Bembine MS., which Umpfenbach absurdly assigned to mediaeval times. But
until the history of the text's tradition has been elucidated, until collations of

1 On Andr. 653 (cum patre altercasti) Donatus remarks : Legitur et c altercatus es ',
non enim ' alterco ' dicimus.
 THE SATURNIAN METRE 5

all the MSS. have been published, until Donatus' commentary has been
utilized, we had better leave Terence, as far as possible, out of our main
discussions and merely indicate the extent of his departure from the
Plautine usage.
Two ancient editions of Plautus' twenty-one plays, the plays declared by
Varro to be universally accepted as genuine, are represented by the Ambrosian
Palimpsest (A) on the one hand, and the minuscule MSS. (P) on the other.
Of one of the best of the minuscule MSS., the (lost) Codex Turnebi, a partial
collation was found at the end of last century (too late to be utilized for the
Teubner or Weidmann editions). It has improved the text of some of the
plays (especially the Rudens) and revealed many features of the second
ancient edition, its arrangement of the Cantica, its designation of Scenes as
D(e)V(erbia) or C(antica), and so forth. A more recent find, N, an
uncial leaf of the Cistellaria (' Palatine* edition) is of little help. Opinions
are divided regarding these two ancient texts, the ■ Ambrosian ' and the
1 Palatine ' editions as they are usually called. At least we may claim for
the consensus A P a greater probability than for the mere evidence of the
minuscule MSS. (especially when the Codex Turnebi is not available). And
the additional testimony of an ancient Grammarian gives weight to a reading,
so that a version marked by the symbol ' {A P) ' or ' (P Gram.) ' should not
be lightly departed from.
Leo's wild but cleverly presented theory, that all copies of Plautus had
been lost until Valerius Probus found one of this play, one (or more) of that
at Beyrout, is now apparently falling into discredit. Leo pushed it so far as
to suggest that our two ancient editions came from no better source than
miserably defective material rescued from the scrap-heap by Probus and
tinkered into one shape by the ' Ambrosian ', into another by the ' Palatine '
editor. Clearly an extreme application of such a theory would be absurd.
We feel that every page of our Plautus bears Plautus' impress, ' Eapse
cantat quoia sit '. But in any form it is liable to weaken unduly our confidence
in the texc's transmission. Leo was too pessimistic. On the other hand his
demonstration of the genuineness of the Prologues was helpful ; and he did
well to remind us that the text established by the labours of Probus is likely
to have been the text on which our two editions are based ; also that
Grammarians were sometimes misled by an inaccuracy in the copy they used.
Yet in nine cases out of ten, where they attest a reading not found in our
MSS., we may believe our MSS. to be at fault, especially if it be an early
spelling, gender, case- or tense-form of which the later equivalent appears
in our MSS. For every scribe (or corrector) was prone to take these for
mere miswritings in the exemplar and alter, e. g. terrai (Gen.) to terrae,
collus (Ma.sc. ; the Indo-European gender, e.g. Germ, der Hals) to collum.
Even in the Palimpsest we find at Stich. 471 illi (Adverb) corrected to illic
against the metre; and the same practice went on all through the Middle
Ages. We need not hesitate to save the metre by substituting terrai ', etc., for
terrae, etc., illic (Dat.), illunc, etc., for illi> ilium, etc. (A'' shows illanc for
the illam of our minuscule MSS. in Cist. 126), med, etc., for me, etc. (Ill 31).
6 INTRODUCTORY
Scribes were, of course, ignorant of Dramatic Metre, so that these alterations
of theirs are (usually) easily detected. And although this modernizing of
Plautus' spelling was so universal, it was done by ignorant scribes athaphazard,
not systematically by learned editors ; and traces survive every here and there
of the older forms : e. g. med before esse was sometimes mistaken for me
desse (deesse) and not changed to me esse ; quoius erat might be mistaken
for quo iusserat and escape the change to cuius erat.
Our two editions seem to represent (ultimately) the editorial work of Aulus
Gellius' time, a work based on Valerius Probus' revival of antique literature.
How scrupulous these editors were to preserve each jot and tittle of the
ancient text we see from Gellius' stories of learned controversy over the old
Latin Genitive fades (settled by an appeal to a copy of Claudius Quadrigarius
in the library of Tibur) or Imperative insece (settled by a ' liber verae
vetustatis ', a copy of Livius' Odyssey in the Patrae Library), etc.
The only editorial re-casting would be at the time of Terence (and later),
when Plautus' * antiquae comoediae ' were re-staged. The ' Revival ' version
is usually, or at least often, preserved by our MSS. along with the Plautine,
e.g. Pseud. 523 Agedum, nam sati' libenter te ausculto loqui (Revival),
Studeo hercle audire, nam ted ausculto libens (Plautine). Even where it has
ousted the Plautine cadence or case-form or prosody, it gives us the cadence
and case-form and prosody of Terence's age (or thereabouts) and is quite
suitable material for our investigation here. We find no justification for the
startling theories advanced from time to time, that this or that usage of
Plautus has been wholly effaced by editorial re-casting. Even in the case of
Terence, a school-author, where re-casting for school-purposes is more
conceivable, we find no cogent reason for believing this. To a scientific
student of the Latin language the diction of the traditional text of Terence
seems suitable for the generation after Plautus and before Lucilius (III 40
A E). It cannot possibly be put on a level with anything even remotely
resembling Dryden's version of Chaucer.
Ritschl collected the earlier inscriptions in order to learn what kind of
language Plautus would be likely to use. He was a pioneer in this investiga-
tion and, naturally enough, made mistakes. These have been corrected by
subsequent investigators (Brugmann and his school), who had not merely
the inscriptions collected by Ritschl but also subsequent discoveries to help
them and who applied the principles of the new science of Comparative
Philology (III 2). "What has hindered the true perception of Early Latin
Prosody has been the neglect of linguistic study.
If any one ventured to edit Chaucer without taking the trouble of learning
the phonetics and history of the English language, and of reading what had
been discovered by the labours of English philologists, we should call him
mad. Unluckily there are plenty of editors of Plautus who have not (and
do not pretend to have) even a bowing acquaintance with Latin Linguistics.
Leo's strange theories of Hiatus, of Elision of -s before an initial vowel, of
the faint utterance before an initial consonant of -m (e. g. ' factum volo '
Aul. 146), and of -d after a short vowel (e.g. 'quod tibi' Stich. 21), would
 THE SATURNIAN METRE 7

never have been printed if he had read Brugmann's work on Comparative

Philology or any manual of Latin Linguistics later than Ritschl's time.
(Buecheler's early work, the Lat.-Deklination, is apparently his sheet-
anchor.) Ritschl had not recognized the Law of Breves Breviantes (a
discovery of C. F. Miiller's), and had absurdly connected Plautus' scansion
ferunt_J\\ke Virgil's viden) with such provincial patois as dedrot and dedro
(for dederunt) on Pisaurum inscriptions (C. I. L. 1, 173, and 177). So he
pictured a decaying language in Plautus' time that was rescued from decay
by Ennius, an untrue picture that is too often reflected in careless school-
manuals. His pupils, ignoring the scientific study of Latin, and forgetting
that no language changes so thoroughly in a single generation (Plautus died
Jin 184; Terence produced the Andria in 166), represent Plautine diction as
/far more archaic than it can possibly have been. Indeed Ribbeck's version
of the fragments of all the early Dramatists in the large Teubner edition often
reminds us of Moth, the antiquary, in Cartwright's Ordinary :
Potluck. Where's your dwelling?
Moth. Yclose by Aldersgate there dwelleth one
Wights clepen Robert Moth, etc.
It is true that Livius Andronicus is cited by the Grammarians for e.g.
dusmo (=dumo), that Naevius used Troiad (Abl.) in his Saturnian Epic
(Bell. Poen. 7) :
Noctu Troiad exibant || capitibus opertis (Troia de MSS.),
but these are Archaisms of Epic (not of Comedy) like the o/ti, aulai of
Virgil's Epic.
To get at what Plautus actually wrote, we have, as a rule, merely to rid
the text of the tangle and barnacles that scribes have attached to this
Glaucus. Now there ate some 21,000 lines of Plautus, and for the greater
number of these we lack the testimony of the Palimpsest, while in a smaller
portion (not merely the Truculentus, but also the Mercator, the second half
of the Miles, etc.) the minuscule testimony is not at its best. Scribes
continually make mistakes and, as Cobet says, ' solent ra avra n-ept ra avra
&fiaprap*iv*. Any theory under the sun, however impossible, could find
support from an appreciable number of such mistakes. That is a danger
we must keep in view when theory after theory comes up in the mushroom-
growth of dissertations. We must ask, ' Is there sufficient support (' A P ' or
1 P Gram.' ?) for the readings on which the theory is based ? Do the readings
admit a more likely explanation ? And is the theory itself probable a priori ? '
And we shall find occasionally, even when this or that usage of metre or
prosody seems reasonable, great difficulty in deciding whether the fewness
of instances is to be referred to its actual rarity or to the mere mistakes of
scribes (e.g. a divided Tribrach like Men. 236 Mare sUptnxm omne
Graeciamque exoticam ; II 45)'

7. Even the pioneers in Latin quantitative verse were no slavish

<
mitators of the Greek. Ennius' Epic is not merely the Homeric
 8 INTRODUCTORY

hexameter in Latin. No more is Plautus' Senarius merely Menander's

Trimeter. Most1 of our second chapter will be concerned with this
difference of type. Wherein precisely does Plautus (we use his name
merely to represent the pioneers in Latin Dramatic verse) differ from
Menander (representative of the New Comedy) ? Why does he avoid
some of Menander's cadences? Why does he ignore some of
Menander's restrictions, while he imposes on himself others which
Menander hardly recognizes ?
It will be easier to point out the difference of technique than to
account for it. For behind Livius and Naevius and Plautus there is
a background of native Latin poetry (not imitative of Greek), of which
so little has been preserved that its true nature is unknown. The
Saturnian Metre has evoked almost as many monographs and magazine-
articles as there are extant lines, but even yet no agreement has been
reached on the most elementary point of all. Was it Accentual (like
most old European metres) ? Or was it Quantitative ? That some
regard for Quantity may be shown in Saturnian lines of Livius and
Naevius proves nothing. In our own country, after the French type of
verse had been introduced, the old native type assimilated itself more or
, less to the new pattern. Alliterative verse lasted for some time after the
importation of rhyming verse. The ' Canterbury Tales ' are in the new
metre, the ' Vision of Piers Plowman ' in the old. Some poets wrote a
mixture of the two, either adding an alliterative element to the new or
a rhyming element to the old. (For details see Saintsbury's ' History
of English Prosody '.)
And can we be certain that Greek Dramatic verse had not been
imitated in Latin before Livius' officially recognized adaptations of a
Greek Tragedy and a Greek Comedy in the thanksgiving service after
the close of the First Punic War ? Did the popularity of the Trochaic
Septenarius go no further back than his time ? Who can tell ?
We feel that Saturnian Verse must have affected the verse of Plautus
and Ennius. How could, let us say, Naevius, who wrote the native
metre at one hour and the metre of Menander or Euripides at another,
avoid a fusion of the two types ? Not to put it so crudely, a poetical
diction for Rome had been formed during successive generations by
Saturnian writers : it was this diction, these phrases, these turns of
speech (e. g. contemptim conterere ; see below) that Plautus and Ennius
transferred to Greek moulds. It is natural to seek in Saturnian Verse
the reason for this and that departure of Plautus from Menander's type

1 The rest will deal with the contrast between Plautus' and Virgil's usage.
 THE SATURNIAN METRE 9

or of Ennius from Homer's type ; and attempts are always being made
to bridge the gulf. These attempts we had better ignore until some-
thing like a firm foundation has been secured on the Saturnian side.
At any rate, they may be relegated to small type.
8. The Saturnian Metre. The Roman Grammarians (or rather the
writers on Latin Grammar, for they are as often Greek as Roman) declared
the Saturnian to be an Iambic Dimeter Catalectic with a Trochaic Dimeter
1 Brachycatalectic * :
W — W— |^ — — || — W — \^|— — .

That they should identify it with some Greek metre or other was inevitable,
since Latin was believed to be a mere dialect of Greek, and manuals of Latin
Prosody were written by the rough-and-ready plan of turning Greek rules
into Latin. The Greek rule that a final short syllable was lengthened before
the rough breathing (representing a lost Digamma or <rf) in Homer (e. g.
'Apvvfifvos tfv re ^vx'!**) was turned into a Latin rule for which an example
was with difficulty hooked out of Virgil (Terga fatigamus hasta. Really a
mere lengthening ' in arsi '). And this wrong-headed rule actually became
the practice of Christian poetry (e.g. per hominem, Vlr humilis). (For
details see Class. Quart. 10, 97 ; Amer. Journ. Phil. 37, 39.)
Before Comparative Philology became a science, this explanation of
Saturnians was accepted. They were scanned quantitatively on the
amateur's principle that * consonants mattered little and vowels nothing
at all'. Any quantity whatever could be supposed to be the * ancient'
quantity of a syllable : e. g. Runcus atque Purpureus, || fiiii Terras ; Quamde
mare saevom ; || vis et (viret MSS.) cui sunt magnae.
When the impossibility of these scansions became patent, various Accentual
theories held the field, e.g. of two types, (A) Dabunt malum Metelli || Naevio
poetae, (B) Prim(a) incedit Cereris || Proserpina piier (see Amer. Journ. Phil.
14, 139, and 305 ; Engl. Philol. Soc. Trans., 1894).
A primitive Greek type of metre, the ' Viersilbler', discovered by Wila-
mowitz — four syllables which took now the guise of an Iambic Dipody, now
of a Choriambus, or again assumed other Ionic or even Trochaic shape —
gave the Quantitative theory another chance ; and Leo's monograph (Der
Saturnische Vers, 1905) is the last (authoritative) word on the subject.1
The Accentualists content themselves with the criticism that the 'Viersilbler '
is a skeleton key which will in heedless hands unlock any metre, even (with
a little oiling) Anna virum|quS cano Tro|iae qui prl|mus ab oris; that all <& t<&~£~
Leo's cleverness cannot furnish a satisfactory quantitative scheme of
Saturnians ; that much of the scaffolding on which his theory rests is
worthless wood. And they await with hope the result of the present
investigation of Irish metre, that is to say, the old, native Irish metre (not
the later imitation of Latin Hymns, like Genair Patraicc in Nemthur || is ed
1 A pamphlet, Zander • Versus Saturnii ', Lund, 1918, has just reached us. See
App. E.
io INTRODUCTORY

atfet hiscelaib. The appeal to such a type is unjustified). One recent

discovery in this field (Berlin. Acad. Abhandlungen, 1913, Phil.-hist. Classe,
no. 6 ; K. Meyer ' Ueber die aelteste irische Dichtung') makes a character-
istic of one type a 'binding' of the beginning of the second hemistich with
the end of the first hemistich by Alliteration, e. g. reemse rig Temro || Aiatha
for slicht slogdai. This recalls Saturnian lines like :
Superbiter contemptim || content, legiones ;
Summas opes qui regum \\ reg\a.s refregit.
The mysterious disyllable Benval^ sole relic of the Pictish language, gave
Sir Arthur Wardour and Mr. Oldbuck a quite delightful freedom of discussion,
since its first syllable (apparently Celtic) and its second (seemingly Teutonic)
afforded equally slender support to their rival theories. The ' ignis fatuus '
of Saturnian Metre will, if we follow it as Klotz has done, lead us into a bog,
if anywhere. And yet there is a fascination in such speculation. How
would the pioneers in quantitative verse be influenced by the older Saturnian
practices? What divergences of Plautus from Menander can be assigned
to such influence? The best case has been made out for Alliteration. It
belongs to the essence of Saturnian, as of other European accentual verse,
and is a characteristic feature of Dramatic (and other) Latin poetry ; e. g.
Poen. 537 (Troch.) || ne nos tarn contemptim conteras. Can we find in the
same quarter an explanation of the strong Diaeresis (with Hiatus or Syllaba
Anceps or even both) in Latin Septenarii, Octonarii, etc. (II 59), e.g. Mil.
1226 (Iamb.), 1239 (Iamb.)?:
Namque edepol vix futt copia || adeundi atque impetrandi (with both),
Si pol me nolet ducere || uxorem, genua amplectar (ditto),
a thing unparalleled in the Greek dialogue-metres. If (we emphasize
the Particle), if the Romans (accustomed to a Dabunt malum Metelli, to a
Virum mihi Camena) visualized the long line of dialogue as two short lines
(Dimeters) rather than one long line, these licences of the Diaeresis are
explained (II 59). Perhaps also Plautus' love for the Versus Reizianus,
i.e. an Iambic Dimeter followed by the Colon Reizianum (IV 15), e.g.
Aul. 415 Redi. quo fugis nunc? tene, tene. || Quid stolide, clamas?, 416 Quia
ad tres viros iam ego deferam || nomen tuum. Quam 6b rem ? (The whole
Scene is composed in this metre.) And there is another straw for our
drowning theory to clutch at. For there are traces, although they are too
few and too faint to be called anything more than possible, barely possible,
of some Diaeresis after the fourth foot of the Senarius (see III 48). If
these traces were to be taken seriously, they would suggest that the Trimeter
was at first visualized as a Dimeter + Monometer.
But— quo, penna, tendis ? This book concerns itself with statistics and
solid facts, and must abjure airy speculations.
 II
PLAUTUS AND MENANDER

i. Renouncing the attempt to guess at Saturnian influence, we may

concentrate our attention on the actual facts. How exactly did Plautus
alter the Greek verse of Menander in his imitation of it? We say
' Menander \ For luckily we are now in a position to estimate
Menander's technique. We are not forced, as were previous writers
on this subject, to compare Plautus with Aristophanes; an unfair
comparison, since it was the New Comedy which Plautus adapted and
not the Old.
Some 1,300 complete Iambic Trimeters are preserved in MSS. of Menander
(edited by A. Korte, Teubner, 19 10), an equivalent number to the Senarii in
three of the best preserved plays of Plautus, the Persa, Poenulus, and Stichus.
These three will be a convenient unit of comparison when a wider survey is
unnecessary. The Menander citations in Stobaeus, etc., we had better keep
apart. For they consist mainly of ypa>nai, and these from their nature exhibit
a more deliberate rhythm than is really typical of Comedy. We should get
a wrong idea of Plautus' lively Senarii if all that we knew of them was
a collection of ' sententiae ' like Trin. 490 sqq. :
Di divites sunt, deos decent opulentiae (Or debs)
Et factiones, verum nos homunculi,
Salillum animai qui quom extemplo amisimus,
Aequo mendicus atque ille opulentissimus
Censetur censu ad Accheruntem mortuus.
And since the 'sententiae' of Plautus are not distinguishable from those of
Terence, hardly indeed from the collection ascribed to Publilius Syrus, we
cannot expect to find in the yvfopai of, let us say, Diphilus any indication of
the peculiarities which there must have been in his dialogue-verse as con-
trasted with Menander's.
2. SPONDEES NOT EXCLUDED BY DIPODY LAW. The first
great and patent difference is that Plautus allows a Spondee in the
1 even ' feet of an Iambic and the ' odd ' of a Trochaic line. No Greek
writer of Iambic or Trochaic Verse ever allows that. (Of course the
Iambic and Trochaic Scazon are a different thing.) The few alleged
Greek instances are illusory and give no reason for believing this licence
to have appeared in Greek Dramatic Verse of Plautus' own time. (See
Hardie, ' Res Metrica', p. 82.) ■
1 Hardie in other passages of his book says (p. 82, pp. 88-9) : ■ Did the native
Saturnian verse set an example of indifference to the quantity of the syllable on
12 PLAUTUS AND MENANDER

Of late much has been said of Latin Literature as the last bloom, the
St. Martin's Summer, of Greek. Greek Epic, Greek Comedy had disap-
peared in Hellas, to emerge — like the fabled river — on Italian soil. These
picturesque phrases should not lead our judgement astray. Not every feature
of Roman Comedy may be claimed for Greek. True ! much that distin-
guishes Plautus' and Terence's from Aristophanes' handling of dialogue-
verse may have already appeared in Diphilus and Menander. When the
Old Greek Comedy was replaced by the New Comedy of character and
intrigue, the verse would probably be modified to catch better the tone of
every-day talk. And it may be doubted whether the Teubner editor is wise
in removing from Menander's text an example of Hiatus at a pause in the
sentence (see III 48), since such a pause seems the justification for a few
instances of Hiatus in the older Greek Drama. But it would be absurd to
say that every Hiatus allowed by Plautus must have been a Greek type.
As absurd as to say that Plautus made agri an Iambus because Menander
so scanned dypoi. Clearly Plautus had to echo not Greek but Roman talk.
He scanned agri because that was his own pronunciation and the pronuncia-
tion he heard on the lips of every one. He made qui am ant an Anapaest
for the same reason. He was not merely seeking the nearest Roman
equivalent for Menander's n olv, an Iambus, any more than he sought
a Roman equivalent for Menander's olhl ep, a Dactyl. To say that Ennius'
agri (Spondee) and (apparent) lengthening 'in arsi' were Graecisms would
not be so absurd (whether correct or not), since Epic diction is no echo
of every-day conversation.
And — to return to our subject — we need not infer that Plautus' admission
of Spondees to the ' even ' feet of his Senarii must have been a Greek
practice of the time. For one thing, the Greeks had a sharper ear for
metrical niceties. For another, they had centuries of tradition behind
them. Plautus was a pioneer, bound by no convention. He had new
material, Roman talk, to pour into the Greek mould. His material must of
itself have influenced the shape taken by his adaptation of Greek metre.
For example, the number of short final syllables ending in a consonant was
limited in his time (III 8). There were twice as many a century later.
3. This outstanding difference has been used as an argument for the
inferiority of Latin Dramatic Verse, for a gross insensibility of the
which the ictus did not fall? Did the incidence of the Latin accent — always on the
penultimate syllable if that syllable is long— accidentally or Kara avfiPtfijjKos aid
and enforce the iambic rhythm, so that less attention to quantity was necessary ?
In a line like
labdrans, quaerens, parcens, illi serviens
the coincidence is obvious. . . . The Latin accent was not strong enough stress to
enable a short syllable to take the place of a long, but when two or three longs
came together it differentiated them. Thus an iambic effect could be obtained
without the quantitative strictness of the Greek trimeter. In shaping the verse as
they did Andronicus and Naevius undoubtedly created something which suited the
genius of the Latin language.'
 PLAUTUS AND MENANDER 13

Roman ear to the finer points of metre. Perhaps the Roman Metricians
of the Empire judged more truly :
In metra peccant arte, non inscitia (Ter. Maur. 2237).
Plautus aims at reproducing in his lines the actual tones of everyday
(educated) speech. By admitting these Spondees he breaks down a
most awkward barrier. Does he do this at too great a sacrifice ? Does
he sacrifice the music of the verse ? Not to our ears at least. For how
can we feel a Spondee to be so much farther removed from an Iambus
than an Anapaest? Apart from mere convention, mere tradition, we
nowadays hardly see what there was to prevent a Greek from admitting
*0 ZeO, 8tLv6v to xprj/jia, etc., as readily as*12 Zev /?a<riA.ev, to xpypQ-i e^c.,
in the opening line of the Clouds. Our ears recognize the true iambic
cadence of the Dipody u> Zev 7raT€p, but hardly find in w ZeG Seivov any
greater departure from it than in w Zev /Jao-iAcv.
4. Senarius for Trimeter. And yet, though our ears may not detect the
difference, there was a difference between & Zev paatieu and &> Zev beivov.
The Greek Comedians could justify the one departure (a> Zev 0<io-i\eO) from
the ' staccato ' form of the « metrum * in Tragedy, 3> Zev Trarep, but not the
other (&> Zev deivov). The Anapaest which they admitted was not quite the
same thing as the Anapaest of Anapaestic Verse. It was bound by various
restrictions (IV 4) which did not bind the other.
Now these restrictions of the Iambic Anapaest are observed by Plautus as
carefully as by Menander, indeed more carefully than by Aristophanes
(48-51). So we find a difficulty in unqualified acceptance of the statement
that the Romans threw overboard the Dipody Law and scanned not by
' metra ' but by feet ; that the Trimeter (three ' metra ') :
*Q Zev (SaaiXev, \ to XPWa r^v I VVKr^>v ocroy,
became a Senarius (six feet) :
Sequer(e) hac | me, gnat(a) |, ut mu|nu' fun|garis | tuum.
The remark of a Metrician of the Empire: Hie Latine Senarius, quod pedes
sex simplices habeat, Graece Trimeter, quod tres o-v(vyias habeat, appellatur
(Gram. Lat. vi, 286, 14), does not imply that the change of name indicated
a change of nature. Ennius gave to his Epic Hexameters the name of
'versus longi' (Cic. Leg. 2, 68), but he scanned the Latin lines precisely as
Homer scanned the Greek.
5. At any rate, if this departure from Greek usage had been ill-judged,
if it were unsuitable for Latin Verse, subsequent writers (e. g. Terence)
would have either discarded it altogether or reduced it within limits.
But, on the contrary, we find it as freely used in the Classical Age (and
later) as in the time of the first pioneers. To any reader unfamiliar
with Latin Dramatic poetry the following passage of Phaedrus, who
follows Virgilian rules of prosody, will best show that these Spondees
do not prevent the easy flow of Iambic verse (preface to Book I) :
i4 PLAUTUS AND MENANDER

Aesopus auctor quam materiam repperit,

Hanc ego polivi versibus senariis.
Duplex libelli dos est, quod risum movet
Et quod prudentis vitam consilio monet.
Calumniari siquis autem voluerit
Quod arbores loquantur, non tantum ferae,
Fictis iocari nos meminerit fabulis.
6. Thus at the very first introduction of this Greek metre to Roman
ears, Roman independence is asserted ; an awkward barrier, the work
(as appears to us nowadays) of mere custom, is broken down. Dramatic
verse is made more capable of echoing ordinary conversation.
Still Plautus observes some restrictions in his new-found liberty.
While he makes free use in the 'odd' feet in iambic lines (where
Spondees were allowed by Greek usage) of Spondees whose metrical
ictus clashes with the accent of ordinary pronunciation (especially in the
fifth, next in the first foot), this clash does not appear in the two ' even '
feet to which he admits Spondees against Greek usage (especially in the
second, next in the fourth). After this restriction was pointed out
(e. g. by Klotz), it was caught up too eagerly by students of Plautine
Metre and made too much of. Any one who reads the plays for
himself to see how it actually works will find that there is in reality
comparatively little clash of ictus and accent in Plautus' Spondees.
(Could one imagine any other arrangement of the words in Aul. 480 ut
indotatas ducant uxores domum ?) In Trochaic verse the harmony is
so easily attained that it escapes our notice. In Iambic verse Plautus
is so bent on securing a Spondee for the penultimate foot (the fifth of
the Senarius) that, when the line ends with an iambic word, he is quite
ready to use a Spondee that (apparently) defies the accent : e. g. in the
first fifteen Senarii of the Miles, five examples : 2 sudum est (sudumst)
solent; 3 conserta manu; 5 conso/an volo; 9 propter virum ; 15
Neptuni nepos. And we have no difficulty in finding spondee-words
in the first foot of iambic lines, a position which brings ictus into clash
with accent : e. g. Mil. 33 Venter creat omnes has aerumnas : auribus ;
140 Nam unum conclave concubinae quod dedit ; 367 || dixit. Dixi
hercle vero, etc., etc. (We assume that Istuc in line 19 was accented
on the final.) But in the second, third, and fourth feet of the Senarius
there is a remarkable unison of ictus and accent. We have no right to
claim it more for the two ■ even ' feet, the second and fourth, than for
the third, an 'odd' foot. Indeed we have no difficulty in finding
examples in the plays of a cadence which Klotz's rule, strictly enforced,
would make impossible, e.g. Trin. 410 :
 PLAUTUS AND MENANDER 15

Quam si tu obicias (ormia's papaverem (A P Gram.),

though the Molossus is not nearly so common as the Cretic in this
place. The unison of accent and ictus in the interior portion, the
heart of the Senarius, seems to be secured by Caesura, which, as we
shall see (41), pervades the Senarius far more than the Trimeter. A line
like Mil. 1095 *s harsh and abnormal:
De concubina, nam nullo pacto potest.
But what precisely is at fault? Is it the clash of accent and ictus in a
fourth foot Spondee {nullo) ? Is it the absence of all Caesura in the
last half? Or is it the sequence of two spondee- words (i. e. a repeated
clash of accent and ictus, nullo pacto) ? For Plautus does seem to take
more pains with his Spondees than with Iambi, etc., in avoiding clash.
It is strange to find even in an Iambic Octonarius, a Canticum-metre,
one instance (Trin. 285 Turbant, miscent mores mali ||) of three
Spondees like the three Iambi of Pers. 410 (Senar.) :
Procax, rapax, trahax : trecentis versibus
Tuas impuritias traloqui nemo potest, 't
where the strident, caesura-less, rhyming trio suits the sense very well.
In Pers. 60 (Senar.) it is rather the sense than the rhythm that demands
the emendation :
Neque his cognomentum erat Duricapitonibus (duris Capitoni-
bus MSS.),
Another restriction (applied to line-endings) may be mentioned here,
but only mentioned, since its proper place is in Chapter IV. (For
details see Appendix D.) When a line ends in a (single) monosyllable
the preceding foot is always 'pure', that is to say an Iambus in an
iambic line, and so on ; e. g. Asin. 713 (Iamb.) :
Atque ut deo mi hie immolas || bovem ; nam ego tibi Salus sum.
7. Molossus (or Spondee with clash of ictus and accent) before third
Dipody of Senarius l :
(a) Molossus:
Amph. 42 Virtutem.Victoriam ; JL . coloniam;
Aul. 576 comrhutet /
100 praefectdst legionibus ; Capt. 91 nonnullum periculumst
103 uxorem fecit suam ; (Or non nullum) ;
490 celetur consuetio ; 93 Aetoli cum Aleis ;
Aul. 86 Dareum, trivenefica 192 subducam ratiunculam ;
(Rather Dareum, like platea) ; Cas. pr. 59 consentit cum 'filio ;
95 pistillum, mortarium ; 447 protollo mortem mihi ;
375 porcinam : cara omnia ; Cist. 616 uxorem duxit domum ;

1 This paragraph (first half) and No. 16 we owe to Mr. F. W. Hall of St. John's
College, Oxford. [See now his article in Class. Quart, of 1921.]
i6 PLAUTUS AND MENANDER

Cist. 630 redducam Selenium ; Poen. 959 monstratust regionibus ;

Cure. 371 subduxi ratiunculam ; 1 34 1 addicar Agorastocli ;
380 quaesivit pecuniam ; Pseud. 456 vicinos inpertio ;
Men. 102 concinnat patinarias ; (544 inter nos convenimus) ;
330 Volcani Violentiam ; 88^ conditum gustaverit ;
702 argento concredidi ; Rud. pr. 14 abiurant pecuniam ;
Merc. 6 humanas querimonias ; 89 abduxi negotiis ;
10 Mercator Macci Titi ; 91 lenonem prehendere ;
Mil. 502 virgarum de te datur ; 461 extraxi praefiscine ;
Most. 656 perturbat paenissume ; 1 195 speravi neque credidi (?) ;
Pers. 160 aediles locaverunt ; Stich. 194 praeconis compendium ;
Poen. 7 fecistis sapientius ; Trin. 163 illius sapientiam (cf. 35) ;
28 infantes minutulos ; 410 formicis papaverem ;
(480 periuras. quid postea ?) 426* pronuper tu exactus es ;
(Rather peieras) ; True. 64 scortorum plus est fere ;
633 tantumdemst periculum ; 658 urbanos amasios (?).
mia;
(b) Spondee (word or ending) :
Amph. pr. 134 ilium censet virum Men. 542 duum nummum stalag-
(Rather illunc) ;
Asin. 800 male dicat. si dixerit ; Merc. pr. 48 illi(u)s augerier (35) ;
Aul. 76 neque quicquam meliust mihi 328 ad portum negotium ;
(On 565, see below) ; 752 quos inter iude datu ;
Bacch. 277 in navem conscendimus ; 796 uxor acerrimastx ; 's
(599 dicam illi periculo) Mil. 820 stertas, quasi sorbeas ;
(Rather illlc. Cf. Bacch. 1018); 857 in cella vinaria ;
856 ilium inventurum te quali' 1095 nullo pacto potest ;
sit (Accent on -rum ?) ; 1 135 exoptabam me maxime
Capt. pr. 1 5 ope vestra censerier ; (Accent on -bam ?)
pr. 40 docte fallaciam ; Most. 595 non debet ? Ne ypv quidem
(8);
85 parasiti venatici (cf.86);
91 (see above) ; Pers. 60 (see above) ;
Cas. 328 odiosast amatio ; 408 inlex, labes popli ;
426 voliiptatem creaverim ; Poen. 155 meretricem maiusculam ;
781 ego ruri cenavero ; 498 meretricem minusculum ;
Cist. pr. 171 meretrici Melaenidi 643 ita vestrast benignitas ;
(cf. 575);
650 tit ad portum processimus ;
Cure. 275 missust in Cariam ; Rud. 872 in nervum conrepere ;
375 belle recogito ; 1246 sapientes aequissimumst ;
649 nee vivam nee mortuam ; Stich. 188 ei verbo vicarium ;
Epid. 400 cave siris cum filia ; 447 ad cenam condicere ;
497 ulla pecunia ; 457 meis dictis deleniam ;
Men. pr. 30 ad ludos convenerant ; 661 Dionysum mihique et tibi ;
294 non nosti nomen meum ? ; Trin. (527 etsi scelestus est)
484 quid multis verbis opust?; (Rather et si) ;
499 quasi nomen nonnoveris; Vid. 30 duces tecum simul.
 PLAUTUS AND MENANDER 17

(Cf. Bacch. 301, Capt. 86, Mil. 481, Trin. 94, where the text is doubtful.)
In Bacch. 844 scan venit, Pres., not Perf. We omit lines like :
Amph. 943 (Sen.) Bis tanto amici sunt inter se quam prius,
where inter would be accented on the final syllable. (Cf. Poen. 775 inter se ;
1 143 inter se.) Spell as two words si quid (Aul. 380 ; cf. Ill 42), praeter
quam (Merc. 95), non vis (Stich. 483). In Bacch. 806 istuc may be oxytone
(and the Hiatus may follow the third, not the fifth foot). The above lists
show that a minor Ionic word (e. g. parasiti) is not preferred to a Molossus
word in this ending. The quantity of the second syllable of sarrapis is
unknown in Poen. 1312:
Deglupta maena, sarrapis sementium,
but the balance of probability inclines towards shortness. For, as has been
already said, a Cretic word is far commoner than a Molossus word in this
position ; also an Iambus word (or word-ending) than a- Spondee word (or
word-ending).
Therefore scan rei (III 30 K) in Rud. 487 :
Nam siqui' cum eo quid rei commiscuit.
We had better complete the scanty list of consecutive Spondee words with
clash of ictus and accent in Senarii (or Trochaics, see IV 16) by adding to
the above examples from the concluding portion of the line (Amph. pr. 134 ;
Men. 294; 484 ; Merc. 752; Mil. 1095; Pers. 408 ; Vid. 50) the following
examples from the rest of the line :
Asin. 100 Venari autem reti iaculo in medio mari (Rather rete, 54) ;
Pseud. 79 Eheu ! ' Eheu ' ! id quidem hercle ne parsis : dabo (Rather
Iambus ; III 42).
The rarity of this rhythm disqualifies any emendation which foists it on the
text, e.g. Poen. 969 ' Cretast, (cretast) profecto ' (Rather Cretast profecto,
(creta). Cf. Plin. N. H. 2, 1, 4 furor est profecto, furor) ; Aul. 565 e(i) vivo
licet (the Dative of is is not Iambus in Early Latin ; III 35). In Trin. 451
the * Palatine ' reading (Mearum rerum me) is corrected by the ' Ambrosian ' :
Mearum me rerum novisse aequumst ordinem.
In Trin. 575 illiic probably (and quam may be in Prosodic Hiatus). In Pers.
408 inlex (cf. • Still watching thro' day-bre'ak, day-light, day-close ').
We take this opportunity of telling our readers, once for all, that they
should refer to the Bibliography in Appendix E whenever they desire fuller
details than those we have thought worthy of their attention. They will find
in that list (s. v. Hingst) a University student's dissertation with extra-
ordinarily full information not merely about a fourth-foot Spondee in Plautus'
Senarii but in the corresponding part of all the Iambic lines of the
Dramatists ; of fourth-feet Anapaests too ; and of Spondees and Anapaests
in the corresponding parts of Trochaic lines, and so on. Hingst's disserta-
tion is a very favourable specimen of this type of literature. (Somehow he
has managed to overlook a Senarius ending of Terence, Phorm. 949 puerili
sententia.) But we prefer that our readers should find his booklet for them-
selves in our Bibliography, and that we should not take the responsibility of
2348 C
 18 PLAUTUS AND MENANDER

referring them to it. For his explanation of the facts does not satisfy us, nor
yet the inferences he draws from them. And, in our judgement, all that is
wanted is to show the reader that Plautus dislikes but does not wholly
exclude a fourth-foot Spondee with clash of ictus and accent in his Senarii,
his lines which echo ordinary talk. If Terence had been properly edited,
we should like to have shown that this dislike is rather heightened in the
I younger poet.
8. REGARD FOR ACCENT. This brings us to another difference
between Plautus and Menander. There is some regard for the accent
of words (i. e. of words in a sentence) in the Latin imitation. This was
pointed out by Bentley, who over-stated it, and now has the majority of
voices on its side, although French scholars, who assure us that their
own language is devoid of stress-accent, form a by no means negligible
minority.
The clearest case is the Roman avoidance of an ictus,1 like face're,
reficere, conficere. (On the frequency of a Tribrach word or word-ending
as a foot of Menander's Trimeters, in contrast to its absence from
*.
jo&/ Plautus, see 58. On its absence from all Roman poetry see Luc.
Mueller De Re Metrica2, p. 167. He disallows Seneca's Fugimus.)
And certain phrases of common occurrence seem to affect one
particular incidence of ictus (although the syllables' quantities would
allow another), an ictus which tallies with what would appear to have
been the accent in pronunciation. Non tu sa's? (a question) usually
begins Trochaic, not Iambic lines. ' I am going ' is at the beginning of
an Iambic line Ego eo\ of a Trochaic, Eo ego. The phrase Quid ego
nuncfaciam Ms a Trochaic line-beginning. On the two occasions when
it opens an Iambic line it is changed to Quid nunc ego faciam. Or
again, most languages accentuate the first of two naturally unaccented
words : e. g. Greek ttoo's //,€, Ital. mel (for me lo), English l in it ', ' for
it '. In Plautus the normal (but not invariable) ictus is id ego, quod ego,
etc. When the second word is emphatic (' I, not you ') the ictus
appears to have become (like the accent) id ego, etc. It certainly
seems as if the variation in Pseud. 1 137-8 (Troch.) expressed the
variation in the tone of an angry repetition (like our Whe're are you ?
and Where are you ?) :
Heus, ubi estis vos? Hi(c)quidem ad me || recta habet rectam
viam. (On hiquidem see III 9.)
Heus, ubi estis vos ? Heus, adulescens, || quid istic debetur tibi ?
And we are inclined to believe (though we cannot prove it) that the
1 Theorists say the ictus fell equally on the two short syllables of the ' rise ',
face're, etc. We adhere to the usual form of symbol. It is at least more convenient
for printing.
 PLAUTUS AND MENANDER 19

ictus on the final syllable of indignant or surprised repetitions reflects

accentuation, e. g. :
Capt. 837 (Troch.) Cedo manum. Manum? Manum inquam ||;
Most. 595 (Sen.) Non dat, non debet. Non debet} Ne ypv
quidem ;
Trin. 940 sq. (Troch.) Ad caput amnis quod de caelo ex||oritur sub
solio Iovis.
Sub solio Iovis ? Ita dico ||
(cf. Mil. 316, 324, etc.).
Studemund first used with effect this clue to the restoration of i
defective lines (in the Palimpsest, etc.), the normal ictus of a phrase. /
Editors cannot afford to neglect it. And yet they must use it with J
discretion, for Plautus was a quantitative, not accentual poet. And /
they must always consider whether the quantitative form of a phrase, a
form often imposed on it by the accentuation of the sentence, allowed
any other arrangement of the ictus. For example in the Pseudolus
couplet Plautus is echoing the pronunciation of heus ubi estis as (1) two
trochees (the ordinary question), (2) a choriambus (the angry repetition).
Does not (1) imply the ictus estis, (2) the ictus estis? Indeed since
Interrogatives in all languages are strongly accented words, it seems
certain that the ictus here does not tally precisely with the accent.1
A stronger witness is the phrase ygl<i sciig. Since volo was here an
Enclitic (a mere Auxiliary like ' will ' or rather ■ would ') the accentua-
tion of the phrase produced (by the Brevis Brevians Law) volo-scire.
The accent fell on the first syllable of the Infinitive. There was
nothing to prevent void scire having ictus on the syllable vo-, but as a
matter of fact the ictus is usually on sci-.
9. (A) In Tribrach Word and Word-ending. Why Plautus should make
a distinction between a Dactyl-word and a Tribrach-word and allow an ictus
to the one which he denies to the other is not clear to us nowadays. For an
ictus like mitte're is not confined to his Anapaestic Verse, where regard for
Accent seems to yield to regard for recurrent Diaeresis (IV 27). It appears,
though seldom (usually in the first foot) in his Iambics (58), e.g. Trin.
54 ; 75 : Omnibus amicis, quod mihist, cupio esse item ;
Omnibus amicis morbum tu incuties gravem.
Terence does not show half a dozen examples in all his six plays : Andr. 236
Hoccingst (or Hoccin est) ; 734 NesciS quid narres (III 38 E)-; Eun. 248
1 Cf. Cist. 649 (Troch.), Capt. 950 (Troch.), Aul. 268 (Troch.), Cure. 162 (Troch) :
Ubi gsti', servi ? occludite aedes || ;
Ubi 6sti' vos ? (heus !) ite actutum ||;
Ubi tu's quae deblateravisti ? || ;
Ubi tu's qui me convadatu's ? ||.
C 2
20 PLAUTUS AND MENANDER

Desine ; Hec. 701 Omnibu' modis ; 866 Omnia omnes. We shall find the
same difficulty with the ictus of a Tribrach divided between words (44)-
While such an ictus as Vendit £as is not unknown in Iambic Verse, an ictus
like Agit eas never (or hardly ever) occurs there. Of the occurrence of the
ictus facere, reficere, conficere the same phrase must be used — either l never
or hardly ever.
The MSS. offer the following examples in Iambic and Trochaic Verse :
Cure. 192 (Troch.) Ebriola persolla, nugae || (Emended to Ebriola's).
The Romance languages point to an accentuation -iola) ;
Trin. 788* (Sen.) Sed quom obsignatas attulerit epistulas
see (Fut.
III 40Perf.
B) j;

Pseud. 147 (Iamb.) Neque Alexandrina belua||ta, tonsiliaa tappetia

Canticum)(in;

Men. 841 (Troch.) || lampadi[bu\s ardentibus ;

Bacch. 893 (Sen.) M inertia, Lato[na], Spes, Opis, Virtus, Venus (see
III 42, s. v. Minerva) ;
Most. 1 100 (Troch.) || vis serere negotium? (Emended to serere vis);
Rud. 1002 (Troch.) || nos facere vis. Viduli (Emended to vis facere) ;
Men. 877 (Sen.) Qui me vi cogunt ut validus insaniam emended)
(Variously ;

True. 510 (Troch.) || iam actum aliquid oportuit (Emended to ali. act. ;
Appendix C, end) ;
Most. 376 (Troch.) Ouaeso edepol exsurge ; pater ad||venit, etc.
(Emended to ed. (te), ex.) ;
Aul. 700 (Sen.) Ibo intro, ubi de capite meo sunt comitia (Or meb with
Hiatus at pause or with ubi. Or capiti) j
Most. 211 (Iamb.) Id pro capite tub quod dedit || (Or capiti ; III 5) ;
Merc. 749 (Sen.) Abi. Quid, abeam ? St ! abl. Abeam f Abi (Rather ;
Anapaest)

leusmatic)
586 (Sen.) Metno ego uxorem, eras si rure redierit (Rather Proce- ;

Cas. 564 (Sen.) Hominem amatorem ullum ad forum procedere (Ditto ?) ;

Merc. pr. 29 (Sen.) Inerit etiam aviditas, desidia, iniuria (Ditto ?) 5
782 (Sen.) Sequimini. Fortasse te ilium mirari coquum (Rather
Sequimini. Fortass' te ; cf. Ill 42, s. v. Fortasse) ;
frag. 102 (Iamb. ?) ... addite || lopadas echinos ostreas (Rather Firstlopadas,
Deck).

Ter. Andr. 265 (Iamb.) Sed nunc peropus est aut hunc cum ip||sa, etc.
(Rather per opus ; III 42, s. v. per) ;
Ph. 955 (Sen.) Hicine ut a nobis hoc tantum argenti auferat? (Or
Hicin; but cf. 47. Or Proceleusmatic with utt)\
Caecilius frag. 232 (Sen.) Egone quid dicam ? egSn quid
omniavelim(Or? Egon)
quae tu;

Accius frag. 381 (Sen.) Reprime parumper vim citatum quadripedum.

1 Never in any metre, unless occasionally in Anapaestic, is the usual verdict.

 PLAUTUS AND MENANDER 21

The above list, into which all conceivable instances have been gathered,
illustrates the previous remark (I 6) that there is no theory under the sun
for which a fair show of support could not be provided out of scribes' errors
in the MSS. of Plautus. The reader will find the same tattered ' men in
buckram ' paraded in support of other theories : e. g. Cure. 192 and Most.
376 in support of one extraordinary theory of Diaeresis (III 48 e) ; Trin. 788*,
Pseud. 147, Men. 841, etc. in support of another hardly more convincing
(III 48 a) ; the Accius fragment as a proof of a by-form repfirimo (III 29).
He may at this early stage of our investigations take to heart the lesson that
all passages cited should be verified, and that one probability has to be
balanced against another. In dissertations by University students (and not
only there) the lists of examples of this or that scansion do not always (or
even often) include a mention of the alternative l scansions.
In Anapaestic Verse (real or supposed) :
Bacch. 1098 Relicuum id auri factum quod ego ei || stultissimus hom5
promisissem (On Bacch. 1106 see 31) ;
Cist. 221 Maritimis moribu' || mecum expetitur ;
Trin. 833 Satellites tui me miserum foede (40) ;
Pseud. 938 || magi' malum et maleficum. Tun id mi ?
? 1 1 33 Lucrifugos damni cupidos, qui se || (Or lucr. damnicup.,
Trochaic Octonar. ?) ;
see below) ;
Men. 361 Animule mi, mihi || mira videntur (an.-mi, a word-group?;

Pers. 761 Quorum opera ea mifacilia factu ||;

787 Siquidem hue umquam eru' redierit eius ||;
Trin. 251 || nox datur, duchur familza tota ;
True. 107 || referimu' gratiam furibu' nostris ;
(In Trin. 837 the MSS. vary.)
Ennius trag. 80 Deformati atque abiete crispa || (Or abjete).
In the Bacchiac line, Poen. 225, Apagesis (ictus on second syllable) should
be printed as one word (see III 42, s.v. sis). In Cretics, Most. 336 anime-mi
1 For instance, Stich. 715 (Troch.) appears in the minuscule MSS. (the
Palimpsest is not in evidence) in this abnormal form :
Bibe, tibicen ; age si quid agis ; bibendum hercle hoc est ; ne nega.
In one dissertation this becomes a prop for a theory (quite unlikely, in our judge-
ment) that siquid mightbe pronounced as a Pyrrhic (Bibe, ti|bicen ; | age siquid | agi' ;
bi||. And yet the usual ictus is age si quid agis ; Appendix A). In another, a prop
for a theory of a very unlikely division of a Proceleusmatic (Bibe, tibicen ; age si
quid agi' ; bi\\ ). In a third, a prop for a theory (quite impossible, in our judge-
ment) of a Brevis Brevians (Bibe, tibicen ; age si quid agis ; || bibendum hercle hoc
est ; ne nega). Everything abnormal vanishes and the flimsy props collapse if
Plautus actually wrote :
Bibe, tibicen ; age, si quid bi||bendum, hercle hoc est ; ne nega
1 Down with it, Mr. Musician : Come ! this drink, of all others, must go down. No
flinching! ', and if a scribe, thinking of the common phrase age si quid agis (e.g.
line 717), added agis carelessly, or else a corrector, scenting an ellipse, supplied
the supposed missing word in a note.
22 PLAUTUS AND MENANDER

(ictus on second syllable) may possibly be a word-group (but see Appendix D) ;

Capt. 207 (Sentio quam rem agitis) is an Ithyphallic Colon (as in 212), and
not a Cretic Dimeter (which would put the ictus on the second syllable of
agitis)', similarly Pseud. 1248 (as in 1249) ; but video ego, hardly a word-
group, has the ictus on the second syllable in :
Pseud. 1286 (of uncertain reading) || quid video ego? (Emended to
quidve vi.).
In the first foot of a Bacchiac line or hemistich an ictus like mittere shows
itself, Cist. 686 :
Wfierdita perdidit me. (Cf. Cas. 673?; Poen. 251 ?).
10. (B) Non tu scis ?, Ego eo, Quid ego nunc faciam ?, Volo scire.
Non tu scis ? :
sarier ;
Amph. 703 (Troch.) Non tu scis? Bacchae bacchanti || si velis advor-

Asin. 177 (Troch.) Non tu scis ? quae amanti parcet, ||eadem sibi parcet
must ;
parum ;
215 (Troch.) Non tu scis? hie noster quaestus || aucupi similli-

Men. 714 (Sen.) Non tu scis, mulier, Hecubam quapropter canem

Graii esse praedicabant ? Non equidem scio ; mali ? ;
911 (Troch.) Non tu scis quantum isti morbo || nunc tuo facias

Merc. 732 (Sen.) Non tu scis quae sit ilia ? Immo iam scio (Hiatus at
change of speaker) ;
Mil. 1 150 (Troch.) Non tu scis ? quom ex alto puteo ||mum
sursumescenderis
ad sum- ;

Stich. 606 (Troch.) Non tu scis quam — efflictentur || homines noctu

in viahie?
Ego eo :
Asin. 108 (Sen.) Ego eo ad forum, nisi quid vis. fietne? ego B)
Ambula (Eo;

Cas. 790 (Sen.) Ego eo quo me ipsa misit. Veron ? Serio ;

Merc. 385 (Troch.) Eo ego ut quae mandata amicus a||micis face;;
Immo tradam.
mane

Mil. 812 (Troch.) E[g]o ego intro igitur. Et praecepta || sobrie ut cures

Most. 853 (Troch.) Eo ego hinc ad forum. Fecisti || commode, bene

ambula (Ego abeo A) ;
Pers. 198 (Troch.) Eo ego. I sane, ego domum ibo||;
217 (Troch.) Eo ego hinc haud longe. Et quidem ego (eo) haud||
longe. Quo ergo (tu is) scelus ? (The second ego is;
emphatic)

Pseud. 169 (Iamb.) Ego eo in macellum, ut piscium || quidquid ibist

pretio praestinem ;
Rud. 403 (Iamb.) Ego eo intro, nisi quid vis. Eas ||;
 PLAUTUS AND MENANDER 23
Therefore scan eo in Cas. 790, and make the first foot a Tribrach in Asin.
108, Pseud. 169, Rud. 403. In Trin. 818 (Sen.) scan :
Mitt(am). Eo eg6 igitur intro ad officium meum.
Quid ego nunc faciam ? :
Cas. 549 (Troch.) Quid ego nunc faciam ? flagitium ||maximum feci miser;
Cure. 555 (Troch.) Quid ego nunc faciam ? quid refert ||;
Men. 963 (Troch.) Quid ego nunc faciam ? domum ire || cupio : uxor
non sinit (Hiatus at pause) ;
Mil. 305 (Troch.) Quid ego nunc faciam ? custodem || me Miaddidit miles ;

Most. 371 (Troch.) Quid ego nunc faciam? iube haec hinc || omnia
amolirier ;
1 1 49 (Troch.) Quid ego nunc faciam? Si amicus Philemoni
j| Diphilo es
aut;
Contrast :
Bacch. 857 (Sen.) Quid nunc ego faciam ? lube sis me exsolvi cito ;
Merc. 712 (Sen.) Quid nunc ego faciam nisi uti adeam atque adloquar?
(We take these examples from Studemund's Studien 2, 55, q. v.)
Void scire :
Aul. 428 (Vers. Reiz.) || void scire. Tace ergo ;
431-2 (Vers. Reiz.) Volo scire sinas an non sinas || nos coquere
hie cenam?
Volo scire ego item meae domi || mean futura ?;
salva

Epid. 507 (Sen.) Volo scire, si scis. Id quod audivi audies ;

Mil. 345 (Troch.) Agedum ergo face. Vol5 scire ut||rum ego id quod
vidi viderim (But egon A> with clash) ;
612 (Troch.) Sed volo scire, ebdem consilio ||;
Stich. 113 (Troch ) Volo scire ergo ut aequum censes || (Clash) ;
True. 261 (Troch.) Sed volo scire quid debetur ||;
779 (Troch.) Hie nunc volS scire ebdem pacto || (Clash) ;
Ter. Ad. 154 (Sen.) Volo scire atque hominem convenire si apud forumst.
11. (C) In Enclitic Group. The ictus id ego, quod ego, etc., is so pre-
dominant in Plautus that examples need not be given here. Almost any
page of the plays will provide them. The less usual ictus, Id e"go (which
corresponds with accent when the Personal Pronoun is emphatic) may how-
ever be illustrated, although the reader should be warned that, instead of
a Tribrach, an Anapaest is conceivable (if unlikely), since ego is not rare in
Plautus (in 31);
Epid. 453 (Sen.) Pol ego magis unum quaero meas quoi praedicem
(Emphatic ego) ;
Merc. 602 (Troch.) Uno verbo eloquere ubi ego sim || (Clash ?) ;
Most. 529 (Sen.) Hercules, ted invoco.
Et ego — tibi hodie ut det, senex, magnum malum
(Emphatic ego) ;
24 PLAUTUS AND MENANDER

Most. 531 (Sen.) Quid ego hodie negoti confeci mali ! (Clash);
781 (Sen.) (Nam muliones mulos clitellarios) ego) ;
Habent, at ego habeo homines clitellarios (Emphatic

Poen. 592 (Troch:) Novistis. Facile. At pol ego eum || qua sit facie
nescio (Ditto) ;
Pseud. 'too (Troch.) . . . quasi Dircam olim, ut memorant, duo nati Iovis,
Devinxere ad taurum, item ego te || (Ditto) ;
773 (Sen.) Neque ego amatorem mi invenire ullum queo ;
872 (Sen.) Item ut Medea Peliam concoxit senem,
Item ego te faciam.Eho, an etiam veneficu's ? phatic
(Em- ego) ;
Stich. 644 (Sen.) Idem ego nunc facio qui proviso Sagarinum (Ditto) ;
Vid. 67 (Troch.) Nisi quid ego mei simile aliquid || contra consilium paro
(Ditto).
We take these examples from Bursian's Jahresbericht 130, 162 sqq., where
a fuller discussion will be found.

12. (D) In various Phrases. In reading Plautus' dialogue (in sermonibus

poscit palmam) we seem to hear actual talk at Rome some two thousand
years ago. We certainly get the actual words and phrases of current use, the
actual quantity each syllable had in the utterance of the phrase, and further,
unless metrical necessity interferes (e. g. mea causa becomes causa mea at
the end of the line), the actual order. It is tempting to believe that the actual
accentuation too is reflected in the ictus. And we ourselves yielded to this
temptation in past years. How difficult to resist it in reading lines like these ! :
Amph. 937 (Sen.) lam nunc irata non es ? Non sum. Bene facis ;
Poen. 403 (Troch.) Non sum irata. Non es ? Non sum. || Da Srgo, ut
credam, savium (cf. Cas. 1007) ;
Stich. 185 (Sen.) veni Mo ad cenam: sic face.
Promitte vero ; ne gravare : est commodum ?
Volo, inquam, fieri : non amittam quin eas ;
632 (Troch.) Egone ? Tune. Mihine ? Tibine. ||;
Merc. 947 (Troch.) Ut valuisti ? quid, parentes || mei valent ? tarn
gratiast :
Bene vocas, benigne dicis. ||eras apud te, nunc domi.
Or the young lover's incoherent protests in Poen. 435 sqq. (Senar.) :
Neque hoc neque illuc neque — enfrn vero serio —
Neque hercle vero — quid opust verbis ?— quippini ?—
Quod uno verbo — dicere hie quidvis licet —
Neque hercle vero serio— scin quomodo ?—
Ita me dl amabunt — vin bona dicam fide ?—
Quod hie inter nos liceat — ita me Iuppiter —
Scin quam videtur ?— credin quod ego fabuler?
We seem to hear the shriek of the enraged wife in Asin. 896 (Troch.) :
am tandem ? edepol ne tu istuc l| cum malo magno tuo.
 PLAUTUS AND MENANDER 25

No doubt the reader may indulge his fancy without inaccuracy in most, or
very many, cases. But a closer view will show him that the alteration of
quantity under the Brevis Brevians Law (19) is a more unerring guide to the
accentuation of the sentence than ictus can possibly be. When Callicles is
afraid that Lesbonicus will detect the trick, his old friend assures him that
the young rake is likely to be drunk. ■ What do you take him for ? ' :
Trin. 81 1 sqq. (Sen.) Diei tempu' non vides ? quid ilium pittas,
Natura ilia atque ingenio ? iam dudum ebriust.
Ouidvis probare poteris.
Since all languages accentuate Interrogatives, the accent must have fallen
on quid\ and this 'brevis brevians' would shorten the unaccented Demon-
strative (31). But there must have been an accent too on the first syllable
of pulast though the quantity does not give the ictus the chance of coincid-
ing, that it gets in e. g. :
Cure. 59 (Sen.) Immo ut Mam censes ? ut quaeque illi occasiost ;
Ter. Andr. 853 (Troch.) Sunt. Cur igitur hie est ? Quid Ilium || censes?
cum ilia litigat ;
Ad. 656 (Sen.) Quid Ipsae ? quid aiunt ? Quid Illas censes ? nil enim.
The question ubi Mast f makes an Anapaest, but it would not be true to
say that this Anapaest begins Trochaic lines only. It is found also at the
beginning of Iambic. Similarly with quid istic ? ' very well ' (' adverbium
aegre concedentis et velut victi ', Donatus), a phrase for which a single
accent, presumably on the Interrogative, is actually attested (Gloss. Lat. v,
622> 55 Quid istic ? sub uno accentu est, profecto vel omnino).
In the following select list (which might be greatly extended) we shall
indicate any alteration of quantity caused by accentuation, but omit mention
of ictus whenever the quantitative arrangement of the syllables would make
a clash of ictus and accent impossible or unnatural.
Quod Hie dixit (the formula for ' Wellerisms '), and the like :
Cist. 14 (Troch. and Cret.) Quod Ille dixit qui secundo || vento vectus
est || tranquillo 163);
mari.
1 Ventum gaudeo ' (cf. Pomponius frag. 1) ;
Asin. 761 (Sen.) Aut quod Ilia dicat peregre allatam epistulam (cf. Mil.

Cure. 634 (Troch.) Nil est quod Ille dicit. Fac me ||certiorem,— obsecro ;
Men. pr. 22 (Sen.) Ut quidem Ille dixit mihi qui pueros viderat (cf. 336).
(Cf. Ter. Eun. 711 quod Iste dicat ; Ad. pr. 15 Nam quod Isti dicunt).
Eho an noti ?, and the like (therefore an had no accent) :
Mil. 301 (Troch.) || proximost. Eho an non domist ? ;
Pseud. 969 (Troch.) Eho an non priu' salutas ? Nullast ||;
Ter. Andr. 766 (Sen.) Eho an non est ? Recte ego semper nuptias
fugi has ;

Ter. Hec. 671 (Sen.) Ego alam ? Quid dixti ? eho an non alemus,
Pamphile ?
 26 PLAUTUS AND MENANDER

Epid. 506 (Sen.) Eho an libera ilia est ? quis earn liberaverit (Volo scire) ;
Merc. 393 (Troch.) || vidi. Eho an vidisti, pater ? ;
Mil. 821 (Sen.) Eho an dormit Sceledrus intus ? Non naso quidem ;
Most. 455 (Sen.) Eho an tu tetigisti has aedes ? Cur non tangerem ? ;
Pseud. 305 (Troch.) Credere autem ! eho an paenitet te ? ||;
Ter. Ad. 389 (Sen.) Est iam intus. Eho an domist habiturus ? Credo,
ut est (cf. 48).
So scan with Hiatus at change of speaker :
Most. 1083 (Troch.) Vendidisse. Eho an negavit ||sibi datum argentum,
obsecro ? ;

Rud. 578 (Troch.) Tu istaec mihi dato : exarescent || faxo. Eho an te

(Also in Pseud. 872. paenitet ?

The first foot of Poen. 11 36 cannot be a Tribrach.
For other examples of eho an, before a vowel, see Studemund's Studien i, 443).
voluptas mea * my heart's delight ■ :—
Cas. 136 (Sen.) Sine tubs ocellos debsculer, voluptas mea.
453-4 (Sen.) Ob Tstanc rem quin te debsculer, voluptas mea.
Quid, debsculere ? quae res ? quae voluptas tua ? ;
Mil. 1346 (Troch.) || ne time, voluptas mea ;
Most. 249 (Troch.) || Philolaches, voluptas mea ;
294 (Troch.) || aufer. sed, voluptas mea ;
Pseud. 52 (Sen.) Minis viginti vendidit, voluptas mea;
Rud. 439 (Troch.) || iam hie ero, voluptas mea;
True. 426 (Sen.) Lucrum hercle videor facere mi, voluptas mea ;
540 (Troch.) || tene* tibi, voluptas mea ;
546 (Troch.) || num nevls me, voluptas mea ? ;
860 (Troch.) || quid agitur, voluptas mea ? ;
To Valerius Probus ictus appeared to reflect the natural accent in dramatic
I
verse (at least the verse of dialogue). For a pupil of his told Aulus Gellius
(6, 7) that in a lecture he declared the accent of exadversum to be properly
on the second syllable and appealed to Terence Ph. 88 (Sen.) :
In quo haec discebat ludo exadversum loco
Tonstrina erat quaedam, etc. ;
further, that he was in the hahit of pronouncing £dfatim, ddmodum (the
current pronunciation of his time being ad fa'tim, ad modum) and claimed
this accentuation for early Latin, so reading to the class Plautus .Cist. 231
(Troch.) :

Potine tu homo facinu' facere ||strenuum ? Aliorum adfatim est (Qui faciant).
And Donatus' notes on pronunciation (in the Terence Commentary) often
seem to have regard to ictus. Thus he says of Ph. 70 (where emphatic me is
presumably in Prosodic Hiatus ; III 53) : 'me' acue. And he points out the
different nuance of siquando from siquando in a note on Eun. 437 (Sen.) :
Scin siquando ilia mentionem Phaedriae (Facit).
 PLAUTUS AND MENANDER 27

Lists of the occurrences of the following phrases will be found in Ap-

pendix A :
age age (with accent on first word), e. g. :
Ter. Ph. 662 (Sen.) Ob decSm minas est. Age age, iam ducat : dabo.
age si quid agis (with accent on the first and third words), e. g. :
Epid. 196 (Troch.) Age si quid agis, di immortales ! ||.
eug(ae) eugae (with accent on second eu-), e. g. :
Aul. 677 (Sen.) Eugae eugae ! di me salvum et servatum volunt.
habe bon(um) animum, * cheer up ! ' (with accent on ha- and an-), e. g. :
True. 525 (Troch.) HabS bonum animum. Savium, sis || (Cf. Loch,
Gebrauch des Imperativus bei Plautus, p. 17).
he u me miserum (with accent on heu and mi-), e. g. :
Merc. 624 (Troch.) Mulier. heu me miserum, etc. (Cf. Studemund's
Studien I, 562 sqq.).
hie iilest (with accent on hie, at least usually), e. g. :
Trin. 43 (Sen.) Hie Iilest senecta aetate qui factust puer ;
hoc era/, hoc est ' that 's why ' (with accent on hoc), e. g. :
Cas. 531 (Troch.) Hoc erat ecastor quod me vir || tanto opere orabat
meus.
mane* mane (with accent on first syllable), e. g. :
Aul. 655 (Troch.) ManS manS. quis lllic est ? quis hie intus ? ||.
miser sum (with accent on second syllable), e. g. :
Pseud. 80 (Sen.) Miser sum, argentum nusquam invenio mutuum.
miserum me (with accent on -rum), me miserum (on mi-), e. g. :
Ter. Ad. 486 (Sen.) Scio. Miseram me ! differor doloribus ;
Andr. 788 (Sen.) Me miseram ! nil pol falsi dixi, mi senex.
quid 8st quod, quid istuc est quod, quid hdc est quod (with accent on
quid),e.g:
Most. 69 (Sen.) Quid gst quod tu me nunc obtuere, furcifer ? (Cf. Ditt-
mar, Lat. Moduslehre 11.)
quid hoc cla?noris ? (with Interrogative and Noun accented), e.g. :
Aul. 403 (Sen.) Sed quid hoc clamoris oritur hinc ex proximo ?
quid tstuc verbi est? (Ditto) begins Trochaic lines; Quid istuc est verbi,
Iambic, e. g. :
Ter. Ph. 343 (Troch.) Quid Istuc verbi est ? Ubi tu dubites ||.
quid opust verbis ? (with accent on Interrogative and ablative), e. g. :
Cure. 79 (Sen.) Solet esse. Quid opust verbis ? vinosissima est (cf.
Bryant in Harvard Studies 9, 122).
sine modo ' just let him ' (with accent on Verb ; but perhaps on Adverb
with a different nuance), e. g. :
Most. 11-12 (Sen.) Quia vivis. Patiar, sine modo adveniat senex.
Sine modo venire salvum quern absentem comes.
tib(i) ego dico (tibi accented), e. g. : '
Mil. 434 (Troch.) Tibi ego dico, heus ! Philocomasium || (cf. Kaempf, de
Pronom. Personal, p. 17).
28 PLAUTUS AND MENANDER

tute tibi (a word-group, with accent on the antepaenultima), e.g. :

Cure. 9 (Sen.) Tute tibi puer es, lautu' luces cereum.
vae capiti tuo, vae {hen, etc.) misero mihi (accent on Interjection and -ti,
-ro), e.g.:
Merc. 661 (Troch.) || abiit. heu misero mihi.
(Cf. Studemund's Studien
2, 76.)
ut fiote qui (a word-group with accent on the antepaenultima), e. g. :
Cist. 317 (Iamb.) Ut pote quam numquam viderim ||.

13. Our own view is that there is a regard for Accent in all Latin
Verse, a result of the Latin accent being a stress-accent, and that this is
most in evidence in Dramatic verse, where the tones of conversation are
echoed. We do not forget that so eminent Frenchmen as Havet,
Meillet, and others deny that the Latin accent was a stress-accent and
thus make it incapable of playing this part in verse. But the view of
the majority seems to us so well established that in the discussion
of differences between Menander's and Plautus' practice, we shall make
a point of asking 'Is the difference due to the Latin regard for
Accent ? '
Of course the Iambic and Trochaic metres lend themselves to a
reconciliation of accent with ictus. Some of the lines of the early
Tragedians read almost like a Christian (accentual) hymn, e. g. :
Ennius trag. 163 (Iamb.) O magna templa caelitum || commixta
stellis splendidis.
How naturally the emphatic contrast of factis and dictis takes the ictus
in Amph. 926-7 !
Nunc, quando/artis me impudicis abstini,
Ab impudicis diclis averti volo.
How naturally the subordinate words, the words we omit in writing
telegrams, slip into the ' fall ' of each foot in (let us say) the first three
lines of the Aulularia prologue, while the accented syllables (e. g./« milia ;
18 A) secure the ictus !
Nequis miretur qui sim, paucis eloquar.
Ego Lar sum familiaris ex hac famnia
Unde exeuntem m(e) aspexistis, etc.
And in Trochaic Verse the reconciliation seems to come spontaneously ;
e.g. Merc. 356; Mil. 280:
Hoccin est amar(e) ! arare || mavelim quam sic amare ;
(tu sali) Solu', nam eg(o) istanc /Vzsultur(am) et || desulturam nil
moror.
But that implies merely that it was easier for a Dramatic poet to avoid
 PLAUTUS AND MENANDER 29
clash of accent with ictus than for other poets ; just as it was also more
necessary for him, since he reproduced the conversation of everyday
life, not (like them) an impassioned self-communing of the soul or
picturesque description of places or events. However easily it was
avoided, it was deliberately avoided in the Latin Drama. Take any
fifty Trimeters of Menander, imagine the Greek words to be Latin
words and give them the accent which the quantity of their penultimate
syllable would give them in Latin ; you will find much more clash than
you ever find in Plautus' Senarii. Plautus' reconciliation of accent and I
ictus is not accidental. It is far from being pushed to monotony;1 but*
whenever a strong contrast between words has to be made, wherever
emphasis has to be laid (let us say) on a Pronoun, there we are nearly
sure to find the ictus at the place required, e. g. :
Amph. 35-6 (Sen.) Nam iniusta ab iustis impetrari non decet,
Iusta autem ab miustis petere insipientiast (cf.
Merc. 263 j Pseud. 791 ; but not Rud. 863) ;
Cas. 369 (Troch.) Huic— immo hercle mihi — vah ! tandem ||;
851 (Sen.) At mihi, qui belle hanc tracto, non bellum facit;
Men. 746 (Sen.) Si me derides, at pol /71um non potes (On pol
ilium, with emphasis, see 31).
To be fair, we must confess that it would not be impossible to find the
same thing in Menander, at least now and then, e. g. Sam. 129 sq. :
To v eh a7ra»/Tas koct/jliov kcu (Tuxfipova
Tous aWorpiovs eis i/xl tolovtov yeyoveVai.
Notice how the ictus points the pun in Bacch. 284-5 (Sen.) :
Quom mi ipsum nomen eius Archidemides
clamaret <&/»pturum esse si quid crederem (pronounce
with long e).

14. While recognizing the Accent as a new factor in Plautus' versifica-

tion, we must not assign to it (as so many do) a leading part. Plautus'
verse is not__accentual ; itjs_quantitative. Even modern accentual verse
occasionally defies accent : ■ F611owed their mercenary calling ' (not
'Pursued'), ' Ihn hat sie weggerissen ' ; and it would be unreasonable
to postulate anything more for Plautus than a strong disposition to
bring ictus and accent into harmony where he conveniently could
without sacrificing things more important. To any quantitative poet
the most important thing is quantity. That must have been Plautus'
first consideration. To a Comedian the next most important thing is
1 We should indeed find ' fungi putidi ' in a successive hundred lines like Bacch. 821 :
Nee sentit, tantist quantist fungus putidus.
3© PLAUTUS AND MENANDER

the natural order of words (e. g. esse oportet, not ' oportet esse ' ;
numquae causa est, not • numquae est causa '), so far as his quantitative
metre will allow it. Accent could not claim any higher than the third
place in his regard. Indeed in his Anapaestic verse Plautus sets before
it a regard for the Greek recurrent Diaeresis and lets accent take care
of itself, more or less (IV 27).
Those who deny Plautus' regard for Accent (and with their denial of
the extreme application of the theory we heartily agree) sometimes think
it enough to remind us that the sixth foot of the Senarius must be an
Iambus and that an iambus-word in that foot brings clash of ictus and
accent. This is not a convincing argument. All that it shows is that
the Greek metre which Plautus imitated put an obstacle in his way. It
was not half so troublesome as the obstacle in Anapaestic lines, which
was more successfully surmounted by poets subsequent to Plautus (e. g.
Ovid reconciles ictus and accent in his Anapaestic Dimeter fairly well,
Medea frag. 2 :
Feror hue, illuc, vae ! plena deo).
Indeed we seem to get rather an argument in our favour when we look
more closely into Plautus' treatment of this difficulty offered by the last
foot of the Trimeter. Senarii like these are rare :

Stich. 260-1 Nullan tibi linguast? Quae quidem dicat 'dabo ' :
Ventri reliqui — eccam, quae dicat ■ cedo '.
(Cf. Capt. 360; Pers. 129; Poen. 324.) Certainly the clash of ictus
and accent would be felt in that pair. But not in the commoner type,
a Senarius like :

Trin. 1 Sequere hac me, gnata, ut munu' fungaris tuum,

where there is no special emphasis on the final iambic word. Tuum
would not be marked in enunciation j indeed its absence would not be
felt. In a line like :
Trin. 729 (Sen.) Ut mihi rem narras, Callicles, nullo modo (Potest
fieri prorsus),

the final word would be enclitic (cf. quomodo or our ' nowise '). And
a word-group of the scansion of a Fourth Paeon (^ ^ w -) at the end
of a Senarius seems to be designed to mitigate the harshness of the close
of the line, as will be apparent from a perusal of the examples in para-
graph 16.
15. Reconciliation of Ictus and Accent of Iambic Words. From a
paper in the Philological Society's Transactions of 1894 (The Accentual
Element in Early Latin Verse) we take the following account of some of
Plautus' methods with these refractory words :
 PLAUTUS AND MENANDER

' (1) In the first place, he avails himself of the tendency of Latin pronun-
ciation to shorten their final syllable, a tendency which asserted itself most
strongly in those ending with a long vowel followed by -r, -t (e.g. amor,
amat) and in those which, ending with a long vowel, were closely connected
with other words in ordinary rapid utterance, e.g. cave-faxis, doml-resto,
domo-veni, dabo-plagam, dedl-plagam, citS-curre, modo-veni.
(2) Or, secondly, he so places them in the sentence that their accent is
diverted from the first syllable, whether to the final (e. g. bonaeque, bonaene,
bonaeve, bonae-sunt, malae-res, mala-fide, fidem-dat) or to a preceding word 3i
(e. g. hoc-modo, aliquo-modo, coctum-dabo, factum-volo).
(3) Thirdly, he elides their final vowel.'
16. Fourth Paeon Word-group Ending of Senarius (i.e. Word-groups
of the type ow, w^).
Here are all the examples in Plautus and Terence :
(PLAUTUS.)
Amph. 37 neque tenent ; Epid. 309 mihi libet (Or mi) ;
44 meu' pater ; 404 nimi' potest ;
104 pater meus (Emended to 421 ita decet ;
meu' pater) ; ? 422 ei volo (Rather ei) ;
Men. 75 modo senex ;
477 meu' pater ;
937 bene facis ; 487 (quid) ais homo ;
1 131 tibi et tuis ; Merc. 91 mihi foret ;
Asin. 112 mihi, tua (oratione)
(Or mi) ; 95 meu'
269 pater ; (?) ;
mihi malum
759 nisi tibi ; Mil. 88 meus erus ;
779 nisi tibi ; ? 105 (mei) eri ;
773 sati' placet ; Most. 40 hara suis (' pig-sty ') ;
72 suom sibi (?) ; 414 sine malo ;
778 neque roget ; 1038 mihi cedo ;
Aul. 709 (ex) eo loco (?) ; 583 abi modo ;
Bacch. in mihi quidem ; 588 meum peto (?) ;
147 cave malo ; Pers. 147 abi domum ;
508 meu' pater ;
866 tibi lubet ; 372 quibu' licet ;
398 tibi libet ;
Capt. 5 (suo) sibi patri ; 691 nisi piget ;
745 ita vale ; Poen. 421 mea salus ;
Cas. 3 date mihi ;
Cure. 82 nisi nevis ; 430
706 abi
ego modo : ;
dare (?)
272 bene facis ;
734 bene putas ;
402 mihi placet ;
1050 tuu' pater;
422 mihin ? ita ; 1 106 mihi placet (Or mi)
461 mora mihi (?) ; 1 146 sine modo;
657 mea soror ; 1308 mihi libet (Or mi)
670 priu' volo ; Pseud. 493 apud erum ;
673 bene facis ;
554 egodabo;
 PLAUTUS AND MENANDER
32
Pseud. 1023 bene volo; Trin. 586 abi modo ;
873 magis. ehem (?) ;
779 tibi dare ;
Rud. 132 neque potest ; True. 5 sine mora ;
504 mihi fuit (Or mi) ; 247 probu' dator (Canticum) ;
842 quasi canem ; 367 cedo bibam ;
Trin. 5 1 ego volo ; 441 bene velim ;
156 suom sibi; 641 neque veto ;
174 face sciam ; 696 cedo manum ;
434 bene facit ; Vid. 82 neque fore ;
508 eum dabo (Or eum) ; Frag. (26 nihil erat) (Rather nil).
576 ita volo;
(Terence.)
Andr. 279 neque pudor ; Phorm. 62 tibi dico ;
458 rei caput ; 365 mihi senex (Or mi) ;
774 magi' dabit ; 596 sibi dari ;
873 male loqui ; 668 mihi dicas (Or mi) ;
Heaut. 457 modo mihi ; 910 tibi darem ;
468 sibi dare : 950 cape, cedo ;
508 tibi darem ; 980 ego scio ;
? 818 mihi (manum) ; Hec. 112 mihi fidem (Or mi);
932 nisi caves ; 189 ubi senex ;
Eun. 57 neque modum; 647 mihi, Laches ;
65 sine modo ; 684 tibi dedi ;
93 mihi dolet ; Adelph. 59 neque placet ;
141 mihi dare (Or mi); 84 neque pudet ;
186 bene facis ; 102 neque fores;
200 neque meo ; 239 satis placet ;
851 tibi placet : 361 ubi siet ;
898 meae fide ;
404 apud forum ;
907 quia pudet ; 463 neque boni ;
Phorm. 41 minus habent; 512 apud forumst ;
61 mihi lucrist :
904 mihi moraest (Or mi).
This list shows that there is a very narrow selection of words for this ending.
With hardly an exception the two words can be described as a blend. (And
from Cicero's story, Div. 2, 84, we can see that cave ne eas was a blend which
fell upon the ear like canneas.) The reading (Rud. 1229) daniint boni is
unlikely.

17. And (a thing to be carefully noticed) we must not confuse with

a regard for Accent an entirely different matter, the scansion of words
(and phrases), according to the quantity which Latin accentuation had
given them. Philologists tell us that the spelling incutio (not ' inquatio ')
is due to the Latin Accent, which (at an earlier period) had fallen on
the first syllable of each word. The accentuation inquatio had produced
the pronunciation (and spelling) incutio, and Plautus took the word as
 PLAUTUS AND MENANDER 33

he found it, with u (not a) in the second syllable. Similarly the accen-
tuation male, bene had affected not the spelling, but the quantity of the
second syllable. Originally male, dene, these words had already, before
Plautus' time, become male, bene. Philologists explain that the reason
why these two Adverbs were never pronounced (in Plautus' time and
subsequently) with the long e of similar Adverbs, e. g. probe, was the
frequency of combinations like malefactum, bene/actum, whereas probe
retained more independence. Plautus took these two words as he found
them, with a short (not a long) final e (III 5). His scansion male, bene,
while it might be called the result of Latin accentuation, has of course
nothing to do with his regard for Accent. No more has his scansion
(12) voluptas-mea (never ' voluptas ' alone, but only in this phrase), where
he gives to the second syllable of the word-group the quantity with
which it was uttered in every-day (educated) conversation, a quantity to
be explained in the same way as calSfacio tepefacio, etc. (22).
Plautine Accentuation.
18. (A) Traces of older Accentuation in Plautus. This seems a suit-
able place for a discussion of some Plautine accentuations, a hazardous dis-
cussion which lacks satisfactory evidence and leaves us open to the charge
of reasoning in a circle. For we have to take for granted Plautus' regard for
Accent.
Quadrisyllables with the first three syllables short seem to have been
accented on the first syllable in Plautus' age and for some time later. When
precisely the change from, e. g., fdmilia to familia (cf. our ' capitalist ' and
* capitalist ') was consummated is hard to say. Since Cicero was not a pro-
fessed phonetician, his statement, newly learned from Tyrannio, that ■ ipsa
natura, quasi modularetur hominum orationem, in omni verbo posuit acutam
vocem nee una plus nee a postrema syllaba citra tertiam' (Or. 58), would
perhaps be consistent with his admission of the exception admitted in
Plautus' time and subsequently. Although it would be wrong to say that an
incidence of ictus on the second syllable of a word of this kind in Plautus
proves the reading to be unsound, his propensity to place the ictus on the
first is unmistakable ; and the same may be said of Terence. The two
Comedians stand in marked contrast to Phaedrus in this respect. (For details
see Philologus 51, 364 and Bursian's Jahresbericht 80, 270). And it cannot
be the mere succession of four short syllables (or three short and one long)
that causes this incidence of ictus. An ictus like Is etiam (Epid. 524), where
ictus and accent would not be in conflict, is common enough, just as Sed ubl
(Interrog.) is common, but never Serere, Facere.
The Latin Thesaurus makes * coemptionalis ' the spelling of the' Old Latin
word for a superannuated slave, ' bought in a lump': but this, the etymo-
logical spelling, has the support only of a mediaeval corrector of the Codex
Vetus in Bacch. 976. In all the occurrences of the word (see Journ. Phil
34, 281) the spelling is comptionalis (corrupted to ' contionalis ' in MSS. of
2348 D
34 PLAUTUS AND MENANDER

Livy 3, 72, 3). This spelling (and pronunciation) reflects the older accentua-
tion coemptionalis
' ', and if the word had remained longer in actual use it
would perhaps have assumed the etymological spelling, with a secondary
accent on the second syllable, ' coemptionalis '.
Were there more of these traces of the earliest accentuation of the first
syllable of each word that still were visible in Plautus' day ? Should we
spell ' cogmenta ' in Most. 829 (Troch.) :
Viden coagmenta in foribus ? Video ||,
or scan (with Synizesis) coagmenta, or admit a Proceleusmatic (a foot much
less frequent in Trochaic than in Iambic verse ; 57) and scan coagmenta (by
the Brevis Brevians Law) ? Should we spell copulonus (hardly copl- ;
III 26) in Pers. 100 (Sen.) :
Terrestri' [te] coepulonu' compellat tuus,
or follow the natural scansion coep-? (The Codex Turnebi seems not to
have inserted te between the first two words.) (On festra for fenestra see 40.
And on co- see III 29.)
(B) Retrogression of Accent. Bentley's theory of the Latin accent's
retrogression may be dismissed summarily, for it seems baseless enough.
In declaring the reconciliation of ictus with accent in the verse of dialogue,
he cited the opening line of the Andria prologue :
Poeta quom primum animum ad scribendum appulit,
and asserted that even the ictus on the first syllable of scribendum tallied
with the accent. For, he urged, the elision of the last syllable would cause
retrogression of the accent, just as in Greek dyadd becomes dydO' epya. For
this theory we find no proof. It is indeed widely accepted, and even stated
as if it were an undoubted fact, by those who hold Bentley's extreme view of
the harmony of ictus and accent, and by others too. But where is there a
hint of such a thing in the Grammarians ? Where is there a trace in modern
Latin ? An Italian puts the stress-accent on the second syllable of ' oscura '
even in the phrase ' oscur' ombra \
There seems to be no justification for allowing by an appeal to the Brevis
Brevians Law an impossible scansion like (Most. 504) Scelestae hae. Rather
Scelestae-sunt aedes (cf. molestae-sunt Mil. 69 ; 29). Nor yet (Merc. 488)
Achfllem orabo. Rather Acchillem (III 28).
(C) Attraction of Accent by Enclitic. When the grammarians of the
Empire tell us that an Enclitic attracted the accent of a preceding word, we
may believe them in part. We may believe that Mosque, illosce were accented
as if single words, illosque, illosce. But we need not believe them when they
apply Greek laws to Latin and tell us that illaque was accented illaque (like
Tavrd re). (Presumably therefore factust, bonust.)
Plautus' scansion raises a doubt whether, when the Enclitic lost its final,
the accent did always retrogress. For viden is inconsistent with an accentua-
tion viden and postulates rather viden.
And while the first syllable of Me when accented (the Pronoun being
emphatic) naturally remains a long syllable even though a short syllable pre-
 PLAUTUS AND MENANDER

cedes (in contrast to ad Ilium, etc., with unemphatic Pronoun), we seem 35

to
find illunc, etc., treated in the same way, e. g.
Amph. 699 (Troch.) Nam dudum ante lucem et Istunc || et te vidi. QuS
in loco?
Mil. pr. 151 (Sen.) Et hinc et illinc mulier feret imaginem ;
Merc. 874 (Troch.) Si hue item properes ut Istuc || properas, facias rectius.
If the accentuation were illunc, etc., why should not the ■ brevis brevians '
have operated and produced et Istunc? If the Neuter (originally illud-ce)
were accented illuc, why should we find in Terence (III 20) illiic?
Even when the Enclitic remains intact, e.g. videsne, prospereque, it would
seem that the accent was not necessarily shifted; e.g. Pseud. 574 (Ana-
paestic) lepide
| omnia prospereque eveniunt ; Asin. pr. 3 (Sen.) Gregique
huic et dominis atque conductoribus ; Trin. 109 Videtque ipse ; Rud. 957*
Feroque ei : Aul. 812 Erumne ego, etc. The ictus magi'que, priu'que (not
1 magisque ', ' priusque ') is normal.
Since the clash of ictus withaccent seems to be avoided by Plautus in
spondee-words (and word-endingsTTn most feet of the Senarius (6), we_are_
tempted to find Enclisis in lines like :
Aul. 372 (Sen.) Ut bene haberem-me filiai nuptiis ;
Bacch. 669 (Sen.) Quid vos maestos-tam tristesque esse conspicor?
(AP ; in a Canticum) ;
Merc. 312 (Sen.) Lysimache, auctor-sum, etc. ;
and even
Mil. 1 135 (Sen.) Nam quos videre exoptabam-me maxime.
(D) Subordinate Words. By the Law of Breves Breviantes (22) a short
syllable may shorten a following syllable which has not the accent (i. e. in the
I accentuation of the sentence). A shortening under this peculiar phonetic
law of La^enables_usJo_detPCt the nnarrpntpri ™- subordinate words of the
Latin Sentence. Since Expletives are sentence-enclitics in most languages
it is natural to find in Plautus, e.g. Epid. 688 (Troch.) || meo hercle vero
atque haud tuo (Meo emphatic). So with Interjections, e.g. sed eccum
(usually. Details in Studemund's Studien ii, 389), Domi eccam erilem
concubinam || (Mil. 470, Trochaic ; Domi emphatic.) So with Particles like
nempe, profecto, etc. Perhaps it is the hurrying of the voice to the following
word which accounts for the (not uncommon) Plautine scansion (the pr6-
operating as ' brevis brevians') profecto, e.g. Mil. 290 (Troch.) Profecto
vidi. Tutin ? Egomet ||. But the scansion of profecto is a puzzle (as of
immo ; III 58. On pirinde, with accent on the first syllable see III 42 s. v.).
Even optume (after a short syllable) seems established by Pers. 544 (sed
optume), Most. 410 (vel opt.), Capt. 345 (hie opt.). And so on.

19. THE BREVIS BREVIANS. This shortening {male, bene, etc.), A

due to the Latin stress-accent {pace our French friends), is produced
only when a short syllable precedes. Thus the e of probe might be
affected, but not of sane. It played a very great part indeed in Latin
D 2
36 PLAUTUS AND MENANDER

pronunciation ; and this law of Latin Phonetics (no mere ' law of
Metre ') is known as the Law of Breves Breviantes or the Brevis Brevians
Law (i. e. short syllable shortening a following syllable). The Germans
call it the Iambenkiirzungsgesetz. It has produced these pronunciations
with which we are all familiar from our reading of Virgil and the other
classical poets : modo beside modo, cito beside cito% ibi beside iH , and so on.
These words too Plautus took as he found them. He scans them
according to their (recognized) double pronunciation, sometimes as
Pyrrhics, sometimes as Iambi ; although, as is natural, with the Dramatic
poet the more deliberate pronunciations are less frequent than with the
Epic poets. Virgil seems to have been more tolerant of such shorten-
ings than Ennius.
20. The instances of modo (Adverb) in Plautus are :
Amph. 644 (Bacch.) Absit, dum modo laude parta ;
646 (Bacch.) || modo si mercedis ;
Asin. pr. 5 (Sen.) Age nunc reside, cave modo ne gratiis ;
897 (Troch.) Dixisti in me. sine venias mo||do domum, faxo ut
scias (Emended to sine revenias || mod6) ;
Aul. 239 (Troch.) Dum modo morata recte || veniat, dotata est satis ;
Capt. 458 (Troch.) Ad fratrem modo captivos || alios inviso meos
(Emended to mo. ad capt.) ;
Cas. 488 (Sen.) Satin docte? Astute. Age modo, fabricamini (Syll.
Anceps in pausa ? Cf. 46) ;
743 (Anap.) Cena modo si || sit cocta ;
Merc. 875 (Troch.) || cape modo vorsoriam ; #
Most. 994 (Sen.) Non equidem in Aegyptum hinc modo vectus fui ;
Pers. 498 (Bacch.) || adlatae modo sunt ;
Poen. 926 (Troch.) || quod modo concreditumst ;
Rud. 1 127 (Troch.) Cedo modo mihi istum vid(u)lum, || Gripe. Concre-
dam tibi (Emended to modo mihi vidu. ist) ;
True. 614 (Anap.) Tange modo ; iam ego (te) hie agnum || faciam et
medium distruncabo (Or Syll. Anceps in pausa).
This list omits the occurrences of the word in positions where a short
syllable may become a Syllaba Anceps (not merely at the end of a line or
hemistich, e.g. Cas. 625, 758, Rud. 951, but also at a change of speaker, e.g.
Cure. 655, Mil. 984, Trin. 587, or at a marked pause in the sentence, e. g.
Asin. 869, Bacch. 638a, Merc. 426). Five of its fourteen instances are in
Canticum Metres (especially Bacchiac ; IV 23) ; and four at least are
doubtful. Some half-dozen instances in dialogue-metres remain ; and against
these we can set some sixty cases of modo. The Pyrrhic is clearly the rule,
the Iambus the exception ; so the Pyrrhic should get the benefit of the doubt
in lines like :
Asin. 876 (Troch.) Sequere hac me mod6, iam faxo ipsum || hominem
manifesto opprimas (With Hiatus at Diaeresis);
 37
PLAUTUS AND MENANDER

(On Pseud. 724 (Troch.) Si modo mihi (mi?) hominem invenietis ||

Hardly in : see 55.)*
Pseud. 689 (Troch.) Meum mendacium, hie modo quod || subito com-
mentus fui (With Hiatus and Syll. Anceps at the pause ?).
Possibly the unnatural order favoured the unnatural pronunciation here and
in Capt. 527 (Troch.), if the true reading is :
Perdidit me Aristophontes || hie modo qui venlt intro.
The instances of ibi in Plautus are :
Capt. 505 (Bacch.) || ibi vix requievi;
Cas. pr. 73 (Sen.) Maioreque opere ibi serviles nuptiae ;
Most. 125 (Bacch.) Nee sumptus ibi sump||tui ducunt esse;
Poen. 718 (Sen.) Ibi quae relicua alia fabulabimur;
Pseud. 1263 (Bacch.) Neque [ibi] esse alium alii odio ||ibi nee molestum;
that is to say, a mere pair in dialogue-verse. (Occurrences at the end of a
line or a first hemistich — Capt. 490 ; Mil. 879 — are omitted.) There are
some eighty of ibi, e.g. :
Pers. 288 (Iamb.) || nam ibi tibi parata praestost.
So we have reason to increase this eighty by scanning : tonat ;
Amph. 1094 (Troch.) Manibu' puris, capite operto. || ibi continuo con-
Asin. 709 (Iamb.) Postidea ad pistores dabo, || ut ibi cruciere currens ;
Capt. 1000 (Troch.) || illic ibi demumst locus ;
Cas. 872 (Troch. ?) || et ibi licet audacius ;
rogat ; .
Cure. 340 (Troch.) Dico me illo advenisse animi || causa, ibi me inter-
Merc. 326 (Sen.) Ad portum propero, nam ibi mihi negotium est ;
884 (Troch.) || quid ibi faceres ? Quod miser ;
Pers. 404 (Sen.) Sed ibi concrepuit forls. quisnam egreditur foras ? ;
Poen. 1 132 (Sen.) Apud aedem Veneris. Quid ibi faciunt ? die mihi ;
? Pseud. 425 (Sen.) Quo in commeatum volui — argentarium
ProficiscI, ibi nunc oppido obsaeptast via (With
Hiatus in pausa ?) ;
(On Men. 187 (Troch.) Uter ibi melior bellator ||, see 55.)
And we may doubt any restoration of Asin. 1 26, True. 69 which makes the
word an Iambus.

21. Such scansions have nothing whatsoever to do with a regard for

Accent in Dramatic verse. They are indeed pronunciations which
Latin words had assumed through the influence of the accent, but they
would find a place in every kind of Latin Verse, even one (if there were
such a kind) in which there was no regard at all for Accent but merely
for Quantity. And they have nothing to do with Ictus. We press this
.point because the argument has been used in all seriousness : ' You say
38 PLAUTUS AND MENANDER

the Latin accent is a stress-accent and that Plautus shows a propensity

to avoid the clash of ictus and accent. This implies1 that the ictus
had the same amount of stress as the accent. Therefore if Plautus
chose (!) to put the Infinitive of venio at the beginning of a Trochaic line,
it would take the pronunciation ventre. For the ictus would fall on the
short syllable ve- and this might shorten the following long syllable -ni.'
One is at a loss how to treat an argument so foolish. If one is to
answer it according to its folly, one might point out that the shortening
of the final of modo, etc., was the result of a long-time operation of the
Brevis Brevians Law and that modo did not acquire the alternative pro-
nunciation modo on the very first occasion the word issued from Roman
lips ; whereas this argument assumes that at the performance of the play
there was a ' hey presto ' transformation (for this occasion only !) of
ventre to venire. But one feels that the reply is as foolish as the
argument j indeed that any reply must be, which takes so foolish an
argument seriously. And yet this argument has been used in all serious-
ness. Nay, is still used. There are many scholars to-day who read
a Trochaic Septenarius in the Truculentus (line 504) so, (with a Tribrach
as first foot) :
Venire salvum — . Scio, sed peperitn(e), ||obsecro, Phronesium ?
Plautus here, they say, puts the syllable ve- ' under the ictus ' and
makes a Tribrach of the word. Surely a moment's reflection should
Ishow that, since Plautus was a quantitative poet, he must have arranged
his words in the line to suit the actual, recognized pronunciation
(quantity) of each word. He could not capriciously give words and
syllables what quantity he chose by merely putting them in this or that
arrangement ; no more than the line :
An A strian army awfully arrayed
could be capriciously arranged so :
' Austrians in army awfiilly were arrayed ',
and be allowed to give a new and incorrect accentuation to three of the
words. (Really, one can only answer fools according to their folly.
There is no help for it.)
1 It does nothing of the kind. When people speak of the ictus as so forcibly
dominating the utterance of the line, we are reminded of Lever's gibe at this
absurd way of reciting verse :
What will you do, love, when / am going,
With white sail /low-ing,
The seas be-yond?

That's WHAT I'll DO.

 39
PLAUTUS AND MENANDER

When Plautus scans modo, cito, voluptas-mea, etc., it has nothing to do

with the ictus of the verse. He heard the words pronounced every day
with this quantity, and naturally and properly put them with this quantity
in his quantitative verse, assigning to the word such a position as the
quantitative scheme of the line allowed the word to occupy. In one of
his plays there is a character who is ridiculed for his mistakes in pro-
nunciatirabonem
on, ' ' for arrabonem, etc., and when this Mrs. Malaprop
(or rather 'Mr.') appears on the stage he is apt to perpetrate some
blunder which gets a laugh from the audience. To this character, but
no other in all the twenty-one plays, would be suitable a mispronunciation
like ' venire ' or * tacSre ' (Poen. 875) or ' arnica ' (Stich. 696) or ' eamus '
or 'dearum ' or ' duorum ' or 'amorem ' and 'nitoribus ' (Cas. 2i7a) or
1 oportet '— to quote only a few of the monstrosities inflicted on us by
this wrong-headed theory of a ' metrical ' Brevis Brevians Law, that the
sledge-hammer blows (!) of the ictus in Dramatic verse transmogrified
Roman words into pronunciations never heard from Roman lips. We
could add any number of arguments, but— against folly even the gods
fight in vain.
22. Besides the Iambic words already cited for the Brevis Brevians
Law, modo, cito, etc., a familiar example is calefacio (with tepefacio,
liquefacio, etc. ; contrast frTgefacio,fe~rvefacio, etc.). The every-day pronun-
ciation ofcalefacio as calefacio (a result of the short syllable ca- preceding)
became universal (like male, bene) and led ultimately to a still more
rapid utterance, calfacio (with syncope of the now short syllable).
Formerly calefacio (or we had better say calefio, if the theory be right
that the primitive type was calens-fio, Masc, Fern., Neut.) was brought
into line with modo, etc., by means of Lucretius' Tmesis (6, 962) et facit
are = et arefacit). Linguists now incline rather to bring modo, etc., into
line with calefio. The Adverb took its shortened enunciation, they say.
owing to its use" in phrases like modo feci, modo vem\ where it was sub-
ordinated to the Verb, the dominating accent of the word-group falling
on the syllable after the -0, modo-fea, modo-veni, like calefio. That is
fwhy modo, Abl. of modus, is always an Iambus in Virgil, etc., while
{modo, the Adverb, is a Pyrrhic. '
In that case the old statement of the Brevis Brevians Law will not
do : 'An Iambic word may become a Pyrrhic through the incidence of
the stress-accent on the short syllable.' It max rather bejhe accent of
the following 1 syllable that_usuallyputs the law_in_operatjpn. At any
1 In fact, both these types are found, e.g. the first in 'ddmum video, non
aedem', e.g. the second in 'domum-vostram, non nostram'. Of the coinage panttn-
loquium, Merc. pr. 31, the first half would be emphasized, not the second.
4o PLAUTUS AND MENANDER

rate the syllablejhat suffers shortening must jjejan tma^t^djyWable

(i. e. according to sentence-accentuation). That is a necessary condition.
We may illustrate it from the Preposition apud, a Preposition (to judge
from Plautus' scansion) very susceptible of the Brevis Brevians Law.
The word lost all its stress in most phrases (like Greek 7rpos, etc.), apud
me'nsam, apud iudices, etc., apud me' (if the Pronoun were emphatic,
' with me, not with you '), etc., but when preceding another Enclitic
(e. g. the unemphatic Pronoun) became apud me (like Greek -n-pos fie).
The short syllable ap} the ' Brevis Brevians,' seems to have made apud
mensam, apud iudices, apud me (with emphatic Pronoun) the recognized
pronunciation, for this is Plautus' scansion. With emphatic Pronoun,
for example, we have in an Iambic Septenarius (True. 163) :
Sed blande quom illuc quod apud vos ||nunc est apud med habebam
(Read med; I 6). With unemphatic Pronoun, which is of course
of commoner occurrence, apud me, apud te, apud se, apud nos, apud
vos.
The Greek loan-words Philippus (3>iA.i7r7ros), the coin, smaragdus
(oyxa/oaySos), the jewel, apparently took the Latin stress -accent on the
same syllable as the Greek pitch-accent. So Plautus invariably scans
Philippus (40); Martial, either smaragdus (5, it, 1) :
Sardonychas, smaragdos adamantas, iaspidas uno,
or (the literary pronunciation) smaragdus (4, 28, 4 ; Hendecasyllabic) ;
Indos sardonychas, Scythas smaragdos.
(For other examples see 40.)
Cretic-words too were liable to the same shortening, as we see
from Catullus' Hendecasyllabic poem where conversational diction is
followed (10, 25) :

• Quaeso ', inquit mihi, ' mi Catulle, paulum

Istos commodd, nam volo ad Serapim (Deferri ').
To have any doubt about this being the Imperative of the Verb com-
modare ■ to lend ' is to doubt unreasonably, in the face of Catullus'
nescio (85, 2), Horace's Pollio, mentio, dixero and the nescioquis of
classical Latin. The last word-group is interesting. The dominating
accent is on the Interrogative (nescio-quis) like our colloquial (or vulgar)
1 L6rd-knows-wh6 '. And this shows us that the shortening of the
finals of cretic- as of iambic-words may often be due to a following
rather than a preceding accent. It was a word-group in Plautus' time ;
since a Senarius ending like nescioquis venit would, if three words were
used, produce an illegitimate division of an Anapaest (50). But Plautus
 PLAUTUS AND MENANDER 41

did not draw a hard-and-fast line between nescioquis venit and nescio quis
veniat (see III 38 E). Just as the origin of the classical Supine is
revealed by Plautus' usage (opsonatum eo, opsonatu redeo, celer cursu,
pulcher spectatui), so the origin of the classical Pronoun nescioquis.
23. The peculiarities of Plautine Prosody which are due to the Brevis
Brevians Law of Latin Phonetics (e. g. apiid mensam, Philippus, voluptas-
mea ; but never ' mea-voluptas ' nor yet ■ voluptas ' when alone, because
the accent fell on the second syllable of the Noun, nescio), are, as we
have said, not to be called examples of Plautus' wish to reconcile ictus
and accent ; nor are they produced by the ictus of the line. They echo
every-day pronunciation. They are to be classed along with Menander's
7rov '(ttiv for ttov coTii', k ov for kolL ov, iywSa for iyot oTSa, with our ' I'll
go ' and ' he's going,' ■ we're going,' etc. A poem in Punch prefers ' I'll ' ;
an Epic poem only ' I will ' ; for the Punch poem tries to echo every-
day talk.
It is these shortenings which are the most un-Virgilian feature of
Plautine verse and seem most to separate Plautine from Augustan
prosody. And yet, as was said before, if we compare Plautus the early
Comedian, not with the Augustan Epic poet, but with late Mime-writers
like Laberius or (slightly earlier) Pomponius the writer of Atellanae, we
shall find no gulf at all. Comedy at all periods echoes the every-day
talk of the time, and these scansions of Plautus are conversationalisms,
not antiquarianisms. They are hardly even uncouth features of early
verse, peculiarities of pioneers which later refinement removed. Rather
they are the true quantitative reproduction of actual (educated) con-
versation, whereas Epic diction and scansion are more or less artificial
or archaic. When Terence wrote (Eun. pr. 8) :
Ex Graecis bonis Latinas fecit non bonas,
it was not because he could not think of any more accurate setting of
the line, nor because his imperfectly trained ear could not detect the
difference between a Cretic and a Dactyl. He deliberately selected the
scansion bonis as the closest echo of the actual pronunciation of the
word in this sentence where it has sarcastic force.
24. (A) In Later Dramatists.
E.g. Afranius (c. 1 00 B.C.) :
togat. 5 (Sen.) Simul limen intrabo, illi extrabunt ilico ;
25 (Sen.) . . . fateor, sumpsi non ab Kilo modo ;
357 (Iamb.) Voluptatem capio maximam ||;
also 91 (viden ut), 95 (mane), etc.
Pomponius (time of Sulla) :
Atellan. 67 (Troch.) || quid abscondisti inter nates ? ;
42 PLAUTUS AND MENANDER

78 (Troch.) || Venu' veget voluptatibus ;

91 (Troch.) Dixi ego Tllud futurum ; in pruna ||;
93 (Sen.) Quid dedT nebuloni quern pertaesum est pauperem ? ;
131 (Troch.) || neque Mo quod faciam scio ;
Publilius Syrus (time of Julius Caesar) :
? mim. 19 (Troch.) Ut quid frgo in ventre tuo Parthi || sarabaras sus-

25. (B) In Ennius. It is a common remark that Ennius madependerunt the law,?
once for all, for elevated poetry, that these shortenings be admitted only in
final vowels. We may guess, as the reason, that final vowels were more
susceptible, (1) partly because any final syllable was more liable to weakening
(the rhetoricians of the Empire denounce mispronunciations like ' causds '),
(2) partly because a final long vowel might be shortened before an initial
vowel (Illi usque, etc., must have had somewhat the same symptoms as
illhts, etc.). At any rate, whatever the reason, we know for a fact that
final a had been shortened before Plautus' time (terra, mensa, etc.) and that
final o was shortened after Virgil's (mando, curro, etc.).
Virgil allows an exception to this rule, viz. viden in Anchises' dignified
conversation with his son (Aen. 6, 779) :
viden ut geminae stant vertice cristae.
Catullus too, who admits many of the every-day forms of speech, has
viden tit in his Epithalamium (61, yy). Presumably (like male, bene) viden ut
(a shortening due to the ' brevis brevians ' ; contrast aiidin) had asserted
itself too strongly in daily speech ' to be claimed definitely for an Iambus.
Neither Virgil nor Catullus allows vides ; although Horace in somewhat
conversational verse once admits paliis, unless we are to disbelieve not
merely the MSS., which do not go for very much, but the testimony of a
number of Grammarians of the Empire. Persius' rogds seems also well
established (5, 134) :
Et quid agam ? Rogas ? en saperdas advehe Ponto.
And Ennius, we are told by Servius, had used viden ; presumably in his
Annals (though Servius does not say so).
Ennius must, we fancy, have had doubts about excluding apiid (templum,
etc.) from his Epic, since, to judge from the Dramatists' practice, this pro-
nunciation held the field. As has been mentioned, he admits it to the more
homely hexameters of his Hedyphagetica (see I 2) :
glaucumque apiid Cumas.
Indeed, if the suggestion be not too audacious, we question his right to select
out of a sentence like ' enlmvero mod6 veni, simiil vidi, citS vici apiid Cumas '
only two of the five Pyrrhics. He allows enim as a Pyrrhic in Ann. 371 :
Non enim rumores ponebat ante salutem. (In Plautus non enim is normally
a Dactyl when a consonant follows ; III. 42.) Possibly if we had the whole
of the Annals we should find apiid, simul, too.
In his Dramas he did as other Dramatists. His dramatic fragments are
1 Cf. Servius' note on the line.
 PLAUTUS AND MENANDER 43
almost wholly from Tragedies. And Tragedy, while on one side akin to
Epic and elevated poetry, is on another in touch with conversation ; though
the conversation it reproduces is not the careless, flippant talk that a
Comedian puts into the mouth of his characters. We do not find the slurred
pronunciations, due to a J brevis brevians ' or other causes, at all to the same
extent in the fragments of Roman Tragedy as of Comedy. Still they are
there, e.g. in Ennius (frag. 49) Adest, adest fax; (143) Quid h6c hic
clamoris?; (265) Quam tibi ex ore orationem ? ; (praetext. 2) vide. Also
perhaps (20) modls ; (315) malam pestem, etc.
In his Hedyphagetica fragment we found (I 2) bonust ; apiid Cumas ;
scariim praeterii.
And even in his Annals it is not quite true to say that he confines himself
to tibi, etc. (tibi 109, 467 ; tibi 45 ; sibi 101 ; sibl 95, 158 ; quast 104; ubf
302,421; ubl 75, 294. Always ibl. The words cito,modo do not occur).
For one thing, since long vowels still retained in his time their length before
a final -r (e. g. clamSr ad caelum ; III 14), his pater (if the original quantity
was pater, like narqp) must be due to the ' brevis brevians ', e. g. :
113 O pater, o genit5r, o sanguen dis oriundum ;
also his Iuppiter (e.g. 457 Iuppiter hic risit ; see III 14). And that these
should be the only two examples in his fragments is not strange. For
linguists tell us that the stages in the shortenings of all finals were probably
the same ; first, iambic words ; next, cretic words ; then, the rest. (Iambic
and cretic words would get another push from the ' brevis brevians '.) Before
Plautus's time final a of Nouns (III 4) had become a. Linguists say that the
shortening first attacked words like fuga (Greek (pvyrj), next words like
transfuga^ advenat till finally the rest (e. g. terra, mensa) were attacked too.
In Virgil's time final o of Verbs was becoming 6 (III 7). The first words to
yield had been sc/o, etc., then nescio (e.g. nescioquis) and the dixerS of
Horace. The shortening of a long vowel before any filial /, r is not a law
to Plautus, who, however, freely admits dedft, etc., perdidlt, etc., feror, etc.,
differor, etc. The shortenings of cretic words hardly get a chance of showing
themselves in Plautus except in Anapaestics (28), owing to the restrictions of
the use of Dactyls in dialogue-verse. We should have no reason to challenge
them in the Annals ; but the instances are not very strong (as indeed the
instances of tity. etc. are few) :
1 5 memini me fieri pavum (Or fiere, attested for Book Xcf.byad Macrobius
Ann. 354);;
73 pars ludicre saxa
lactam, inter se licitantur (Or Ace. Neut. Sing, of I-stem ?) ;
' 2, 536)
536 SicutI siqui' ferat vas vini dimidiatum (Lachmann, in his peremptory ;
way, denies the possibility, and substitutes sicut ; ad Lucr.

10 1 Virgines; nam sibi quisque domi Romanus habet sas (Emended to

Virgini' or the like).
(381 Hannibal is also open to challenge.) Still, there could be no harshness
in making Dactyls of such words ; or Horace would not have scanned Pollio,
44 PLAUTUS AND MENANDER

dixero, mentio, nor Catullus commoda, nescio. If Serv. auct. (i. e. Donatus ?)
is right in deriving carinor from carina, we have a shortening (of the
calefacio-type) in
564 neque me decet hanc carlnantibus edere chartis,
the unshortened form appearing in 563 Contra carfnantes verba atque
obscena profatus (Hardly Contra carln., with Anapaest for Dactyl or with
Verba beginning the next line).
26. (C) In the Tragedians and Lucilius.
From the fragments of the other Tragedians we take as specimens of the
law, from Pacuvius : (trag. 58) Quid Istuc est ; (236) Possum ego Istam.
From Accius : (trag. 81) Sed angustitate inclusam ; (133) Vel hie qui ; (147)
Quid est cur; (290) datiir bona pausa ; and apparently (538) neque affari.
The violent departure from the literary pronunciation in a Senarius of Livius
Andronicus (trag. 11) :
ClytaSmestra iuxtim, tertias natae occupant,
has been challenged, and defended by comparison of e.g. calefacta. Naevius'
timos (trag. 40) is not quite certain.
Lucilius* trochaic fragments offer: e.g. (603) vide* ; (818) deierat enlm
scripse (more likely than enlm ; cf. 651 At enim dicis clandestino ||; 654 Ego
enim contemnificu' fieri ||) ; (722) Facit idem quod Illi qui inscriptum ||. His
Hexameters, e. g. :
? 137 malas tollimu' nos atque utimur rictu (Emended to ut. (ociu') ri.) ;
377-8 R, non multum abest hoc cacosyntheton atque canina
Si lingua dico ' nil ad me.' nomen hSc illi est (scil. ■ littera
canina ') ;

1216-7 Sic 'apud se ' longe esse aliud, neque idem valet ' ad te ' :
Intro nos vocas, at sese tenet intus (on the distinction of afiud
from ad, and intro from intus).
Lachmann demanded (without justifying his demand) the substitution of
Sicut for SicutT (198; 1029; 1298). With more show at least of reason
situst has been challenged in Lucilius' epitaph on his old butler :
580 (Elegiac) Lucili columella (' pillar of the house ') hie situst Metro-
phanes (situs, MSS. of Donatus on Ter. Ph. 2, 1, 57 ; situs,
situ est, situs est, MSS. of Martial 11, 90, 4).
Certainly ' miserrimum ' (impossible under the Brevis Brevians Law, since
the second syllable is accented) must go from a trochaic line where the
archetype of the Nonius MSS. had miserfnum (like Ital. poverino, etc.) :
733 Ardum, miserinum atque infelix || lignum ; sambucum vocat.
And Aristippum is not more probable than mis[is]se in another trochaic line
(742 || mis[is]se Aristippum autumant).
Linguists tell$ us that Smitto is a weakening of ommitto (i. e. obmitto), and
appeal to mamilla, ofella, etc., as evidence that a double consonant before
the accented syllable was liable to weakening. If so, how can we doubt
Consentius' statement (Gramm. Lat. v, 400, 4) that Lucilius scanned ore
corrupto (corupto) ? The ' brevis brevians ' co-operating with the natural
 45
PLAUTUS AND MENANDER

tendency to weakening would be irresistible. Editors of Lucretius should

not alter :
6, 1 135 An caelum nobis ultro natura corruptum (an imitation of Ennius?).
27. (D) In Plautus. We could not frame a hard-and-fast rule for the
use of * I 'IF and ' I will ', ' he 's ' and ' he is ' (or * he has ') in English ; and
a foreign dissertation which should attempt it would make amusing reading.
But we may perhaps venture to elicit from Plautus' practice at least some
general hints at the actual operation of the law in the slurred pronunciation
of Latin some two thousand years ago. Two thousand years hence an
English Musical Comedy (the Plautine type) will show the studious inquirer
that ' often ' and ( orphan ' were so slurred in Victorian pronunciation as to
be indistinguishable.
(1) In iambic words the most constant examples are those recognized by
Ennius and Virgil, modo, cito, ubf, ibf, tibi, sib?, etc. (see above, 20). Also '
scio (recognized by Virgil), vide (recognized by Phaedrus), cave (recognized
by Catullus, Horace, Propertius, Ovid), volo (recognized by Catullus), dabo
(recognized by Catullus). And the shortening effect of -t, -r which fully
sserted itself a little later is acknowledged in the frequency of fult, amat, I
etc., pater (III 14), amSr, etc. Also preconsonantal enlm and, as was re-
marked before, apud. (On viden see III 13.)
(2) In polysyllables there is no more shortening before a double consonant
(supellectilis, etc.) than before other consonant groups, e.g. pt (voluptatis,
ii^-tem, etc. is the most frequently-shortened qfjdl the polysyllahles). The
theory of Ritschl's time, that the old Roman practice of writing double con-
sonants single (e. g. fajalumed ' facillime ') implied that they were not quite
pronounced as two consonants, is now dead and buried. Nor is the shorten-
ing most favoured before a consonant-group capable of beginning a syllable,
since guber-nator is as common as ege-statis, etc.
(3) A vowel long by nature is rarely shortened in a polysyllable (not a
Cretic; see no. 5). Clear cases are (besides calefacio, etc.) : e. g. Amph. 930
ind Epid. 405 pudlcitiam ; Merc. 846 amlcitiam ; Amph. 903 verecunda (30) ;
Men. pr. 37 Syracusas ; Asin. 552 cicatrices; Poen. 378 verberetillum. Less
clear cases in Terence are Ph. 902 verebamini, etc. Rarely in a word-group
like ' quid amisit ? ' : e. g. Merc. 774 tibi evenit ; Poen. 1078 bene evenisse ;
Most. 131 ubi emeritumst; Mil. 623 tibi In senecta, Stich. 577 lupum In
sermone (Ins-, Inf- was the Roman pronunciation ; on Capt. pr. 49 see below,
Jo). More often in phrases like 'quis haec dixit ? '.
(4) A short vowel before a Mute and Liquid very rarely is a 'brevis
brevians '. (Yet putrefactus seems to have been the later pronunciation.)
Examples (but all have been challenged) are Bacch. 404 Patrem sodalis
(IV 21); Aul. 715 (Anapaestic) obsecro ; Bacch. 1167 (Anap.) probrl (or
probriperlecebrae, a Plautine Compound) ; Bacch. 1041 utram tu. Apparently
1 The ' metrical ' view of the Brevis Brevians supposes that a Roman schoolboy,
announcing that he was at words of three letters, said always ternas scio tarn, and
never ternas scio ianx. What an unnatural apprehension of Merc. 304 (Sen.) !
Lysimache, ternas sci6 iam. Quid 'ternas'? [A,M*0.
46 PLAUTUSAND M ENANDER

qu also acts as an obstacle (and yet liquefacta is used by Ovid), for examples
are almost as difficult to find : e.g. Cist. 698 (Anapaestic) persequar ; Poen.
1403 sequar; Cure. 161 dat aquam quam bibant ; Pers. 792 (Anap.) ferte
aquam pedibus. Examples of equum (ecum), loquor (cf. locutus ; cotidie and
quotidie) are easily found, but are less convincing. (Add, from Terence,
sequar Andr. 819; loquar Ph. 186).
(5) If cretic words become Dactyls not nearly so often as iambic words
become Pyrrhics, we must remember that the sphere for these Dactyls is
almost limited to Anapaestic Cantica (see 28). Very iarely does a Fourth
Paeon (e. g. maritimis, 9) become a Proceleusmatic. Indeed some scholars
declare that such a shortening is inconceivable under a phonetic Brevis
Brevians Law. And yet if * balneis ', pdssimis were occasionally pro-
nounced with -Is, why not l balineis, maritimis ?
(6) Was the shortening excluded from this or that part of the line ?
Various restriction-theories set up by Plautine scholars seem to have dis-
appeared, one after another, under closer investigation. Klotz's (p. 56) rule
for the exclusion of this shortening from the ' even ' feet of Iambic lines and
the ' odd ' of Trochaic has too many exceptions, e. g. Capt. pr. 21 (Sen.) :
Hie nunc domi servat sub patri, nee scit pater.
Nor can the theory that two shortenings were not tolerated in the same
foot (a Proceleusmatic, etc.) be maintained in the face of e.g. Cas. 882
(Anap.) senex abest; Pers. 288 (Iamb.) nam ibi tibl parata praesto est;
Pseud. 569 (Sen.) Novo modo novum aliquid ; 1096 (Sen.) Vide modo ne
illic sit contechinatus quidpiam.
The latest theory (see Appendix E, s. v. Jachmann), that the Brevis
Brevians does not shorten a disyllable before the Caesura, seems to have no
firmer footing than the rest. It proposes to sacrifice the traditional form of
Ter. Eun. 832 (Sen.) :
Scelesta ovem lupo commisisti. dispudet,
to a less pleasing form which lacks Caesura (Scelesta ovem lupo commisti.
dispudet).
It is true that the shortening rarely shows itself in the seventh foot (a foot
from which the Proceleusmatic may be excluded ; IV 10 B) of an Iambic
Septenarius (e.g. Pers. 316 abi atique cave sis; Cure. 520 vendidi ego te).
1 If the y metrical ' Brevis Brevians in its war against the true (phonetic) Brevis
Brevians can still resist all the assaults by facts and common sense, these rare
Plautine shortenings (maritimis, etc.) will be its Mast ditch'. We acknowledge
their difficulty. For pessimls, bonis we can appeal to literary usages of classical
Latin, like Catullus' commoda, nesci6, Horace's dixerd, puta, palus. But not for
maritimis. And why ? Because in the Classical Age the accentuation was mari-
timis, not (as in the time of Plautus and Terence, maritimis ; 18 A). Now Anapaest-
words (e g. variis) or word-endings never became Tribrachs. There is no trace of
e. g. ' Simula ' beside Catullus' commoda, nor of e. g. * aberfi ' beside Horace's
dixerft. Anapaest-words kept their final syllable unshortened, like spondee-words
(simula like manda, aberS like mando) in the Republican Drama as in classical
poetry.
 PLAUTUS AND MENANDER 47
But this is merely the result of two other rules: (1) that a monosyllabic
ending in this metre postulates an Iambus (rarely a Tribrach ; never a
Dactyl) in the seventh foot (IV 10) ; (2) that a pyrrhic word does not take
ictus on the final in Iambic Verse (44), so that a Trochaic Septenarius
ending like lace modo cannot end an Iambic Septenarius. But this has
nothing to do with the Brevis Brevians. Age te is as rare as ego te ; age
modo as impossible as lace modo. On Pers. 290 (uti liceat), Ter. Heaut. 737
(ego hie maneo), see IV 10 B.
It seems as if these shortened syllables were admitted wherever an ordinary
short syllable was admitted. And this means that they were actually short
syllables, that mod5 (a Pyrrhic) was one form of pronouncing the word and
modo (an Iambus) another; just as with us 'I'll' (a monosyllable) is one
pronunciation, quite distinct from ' I will ' (a disyllable). To write modo and
describe the second syllable as * half-long ' is not strictly accurate. Still it is
clear that some shortenings (e.g. Ennius' select list for Epic Verse) had
established themselves more than others.
(7) The progress of calefacio to calfacio in Latin becomes intelligible when
we consider the appearance of these compounds in Plautus. The short
pronunciation seems normal, if not invariable. Here are all the occurrences :
calefacio Epid. 654 (calefieri) ; 674 (calefacit) ; Pers. 105 (calgfieri) ; no
(calefierent) ; calefacio Cas. 400 (calefactabere) ; Rud. 411 (calefactat) ;
commonefacio Stich. 63 (commonefaciam) ; labefacto Merc. 403 (labefacto) ;
madefacto (-cio ?) Pseud. 184 (madSfactatis or madefecistis) ; olefacio Men.
163 (olefeceris) ; Mil. 1255 (olefactu ?) ; olefacio Men. 167 (olefactare) ;
169 (olefacta) ; palefacio Most. 1046 (patefeci) ; permadefacio Most.
143 (permadefecit) ; perpavefacio Stich. 85 (perpavgfaciam). Thus, out of
sixteen occurrences, only one, Men. 167 (for we cannot admit Most. 112
piit[r]efacit) shows long e ; and since the reading of the line is not above
suspicion, it may be that these compounds had become shortened in every-
day talk as definitely as bene and male. (Still, Terence allows obstupefecit.)
If so, they had actually reached a further stage than ego, for ego is by no
means uncommon in Plautus (III 31). While, e. g., amaloris (-i, -em, etc.)
was continually being rescued from the 'brevis brevians' by the help of
amalus, amare, etc. (not to speak of amator Norn.), there was nothing to save
calefacio and similar compounds. In True. 325 the ' amatores ' of two MSS.
seems a wrong reading, for the best MS. has amantes ; and in another line
True. 46 ' amatori ' can be avoided without defiance of the MSS. (38). The
'inafidivi ' of Mil. 211 cannot stand ; we must substitute the Plautine indau-
divi. An example of the calefacio-type occurs in Ennius1 Epic :
Ann. 558 Inde patefecit radiis rota Candida caelum.
(8) Perhaps we might use as a test of the shortenings most fully recognized
by Plautus his admission of them to the final foot of acatalectic lines, e. g. :
Mil. 925 (Iamb.) Qui noverit me quis ego sim || (Cf. Cure. 104) ;
Epid. 332 (Iamb.) Alicunde ab aliqui aliqua libispes ||;
Bacch. 1068 (Sen.) Hoc est incepta efficere pulchre : vel uttmi (Evenit)
(cf. Pseud. 164) ;
48 PLAUTUS AND MENANDER

True. 163 (Iamb.) Sed blande, quom illuc quod afiiid vos || nunc est
apud med habebam ;
Epid. 94 (Cretic Monometer) At enim tu ;
and of catalectic, e. g. :
Poen. 298 (Troch.) || meretricum aliarum, *'£?tibi ;
Mil. 1 138 (Troch.) || volumu' conventum. Et ego vos (cf. Mil. 1429) ;
? 1319 (Troch.) || homini pietas— Sao; sapis (Of doubtful reading) ;
Capt. 558 (Troch.) || gliscit rabies. <r^/tibi (cf. Men. 934; Pers. 316) ;;
Mil. 1335

Men. 416 (Troch.) || limen. Quin tu tac$ modo (cf. Poen. 906) ;
True. 952 (Troch.) || Philippicense est. tenetibl ;
True. 832 (Troch.) || qui improbust, si quasi bibit.
But we doubt the wisdom of extending this test to the penultimate foot (e. g.
to take two examples from Senarius-endings in the prologue of Plautus' first
play: Amph. 125 in exercitum ; 140 ab exercitu) or of going so far as to
exclude the possibility of e. g. tuo, mihln in the endings (Cas. 230, Iambic)
tarn tristem tub Iovi ; (Cure. 422, Sen.) mihin ? Ita. And we doubt
whether Marx's elaborate details (Zwei Auslautsgesetze ; see App. E) of this
shortening in the antepenultimate foot prove anything more than the fre-
quency ofad Ulum, etc. (31) in the Roman Comedians' lines. (On the great,
not to
IV 27.) say excessive, use of the Brevis Brevians in Plautus' Anapaestics see

28. (E) In Terence. There are only half-a-dozen plays of Terence, but
a score of Plautus. So it will not do to compare the lists of words shortened
under this law in Terence and in Plautus and to assert that each and every
shortening in the Plautus list (even though it appear only once or twice in
the twenty-one thousand lines) which is not in the Terence list must have
gone out of fashion in Roman pronunciation between Plautus' and Terence's
time. We must use discretion in making the comparison. Perhaps it is not
out of place to remark that the younger poet has another device for echoing
talk, the use of incomplete sentences (e. g. Andr. 149 non tu ibi gnatum ?).
The Plautus list shows far more examples like pessimis, ' balnels '. But
they come mainly from Plautus' Anapaestic Verse, a metre eschewed by
Terence for some reason or other. There is hardly room for them in Iambic
or Trochaic verse where such a division of a foot (• • ^ w,— ), as * balnets est\
etc., is not allowed (50; 55); e.g. not ■ MaorzV vortit barbare'. When
Terence writes Cantica in other metres he uses (Ad. 610 sqq.) objid (Ini.),
consili (Gen.), suspicio (Norn.). Even in the Trochaic Hec. 281 Nemini.
The example already cited (Eun. pr. 8) :
Ex Graecis bonis Latinas fecit non bonas,
shows that he is far from limiting himself to the select list admitted by
Ennius to Epic Verse. In Terence's as in Plautus' time pyrrhic scansion
was felt to echo the actual sound of the emphatic Iambus-word, even when
it ended not in a vowel but in -s (or any other consonant). Terence's bonis
is like Plautus' :
 PLAUTUS AND MENANDER 49
True. 677 (Sen.) Novos omnes mores habeo, veteres perdidi ; Au l. 234) ;
Pseud. 812 (Sen.) Boves qui convivas faciunt herbasque oggerunt (cf.

860 (Sen). Manum si protollet, pariter proferto manum ;

Merc. 694 (Sen.) Decern si vocasset summos ad cenam virosobsonavit),
(Nimium

where the scansion seems to mark the emphasis sufficiently without the help
of the ictus.
And this bonis of Terence shows the absurdity of comparing the shortened
words, all and sundry, of the two Comedians. It is merely the emphasis
attached to the word in this sentence which has produced this scansion. If
Terence had been speaking of a Latin improvement, not deterioration, of
Greek plays, quite another word would have appeared in our list, malls.
The words suitable for comparison are those which by virtue of their own
nature invite shortening, e. g. Adverbs like mod5, Imperatives like cave" (used
as a mere Prohibitive Particle in a sentence like Plaut. Poen. pr. 113 Cave
dirumpatis), etc. The point to notice about this line is not the shortening of
the word bonus but of the syllable -Is in an emphasized disyllable, any
(iambic) disyllable. For we can imagine some Comedians hesitating to
shorten a syllable of this kind.
His few examples of the calefacio type conform to "Plautine usage (e. g.
patefecit Hec. 303 ; patefit Ph. 825), except obstupefecit in a line of not quite
certain reading :
Ph. 284 (Sen.) Ita eum turn timidum ibi obstupefecit pudor.
29. A distinction of Quantity is more a feature of some languages than
others. Dr. Bridges' experiments in English Quantitative Verse have
brought home to us the difficulty of deciding whether this or that
English syllable is a long syllable or a short. Had the pioneers in
Latin Quantitative Verse anything like the same difficulty ? Something
of the kind probably, but not nearly so much. That they made an
honest attempt to conform their scansions to actual utterance is shown
by their fastidiousness. Plautus seems not to favour agro, etc., aquas,
etc., verebamini, etc., quid amisit ?, etc. ; Ennius admits only a select
list, and even these sparingly, to the ' large utterance ', the deliberate,
dignified tone of Epic poetry; Lucilius makes a compromise between
the two. It is a thousand pities that the non-Epic hexameter verse of
Ennius has not been more fully preserved (with its apud Cumas, etc.).
It would have given us a glimpse at these quantitative pioneers in their
task of measuring the Roman syllable. And it would probably have
spared us the infliction of that silly theory of a ■ metrical ' Law of
Breves Breviantes, that syllables were shortened in Dramatic verse by
the force of the metrical ictus, surely the silliest theory that ever led
respectable scholars astray.
2348 E
5o PLAUTUS AND MENANDER

How ever did such a theory arise ? Perhaps we can account for it by
the great unison of metrical ictus and natural accent in the verse of
dialogue. The phrase apiid Cumas had the accent on the third syllable.
At the beginning of an Iambic (not Trochaic) line the ictus falls on the
same syllable. The practice (now going out) of using the accent-
symbol to indicate the ictus in editions of the Dramatists presented to
the reader, e. g. :
Apud Cumas adstat imperator imperans,
and suggested that ictus had the same stress and the same dominance
of a word as had accent. There seemed to be two equally possible
ways of explaining the short second syllable of the line, ictus or accent.
And in an evil hour the wrong choice was made. And there were other
prevalent errors which lent strength to the delusion. It was not recog-
nized that Plautus avoids a Pyrrhic Caesura of a Tribrach in Iambic
verse (see 44) like * modus erdX \ * modtf fuit ', with ictus on the second
syllable ; hence the absence of a scansion like ' modo fuit \ with ictus on
the second syllable, was thought significant. The presence of ictus
on the second syllable was believed to produce modo, and its absence
(or rather its presence on a neighbouring syllable) modo. Again,
support seemed to be offered by Anapaestic Verse, where such a phrase
as tanius ibi would be nearly sure to exhibit the scansion ibi. The
reason was not detected, viz. that a division of the Anapaest like tantus
ibi c/ientsirum (Poen. 1180) || clues, at ego cu/inae clueo (True. 615),
is hardly ever found in this metre owing to the recurrent Diaeresis
(IV 27). But perhaps the chief fault lay with the early editors of
Plautus. Before Leo's monograph (and, alas ! not only before) Cantica
were often analysed and scanned most absurdly. An Anapaestic line
was found, e.g. in Cas. 217*:
Omnibus rebus ego amorem credo et nitoribus nitidis antevenire
(nee potis).
The only Procrustes that could distort amorem into ' amorem ', nitoribus
into ' nitoribus ' seemed to be the ictus. Once let its sledge-hammer (!)
fall on the last syllable, ' amow// ', and the second last syllable,
' nitor/bus ', and the thing seemed to be done. And indeed so long
as there appeared no other way (IV 32) of scanning the line (for
Plautus' wealth of metre — his ' numeri innumeri '—was not recognized
then), even saner scholars had to acquiesce. For undoubtedly amorem,
nitoribus was the only conceivable Roman pronunciation. If the actor
had to articulate ' amorem ', ' nitoribus ' in this line, the inference must
be that actors were forced to mispronounce Latin words. The result
 PLAUTUS AND MENANDER 51

was that even in the verse of dialogue, where such mispronunciations

should be inconceivable, any shortenings were tolerated so long as a
short syllable happened to precede, even though it belonged to a different
speaker ; therefore the traditional text was left unemended of a Senarius
like Merc. 683 :
sy. Dorippa, mea Dorippa. do. Quid clamas, obsecro ?
(Read Quid nunc, obsecro ?, or the like.)
The fourth foot was allowed for an Anapaest -pa. Quid cla-.
Of course only a (supposed) ' metrical ' Law of Breves Breviantes
could effect such a shortening. In actual Roman pronunciation the
phonetic Law of Breves Breviantes would operate only in single words,
in word groups, in phrases where one accent dominated and reduced
the other words to subordinate stress, e. g. modo, voliiptatem, voliiptas
mea, quis accurrit?, quis hue currit?, etc. So that this line of the
Mercator is a good test for discriminating a scientific and rational
application of the Breves Breviantes Law from a mere 'theory on
paper' (see 32, end) that has no regard to the actual operation and
conditions of phonetic change.
In Cas. 242 omnia quae tu is hardly possible. The line is either Ana-
paestic or corrupt. In Bacch. 725 belhts atque ut esse is hardly possible.
Read bellust atque ut esse. In Cist. 516 Troch. || patruus [et] summus Iup-
piter ? In Pers. 769* manibus, [ap]ponite.
This rouses an uneasy suspicion. How many other lines are there
where our MSS. have gone astray and left a scansion which, though it
can possibly be adapted to a phonetic Law of Brevis Brevians, would
have made Plautus ' stare and gasp ' ? If we had an immaculate text of
Plautus we should have a reflexion of the intonation of the Latin sen-
tence inhis time, its enclitic or subordinate words, etc. Now subordinate
words are not always what they seem. Phoneticians tell us that the
word ' part ' is an Enclitic in a phrase like ' some parts of England \
The word 'some' dominates it (as it dominates 'thing' in 'something').
Who would think of a Noun like ' part ' being an Enclitic like ' and ',
'or', 'is'? In the phrase 'and, Sir' the Noun ' Sir ' is enclitic, the
Conjunction takes the accent. Who could have guessed this ? Latin,
too, must have had words (18 D) which, though not included among
Enclitics (like -que, -ve) in Grammars, really were Sentence-enclitics (at
least in certain phrases). Now and then we are pulled up by a
suspicious shortening in a line of Plautus and have to ask ourselves :
' Was this a Sentence-enclitic to a Roman ? Or is the text unsound ? '
For example, what of Mil. 55 (Senar.) :
Quid tibi ego dicam quod omnes mortales sciunt ?
e 2
52 PLAUTUS AND MENANDER

In the phrase ' every mortal thing ', ' all mortal men ', could the word
• every ' or ' all ' be subordinate in Latin ? Apparently it could ; though
it could not in a phrase like 'some, not all', 'not all, but only a few ',
e.g.:
Pseud. 534-5 (Sen.) non unum diem,
Verum hercle in omnes quantumst ;
Rud. 1099 (Troch.) Ut tute es, ita omnes censes || esse, periuri
caput.
And what of the defence of simillumae in Asin. 241 (Troch.) ?
Portitorum simillumae sunt || ianuae lenoniae :
Si adfers, turn patent, si non est || quod des, aedes non patent.
It has been ascribed to the shifting of the accent in the word-group
simillumae-sunt. The shifting would, it is said, take the form simillumae-
sunt and leave the second syllable with only a secondary accent and
therefore capable of being shortened by a ' brevis brevians ' ? A scan-
sion like Quid abstulisti? (cf. Ego opsonabo Bacch. 97, with emphasis
on Pronoun) suggests that a mere secondary accent could not always
resist a ' brevis brevians '. That the Substantive Verb was an Enclitic
and that the accent would or might be shifted in such a phrase is not to
be denied. But the justification seems somehow more acceptable in a
case like Mil. 69 (Senar.) :
Mol&stae-sunt, orant, ambiunt, exobsecrant (cf. Trin. 830),
and editors usually leave this traditional reading unchanged, molestae-
sunt like voluptatem. Sometimes an objection is raised : Why then do
you allow, e.g., 'nam molestae sunt mihi' to end a Senarius? The
answer is : Because these shortenings are not the invariable pronuncia-
tion, but only an alternative pronunciation. Both voluptatem and
voluptatem are admitted by Plautus, both modo and modo, etc. Why
precisely Plautus chose the one scansion here, the other there, we can
seldom guess. Why do English or German writers use now ' He 's
going ', ' Er hat's gethan ', now the fuller forms ?
30. If we consider how our Dramatic Verse would suffer if conversa-
tional forms like ' I '11 ', ' he 's ', etc., were excluded, we shall realize what
a help to Latin Comedians must have been this variety of scansion
which is so great a hindrance to their modern readers. The nuance of
thought, the touch of flippancy, the tone of excitement expressed by
these shortenings are, no doubt, very often lost to us. But not always.
The Iambus iube in Capt. 843 clearly is appropriate to Ergasilus' pause
for deciding what he wishes to be ordered. We can realize what a
shout there must have been at the word magna in the Trochaic line,
 PLAUTUS AND MENANDER 53

True. 702, where no less than three preceding words seem to be hurried
over in the effort to reach it :
Ita ad me magna, nuntiavit || Cyamus hodie gaudia,
a line which forms an exception to the usual rule that the accent falls
on the neighbour to the shortened syllable. (There was, however,
presumably a secondary accent on Ita, like the secondary accent on the
first syllable of calefacio, tepefacio, etc.) We feel a somewhat different
effect from the more usual type of emphasis, e. g. Pers. 380 (Sen.) Et
ut vi surrupta fueris, etc.
It is hardly right to find laxity in Plautus' imitation of Menander
because the Latin lines admit such a variety of scansion. Lines which
seem to us capable of being scanned variously, would perhaps not so
seem in Plautus' time. Indeed, our investigation in this book of the
minuter details of Latin Verse removes many an alternative ; e. g. shows
that only a Tribrach (not an Anapaest) is to be found in Pseud. 645
(Troch.) :
|| res agitur apud iudicem (cf. 48) ;
that the only scansion of Amph. 903 (Sen.) is with the first foot an
Iambus :
Nimis verecunda es. Potin ut abstineas manum ? (cf. 44)
and so on. (Index B contains a list of the lines whose scansion has
thus been determined.) When we remember how the Latin of Plautus'
time subordinated the Antecedent to the Relative (e. g. Amph. 1009
Naucratem quern convenire volui in navi non erat. As if 'quern
Naucratem') we see how natural is this shortening in Gymnasium's
emphatic protest 'at its very source' (with strong stress on oritur),
Cist. 62 (Troch.) :
Indidem unde oritur facito facias || stultitiam sepelibilem.
(Cf. Epid. 448 ; Cure. 419 Sed istum-quem-quaeris.)
31. Illustrations of the Brevis Brevians To go through all the
doubtful cases of the Brevis Brevians Law in Plautus and decide whether
this or that shortening is possible or the text corrupt ' is rather the duty of
an editor or a grammarian (in his chapter on the Latin Sentence-accent).
We prefer to give instead some illustrations of the way in which the opera-
tion of the law is to be seen in dialogue-verse and how, in our opinion, some
typical lines should be scanned. The list is confined to the first eight plays.
1 One emendation is so obvious that it may be mentioned here. Omit the
recalcitrant l Philoxene*' in Bacch. 1106 (Anap.) :
[Philoxene,] Salve. Et tu. Unde agis? Unde homo || miser atque infortunatus.
How the omission improves the line ! We get a true Plautine symmetry of a pair
of Spondees with a pair of Dactyls.
54 PLAUTUS AND MENANDER

Amph. 89-90 (Senarius) Quid ? admiratin esti', quasi vero novum

Nunc proferatur Iovem facere histrioniam ?
(Emphasis on Iovem) ;
410 (Trochaic) Quid, domi'tm-uestram. Ita entmvero || (with
Hiatus at change of speaker. Vestram has the stress) ;
518 (Troch.) Carnifex, non ego te novi ? || abin e conspectu
meo ? (With Hiatus at Diaeresis. Abfn is normally
Inf.);
a Pyrrhic, like viden) ;
721 (Troch.) Verum tu wa/tfw-magnum habebis ||;
751 (Troch.) Specto. Vera void loqui te || (Volo normal with

989 (Iamb.) Ego sum Iovi dicto audiens ||*(Emphasis on Iovi) ;

1050 (Troch.) Seu patrem sive aviim videbo || (Plautus does
not favour patrem, etc. ; 27) ;
Asin. 32 (Sen.) Quid istuc est ? aut ubi istuc est terrarum loci ? (The
stress of Interrogatives made the accentuation of
such phrases quid-istuc, ub(i)-istuc) ;
509 (Troch.) || matris-imfierium minuere ;
720 (Iamb.) Opta id quod-/?/-contingat tibi || vis. Quid siEveniet
optaro ?;

752 (Sen.) Lenae dedit-dono argenti viginti minas ;

Aul. pr. 32 (Sen.) Sib(i)-&tw£w-poscat, etc. ;
47 (Sen.) Ut-incedit/, etc. ;
53 (Sen.) Oculos hercle ego istos, improba, effodiam tibi (Accent
on the first, ego, of the enclitic group, ego istos) ;
87 (Sen.) Araneas mihi ego Was servari volo (Ditto) ;
198 (Troch.) Ego ft/to novi polypos qui ubi || quidquid
tenenttetigerunt
(Ditto) j

814 (Troch.) Credo ego ilium, at iussi, eampse anum adiiss(e)

(Ditto) ||
;

234 (Troch.) Asini me mordicibu' scindant, || doves incursent

cornibus (Emphasis on boves) ;
321-2 (Sen.) Sed uter vostrorum est celerior ? memora mihi. (Em-
phasis on Interrogative) ;
Ego, #/-multo-melior, etc. (Main stress onon Ego melior)
and ;

460 (Troch.) Illic hwc-abxit. di immortales ! (Accent on the first,

Illic, of the enclitic group, Illic hinc) ;
673 (Sen.) Nunc hoc ubi-dbstrudam cogito solum locum ;
Bacch. 44 (Troch). Id, amabo te, huic caveas. Quid isti || caveam ?
(See above, on Asin. 32) ;
885 (Sen.) Quid ?7/z molestu's ? quid ilium morte territas ? (Ditto);
97 (Troch.) Ego-obsonabo, etc. (Emphatic Ego) ;
46 (Troch.) Nam si haec habeat aurum quod till || renumeret
(See above, on Aul. 53) ;
897 (Sen.) Neque osculatur neque illud quod dici solet (Ditto) ;
 PLAUTUS AND MENANDER 55

Capt. 69-70 (Sen.) Iuventus nomen indidit ' Scorto' mi, eo

Quia invocatus soleo esse in convivio.
(Since the in- of invocatus could hardly be slurred without injury to the pun,
this seems a better arrangement than E6 quia invocatus ; IV 3 D) ;
617 (Troch.) Nunc ego tater-sacrum saxumque || sto (Main
stress on sacrum and on saxum) ;
716 (Sen.) Quia ilii fuisti quam mihi fidelior (Not quia illi, for
the Demonstrative is emphatic ; though Illic is
possible ; 18 C) ;
Cas. 463 (Sen.) Ut tibi morigerus hodie, ut voluptati-i\x\ ! ;
488 (Sen.) Satin docte ? Astute. Age modo, fabricamini (With
Hiatus at second change of speaker. Satin is normal);
536 (Troch.) Sed tecum egreditur, senati || eolumen (With Hiatus
after Interjection. Sed Sccum is common) ;
762 (Sen.) Seni-nostro et nostro Olympioni vilico ;
764 (Sen.) Senex in culina clamat, hortatur coquos (Common Iam-
bus-words like senex, pater are very often Pyrrhics) ;
970 (Troch.) Nunc ego inter-sacrum saxumque || sum (See above,
on Capt. 617) ;
Cist. 39 sqq. (Iamb.) Meretrices fuimus : ilia te, || ego hanc mihi educavi
Ex patribu' conventiciis. || neque ego hanc super-
biai (Unless emphatic, as in preceding line)
Causa pepuli ad meretricium ||quaestum, nis(i)-#/-
ne-esurirem ;
120 (Sen.) Idem miMst quod magnae parti vitium mulierum (Em-
phasis on mihi) ;
136 (Sen.) Recens-nzXum, eapse quod sibi supponeret ;
292 (Sen.) Qui equum me adferre tubes, loricam adducere (Adferre
tubes would be a false Anapaest ; 49) ;
366 (Iamb.) //^Vperditui et praedatui ||;
Cure. 193 (Troch.) Quod quidem mihi polluctu' virgis || servu' sermonem
serat (Emphasis on mihi) ;
639 (Sen.) Et istae me heredem fecit. O pietas mea ! on (SeeAul.
above,
53).

It will be seen how normal is the shortening tile, iste, when these pronouns
are not emphatic and are preceded by a short syllable (et, sed, ad, in, pol,
ut, etc.). Contrast the emphatic :
Men. 746 (Sen.) Si me derides, at pol Ilium non potes ;
Cas. 431 (Sen.) Ne ea mihi daretur atque ut Illi nuberet.
(Other instances in III 35.)

32. (A) Lastly, before leaving the Brevis Brevians, a word of caution
against misconceptions. The old theory of Ritschl's time, that a pyrrhic
boves, canes, fmt, amat, pater, amor, etc., attested a Roman Nom. Plur.
suffix -es like the Greek and proved that the shortening of a long vowel
56 PLAUTUS AND MENANDER

before -t, -r had already asserted itself by Plautus' time, need hardly be
mentioned. It may be left to slumber in its grave (though readers of
Neue's Formenlehre should be on their guard against it). But we
suspect that the widely spread notion that ' ille ' was a by-form of ille
(just as the first syllable of toiovtos may be long or short) is due to the
frequency of Plautine scansions : et ille, sed ille, ad ilium. People
forget that this shortening is found only when a short syllable (a * brevis
brevians ') precedes the Demonstrative (III 33).
The phantom of a ' metrical ' Law of Breves Breviantes, whereby the
mere metrical ictus ctmld produce these shortenings, has led some
editors who forget this necessary condition— after a short syllable — to
accept some strange scansions. For example, Leo has perpetrated
(Stich. 21) Fac quod tibi (a Proceleusmatic), apparently supposing that
the absence of ictus from fac allowed the word to be scanned as a short
syllable. (On his curious idea of quod see III 10.) In the second
issue of his Forschungen he has recanted his heresy1 of pgrdidi
(p. 318 «.), etc., in favour of the correct perdidi, etc. Had his text
reached a second edition in his lifetime it would probably have
discarded such strange scansions as (Cist. 463) At ego nee do neque te
amittam hodie (' dimeter anapaesticus ').
The theory that a word retracted its accent when the last syllable was
elided, so that e. g. molest(ae) would become accented molest(ae) and
might be reduced by the 'brevis brevians' to molest(ae), we have
already seen to be baseless (18 B). This theory defends the manuscript
reading Arnica (Stich. 696).
Baseless, too, is the (strangely wide spread) belief that in the first
foot of a line an accented syllable may be shortened by a Brevis
Brevians. Why should a word be mispronounced in the first foot?
This theory takes no umbrage at the manuscript readings Venire (True.
504), Dedisse (Amph. 761), etc.
And statistics (especially in University students' dissertations) of the
Brevis Brevians Law have often been vitiated by the curious belief that
it was the metrical ictus (not the pronunciation of the sentence) which

1 In his first edition (p. 292 n.) he says of such Cretic words that 'the third
syllable has a secondary accent '. This statement is often made. It has been used
as an argument against any accentual theory of Saturnian Verse which ascribes two
(and only two) accents to the second hemistich ' Naevio poetae ' (like ordine
pomintur', < insece Camena'). But what shred of evidence is there for this
secondary accent? In Catullus' commoda (Imperative clearly), nescifi, Horace's
dixero, mentiti (to say nothing of the older poets' scansions) is proof positive that
there was no secondary accent on the last syllable of Cretic words in Latin
pronunciation.
 PLAUTUS AND MENANDER 57

determined the law's operation. Suppose a trochaic line to begin with

Et modo properat, modo cessat. This would be claimed for an instance
of modo followed by another of modo. Since the ictus fell on the third
syllable of this line, Et modo was assumed (without any proof; 55) to
be the only possible pronunciation. No reason was produced against
Et modo (with the second foot a Tribrach, not a Dactyl) like Et bene
properat, like Femina properat, like O bone, properas. The same
assumption was, still more unreasonably, extended to words ending in -s
(preceded by a short vowel), e. g. Et bonus properat. The possibility
of a Tribrach for the second foot (Et bonu' properat) was absurdly ruled
out (see III 17 C).
A final -u', -i' produced by the faint pronunciation of final s is never
a ' brevis brevians ', so that e.g. surdu' sum (Pseud. 255) is not possible.
The line is a Bacchiac Tetrameter : Surdus sum profecto ||inani (' empty-
handed ') logistae. Read in Stich. 686 (Troch.) Quisq' praetereat
(Quisquis MSS., instead of Quisque; III 38 D). In Cure. 170 (Troch.)
Ipsu' se excruciat, with Dactyl in first foot, is impossible. In Poen. 823
(Troch.) ||ei(u)s neglegere gratiam (not ' eiu' negl.'). Stich. 67 cannot
be trochaic, with Siqui' me quaeret.
And need we add that a syllable, itself shortened by this law, cannot
proceed to play the part of a ' brevis brevians ' ? Quis accurrisset ? never
produced a • quis accurrisset ? '.
Anapaest-words never shortened their final syllables as did Iambus-
and Cretic-words. Never e. g. * redeo ' but only eo, exeo. Thus in
Terence's Cretic hemistich (Andr. 629) :
||genus hominum pessimum,
the scansion hominum is inconceivable (IV 25). (See also IV 8.)
Nor anapaestic word-endings, syllable-groups, &c. For example,
attolle is hardly conceivable in Cas. 815 (Vers. Priap.) :
Sensim super attolle li||men pedes, nova nupta (IV 33).
But no objection can be taken to e. g. et ego vos ; at enim tu ; vel utf
mi, &c, for these are shortenings of iambic words rather than of ana-
paestic word-groups. Nor presumably to e. g. interibi (like ibi). The
phonetic justification of redeo beside eo lies outside our province. Any
investigation must take account of the different treatment of it&ne beside
slcin(e), although ita^(^) is of course not unknown ; also of redeamus
beside eamus (or eamus).

(B) Brevis Brevians in lines divided between two speakers. A moot point
is whether, when a line contains the remarks of more than one speaker, the
final syllable of a remark can be a ' brevis brevians '. Our own view is that
58 PLAUTUS AND MENANDER

such shortening is unreasonable, and we would emend or explain away the

sust) ;
apparent examples in the traditional text of Plautus, viz. :
?? Amph. 429 (Troch.) || hirneam. Ingressus est viam (Certainly ingres-

? Asin. 483 (Iamb.) Tibi quidem supplicium, carnifex, || de nobis detur?

Atque etiam (False Proceleusmatic ; 52 ; IV 10) ;
? Bacch. 78 (Troch.) Sci5 quid ago. Et pol ego scio quid || (metud)
(Rather et) ;
?Capt. 231 Scio. At scire memento quand(o) || id quod vales habebis
(Rather scio at ; Anap. Dim. Cat. with Iamb. Dim. Cat.)
(Nam fere maxima pars morem hunc || homines habent :
Ac at.) ;sibi volunt) (Anap. Dim. Cat. with Iamb. Dim.
quod

? Cas. 634 (Choriamb.) Vae tibi. Immo istuc tibi sit. Ne cadam,
amabo, tene me (Immo is often Pyrrhic ; III 58) ;
75 5a (Col. Reiz.) (OL. quin tu i modo mecum)
Domum. LYS. At pol malum metuo (AP). (Possibly A, the
1 nota personae ' of Lysidamus, was miscopied at) ;
? Cist. 95 (Troch.) Coepi amare contra ego ilium et || ille me. O mea
Selenium (Rather ill' me ; 36) ;
Merc. 683 (Sen.) Dorippa, mea Dorippa. Quid clamas, Anapaest obsecro ? ;(False
49) ;

751 (Sen.) Sed — Interii ! Quid ais tu ? etiamne haec illi tibi
(Possibly sed is a mis writing of st 'hush ' used already by
Lysimachus
in line 884) ; in line 749, and here put ' extra metrum ' as
Stich. 660 (Sen.) Stiche. Hem ! Quid fit ? Eugae ! Sagarine lepidissime
(Since Interjections have often no fixed quantity, hem may
be a mere spelling of he in this line, unlike Ter. Phorm.
682. And the Interjection em is absorbed in ellum for em
ilium ; III 52. Cf. Ter. Heaut. 380.) (Or rather Stiche
may be ' extra metrum ', like Dorio in Ter. Phorm. 485.)
(On Ter. Heaut. 978 see III 49.)
But we acknowledge that something may be said on the other side. It
may be said that this is a literary convention, and conventions are not always
reasonable. Is it reasonable that the second syllable of nullas spes or nulla
spe should be no longer than the second syllable of Catullus' nulla spes
(64, 186)? Yet literary convention assigned to each of them precisely two
* morae '. If we turn the cold light of reason on the convention of Greek
Dramatic lines divided between two speakers, which requires Elision wherever
possible, its ludicrous side is revealed. Was a speaker not allowed to finish
his last syllable? Irresistibly one is reminded of that passage in The
Critic :
For all eter —
— nity — he would have added, but stern death
Cuts short his being, and the noun at once.
 PLAUTUS AND MENANDER 59

To insist on reasonableness would be here unreasonable. When Greek

dialogue-verse was published in book form, the lines were adapted to the
reader's point of view as much as to the actor's. When read (aloud) as lines
of poetry the elisions became natural enough. May not the same be true
of the Brevis Brevians in these lines of Plautus ?
To this argument we reply that the Greek convention was not followed by
Plautus. He allows Hiatus, as well as Elision, between the remarks of
different speakers. And elsewhere (III 49) we state that Hiatus implies
something of a pause, since Syllaba Anceps (i.e. the pause over a short
syllable which adds to it another ' mora *) is used in the same circumstances.
This statement, however, may be turned against us here. If in Cas. 755* :
OL. Domum. lys At pol malum metuo,
the elision of the final syllable of Olympio's remark indicates that Lysidamus
made no pause before replying, may not the operation of the Brevis Brevians
be Plautus' device for showing that the reply was snapped out instantaneously
— in fact, as if it were part of Olympio's utterance ; just as a sharp retort
often continues the utterance of another person, e. g. Capt. 885 :
vae aetati — Tuae,
Quippe quando mihi nil credis,
where Ergasilus (to use a homely phrase) takes the words out of Hegio's
mouth.
Our readers have heard the plea and the counter-plea, and we leave to
them the verdict.
33. OTHER CONVERSATIONAL PRONUNCIATIONS. Before
resuming our comparison of Plautus' verse with Menander's, we had
better finish this topic, the conversationalisms (like voliiptatem, ad ilium)
of Plautine Prosody, and get them out of the way.
(A) Synizesis. In Menander a short e is sometimes fused with
a following o-sound (0ca>v and 0eovs, monosyllabic ; 7rpa^€<oi/, disyllabic ;
cf. la, monosyllabic), a Greek parallel to the conversational pronuncia-
tion of eosdem, eodem, easdem, &c, as dissyllables. This was the
pronunciation of the Augustan Age as well ; witness Propertius' famous
lines on the ghost of Cynthia (4, 7, 7-8) :
eosdem habuit secum quibus est elata capillis,
eosdem oculos : lateri vestis adusta fuit.
Even Virgil did not fear to recognize it, although, probably by design,
he left a possibility of appeal to the Law of the Brevis Brevians (Eel. 8,
81 Uno ebdemque igni ; Aen. 12, 847 Uno ebdemque tulit ; 10, 487
Una eademque via). Ennius admits it to his Annals :
200 eorundem me libertati parcere certumst ;
Lucilius to his dactylic verse :
1191 Hunc catapiraten puer ebdem devoret unctum
60 PLAUTUS AND MENANDER

(see Class. Quart. 5, 97), as well as to trochaic (885 ebdem uno hie
modo rationes ||) and iambic :
754 Adde ebdem tristes ac severos philosophos;
Porcius Licinus in his trochaic lines on Terence offers an example :
ebrum ille opera ne domum qui||dem habuit conducticiam ;
Lucretius, many ; e. g. r, 306 :
Uvescunt, eaedem dispansae in sole serescunt,
— evidence enough surely that the slurred pronunciation of Plautus' time
persisted as late as Virgil's. Examples need not be given of this
scansion in Plautus and the other Dramatists ; it is so common. But
we may mention the Synizesis in eum in a fragment (inc. 3) of Accius'
Annals, preserved by Macrobius :
elimque diem celebrant per agros urbesque per omnes,
where the Naples MS. too of Macrobius (not used for the Teubner text)
has Eumque and not Cumque, and where the context seems to allow
the Demonstrative.
The ordinary Greek ui-diphthong is parallel not only to the Latin ui-
diphthong of cut, &c, but to a Synizesis (see III 24) like Horace's
pltuita (Sat. 2, 2, 76 ; Epp. 1, 1, 108), a word which makes two Trochees
in Catullus' Hendecasyllabic (23, 17 Mucusque et mala pituita nasi),
but which is shown by Aelius Stilo's derivation (' quia petit vitam ') and,
more cogently, by its Romance descendants to have been actually
pronounced as Horace scans it. In the Silver Age the monosyllable
cut (in Plautus quoi) became too short syllables, cm. How far this was
a law of phonetic change in the Empire (cf. fortuitus, gratiiitus) has not
yet been investigated ; but it is perhaps significant that Servius (in
a note on Aen. 1, 575) explains Virgil's ebdem wrongly, as if it were
eodem. We must not regard as decisive * (though we see no reason to
impugn it) the statement of Aulus Gellius (10, 24) that the early word
diequinte (cf. dieseptumi Pers. 260, Men. n 56) was pronounced die- and
not die-, a statement which has been unfairly pressed by some scholars
who discard Synizesis in Plautus for the Brevis Brevians and actually
prefer to scan ■ eodem ' than ebdem. Now, though ibi and ibidem hang
together, the pair makes hardly a precise parallel to eo and eodem ; and
since we have proof positive of ebdem, there seems no need for recog-
nizing the doubtful claim of a rival, ' eodem '. The rest of the phantom
brood of a ' metrical ' Brevis Brevians Law (21) ! eorum ', ' eamus ',
1 duorum ', ' dtiobus ', and similar freaks, have no claim at all.
That eb and eo were both used is probable enough. The nature of
1 We suspect that Gellius (13, 23, 18) misunderstands Ennius1 Nerienem (Ann.
104 ; cf. III. 23).
 PLAUTUS AND MENANDER 61

dialogue-metre makes it difficult to find proof of either type. Meo is

elided in Stich 39 (Anapaestic) :
Quia pol m(eo) animo omnes sapientes (AP; cf. Aul. 782).
In this line therefore meo must have been the pronunciation. The
Verb eo is a Pyrrhic in the Canticum (ww- w-) at Cas. 715 :
Ed nunciam,
Nisi quippiam
Remorare me,
and is probably never a monosyllable in Plautus, although eamus is the
predominant form. A pyrrhic Possessive is required for Cas. 629
(Choriambic) :
Eripite isti gladium ||qui suist impos animi,
and probably Bacch. 638 a (Dactyls and Trochees, the ' Glyconic ' class) :
Tace modo : deu' respiciet
Nos aliquis. Nugae.
Mane. Quid est ? Tuam copiam eccam
Chrysalum video (Or Eccam Ch. vid.).
But a monosyllabic for Mil. pr. 136 (Senarius), 262 (Troch.) :
ltaque illi amanti s(uo) hospiti morem gerit (The accented
syllable hos- would not succumb to a ' brevis brevians ') ;
Nam ill' pon potuit quin sermone ||s(uo) aliquem familiarium ;
and probably Mil. 391 (Iamb.), Poen. 860 (Troch.) :
Quom ilia osculata mea soror ||gemina esset submpte amicum,
Neque erum m(eum) adeo. Quem ament igitur? || Aliquem id
dignu' qui siet (III 46 and 53).
In Trin. 665-6 (Troch.) we seem to have two varieties :
Pernovi equidem, Lesbonice, in||genium t(uom) ingenuum ad-
modum ;
Scio te sponte non tuapte er||rasse, sed amorem tibi.
If we may follow the analogy of other words, we may believe the
unaccented forms (mei, meo, meum, tui, tuo, tuum, sui, suo, suum, etc.).
7>-£i -'
to have been monosyllables (with Synizesis), the emphatic Possessives to
have taken Pyrrhic scansion in the same circumstances (28) as Terence's
bonis (Eun. pr. 8 Ex Graecis bonis Latinas fecit non bonas), whereas at
the end of a line or sentence (i. e. ' in pausa ') and in deliberate diction
iambic scansion would appear. The argument against Synizesis, that
meo, etc., never appear at the end of a line is not convincing. That is
just the place suitable for the full forms of deliberate utterance (III 59).
Of ie (see III 23) clear examples are : (Poen. 869, Troch.)' Diespiter
62 PLAUTUS AND MENANDER

me sic amabit ||(but a quadrisyllable, Poen. 739, etc.) ; 1207 (Troch.) ||

diebu' paucis liberas. Less clear are sies, siet} where a monosyllable is
needed ; for this (a feature of MSS. of Terence still more than Plautus)
may be merely a trick of spelling. Editors usually print sis, sit in such
cases (see III 40). Nor are the traces of guiesco, quietus (e. g. Epid. 338)
quite free from doubt (see III 42). (On lien see III 42). Notice the
pun in Cure. 352 (with demorari).
The Synizesis uj of fuisse. etc., is frequent in Plautus, e. g. :
Stich. 628 (Troch.) || fuisti, rem confregimus ;
Merc. 470 (Troch.) Fuisse credo praefit quo pacto ||;
also fuere, etc., like Ennius (Ann. 192) :
Qui invicti antehac fuere viri, pater optime Olympi.
The scansion « fuisse ' could hardly rest on anything but that broken
reed, a ' metrical ' Law of Breves Breviantes in- the Dramatists. And,
apart from them, Lucilius offers :
91-2 Maluisti dici, Graece ergo praetor Athenis,
Id quod maluisti, te, cum ad me accedi', saluto ;
542 Compernem aut varam fuisse Amphitryonis o.koitlv.
So we may allow adhibuisti in Poen. 131 7 (Sen.) :
Cur non adhibuisti, dum istaec loquere, tympanum ? (With -hi-
shortened by the Brevis Brevians ; see III 23).
Skutsch, a vigorous opponent of Synizesis, censured the wastefulness
of two explanations (Synizesis and Brevis Brevians) when one (e. g. eo,
suo) would serve the purpose. But, as we have seen, the Brevis
Brevians Law (except the fictitious ' metrical ' variety) will not suit
every instance. And the Latin language exhibits clear cases of such alter-
natives, e. g. in the Fifth Declension Gen. Sing, fidei, rei (earliest form) ;
fidel, rei (the latter, e. g. Epid. 203) ; fidei (afterwards written fidi), rei ;
cf. el and ei (afterwards written i) Nom. Plur. ; els and eis (afterwards
written is). Skutsch made a better case against scib (cf. Charisius Gramm.
Lat. i, 16, i5)vgratias, etc. Still, if Ennius (Ann. 163) allowed Iunis
(for Iuniis), we cannot declare gratfis (class, gratis) to be impossible,
even though the spelling of Plautus^and EnniuaUime was gratids (cf.
Ill 42), Radford (Amer. Phil. Assoc. Trans. 36, 158 ; 37, 15), in a
reply to Skutsch, pleads that Synizesis is the natural accompaniment
of the Latin Phonetic Law ' corripitur vocalis ante vocalem '. While
a long vowel before another vowel is shortened, a short vowel is reduced
to the Hebrew Sheva-vowel ; so that e. g. eamus might be written with
a very small £_(to mark the evanescent sound) above the line. He
associates with this law not only sorsum beside seorsum (III 42) but also
 PLAUTUS AND MENANDER 63

dudum for 'diudum' (cf. Most. 293 with din elided), and compares
Vulgar Latin quescas for quiescas, &c. He argues that the word-group
ea-res, which has for Dat. el-rei, would naturally have for Ace. and Abl.
eam-rem, ea-re; that the pronunciation Iuno Lucina tuam-fidem best
agrees with Di vestram-fidem (cf. below, 48). He interprets (Cure. 236)
renes (cf. Priscian Gram. Lat. ii, 149, 7) as rienes, on a par with lienem
(cf. Horace's vietus Epod. 12, 7).
The normal scansion seems to be di deaeque, trisyllabic, but quadri-
syllabic sometimes, e.g. Most. 655 (Sen.) :
Malum quod isti di deaeque omnes duint (borrowed by Ph.
Terence
976).

Debrum is normal ; deos rarely an Iambus (Class. Quart. 3, 8). Simi-

marly mebrum, mearum, &c. Whether the monosyllabic pronunciation
of the Possessive was indicated by spelling (e. g. miis in Ter. Heaut. 699,
a monosyllable; III 58) is doubtful. The Scholiast on Persius i, 108
(sis apud antiquos pro suis ponebatur) may refer to Ennius' usage in
the Annals. (For details see our Latin Language eh. VII, p. 12.)
From Ennius (Ann. 149) Lucretius takes the monosyllable (3, 1025, etc.),
a scansion frequent in the Tragedians' fcagments. Ennius seems to
have made quattuor a disyllable, quattubr (Ann. 92 ter quattuor corpora
sancta : 593 Iamque fere quattuor partum (if this comes from the
Annals), with the same Synizesis as is so often found in the various
cases of duo e. g. dubs in a Praetexta of Accius, frag. 2 1 ; dubdecim,
like Ital. dodici, in Epid. 675). For this pronunciation of every-day
talk was recognized not merely in Comedy but in all poetry of the
Republic, and even (in the case of eodem, etc., at any rate) of the
Augustan Age. Disyllabic eamus, Plautus' usual scansion, e. g. :
Men. 387 (Troch.) eamus intro ut prandeamus ||
but never ' abeamus \ etc. ; 32, end), may have been peculiar to Comedy ;
but the evidence is insufficient for certainty. (On duellum, perduellis
see III 42.)

34. These specimen-passages will illustrate our suggestions for the scansion
of these words :
Poen. 1274 sqq. (Trochaic):
Di deaeque omnes, vobis habeo || merito magnas gratias
Quom hac me laetitia adfecistis || tanta et tantis gaudiis,
Ut meae gnatae ad me redirent || et potestatem meam.
Mi pater, tua pietas plane || nobis auxilio fuit.
Patrue, facito in memoria habeas || tuam maiorem filiam
Mihi te despondisse. Memini. Et || doti' quid promiseris. '
64 PLAUTUS AND MENANDER

Poen. 1338 sqq. (Senarii) :

Decipitur nemo, mea quidem sententia,
Qui suis amicis narrat recte res suas ;
Nam omnibus amicis meis idem unum convenit, (Or mefe, emphatic ?)
Ut me suspendam ne addicar Agorastocli.
Leno, eamus in ius. Obsecro te, Agorastocles,
Quid tibi mecum autem ? Quid ? quia hasce aio liberas ;
Eae sunt surreptae cum nutrice parvulae.
Meae quidem profecto non sunt. Leno, in ius eas.
Trin. 181 sqq. (Senarii) :
Neque adeo hasce emi mihi nee usurae meae :
Illi redemi rursum, a me argentum dedi.
Em mea malefacta, em meam avaritiam tibi ! (Or Hiatus at pause ?)
Uava-ai : vicisti castigatorem tuum.
Numquid vis ? Cures tuam fidem. Fit sedulo.
Sed quid ais ? quid nunc virgo ? nempe apud test ? Itast,
Iuxtaque earn euro cum mea. Recte facis (if iuxta is right ; III 41).
Di deaeque. With Synizesis : Aul. 778: Cas. 279 ; Cist. 512; Epid. 396 ;
Most. 192; 464; 684; Pers. 292; 296; 298; 666; 832; Poen. 668; 859;
1274; Pseud. 37 ; 271. Without: Capt. 172; Cure. 720 ; Merc. 793 ; Most.
655 ; Poen. 460. So scan with Synizesis Mil. 501 and 725.
A curious example of ui is
Naevius com. 21 (Sen.) Quis heri apud te ? Praenestini et hospites, Lanuini

which may or may not be an imitation of local patois (like Plautus' Campans,
Trin. 545, True. 942; see III 27. Cf. also Varro Menipp. 356 Pacvi). Since
-it of 3 Sing. Perf. Ind. had a long vowel in Plautus' time, fuit (as well as
fuft, by the Brevis Brevians Law) is quite likely even before a vowel. Is it
the monosyllable which occurs in Lucilius 468 ?
In terra fuit, lucifugus, nebulo, id genu' sane.
(For other examples of Synizesis in Possessive Pronouns, see III 32 ; in
Demonstratives, III 37 ; in Compounds with co- and de-, III 29.)
35. (B) Slurred Gen. Sing, of Pronouns. The Latin Pronoun
Declension offers many problems to Comparative Philology, and none
harder than the unparalleled Gen. Sing, suffix of illius (reduced to illius
in Augustan poetry, but perhaps not in Plautus), cuius (in older spelling
quoius\ etc. Small wonder that the scansion of a colloquial slurred
form of illius (istius) and cuius (huius, eius) should be contested. Some
declare for cuius a Pyrrhic and appeal to Greek toiovtos, 7roiw, etc., with
first syllable short as well as long. This explanation, however, will not
do for the slurred form of illius, as in these lines :
Capt. pr. 39 (Sen.) Huius illic, hie illius hodie fert imaginem
(with emphasis on illic and hie) ;
 PLAUTUS AND MENANDER

Epid. 447 (Sen.) Silas pugnas, de illius illae fiunt sordidae

J, „ (Or d(e) illius) j
Cure. 716 (Troch.) || est, haec autem illius soror;
Merc. pr. 48 (Sen.) Lacerari valide suam rem, illius augerier.
The only scansion that will suit is a disyllable illis, istis. Hence the
accepted explanation of cuius, etc., as cuis, huis, 8s, forms which we
shall print here as quoi(u)s, etc.
This conversationalism we can trace as late as Virgil's time. In the
Catalepta (9, 35) there is a reference to Lucretia :
Non cuius ob raptum pulsi liquere penates
Tarquinii patrios, films atque pater.
It was not deemed unworthy by Lucretius of his serious verse (see
Munro on Lucr. r, 149). It occurs in the opening line of Laberius'
dignified protest to Julius Caesar :
Necessitas, cui(u)s cursus transversi impetum,
and is frequent in the Tragedians (e.g. Pac. i6q ; Ace. 153, 294 ; Pac.
36, 391 ; Ace. 517 ; Naev. 2, etc.). We may quote:
Ace. praet. 23 (Sen.) Deinde ei(u)s germanum cornibus conitier ;
Lucilius 905 (Sen.) Cui(u)s si in periclo feceris periculum.
Illius (istius), cuius (huius, eius) seem to be the emphatic or deliberate
forms ; cui(u)s, hui(u)s, ei(u)s are sentence-enclitics or subordinate words ;
while cuius, huius, eius only appear after a ' brevis brevians ' ; e.g.:
credidi ;
Men. 1 07 1 (Troch.) Ego quidem huius servu' sum, sed || med esse huius
meum,
Poen. 392-4 (Troch.) Obsecro hercle te, voluptas || huius atque odium

Huius arnica mammeata, || mea inimica et male-

vola, (with Dactyl in first foot.)
Oculus huius, lippitudo || mea, mel huius, meum;fel

1257 (Iamb.) Huiusce fratris filius ||;

1367 (Sen.) Habeamus hilare, huius malo et nostro bono ;
Trin. 178 (Sen.) Qui emisset, eius essetne ea pecunia? ;
122 (Sen.) Malumque ut eius cum tuo misceres malo ;
Rud. pr. 1-2 (Sen.) Qui gentes omnes mariaque et terras movet,
Eius sum civis civitate caelitum (Dignified utterance);
Poen. 157 (Sen.) (differor) Cupidine eius. sed lenone istoc Lyco (In
pausa) ;
155 (Sen.) Lenonis huius meretricem maiusculam (Pointing to
a house, ' who lives here ') ;
Aul. pr. 25 (Sen.) Dat mihi coronas, eius honoris gratia (On the divided
Anapaest see 48. Cf. met honoris III 42, s. v. Honor) ;
66 PLAUTUS AND MENANDER

Poen. 1 1 36 (Sen.) Eho an huiu' sunt illae filiae? Ita ut praedicas

(Hiatus at change of speaker) ;
Pers. 830 (Troch.) Tace, stulte: hie eius geminust frater ||;
83 (Sen.) Sed eccum parasitum quoi(u)s mihi auxiliost opus ;
The last is the commonest type. Two Senarii of the Amphitruo prologue
illustrate the different treatment of the (accented) Interrogative and the
(unaccented) Relative :
17 Nunc quoiu' iussu venio et quam ob rem venerim (Dicam) ;
26 Etenim ille quoi(u)s hue iussu venio, Iuppiter.
Quoius (-a, -urn), the Possessive, a Trochee (III 32), appears in Amph. 346
(Troch.), with a most improbable scansion:
Possum scire quo profectus, || qufiiu' sis aut quid veneris (Read
quoiu's aut. ? Indie, beside Subj.)
Illius, istius, etc. It will be best to give a full list of all the occurrences
in Plautus of the Gen. Sing, of ille, isle (Gen. Sing, of ipse, only once,
Capt. 287) :
Amph. 896 (Sen.) Illius ira in hanc et maledicta expetent (cf. 473) ;
Asin. 77 (Sen.) Volo amari obsequium illius, volo amet me patrem ;
Aul. pr. 35 (Sen.) Is adulescentis illi(u)s est avonculus (Or illiust) ;
Bacch. 252 (Sen.) Istius hominis ubi fit quaque mentio ;
487 (Troch.) Ut opino, illi(u)s inspectandi ||;
494 (Troch.) Mnesiloche, hoc tecum oro ut illi(u)s || animum
atque ingenium regas ;
851 (Sen.) Vir hie est illius mulieris quacum accubat (cf. 601) ;
1044 (Sen.) Miseret me illius. Tuus est, non mirum med facis (Or;
illi(u)s)

Capt. pr. 39 (Sen.) Hui(u)s illic, hie illi(u)s hodie fert imaginem ;
287 (Troch.) Immo edepol propter avaritiam ip||sius atque
audaciam (Or || ipsi(u)s) ;
Cas. 550 (Troch.) Propter operam illius hirci ||;
Cist. 515 (Troch.) Itaque me Ops opulenta, illius || avia, etc.
mater ;
745 (Iamb.) Quid istuc negoti est ? aut quis es ? || Ego sum illiu'
Tstanc) ;
766 (Sen.) Illi(u)s ego istanc esse malo gratiam (Or Illius ego

Cure. 413 (Sen.) Libertus illi(u)s, quern omnes Summanum vocant ;

716 (Troch.) Libera haec est, hie huiu' frater || est, haec autem
illi(u)s soror ;
Epid. 119 (Troch.) Malim isti(u)smodi mi amicos || (Or istius modi) ;
446-7 (Sen.) Nam strenuiori deterior si praedicat,
Suas pugnas, de illi(u)s illae fiunt sordidae ;
629 (Troch.) || siquidem istius gratia ;
717 (Troch.) Ain tu te illi(u)s invenisse || filiam? etc. ;
Men. pr. 42 (Sen.) Illiu' nomen indit illi qui domi est ;
pr. 45 (Sen.) Propterea illius nomen memini facilius ;
 PLAUTUS AND MENANDER 67
Men. 904 (Troch.) Sed ego stultus sum qui illius ||esse dico quae meast ;
Merc. pr. 48 (Sen.) Lacerari valide suam rem, illi(u)s augerier;
144 (Troch.) Apage isti(u)smodi salutem || (Or istius modi);
236 (Sen.) Ait sese illius opera atque adventu caprae ;
276 (Sen.) Atque illi(u)s haec nunc simiae partes ferat (Pi Ac
metuo ne illaec s. p. f. A) ;
443 (Troch.) Sanu' non est ex amore il||lius.
(Orquod
amoreposces, feres;
|| illi(u)s)
534 (Iamb.) Nunc quando amicum te scio || esse illi(u)s, indi-
cabo (Or es||se illius) ;
657 (Troch.) Adeo dum illi(u)s te cupiditas || ?
739 (Sen.) Nil hercle istius quicquam est. Numero purigas ;
Mil. pr. in (Sen.) Quam eru' meus amabat. nam is illius filiam ;
589 (Sen.) Nam illius oculi atque aures atque opinio;
987 (Troch.) Quae haec celox ? Ancillula illi(u)s || est quae hinc
egreditur foras (Or illi||ust) (cf. 986) ;
1 163-5 (Troch.) Nempe ut adsimulem me amore is||tiu' dififerri.
Tenes.
Ouasique istius causa amoris ||ex hoc matrimonio
Abierim, cupiens istius ||nuptiarum. Omne ordine ;
1 1 70-2 (Troch.) Ita volo adsimulare prae Illi(u)s || formam quasi
spernas tuam
Quasique eius opulentitatem || reverearis, et simul
Formam, amoenitatem illius, || faciem, pulchritu-
dinem (Collaudato) ;
1238 (Iamb.) Istuc curavi ut opinio||ne illiu' pulchrior sis ;
1299 (Sen.) A matre illius venio. si iturast, eat ;
Most. 242 (Iamb.) Pro illiu' capite quod dedi || ;
612 (Sen.) Illius ; is tibi et faenus et sortem dabit ;
746 (Iamb.) Patrone salve. Nil moror || mi isti(u)smodi clientes
(Or istiu' modi) ;
Pers. 36 (Troch.) Ut mihi des nummos sescentos || quos pro capite
illiu' pendam ;
Poen. 158 (Sen.) (Lyco) Illiu' domino non lutumst lutulentius ;
895 (Troch.) ebdem quo soror illius ||;
Pseud. 1 09 1 (Sen.) Memini. attulit ;
Em Illi(u)s servus hue ad me argentum

1 169 (Troch.) Sequere. quid ais ? nemp' tu illius ||Planissime

servus es?;

Rud. pr. 51 (Sen.) Is Illiu' laudare infit formam virginis (Or illi(u)s) ;
pr. 77 (Sen.) Ad villam illius, exul ubi habitat senex ;
321 (Iamb.) Cum isti(u)smodi virtutibus || (Or istiu' modi),
1094 (Troch.) Si scelesti illius est hie || quoiu' dico vidulus ;
Trin. 157 (Sen.) Siquid eo fuerit, certo illius filiae;
163 (Sen.) Quid tibi ego dicam qui illius sapientiam
(Et meam fidelitatem et celata omnia) ;
F 2
 68 PLAUTUS AND MENANDER

Trin. 552 (Sen.) Aequum videtur, qui quidem isti(u)s sit modi (Or
qulquidem istius ?) ;
894 (Troch.) Pater istius adulescentis ||;
965 (Troch.) Atque etiam Phillppum numeratum il||lius in mensa
manu (Or || illi(u)s) ;
True. 656-7 (Sen.) Fult edepol Mars meb periratus patri,
Nam oves illius haud longe absunt a lupis ;
930 (Troch.) || quae ames hominem istimodi ? (See above, on Rud.
321.)
Thus of some sixty occurrences none require the scansion illius (istius) and
only twelve admit it : Aul. pr. 35 ; Cist. 766 (Dactyl-word as first foot of
Senarius) ; Epid. 716 and Merc. 657 (Dactyl-word as second foot of Troch.
line) ; Capt. 287, Merc. 443 and Trin. 965 (Dactyl- word as first foot of Troch.
hemistich) ; Mil. 987 and 11 70 (Dactyl- word as fourth foot of Troch. line) ;
Most. 746 and Rud. 321 (Dactyl-word as first foot of Iamb, hemistich) ;
Trin. 163. None of these twelve but the last seems to favour illius; and it
cannot be said to demand this scansion when we remember the number of
occurrences of a Molossus at this part of the line ; 7). The material therefore
gives no strong support to illius (istius), a scansion which otherwise would
\ have gone unquestioned since the shortening of I before -us, etc., is carried
even farther by Plautus than classical poets (e. g. Chlus Adj. in Cure. 79 ;
see III 25). For alterius Capt. 306, some prefer alteri (like ulli True. 293).
In Terence, so far as can be determined in the lack of a proper edition,
there is no more evidence of ' illius ■ than in Plautus ; nor yet in any of the
Dramatists' fragments. A Lucilius fragment has either illius or illi(u)s :
158 Usque adeo, studio atque odio illius efferor ira. Examples of illi(u)s,
isti(u)s in Terence seem to be :
Andr. 810 (Sen.) Quae illius fuerunt, possidet; nunc me hospitem ;
Eun. 370 (Iamb.) Responde. Capias [tu] illius ves||tem. Vestem ? quid
turn postea ? ;
620 (Troch.) Id faciebat retinendi illius || causa : quia Ilia quae
cupiebat ;
741 (Troch.) Usque adeo ego Illius ferre possum in||eptiam et
magnifica verba ;
Heaut. pr. 33 (Sen.) De illius peccatis pluia dicet, cum dabit ;
203 (Iamb.) Huncine erat aequum ex illius mo||re an ilium ex
huius vivere ? (Emphatic ; emended to aeq. ill.) ;
387 (Troch.) Et vos esse istius modi et nos ||;
544 (Sen.) Abeat, cum tolerare illius sumptus non queat ;
Ph. 648 (Sen.) Ut ad pauca redeam ac mittam illius ineptias ;
969 (Sen.) Non hercle ex re istius me instigasti, Demipho ;
liec. 232 (Troch.) Illius dices culpa factum ||;
589 (Iamb.) Illius stultitia victa ex ur|[be tu rus habitatum
migres ? ;
Ad. 261 (Iamb.) Quid est ? Quid sit ? illius opera, Syre, || nunc vivo :
festivum caput ;
 PLAUTUS AND MENANDER 69
Ad. 441 (Sen.) Ne illius modi iam magna nobis civium;
572 (Troch.) Illius hominis, sed locum no||vi ubi sit. Die ergo
locum.
An example of ipsi(u)s is :
Heaut. 576 (Iamb.) || apud alium ipsius facti pudet (Cf. Ph. 725).
Editors substitute soli for the disyllable in Heaut. 129 (solius MSS.) like nulli
(Andr. 608 tarn nulli consili sum). Priscian (Gramm. Lat. ii, 227) attests
Terence's nulli and Plautus' ulli along with uni of Titinius (frag. 7), toti of
Afranius (frag. 325), ipsioi Afranius (frag. 230), and so on.
For the recognition of illi(u)s, etc., by the Tragedians we may cite :
Accius 201 (Sen.) Qui illius acerbum cor contundam et comprimam (cf.
185, 268, 465, 579 ; isti(u)smodi 136).
While the word ' his ' is clearly unaccented in the phrase ' his book ', etc.,
and clearly accented in ' mine, not his ', we should often be in doubt how to
classify it. The same doubt attaches occasionally to the Gen. of ille {isle),
hie, is. Still Terence's hui(u)s in such lines as Andr. 480, 888, where the
Pronoun is clearly emphatic, suggests either that our law is no strict law or
that in Terence's time it was honoured in the breach as well as in the
observance.
Andr. 480 (Sen.) Nunc hui(u)s periclo fit, ego In portu navigo (Cf. 49) ;
888 (Sen.) An ut pro hui(u)s peccatis ego supplicium sufferam?
And the Trochee in Plaut. Asin. 393 (Iamb.) is puzzling, for emphasis is
absent :
|| quid eius atriensis ? (Cf. 402).
In Cist. 138 (Sen.) :
Feci eius ei quod me oravit copiam,
some will prefer to print Feci ei(u)s ei quod med oravit copiam (but cf. 6).
At any rate we are on safe ground when we require an Interrogative to be
accented, quoius, while the Relative (unless it precedes an enclitic word)
must be unaccented, quoi(u)s. Cf. 'whose is it?', 'the m<in whose book
it is '. Editors should not disdain this clue to the true alternative in the
scansion of such lines as :
Stich. 3a Nam nos ei(u)s animum (see IV 44) ;
Men. 812 (Troch.) Debsque do testes. Qua de re aut || quoiu' rei rerum
omnium ? (Not || aut quoi(u)s) ;
1 135 (Troch.) Hoc erat quod haec te meretrix || huiu'
nomine vocabat ;
(Emphatic)
Most. 961 (Troch.) Quoi(u)s patrem Theopropidem esse o||pinor, etc.
(Contrast line 970 quoius || est p. Th.) ;
Epid. 356 (Iamb.) Ut quom rediisses ne tibl || ei(u)s copia
(Notesset. Eugae
tibi e||ius);
Capt. 887 (Troch.) Sed Stalagmus quoius erat tunc || nationis quom hinc
abit ? (Erat spoils the correct Tetrameter)
7© PLAUTUS AND MENANDER
and perhaps
Rud. 285 (Canticum Iambics ?) Veneris fanum, obsecro, hoc est ?
Fateor. ego hui(u)s fani sacer|dos clueo|
(Last Dipody a Choriambus ; IV 12).
The Amphitruo fragment (n° x) is rather trochaic (part of a line) than a
Senarius, if its true form is :
Nonius).
Quoius? quae me absente corpu' vulgavit suum? (Cuiusque MSS. of
The utterance of the god in Men. 848 requires the trochaic form, and the
same form should be printed at the Diaeresis of line 855 (Troch.) :
Ita mihi imperas ut ego huius || membra atque ossa atque artua (Not
egS hui(u)s).
No importance can be attached to the spelling 'huis' in some MSS. at
Mil. 908 (unemphatic). (Cf. Ter. Ad. 452.) The reading varies in A and P
at Poen. 316 (Ut quidem tu A : Ut tu quidem P), but neither variety allows
the trochaic form for an (apparently) emphatic huius. The traditional
reading at Amph. 141 (Sen.) :
Et servu', quoi(u)s ego hanc fero || imaginem,
has the proper unaccented form, while editors, justly doubting the Diaeresis
(III 48), substitute the accented (quoius ego fero hanc) : though indeed the
emendation admits the unaccented too, quoi(u)s egd fero hanc, but at the
cost of dispensing with Caesura. Similarly at Rud. 1200 (Sen.) :
Iussique exire hue servum ei(u)s, ut ad forum (Iret),
where the transposition (eius servum) saves the metre at the expense of
wrongly substituting the accented form of the Pronoun.
Examples of the trochaic ' in pausa ' are :
at the end of a line or hemistitch, e. g. :
Pseud. 986 (Troch.) Nosce imaginem : tute eius ||nomen memorato mihi ;
Rud. 1066 (Troch.) Lenonem extrusisti, hie eius ||vidulum eccillum tenet ;
Trin. 671 (Troch.) Quom inopiast, cupias, quando eius || copiast, turn ;
non velis

Most. 970 (Troch.) Philolaches hie habitat, quoius ||.est pater Theopro-
pides ;
Men. 190 (Troch.) Interim nequis quin eius || aliquid indutus sies ;
at a change of speaker, e. g. :
Pseud. 651 (Troch.) Nam hicquoque exemplum reliquit || eius.rem Omnem
tenes ;
Men. 222 (Troch.) Ego et Menaechmus et parasitus || eius. lam Isti
sunt decern ;
Rud. 1 27 1 (Troch.) Quid matri eius? Censeo. Quid || ergo censes?
Quod rogas (But cf. 1276);
at a pause in the sense, e. g. :
Ter. Hec. prol. 8 Alias cognostis eius : quaeso hanc noscite.
Other examples in III 34.)
 PLAUTUS AND MENANDER 71

36. (C) Dropping of short -e. This is a feature of most languages,

e. g. Greek 7rav(£), German im Haus(e), Latin calcar(e), animal(e), etc.
The examples which call for mention here belong to the same class as
nee (for neque), ac (for 'atq', atque), neu (for neve), seu (for sive),
dein, proin (for ' deind ', ' proind ', deinde, proinde, etc.). Skutsch in a
thorough and convincing treatment of the whole question (Forschungen
zur lateinischen Grammatik I) showed that this was the secreTofthe
puzzling appearance of nempe in the Latin Drama (always of two
1 morae ', whether before an initial vowel or consonant), nemp(e) eram,
nemp(e) sum, and that the old explanation ' nemp(e) eram, nepe sum '
(cf. Trin. 328, 966) could not stand. Similarly the Demonstrative ille,
viz. ill(e) est, ill(e) fuit (not ■ ill(e) est, ille fait'). Also quippe, etc., etc. j/j
If Aulus Gellius (10, 24) is right in saying that the early word diequinte,
' on the fifth day ', began with two short syllables, we must pronounce
esse as ess' in Pers. 260 (Troch.) :
|| dixit ess' dieseptumi.
And if the Gellius MSS. are right in their quotation of Most. 280
(Troch.), we must scan :
Verum illuc ess' maxima adeo || pars vestrorum intellegit
(illud est MSS. of Plautus),
although a phrase like this hardly satisfies the conditions which a
phonetician would require, viz. that the word which drops -e shall
be closely joined to, 'jammed up against', the following word. Of
nempe we are told that its accent was that of a subordinate word (Gloss.
Lat. IV 123, 49 Nempe: recte vel certe, sed gravi pronuntiatione.
Perhaps from Donatus). In a sentence like Nempe tu dixisti, 'yes, it
was you who said it', the voice would hurry from the subordinate
nempe to the emphatic tu, and the -e would become inaudible, ' squeezed
out '. That seems the phonetic explanation. And it suits ill(e), ist(e\ /
nemp(e), quipp(e), ind(e), proind(e) (= proin), deind(e) (= dein), und(e),
but hardly esse quite so well (at least in some contexts). The dropping
of -e in atque, neque, neve, sive before a consonant-initial (ac, nee, neu,
seu) could receive the same explanation. The full forms of these four
Conjunctions seem occasionally to be written by scribes where the
shortened forms were pronounced; though the full pronunciation is.
required in such lines as Merc. 332 (Sen.) ; Pseud. 349 (Troch.) :
Ut illam vendat neve det matri suae ;
|| qui hunc occidam — atque 1 me,
1 Is it fanciful to find in this Trochee an echo of the slow utterance of the threat
' ay ! and myself too ' ? And similarly in Capt. 942 atque te (Nolim suscensere),
of Hegio's reluctant admission of his cruelty to Philocrates' servant ? -
72 PLAUTUS AND MENANDER

a misuse which may be set beside seu istuc (if that is the right reading)
in Merc. 306 (Sen.) :
Si canum seu istuc rutilumst sive atrumst, amo (AP).
To the rule which restricts namque to usage before a vowel-initial, there
is similarly an exception :
Ennius trag. 370 (Troch.) Namque regnum suppetebat ||.
Another Plautine pair of 'doublets' (like namque and nam) is anne
(before a vowel only) and an (III 13). Another, nonne (prevocalic) and
(Interrogative) non (preconsonantal) ; thus nonne aspicis ? or non vides ?
'do you not see'? (but Ter. Andr. 238, 869). And they go with
Plautus' restriction of hisce, illisce Nom. Plur. to prevocalic use, e.g./iisce
homines, hi viri (III 37). These and similar niceties of usage, rarely or
never ignored by Plautus, teach us how closely his language follows the
actual (educated) speech of his time and encourage us to collect statistics
of what at first sight appear as trivialities. (Cf. Ill 42 on dice contrasted*
with die mihi; also III 40 G.) The chance of getting a phonetic analysis
of the Latin spoken about 200 B.C. is a golden opportunity, not to be
neglected. (On the substitution of -n for -ne see 51, 58. The MSS.
point to quodrf vobis in Mil. 614. But at the end of a line, e. g. Cist.
580 sicine; or hemistich, e.g. Cist. 746 hlcine.)
Since full statistics of each word are given in Skutsch's book we need
not repeat them here. But attention may be called to his list for unde,
which suggests that (before a consonant) the Interrogative (i. e. accented
unde) was normally a Trochee, the Relative (i. e. proclitic unde) a mono-
syl able, e.g :
Asin. 258 (Troch.) Unde sumam? quern intervertam? ||;
Rud. 273 (Cret.) Unde nos hostias || agere voluistis hue?;
Trin. 937 (Troch.) Qui egomet unde redeam hunc rogitem || ;
Capt. 109 (Sen.) Und' saturitate saepe ego exivi ebrius ;
Pers. 494 (Anap.) Und' tu pergrande lucrum facias ||.
(Therefore make unde a Relative in Pers. 150, with parentes as Ante-
cedent.) The full forms are, of course, appropriate to the end of a line
or hemistich or to any marked pause in the sentence, e. g. :
Capt. 490 (Troch.) Nunc redeo inde, quoniam me ibi || video
ludificarier ;
Rud. 600 (Sen.) Neque eas eripere quibat inde : postibi.
(So we may retain the nonne of the MSS. in Amph. 405 (Troch.) :
Venit, quae me advexit. nonne || me hue erus misit meus ?)
And they are the forms of deliberate utterance, e. g. :
 PLAUTUS AND MENANDER 73

Asin. 777 (Sen.) Neque, quom descendat inde, det quoiquam

manum (In a legal contract).
Quippe seems to have had its full pronunciation in the phrases quippe
qui (but Asin. 66), quippe quom, quippe quando, e. g. :
Am ph. pr. 22 (Sen.) Scibat facturos, quippe qui intellexerat ;
Capt. 886 (Troch.) Quippe quando mihi nil credis ||.
(On lacte see III 20 ; poste and post, III 42 ; fort\ III 42.)
Finally, it should be mentioned that Skutsch's attempt (pp. 116-117
of his book) to include with ill' (= ille) also an ill' (= ilia) and an ill'd
(= illud) is certainly wrong: Pers. 232 (Ilia militia or Ea militia);
Stich. 159 (Nam ilia me or Nam illaec me); Trin. 809 (Lepida illast
causa); Asin. 123 (Nam ego illuc); Mil. 757 (Fit pol illuc ad illuc) ;
Most. 281 (see above) ; Most. 626 (Quod illuc); Trin. 259 (^Quam illud
est dulce). On illic see III 33.

37. (D) Shortening of Long Monosyllable, as in siquidem.

Why does si appear with short vowel in siquidem ? This question has
been treated in a learned monograph by Wackernagel (Beitrage zur g"»*t
Lehre vom griechischen Akzent, p. 22), who explains it in the technical
language of German phoneticians by the ' stark geschnittener Akzent '.
Our readers need not be startled by this phonetic mystery. Let them
ask themselves why ' sheep ' takes a short vowel in ' shepherd ', ' know '
in ' knowledge ', * goose ' in ' gosling ', ' goshawk ', and apply the same
explanation to si and siquidem. In Augustan poetry only siquidem is
recognized, but the Dramatists and Lucilius have also miquidem,
mequidem, tuquidem, tequidem ; e. g. :
Epid. 99 (Troch.)
Tuquidem antehac aliis solebas || (Or Tu
quidem antehac, if ant- was unaccented) ;
Lucilius 475 Quod viscus dederas tuquidem, hoc est, viscera largi.
Also apparently quiquidem, e. g. :
Poen. 1 2 13 (Troch.) || quiquidem inimicus non siet.
And quandoquidem (where a disyllable ending in a vowel is joined to
quidem) presumably obeys the same phonetic law, a law, however, which
must not be extended to words ending in a consonant. The true
explanation of hicquidem seems to be that the Particle -ce was dispensed
with (as in il/equidem, istequidem). The word should be printed hi'c)-
quidem (III 9). (On the illusory 'idquidem ' see III 33 A.)
Nor may we infer 'siquis' from siquidem nor 'nequis' from mequidem,
at least until more satisfactory evidence of pyrrhic pronunciation is
found; nor 'numquis'; nor yet 'ecquis' (see III 38 A). 'Nor Leo's
 74 PLAUTUS AND MENANDER

'quidquid': scan Mil. 311. Hercle quidquid est (cf. Ill 10). It is
certainly tempting to find in quasi and quam si the same phonetic
variety as in quiquidem and qui quidem. Still there is another explana-
tion possible, that qua- is Neut. Plur., quam- Fern. Sing. The evidence
of a 'nesi' (Most. 1006) beside nisi is still less convincing, and of a
' hodie ' beside hodie (III 42). For the present at least we cannot
safely go beyond siquidem, miquidem, mequidem, tuquidem, tequidem,
quiquidem, quandoquidem. We mark these vowels as either short or
long, since the forms of deliberate utterance are also found, e. g. :
Pseud. 1 1 54 (Troch.) || mihi dares. Si tu quidem es (Leno Ballio).
The substitution of equidem for quidem, e. g. ' tu equidem ' instead of
tuquidem, * te equidem ' instead of tequidem, is a now abandoned device
of editors. (On egoquidem see III 31.)

38. (E) Enclisis of (e)st. Amatust, amatast, amatumst (or-ust),

bonust, bonast, bonumst (or -ust), etc., can hardly be called conversa-
tionalisms (like our ' he 's going '), since the oldest and best MSS. show
them in Epic Verse of the Augustan Age and later. The Second
Singular, amatu % etc., would have a better claim to the name. (The
scansion amatu 's reflects the scansion es. Plautus spelled amatu'ss ;
III 20). Leo's use of bonust, amatust, etc., as evidence for his impossible
theory that Plautus elided -us (like -um) before an initial vowel (III 17)
will not do. Only an amateur linguist would seek to elicit laws of
language from slurrings like our ' good-bye ' (for ' good be with ye ').
Alternative explanations, e. g. (a) prodelision ,ubi 'st ?) like ttov 'cm ; (b)
suppression of one of two similar neighbouring syllables, as in por(ti)-
torium, mis(is)ti, ters(is)ti, are quite as probable.
tfl In the Palatine archetype 'st, 's were tampered with by some corrector
(see the preface to the Oxford text, p. viii), so that an editor need have
Z$ri*71 no scruple in preferring the reading of A (Idem mihist, etc.) in Cist. 120
to that of P (Idem mihi magnae quod parti est) or in re-casting the P-
version of True. 46 (Si iratum scortum fortest amatori suo) into this (or
any like) form : Si iratum scortum forte amatorist suo. The MSS. fail
us very often indeed when we seek evidence of Plautus having used the
full or the contracted expression, and we have to look for other indication.
Since a monosyllabic line-ending would require an Iambus in the seventh
foot of the Iambic Septenarius, Asin. 710, we see that Plautus must
have written nequam 's :
Adsta ut descendam nunciam in || proclivi, quamquam nequam 's,
just as the Roman MS. of Catullus offers omniums in the poem to
Cicero (49, 7 Quanto tu optimus omniums patronum). (See App. D,
 PLAUTUS AND MENANDER 75

end.) Since me is emphatic in Rud. 281 and the Brevis Brevians Law
does not operate so freely in Bacchiac as in other metres, we may
perhaps print :
Misericordior nulla mest feminarum,
rather than me est. The sly innuendo of the servant after the master's
hasty question is best expressed by printing Men. 434 (Troch.) so :
Quid eo opust ? Opus est — Scio, ut ne || dicas. Tanto nequior,
and this is actually the archetype's spelling. (The Palimpsest has not
this portion.) Were there no manuscript support, few editors would
think of printing so subtle a distinction, since the line can be scanned
without it.
All such clues, it is clear, cannot bring us to more than the merest
fraction of the actual Plautine instances. To print the slurred forms
wherever the metre allows would almost seem a wiser course for an
editor of a comedy than to print them only where the MSS. (especially
the minuscule MSS., in the absence of A) have happened to preserve
them. Some editors take this course, while others shrink from making
the lively lines of Plautus so uniform and monotonous, and from losing
the chance of indicating a nuance of expression by this variety of form
(as in Men. 434).
At any rate one thing is clear. Wherever the MSS. offer 'st we must
accept it. How can we afford to lose rest (for res est) attested for the
minuscule archetype in Merc. 857 (Troch.) :
|| quaeritatum. Certa rest (rest B : res est CD),
and apparently in Merc. 351 ? In Cas. 578 (Sen.) the metre requires rest,
not res est :
Te ecastor praestolabar. Iamne ornata rest,
and it is possible that the (illegible) Palimpsest has rest here, though it
may agree with the minuscule MSS. in res est. In Stich. 473 (Sen.) the
minuscule MSS. again offer res est, where the metre requires rest :
Promitte. Certumst. Sic face, inquam. Certa rest,
but here the Palimpsest offers res. Now although it is true that certa res
(without est) may conceivably have been an alternative expression, is not
the probability much greater that the exemplar of A had here rest ? For
while -ust was familiar enough to an ancient scribe, -est was not, and
rest would readily be regarded as a mere mistake for res, (For a fuller
discussion see Havet in Revue Philologie 29, 185.) In Ter. Eun. 268
(Iamb.) the Bembinus version of the line requires rest :
MSS.).
Rivali' servum : salva rest. ||nimirum hie homines frigent (res est
76 PLAUTUS AND MENANDER

One type of -est has fortunately left clear traces. Nonius attests for
early Latin similest for similis est, qualest for qua lis est, also (in all
probability) talest for talis est ; and the minuscule MSS. win our respect
by their indications of this spelling in their archetype (e. g. similest True.
170, consimilest Men. 1063, where A seems to show -le or else -les).
Nonius calls it ' Neuter for Masculine ' ; and this may not be so absurd
a rough-and-ready label as it seems. For while we have thesft Hear
traces_of -est for -is est in the case of Adjectives, we have no trace of any-
thing but -ist in the case of other words, unless we allow in Merc. 880
the variants (ut est splendore B : ut splendore est CD) to establish
this version of the Trochaic hemistich :
Caelum uti splendorest plenum,
and find here the normal construction with plenus, i. e. the Genitive, not
the Ablative (III 17). It may therefore be that the existence of a Neuter
in -e led in the case of an adjective to the spelling -est, while in the case
of other words the spelling was -ist. Our Terence MSS. write fortis for
forti 's (?) Andr. 702, but usually -is est, (es) in full, where the metre
requires these slurred forms, e. g. Hec. 352, Eun. 273. Nonius attests
simile's (for similis es) in Novius 62.
These, however, are matters for a book on Early Latin Orthography,
not on Early Latin Verse. What more concerns us here is a question
like this : When Plautus begins an Iambic line with (let us say) Quid-
quid est, Istic est, did he scan Quidquidst, Isticst or Quidquid est, Istic
est (by the Brevis Brevians Law) ? It is true that we have no clear
traces in the MSS. of Quidquidst. But would so unfamiliar a form have
much chance of surviving in transcription ? Leo's denial (Plaut. Forsch.2,
p. 281) of the possibility of anything except -ust, etc., -est, -1st seems
too bold (e.g. penest Amph. 653 ; hicst, Poen. 1333).
39. (F) Short Vowel not lengthened before Mute and Liquid.
This feature of the Roman Drama (Tragedy and Comedy) is also
a feature of Greek Comedy. A lengthening like aypoi, agri was
apparently a licence of elevated poetry like our pronunciation of l wind '
so as to rhyme with ' kind '. When we remember that Greek was a
quite different type of language from Latin, we cannot avoid a sus-
picion that this length is a Greek scansion slavishly imitated by Ennius
(cf. Italia, a First Paeon). But we hesitate to declare the Roman
pioneers in quantitative Dramatic verse to have been seduced by Greeks
from the true scansion of conversational Latin or to have allowed, e. g.,
ottXov to have decided the quantity of popli, e. g. :

Most. 15 (Sen.) Tu urbanu' vero scurra, deliciae popli,

 PLAUTUS AND MENANDER 77

however strange it seems to us that popli (Iambus) should have a ' mora '
less than populi (Anapaest), e.g. Poen. 524 (Troch.) Praesertim in re
populi placida ||.
(For a fuller treatment of the scansion with Mute and Liquid see
ch. Ill 58.)
40. (G) Certain Words. In the paragraph on the Brevis Brevians
Law (22) we have already mentioned Philippics as the invariable pro-
nunciation ofthe name of the coin 'a gold Philip' and shown that it
was the Roman accentuation of the first syllable (in imitation of $6'A.t7r7ros)
which allowed the operation of the ' brevis brevians.' From what lan-
guage the word for an arrow was borrowed (Etruscan ?) is unknown. It,
too, seems to have been accented on the first syllable, since the Plautine
scansion (in all three occurrences j III 42) is sagitta. On satelles (an
Etruscan loan-word ?) see III 42. Whether the name of King Philip
was treated like the name of his coin is doubtful, e. g. Pers. 339
(Senarius) :
Mirum quin regis Philippi causa aut Attali (Or regi' Phillppi).
If regi'-Philippi was a word-group the divided Anapaest would be
legitimate (as in propter-amorem, 48). All languages are apt to take
liberties with coin-names (' twopence ' pronounced ( tuppence ' ; ' half-
pen y pronounced
' ' ha'penny ') ; and perhaps Phillppus for the king,
and Philippics, for the coin, were the pronunciations in Plautus' time.
On the doubtful talentum see III 42. Fenestra too is doubtful. It also
has been declared to be a loan-word (Greek?) with accent on first
syllable. But the Plautine form may befestra (III 42), like mo(ti\strum%
and be explained in the same way as monstrum (for c m6nestrum ', as
iustus for i6vesto-), namely by the older accentuation of the first syllable
of each word (18). Quadrigenti, the etymologically correct spelling}
seems to have been the only form known to Plautus, since the first
three syllables invariably make an Anapaest (III 42). Quadringenti is
a later coinage (on the type of septin-genti, etc.). If it had been Plautine
the second syllable would have occasionally been long j or rather
normally, since Plautus was averse to allow a ' brevis brevians ' before
a Mute and Liquid (27).
An unmistakable conversationalism is aunculus for avunculus, a curious
anticipation of French oncle (III 42). Metathesis of r was a common
trick of careless Latin speech. Thus TpaTre&Ttis appears regularly in
Plautus' pages as tarpessita ; phrygio seems to appear once as phyrgio
(Aul. 508) ; from crocus is derived corcotarii (Aul. 521) ; from porcus,
proculena (Mil. 1060 ; III 26), and so on.
(Examples of all these words are given in ch. iii § 42.)
78 PLAUTUS AND MENANDER

41. CAESURA. On any page of Menander you will probably find an

example (one or more) of a Trimeter which wholly lacks Caesura (e. g.
Epit. 234 Tavpo7ro\ioLs d7rwXe(r€v tovtov 7tot€). In Plautus such Senarii
are few and far between. It is patent that the Roman writer insisted
on Caesura far more than the Greek. Sometimes the nature of the
words made Caesura impossible for him, e. g. Aul. 510-11 :
Flammarii, violarii, carinarii,
J\ut manulearif, aut murobatharii.
Sometimes a special effect is aimed at by absence (complete or almost
complete) of Caesura, e. g. :
Trin. 1094 O Callicles, O Callicles, O Callicles;
Pers. 410 Procax, rapax, trahax — trecentis versibus
(Tuas impuritias traloqui nemo potest. Cf. 6) ;
movebor) ;
373 Dicat quod quisque vult : ego de hac sententia (Non de-
dem ;
Bacch. 257 Quamne Archidemidem ? Quam, inquam, Archidemi-

Men. 750 Negas novisse me ? negas patrem meum ? ;

Cas. 510 Nostro omine it dies : iam victi vicimus.
Where these two causes are not visible, editors are faced with the
alternative of declaring a line with unsatisfactory Caesura to be corrupt
or of finding in it a trace of the ' non astrictus soccus ' of Plautus.lines ?) ;
? Frag. 148 In pellibus periculum portenditur (Or trochaic 1 Or two

Amph. pr. 95 Et ego una cum

illo. nunc animum advertite
(Emended to nunc(iam) an.) ;
Bacch. 344 Id mi haud utrum velim licere intellego ;
Cure. 10 Egon apicularum opera congestum non feram ? (Emended
to op. (opu') con.);
Epid. 477 Produci intus iubes? Haec ergo est fidicina (Emended
to Int. iub. prod.) ;
Merc. 691 Ecastor vero istuc eo quantum potest ;
777 Drachumam dabo. Dabitur. Dari ergo sis iube (Emen
ded to si vis) ;
796 Concivit hostes domi : uxor acerrimast (Emended to
domi (mi) ux.) ;
Pers. 456 Igitur proventuram bene confido mihi (Emended prov.)
to be.j

Poen. 980 Atque, ut opinor, digitos in manibus non habent (Atq.

hercle op. A ?) ;
 PLAUTUS AND MENANDER 79

Rud. 1 195 Ego hodie neque speravi neque credidi Trib

(Anrach
impossible
j 45) ;

1 34 1 Isque in potestatem meam pervenerit (In a prayer) ;

1353 Si maxime mi ilium reddiderit vidulum (Or mihi illunc) ;
Stich. 157 a Quam ego meae matri refero invitissimus (Ma. me. P) ;
Trin. pr. 1 5 Dedi ei meam gnatam quicum aetatem exigat (Or el) ;
425 Tarpezitae mille drachumarum Olympico.
(In Frag. 24, where venter is unlikely, read Nam (unum) me puero
venter erat solarium.) (On Stich. 300, a Canticum-line, see IV 9.)
The rarity of such lines warrants every device to avoid them. So we
may read est with the MSS. rather than -st in Most. 475, Trin. 56 :
Capitale scelu' factum est. Quid est ? non intellego ;
Vivit victuraque est. Bene hercle nuntias ;
(though whether we should depart from the MSS. in Merc. 287, Trin.
734 is a debatable point ; 38).
And we must scan the following lines thus :
Asin. 755 Addone ? Adde, et scribas vide plane et probe (with
Hiatus at pause) ;
Cas. pr. 41 Puellam exponi. adit extemplo ad mulierem (ditto) ;
Cist. 156 Fuere Sicyoni iam diu Dionysia ;
Mil. 584 Nam uni sati' populo impio merui mali j
True. 656 Fult edepol Mars meb periratus patri (44).
(On Bacch. 806 see 7 b.)
42. The normal Caesura in Latin, as in Greek, is either in the third
(' penthemimeral Caesura ') or the fourth foot (' hepthemimeral '), e. g.
Amph. pr. 1-2 :
Ut vos in vos/m volti' mercimoniis (Also in fourth),
Emundis vendundis^^ me laetum lucris.
The Caesura may be ' latent ' (i. e. produced by Elision), e. g. Amph.
pr. 3 :
Adficere atque adiuva^ in rebus omnibus.
In the longer Iambic and Trochaic lines the break (caesura) is the
Diaeresis after the first hemistich (59).
Some would find in the Greek Trimeter a third type of Caesura, viz.
a Diaeresis after the third foot (' the medial pause ' of White, p. 62)
and so efface this distinction between the Trimeter and Tetrameter.
But not without challenge. It is true that of the many caesuraless
Trimeters of Menander some present a form which would suit this
theory, e. g. Epit. 9 :
Tt ow e/xot fxiXu j K.pLTrjv tovtov riva.
 8o .PLAUTUS AND MENANDER
Yet what are we to make of others ? E. g. these three in the first
eighteen lines of the Periceiromene :
3 (Tavrrjv) KaTOLKOvo"f], Seofxevr) 7rai8tov,
9 Tovtov veavto-Kov, yeVet KopivOiov,
18 Et 7tot€ SerjOcLT) fiorjOtias twos.

These have a Spondee at the ' Diaeresis \ whereas in the Tetrameters

an Iambus is required here. And have we any more right to sanction
this fortuitous division of a caesuraless Trimeter than the fortuitous
division by dipodies of e. g. Menander's Trimeter quoted in the last
paragraph ?
Epit. 234 Tavpo7roXtoL<s aTruXtarev tovtov 7rore.
At any rate there seems no justification for finding ' mid-line Diaeresis '
in Aul. 510 nor yet 'dipody Diaeresis ' in Aul. 511 (the couplet quoted
at the beginning of the last paragraph). The Hiatus in 5 1 1 is due to
the pause (III 56), not to the (imaginary) Diaeresis. (On the Iambic
Octonarii of Terence see IV 11.)
43. DIVISION OF RESOLVED FEET BETWEEN WORDS. The
Caesura of the line leads us to the Caesura of the foot. The feet con-
cerned are what are technically known as ' resolved ' feet, in which
a long syllable has been resolved into two short syllables. In the verse
-of dialogue the resolved feet are the Tribrach (w ^ w), the Anapaest
(\j v -), the Dactyl (- w kj) and (apparently not in the Menander MSS.)
the Proceleusmatic (w ^ ^ w). The rules which Menander follows in \
dividing these feet between words are, in general, the rules followed by
Plautus. And they are rules of which the reasonableness is sometimes
apparent, sometimes not. In an Iambic line the natural foot is the
Iambus (w -). When an Anapaest is admitted to an Iambic line, two
short syllables take the place of the one short syllable of the Iambus.
When this pair of short syllables is divided between words, the first of
the pair standing at the end of one word and the second of the pair at
the beginning of another word, the rule is (and it seems reasonable)
that the two words should be closely connected— an Article and Noun,
J u „,/ ftiijf- Preposition and its Case, etc. — otherwise the resolved foot would be
too lengthy, would occupy too many atoms of time in its pronunciation.
vuJu But on the other hand a Tribrach contains only three ' morae ', precisely 1
the length of an Iambus ; so the reasonableness of a rule for the division \
of a Tribrach between words is not so apparent.1 The rule for the (
1 Hardie 'Res Metrica', p. 74 : 'The strict rule, observed by Archilochus and
i the Greek Tragedians, was that the two syllables which stood for the long should
be in the same word.'
 PLAUTUS AND MENANDER 81

Tribrach in Iambic verse interferes only when the foot begins with
a Pyrrhic word (<-^,) or word-ending (. . <^,). When the foot begins
with a final short syllable (or a monosyllable or elided disyllable) the
Tribrach is free to take what form it pleases,

44. (1) In Iambic Lines.

(A) Tribrach. Here there is a marked difference between Plautus'
usage and Menander's. While Menander allows a line-opening like
(Epit. 363) Ila/oa rt|i/os ovtos, (Cith. 58) Kara Xd|yov, although it is by
no means a favourite opening of his, Plautus substitutes a Proceleusmatic
for Menander's Tribrach. Our faith in Klotz's guidance is shaken when
we find him stating (Grundziige, p. 255) that the first three syllables of
e. g. mare superum are a quite legitimate (' ganz regelm'assig ') first foot
of a Senarius. Every one who reads Plautus with an attentive ear has
his suspicions aroused by Klotz's example (Men. 236) Mare .raperum
orrine Graeciamque exoticam. He feels that Plautus is likely to have
written another opening foot, a Proceleusmatic : Mare j^rum(que) omne
Graeciamque exoticam. Openings like Mare .rwperum (Tribrach) in the
traditional text can almost be counted on the fingers of one hand.
Plautus either never or hardly ever allowed them.
Menander admits a Tribrach so divided to the second foot as well as
to the first, but no more freely (e. g. Epit. 443 Twat, | 7ro0a> 2|x«s, ««r€
/mm, tov 7rat8a crv ; (And the fourth ? Phas. 30.) In every hundred lines
there is one example, either in the first foot or the second. In Aristo-
phanes too there is about one example in every hundred lines (in first,
second, or fourth foot). It is the fourth foot of the Senarius in which
a Tribrach, beginning with a pyrrhic word or — more often — word-ending,1
is_clearly recognized in Plautine usage and only in one peculiar cadence
that is familiar to all readers of the plays : e. g. (pyrrhic word) Asin. no
Ubi eris ? Ubicumque lubitum erit ammo meo ; (pyrrhic word-ending)
Capt. 159 Multis et multigener/^j ^pus est tibi.
Now this ending of a Senarius is abnormal. When the last foot of
the line is an iambic word (meo, tibi), the fifth foot ought not to be an
Iambus (58). We should expect in the first example the fifth foot to be
the Anapaest animo ; in the second, the Anapaest opus est. It is just
possible therefore that this form of Tribrach {y\j^ ^ or . . ww, w) must be
driven from the single sure foothold which it seems to have in Plautus'
Senarii, (or let us say his ' iambic lines ', for it has the same right to the
corresponding part of Octonarii or Septenarii). But since this is

1 A pyrrhic word-ending cannot form the beginning of a Proceleusmatic (or

Anapaest) in Latin Iambic (or Trochaic) Verse (see 50, 52, 55).
2348 G
 82 PLAUTUS AND MENANDER

a minor point and requires an unduly long discussion, it may be

relegated to small type. At least so much is certain, that Plautus in his
Iambic Verse shows no love for a Tribrach which begins with a Pyrrhic,
or (to put it in another way) for a pyrrhic word (or word-ending) on
whose final the ictus falls. That is why we (perhaps) never find in his
Iambic verse a scansion like ego, tibi, cave, modo, ubi, with ictus on the
I final ; a fact which has been perverted into an argument regarding the
I Brevis Brevians Law (29), but most unfairly. Since e. g. modus with
ictus on the final is banned just as much as e. g. modo with ictus on the
final, it it clear that the Brevis Brevians has nothing to do with the
matter (cf. 27, [6]).
45. First we give statistics of this Tribrach (^w, w) in the first foot. The
alleged examples are :
Cist. 50 || sine meo saepe eri' sumptu ;
151 Ita flroperavit de puellae proloqui ;
Men. 236 Mare j^perum omne Graeciamque exoticam (Emended to
Mare superum(que)) ;
Most. 685 Ita mea. consilia undique oppugnas male (Emended to Ita
mea consilia (tu). Cf. line 656 Ita mea consilia perturbat
matic) ;;
paenissume)
Poen. 467 Mina mibi argenti dono postilla datast (Or mi hi, a Proceleus-

Rud. 166 Neque ^wbernator umquam potuit tarn bene (AP) ;

920 Nimis homo nihilist quist piger || (Emended to Nimis (is)).
Even if an editor be unwilling to remove them, he must disallow emenda-
tions which foist on the text so doubtful a usage ; e.g. Men. pr. 21 (Puer
aberravit) ; Rud. 291 (|| neque didicere) ; Poen. 1272 (Quia numero) ;
Amph. 978 (Face iam Amphitruonem) ; Cure. 517 (|| cures, bene siet isti).
And when the traditional reading can be scanned otherwise, he must not
favour this form of Tribrach : e. g. Rud. 596 and Merc. 227 (Velut ego. Not
'Velut'); Pers. 162 {Neque quadruo\Aax\. Not ' quadruplari ' ; III 42);
Cure. 88 (Ita factum. Agite, etc. With Hiatus at change of speaker) ; Pers.
255 [Quia meo amico. With Hiatus after emphatic Pronoun, III 54) ;
Amph. 903 (Nimis. An Iambus) ; Cist. 296 (Modo. An Iambus) ; Cure. 395
(Apud. An Iambus ?). The scansion pater (III 14) makes Mil. 373, Pers.
57 and 355 normal. The substitution of istanc for istam in Cas. 311 is hardly
an interference with the traditional text (I 6). (On Merc. 586 see 9 ; on
Mil. 481 see III 33.) In Poen. 11 23 Erus meus hi(c)quidem est (Iambus
followed by Proceleusmatic) is the Palatine reading. Sometimes the editor
is faced by an alternative that seems likewise abnormal (e.g. Poen. 1137 Tua
pletas ; True. 698 Ubi | male accipi|ar) and will prefer to emend the tradi-
tional text (e.g. Rud. 1195 Ego hodie (qui) neque speravi neque credidi).
46. Next we give statistics of this Tribrach (^^, ^ or * ' v-/w, <j) in the
fourth foot of the Senarius. The strongest examples are :
 PLAUTUS AND MENANDER 83

Amph. 102 Is priu' quam hincabilt vpsemetin exercitum (48, beginning) ;

Asin. 1 10 Ubi erls ? Ubicumque lubitum erit ammo meo ;
775 Neque illaec ulli pedepedem ^mini premat ;
Aul. 378 Ita fllis impuris omnibus ad\\ manum ;
Capt. 143 Quom quae in potestate habuimus, *a amisimus ;
159 Multis et multigenento #pus est tibi;
362 Vobisque, vult te novus erus ^peram dare ;
373 Sequere. em tibi hominem. Gxatiam hahzo tibi (cf. Ill 55) ;
Cas. 134 Quom mihi Ilia dicet ' mi ammu/e, mi Olympio ' ;
Cist. 753 Istic quidem edepol mei viri hahxtat gener ;
Men. 887 Utrum me dicam ducere w^dicum an fabrum ?
Merc. 693 Ni sumptuosus msuper etiam siet ;
Mil. 27 Quid ' bracchium ' ? Illud dicere vo\xn ' femur ' ;
Poen. pr. 93 Is ex Anactorio, ubi prius ^^bitaverat ;
628 Eum oportet amnem quaerere ^mitem sibi ;
1052 Haec mi hospitalis tessera cum illo fuit ;
Pseud. 59 Haec praestituta est, oroxuma Z>zbnysia ;
True. 333 Quid iam revocabas, xmfirobe whilique homo ? (On nmlll, the
Plautine form, see III 11) ;
Vid. 87 Quom mihi qui vivam copidm inoo'x facis.
Since a pair of Iambi seem always to follow, we find argument for (Capt. 143)
ea amisimus (an operation of the Brevis Brevians which is not favoured ; 27).
(In Bacch. 146 read i rusum domum.) Others exhibit a pause in the sense
after the Pyrrhic :
Amph. 882 Durare nequeo in aedibus. ita me probri ;
Bacch. 134 Ibidem ego meam operam verdidi, #bi tu tuam ;
Cas. 427 Quid opus est, qui sic mortuus f ^quidem tamen, etc.
These are much weaker examples. For, as we shall see later (III 50), Hiatus
and Syllaba Anceps are freely allowed ' in pausa ' ; whether at this part of
the line (e.g. Capt. pr. 11 Negat hercle illic ultumus. accedito; Mil. 848
Numquam edepol vidi promere. verum hoc erat) or elsewhere. So why
should we not scan (Amph. 882) aedibus, and (Cas. 427) mortuus ? (We
ignore lines like Epid. 471 where there is a change of speaker : Estne empta
mi istis legibus ? Habeas licet ; III 49). Indeed we might also class with
this trio some of the 'strongest examples ', e. g. Capt. 143, Cas. 134, Men.
887 (utrum ' which thing ? ', ' which alternative ? '). (In Amph. pr. 94 scan
Iuppiter; Most. 541 citd ; Poen. 988 perhaps plurimin', a question.)
47. If the theory already mentioned (1 11), that some older usage had
allowed Diaeresis here, had stronger claims to probability than it seems to
have (III 48), there would be a way out of the difficulty. Just as the
Diaeresis in the middle of a Septenarius, etc., allows Syllaba Anceps or
Hiatus or both (e.g. Mil. 1239 Si pol me nolet ducere || uxorem, genua
amplectar), so, if this theory were right, we should print, e.g. (True. 333)
improbe || nihilique homo. The claims of this Tribrach to a place in the
Senarius are therefore not overwhelmingly strong. If this theory of Diaeresis
G 2
84 PLAUTUS AND MENANDER

were ever definitely established, our verdict would have to be : ' If Plautus
allows this Tribrach here, it is through old associations, because he has not
yet entirely forgotten the earlier practice of Diaeresis. That is why it shows
itself in the fourth foot of the Senarius, but in no other.' Before the Diaeresis
in the middle of Septenarii was definitely established, editors used to allow
alternative scansions like Rud. 339 (1) Sed Plesidippus tuus erus || ubi,
amabo, est ? Heia vero ; (2) Sed Plesidippus tuus erus «||bi, amabo, est ?
Heia vero). Nowadays every one accepts the first (III 58).
On the other hand, it may be argued, this cadence exhibits other forms
against which no exception can be taken (e.g. with a Dactyl, Pseud. 524
Priu' quam istam pugnam pugnaoo, ego <?tiam prius), where a Diaeresis is
out of the question (see IV 8). Why then should we invoke Diaeresis to rid
us of the abnormal Tribrach ? So the matter is not clear.
Hitherto we have spoken only of the Senarius. But, naturally, the same
thing is to be looked for at the corresponding part of the Septenarius and
Octonarius ; and if this part be in the first hemistich, the Tribrach will be
the second foot of the whole line :
Asin. 382 Demaenetus «bi dicitur || (Emended to Ubi di. De.) ;
631 Quia ego hanc amo, et haec med amat ||( (Or with Hiatus ' in
Poen. 239 Nimia omnia «*mium exhibent ||;
Pseud. 160 Numqui minus ea. gratia ||; pausa ') ;

An example from a second hemistich is

Epid. 60 |j servum hominem. ea sapientiast (Or with Hiatus ' in pausa ').
Editors must not admit this Tribrach into other parts of Iambic lines.
Thus they must scan Bacch. 168 Istac tenus tibi (with Iambus as second
foot; cf. 52 ; III 42); Asin. 831 Pietas, pater, oculis (III 14); Epid. 518
Eamne ego sin(am) impune (Or with Proceleusmatic ?). And they must
doubt the traditional text of Asin. 762 Ne epistula quidem ulla sit in aedibus
(see III 30) ; probably also (unless omnium hominum is a word-group ;
III 55) of Most. 593. An awkward obstacle is the Senarius : Most. 675
Atque evoca aliquem intus ad te, Tranio (AP), for the Prosodic Hiatus
(III 55) is supported by the Anapaestic line :
Pseud. 1 121 || atque aliquem evocem hinc intus (AP).
The ^4 -reading seems preferable in Trin. 538. Scan Aul. 135 (Iamb. Dim.
Acat.) Da mi, optima || femina, manum (with Dactyl in third foot, as in line
137, and with Syll. Anc. at Diaeresis).
In Terence we find the same usage as in Plautus. A pyrrhic word forms
the beginning of a Proceleusmatic (or Anapaest) in Iambic Verse, not of a
Tribrach. For example, the Iambic line-opening Ita faciam (-cio, etc.)
appears four times, and in each of the four the first foot is a Proceleusmatic :
Andr. 46 Ita faci&m. hoc primum in hac re praedico tibi ;
Eun. 188 Ita facers certumst : mos gerendust Thaidi ;
 PLAUTUS AND MENANDER 85
Haut. 140 Itafacio prorsus ; nil relinquo in aedibus ;
Ph. 776 Ita factum ut frater censuit ||.
(So that we are justified in finding Hiatus in Plaut. Cure. 88, cited in 45.)
The two opening Senarii of the Hecyra's first prologue are :
Hecyra est huic nomen fabulae : haec cum datast (With Hiatus at

Nova, novum intervenit vitium et calamitas. pause)

Surely we may find Syllaba Anceps (Nova) at the pause in the second
(rather than suppose Hiatus after novum), as we find Hiatus at the pause in
the first. Similarly in Andr. pr. 23 (Sen.), Eun. 601 (Septen.) :
mal. ne) ;
(desinant) Maledicere, malefacta ne noscant sua (Emended to Mai. (ut)

Interea somnus virginem op||primft. ego limis specto,

we may find Syllaba Anceps (Maledicere, opprimlt) at the pause (III 50).
(Cf. Andr. 598.)
And the retention of a long vowel before -t in other lines (e. g. Ad. pr. 25
augeat industriam ; III 19) favours these scansions in Ph. 160 (Octon.), 250
(Octon.), 297 (Sen.), Hec. 576 (Octon.), Ad. 839 (Sen.) :
At non cotidiana cu||ra haec angeret animum. Audio ;
Opu' ruri faciendum : horum nil || quicquam accidet animo novum ;
Dotem daretis, quaereret alium virum ;
Celare, cum sciat alie||num puerum tolli pro suo ;
Exporge frontem. Scilicet ita tempu' fert.
But in the fourth foot of the Senarius we have one good example :
Haut. 803 Et simiil conficiam faci/ius ego quod volo ;
And these in the corresponding parts of long lines :
Andr. 535 Meam nubere tu6 gnato ; id vi||so tun an illi insaniant ;
Ad. 343 || mea Sostrata, vide quam r(em) agas ;
598 Sed quaeso ut una mecum ad ma||trem v'wginis eas, Micio ;
Ph. 162 Aliis quia defit quod amant ae||grest : tibi quia jwperest dolet ;
not to mention these :
Eun. 264 (habent) Vocadu/a, ^rasiti ita ut || Gnathonici vocentur (Or
Syll. Anceps at pause) ;
Ad. 262 Qui(n) omnia sibl post puta||rit esse prae meb commodo (Qui
ignominia(s) sibi A).
(Cf. Ph. 725.) When Terence is at last edited, the editor will presumably
print in Ph. 69 supererat ; Haut. 216 libidini ; Ad. 315 Satis mi ; 346 virgini ;
and scan in And. 950 pater ; Eun. 834 tace ; Hec. 503 prSterve (cf. Ill 42).
He will have to emend Haut. 217 (Octon.) :
Mihi si umquam fi/nw frit, ne il||le facili me utetur patre,
where indeed the second hemistich seems to be || ne ill' facili, etc.
(We ignore lines like Ph. 996, Ad. 521, where there is a change 'of speaker.
 86 PLAUTUS AND MENANDER

On Syllaba Anceps at the Diaeresis of long lines see III 58. On Ph. 955,
see above, 9.)
The fragments of Livius Andronicus are so scanty that we can hardly claim
the Greek usage for him on the mere strength of trag. 38 (Sen.) :
Quern ego /^frendem alui lacteam immulgens opem (v. I. non frendem).

48. (B) Anapaest. Since the Anapaest substitutes a pair of short

syllables for the single short syllable of the Iambus, it is reasonable to
regulate the division of this pair between words in Iambic Verse. Take
a line like Mil. 1238 :

. Istuc curavi, ut <?//nio||ne illiu' pulchrior sis.

It is plain that if ut opi- is to play the part of opt- (and indeed it
1 happens that the sense is complete without ut) there must not be any
(perceptible interval between the two short syllables. Otherwise the
Anapaest will not be a fair exchange for the Iambus. This is the rule
observed equally in Menander's Trimeters and Plautus' Senarii. The
two words which divide the pair must be closely connected words,
e. g. (Article and Noun) to Slkolw, (Preposition and Case) kclt' ifiavrovy
(yj)er amorem. These are the normal types, freely used by Menander
and Plautus; and only a step removed from them stood types like
(Epit. 220) Outos, I re togtov\tov ; (Peric. 165) Kctt ttjv o-n-dO-qv, | Zv i8w |,
quid istuc?, ut earn. If Plautus' Anapaests sometimes appear less
smooth than Menander's, it is because we do not realize that the Brevis
Brevians Law is an expression of the actual Latin pronunciation. In a
line like Amph. 140 :
Nunc hodie Amphitruo veniet hue ab exercitu,
we must not say that ab exer- adds another l mora ' to exer-. In Roman
talk exercitu did indeed begin with a spondee, but not when it followed
the Preposition ab> etc. In that phrase it rather began with an Iambus,
just as the phrase cito currit with a Pyrrhic.
The same misunderstanding makes us imagine that Anapaests divided
between words are far more frequent in the plays than they really are.
Klotz's list of so-called Anapaests (Grundzlige, pp. 257 sqq.) in Trochaic
lines is nearly all Tribrachs. The Preposition apud lost its accentuation
before a Noun, and Plautus represents the pronunciation by his scansion
lapud^mensam, apud iudices, etc. The 'brevis brevians' ap- made the
word a Pyrrhic in the every-day pronunciation of such phrases. So in
the sentence || res agitur apud iudices (Pseud. 645) we have a Tribrach,
not an Anapaest. Similarly, the Demonstrative was subordinated to
the Interrogative in a sentence like ubiillast?, and the ' brevis brevians '
ub(i) turned the next word into an Iambus. So we must make the
 PLAUTUS AND MENANDER 87
second foot a Tribrach in Merc. 900 Die \g\tur ubi z/last. Klotz is
[quite wrong in citing these as Anapaests divided between words ; but
kince they are feet of trochaic verse, they do not concern us here.
The only examples which belong to this paragraph are such iambic
lines as :
Mil. 1 2 Neque aequiperare suas virtutes ad tuas (With emphatic
suas, like emphatic bonis of Terence ; cf. 23) ;
Poen. 1045 Siquidem Antidamai quaero tfafo/Haticium (With
adoptaticium) ;

Aul. 395 Confine sagittis fures thensaurarios (With sagittis ; cf. 40) ;
Pers. 339 Mirum quin regis Philippi causa aut Attali (Perhaps
with Philippi ; cf. 40) ;
Rud. 1256 At ego debs quaeso ut, quidquid in il\o vidulost (With
in illo; like 11 09 in isto vidulo; 1133 m isto vidulo; 1228 de
illo vidulo. Cf. 31 end) ;
mensam) ;
Trin. 478 Verecundari neminem apud mensam decet (With apiid

Asin. 36 Modo pol percepi, luibane, quid wtuc sit loci (With esquid
t ?) ;
istuc, the normal pronunciation. Cf. e.g. line 32 Quid istuc
est ? aut ubi istuc est terrarum loci ? ; line 50 quid istuc novi

Cas. pr. 68 Quaeso herr/<? quid zVtuc est ? serviles nuptiae ? (Ditto) ;
Bacch. 507a Id isti dabo. Ego /.ftanc multis ulciscar modis (With
ego istanc, the usual pronunciation. Cf. 31. The next line
begins Adeo ego illam cogam) ;
Capt. 536 Res omnis in incerto sita est || (With in incerto ; like 83
in occulto, etc., etc. ; cf. 27) ;
Cas. 114 Ex sterculino effosse, tua //laec praeda sit? (With tua
illaec. The Possessive has the emphasis ; cf. 31) ;
Cist. 766 l\\\us ego zVtanc esse malo gratiam (With ego istanc?
With Illl(us) ? Cf. 35) ;
Epid. 392 Malefacta mea ^sent solida in adulescentia (With mea
essent. The Possessive has the emphasis).
When the first short syllable is a monosyllable (e. g. ut eratis) or a
disyllable with elided final (e. g. ubi eratis), the close connexion of the
two words is usually secured. Difficulty arises when it is the final
syllable of a word. Here, too, a close enough connexion appears in,
e.g. Menander's Ovk la\n hiKai\ov (Epit. 131), Plautus' propter aworem
(Mil. 1284 and 1288. Cf.Ter. Andr. 155). But, as a rule, this arrange-
ment is avoided. It is incredible that in Aul. 692 :
Iuno Lucira, tuam fidem ! Em, mater mea,
88 PLAUTUS AND MENANDER

the third foot can be an Anapaest. Plautus must have pronounced

tuam (33), or, less probably here, tiiam (III 33). (Cf. Men. 736 ; Amph.
477 ; Capt. pr. 101.) So in Amph. pr. 31 :
Contagion mei patris metuo malum,
the third foot is probably a Tribrach, with emphatic mei (or, less
probably here, mei, making the foot an Iambus). It is natural to
pronounce minu' (not minus ; cf. Ill 16) in Aul. pr. 19 :
Curare minusque me impertire honoribus,
so that here too we have a Tribrach, not an Anapaest.
In Aristophanes there are some examples of ■ false ' division of the
Anapaest, e. g. :
Ach. 6 TotS 7T€VT€ TdXaVTOlS OtS KA.€0t)l/ i£yfJL€<T€V,

and in the Birds, line 114 "On 7rpQyra jxkv rja-Oa, 144 'Arap 1<ttl y biroLcw
Xeyerov, 843 K^ovko, Sk Tre/xif/ov, 1024 vE7re/>t^€ Se tis; 1495 Hov IIcio-^c-
raipos io-Tiv ; vEa (?). So far as we can judge from the Menander
fragments, the New Comedy was stricter in this matter than the Old.
Terence seems to stand with Plautus. (In Ad. 337, 913 the MSS. are
at variance.)

49. Doubtful cases may now be discussed. Since the MSS. of Plautus err
sometimes by omitting small words, sometimes by inserting them, an editor
has often to ask whether a (seemingly) less smooth Anapaest is Plautine :
e.g. True. 240 and 251 :
|| neque iimquam ulla. sati' poposcit (A : om. umquam P).
\\prope hdsce «*des adgrediri (P: om. hasce A).
He will find no harshness, but rather an exact echo of talk, if he compares
with Romance usage (e. g. French je l'sais) such a phrase as Amph. 922,
Men. 246 :
Ego ilium scio quam doluerit cordi meo ;
Ego ilium scio quam cordi sit carus meo ;
(Cf. Men. 535-6), or quid est quod of True. 238 (|| quid est quod male agimu'
tandem ?) And he may make a rough-and-ready rule that lenity should be
shown to claimants which begin with a monosyllable or a pair of mono-
syl ables, e.g. :
Merc. 699 Sed hinc quina.m a nobis exit ? aperitur foris ;
Asin. 720 Opta id quod ut ^«tingat tibi || vis ;
Mil. 28 At indiWgenter iceram. Pol si quidem (Cf. Merc. 751) ;
Capt. 71 Scio absurte dictum hoc derisores dicere (Cf. Merc. 726) ;
Amph. pr. 107 Is amare occepit Alcumenam clam virum ;
Merc. 5 13 ||quod ego omnes scire credam (With emphasis on omnes ; 29) ;
Asin. 672 \\fer aman\\ ero salutem ;
Cas. 618 Quoi sic tot amantl mi obviam eveniunt morae ;
 PLAUTUS AND MENANDER 89
et);
Trin. 421 Et Ule #*des mancupio abs te accepit. Admodum (Cf. Merc.
567, 828) ;
and perhaps even
Merc. 536* Ego cum viro et UP cum muliere || (Emended by deletion of

Mil. 552 Aqua aqua'x sumi, quam haec est atque ista hospita.
But, on the other hand, that where the opening syllable is the final of a word,
the utmost severity is proper. Luckily we have the Palimpsest to correct
the false Anapaest in Men. 268 : Tu mzgnus amator mulierum es, Messenio
(Tu magis amator A) ; while P corrects A in Stich. 432 fra/m «//*//l[ul]am.
And though the traditional reading may^ possibly be retained in other lines, it
is only by the expedient of making the foot a Tribrach, not an Anapaest, e. g. :
Amph. 873 Nunc Amphitruonem me///*/ ut &rcepi semel ; of et ) ;
Cure. 271 Petas, ne for/* tibi *veniat magnum malum (Or fort' ? ; 36);
Asin. 59 Bene hercle facitis et a me initis gratiam (Emended by deletion

or by some other scansion which removes the illegitimate type of foot (e. g.
by ill', not ille, in Bacch. 885 ; by ecqui or a pyrrhic ecquis in Rud. 413
Heus ecquis in villast ; Bacch. 583 Fores pultare nescis : ecquis in aedibust ?
Cf. Ill 38 ; by hasc', not hasce, in Bacch. 787 ; perhaps by atq' ' forthwith ',
not atque, in Men. 508. Cf. Ill 50). Usually the reading must be emended :
Capt. 94 Nam Aetolia haec est, illi est captus [in] Alide ;
Epid. 532 Ita gnato men Ziostiumst potita (Emended to Ita mea gnata) ;
Asin. 405 || animisque expletus [in]ced\t (cf. Mil. 897 ; Poen. 577 incedit
P : cedit A ; see III 17) ;
Rud. 888 Nam in columbari coitus hau multo post erit (Some MSS. of
Priscian have Nam collus in columbari. This seems the
right order) ;
Merc. 773 Cur hie astamus ? quin abimus ? incommodi (AP).
(In this last example there is no room for the operation of the Brevis
Brevians Law. The -mus cannot be a ' brevis brevians ' ; 29.)
Finally may be mentioned a curious case where the first short syllable is
indeed a monosyllable (or rather its equivalent, a disyllable with elision) but
is followed by a change of speaker :

Merc. 523 Operam accusari non sinam || meam. Em z'jtaec hercle res est.
We find here a Tribrach, with em fetaec (III 52). In Merc. 580 (Apud ted
hie usque ad vesperum. Em &tuc censeo) the Tribrach saves the similar
situation. Emendation (e. g. quid nunc, obsecro ?) is the only cure for Merc.
683 (Dorippa,mea Dorip^ta. Quid clam&s, obsecro ?), where indeed an Anapaest
cannot be found, since there is no room for the Law of Breves Breviantes (29).
In an Atellan farce of Pomponius (time of Sulla) this false Anapaest may
be due to the necessities of the phrase amicus amici ' a true friend ' (III 17) :
146 (Sen.) (ut siquis est) Amicus amici, gaudet sicui quid bohi (Evenit) ;
9Q PLAUTUS AND MENANDER

though Terence and Accius prefer to depart from the usual order of the
words :
Ter. Ph. 562 (Troch.) Solus est homo amico amicus f| ;
Ace. 132 (Troch.) Qui neque amico amicus umquam ||gravi' neque hosti
hostis fuit.
Of course Iambic lines of Cantica are not subject to the same strict rules
as the lines of dialogue ; so we may tolerate in Cas. 751, if it is an Iambic
Dimeter Acatalectic :
Gladium Casina in||/^j habere ait,
though some find here an Anapaestic Dimeter (with ait).
Elision (of the final of a trisyllable or polysyllable) often invalidates rules
for divided feet. It does not invalidate this rule. An Anapaest like ' omni(a)
habeb&s ■ or ' msomni(a) nabebas ' or ' negoti(u?n) kabebas ' is as strictly for-
bidden as one like ' omne tenebas '. Editors change paululum to paullum in
an Octonarius of Terence (Adel. 950) :
Huic demu' qui fruatur.
Pau||l[u]lum id autemst. Si multumst, tamen
(similarly vah to ah, Ad. 405).
In Ter. Eun. 830 some scan Istucine (III 20) ; some read Istucne. In
Hec. pr. 34 scan ebdem ; 331 eri; 769 ebria eri' (or erls). In Poen. 1231
(Iamb. Sept.) A offers :
Sed fllud quidem volui dicere — immo herc||le dixi quod volebam (immo ||
dixi he. P ; III 58).
In Cure. 520 (Iamb. Sept.) the seventh foot must be a Tribrach (IV 10).
(On the evidence of this division of an Anapaest in trochaic verse also
see 55.)
50. But there is another rule for the division of an Anapaest between
words, a rule as strictly observed by Plautus as by Menander, and in
trochaic verse as well as iambic. If-the two short syllables stand in!
Jone word, it must be a pyrrhic word. An Anapaest must not begin I
*with a pyrrhic word-ending. The rule is well known through the
discussion aroused by the unlucky proposal to alter the traditional
reading in the Asinaria prologue, Maccus, to Maccius (pr. n):
Demophilu' scripsit, MacaV vortit barbare.
This gives an illegitimate Anapaest, where two short syllables are a
word-ending.
This type of Anapaest is of course quite legitimate in Anapaestic
verse. That Plautus, while using it with the utmost freedom in his
anapaestic Cantica, should have restrained himself from its use in
dialogue, shows that he was no rude versifier. Our ears nowadays
cannot detect a reason why, e. g. || video. Ut tremit ! atque extimuit
(Mil. 1272) is legitimate, but not e.g. '|| video. Attremitjaque exti-
muit \ The reason why a Dactyl was not equally disliked when its two
 PLAUTUS AND MENANDER 91

short syllables stand in a different word (e.g. Mil. 58 Amant ted omnes
mu/ieres neque iniuria), an extremely common type of Dactyl, is thought
to be that the resolved syllable is the equivalent of the long syllable, not
the short, of the normal foot, the Iambus.
Aristophanes has some Anapaests so divided, e. g. in the Birds, line
1022 'E7rtcrK07ros r/KU), 1228 'A/cpoaTcoy vfXLv, 1 363 'AAA.' oldirep avros. So
we are left in doubt whether the New Comedy made the law stricter or
whether, when more of Menander is discovered, we shall find examples.
The examples in the traditional text of Terence are usually removed by
scholars, e. g. Ad. 266 (Iamb.), Ph. 266 (Sen.) :
Nil video. Ehem opportune : te ip||sum quaer[it]o : quid fit,
Ctesipho ? ;
Hie in nox[i]ast, ille ad defendendam causam adest ;
But whether rightly or wrongly is a question that cannot be answered
until there is a proper edition. Notice that the iambic line-opening,
Hec. 604, is in the Bembinus Si cetera sunt ita, but in the other MSS.
has the legitimate form Si cetera ita sunt (with Tribrach, not Anapaest,
in the second foot).
51. This type of Anapaest is rarely defended by an editor, and the emen-
dation isin most cases certain. Examples are :
omnis) ;
Amph. pr. 55 Comoedia ut sit omnibus iso\tm versibus (Emended to
of sibi) ;
1061 || nam ubi parturit deos sibi invocat (Emended by deletion

Asin. 469 Nemo z.ccifiit ; aufer te domum (Emended to te aufer) ;

Bacch. 142 Praesen//^j tills paedagogus [una] ut siet (Praesente illis is
the usual old Latin phrase) ;
Capt. 532 Nugas, meptias inclplsse. haereo (in a Canticum).
Cas. 335 Sed tandem si tu luppiter sis [ejmortuus ; adicere) ;
Merc. 282 I et hoc memento dlcere. Numquld amplius ? (Emended to
rustica) ;
Most. 40 Germana illuvies rusticus hirews hara suis (Emended to

Poen. 646 Suspendant omnes mmciam sese haruspices (sese A : seP);

Pseud. 149 Verum ita vos estis pr&editi w^glegentes ing enio impmetre)
Canticum robi (a;

805 Nemo ilium quaerit qui optumus [et] carissumust (' seeks the
best who is dearest ') ;
Rud. 304 Nisi quid concharum c&psimus, [z'«]cenati sumu' profecto ;
True. 78* Fhronesium, nam phronesis est sapientia (A gloss ?) ;
447 Quam mihimet omnia qui mihi facio mala ;
19 Quo citiu' rem ab eo auferat cum pulvisculo (Emended averrat).to
92 PLAUTUS AND MENANDER

The reading may be retained, but the scansion altered in

Aul. 297 Pumex non aeque est aridus «/que hie est senex (Not aridus,
but ardus ; cf. ardeo) ;
pr. 35 Is adulescentis Wlius est avonculus (Rather illiust ; or with
disyllabic illius of 35).
And the final e of haecine, istocine, zsfaeine, isticine must be suppressed in
Poen. 1 166 ; Pseud. 83 ; 847 ; Rud. no (and therefore presumably through-
out the plays). Print haec quidem in Mil. 1259 as two separate words.
The only cases where any strong defence seems possible are some lines of
Cantica :
Pers. 284 Video ego te : iam incubitatus es. || Ita sum. quid id dXttnet
ad te ? (The anapaestic flavouring is very pronounced) ;
Stich. 769 Qui lonieus aut cinaedicust ? ||;
Pers. 854 Sati' sumpsimu' suflpMcl iam (Emended to iam supplici) ;
Aul. 429 Quia venimu' eoetum ad nuptias. || Quid, tu, malum, curas ?
(A versus Reizianus) ;
157 His \egtdu' quam dare vis ? cedo : || nuptias adorna (Ditto).
(On Trin. 247 see IV 12.)
We may then accept this law for dialogue without any exceptions and so
determine the quantity of Ampsigura in Poen. 1065 (Ampsigura mater mihi
fuit, Iahon pater) and fustitudinas (from tundd) in Asin. 33 (Apud fustitu-
dinas, ferricrepinas insulas).
But a polysyllable with elided final is exempt (cf. Menander Peric. 58 Tdp-
ravB' aKOvo-rj yi\v6yL€v' €K\Tre7rop<f)€ fie)} e. g. :
Mil. 847 Non edepol faciam, age tloquere awdacter mihi ;
Pseud. 14 Prohibessit. Nil hoc Iovis ad zxbitrium «/tinet;
Aul'. 49 Testudineum zstum tibi ego grandibo gradum ;
Pseud. 105 Tibi inventurum esse zuxifium argentarium ;
Amph. 196 || ductu, imfierio ##spicio suo.
So we may print in Mil. 483 :
Certo Wlaqnidem hie nunc intus est in aedibus (If hie is emphatic, a
Tribrach is not to be found here) ;
and we may leave unaltered Amph. 463 :
Bene prospereque hoc hodie open' processit mihi,
where the succession of short syllables suits the joyful tone of the line.
Examples from the (unedited) text of Terence are : Andr. 613 ; Eun. 197;
Ph. pr. 14 ; 266 ; 665 ; Hec. 134 ; 573 ; Ad. 266 ; 949. (Cf. Hec. 604.)
Add from the Dramatic Fragments :
Accius 100 (Sen.) Quot luna circ[ul]os annuo in cursu institit (III 26).

52. (C) Proceleusmatic. The Proceleusmatic (unknown to the

Menander MSS., unless Epit. 22 has one) is a favourite foot of Plautus'
Senarii. We have already seen that while Menander tolerates as an
Iambic line-opening a Tribrach which begins with a pyrrhic word,
 PLAUTUS AND MENANDER 93

e. g. I Kara \6\yov, Plautus substitutes a Proceleusmatic. His line-

opening would be I KaTtt Xoyov \ . And not merely in the first foot
of the Senarius, but elsewhere (with the apparent exception of the fourth
foot) a pyrrhic word does not become the beginning of a Tribrach, but
rather of a Proceleusmatic or Anapaest (e. g. Poen. 1069 Pater tuus, is
exdX. Not * tuus tfrat '). It is natural to refer this difference to Plautus
regard for Accent, although the limited number of such Tribrachs in
Menander (about one in a hundred lines) shows that even the Greek
1 ear found something unattractive in them. At any rate these Proceleus-
/ matics contribute greatly to the harmony of accent and ictus in Plautus'
lines. To English ears, accustomed to accentual verse, an Iambic line
opens smoothly with Ita mea consilia ; harshly with Ita mea, consilia.
Not that the ictus of the Proceleusmatic falls with unvarying monotony
on the accented syllable of the quartette. Occasionally it is the first of
the two Pyrrhics that bears the emphasis, e. g. True. 388 Tibi mea
consilia semper summa credidi (with emphatic Tibi, ' it is to you I ', etc.) j
but the smoothness remains for all that. And the conglomerate of short
syllables, grouped under a single foot, seems to echo admirably the
patter of talk ; True. 172 Ego/ateor, sed etc. ; Cist. 298 Video ego; Trim
215 Tibin ego rationem reddam?; Asin. 844 \\ouia tibi non cupiam ;
Rud. 401 At ego etiam ; Rud. 1199 Ego eum adeo arcessi hue ad me
quam primum volo ; True. 188 ||utut atiis, tibi quidem intust (* she is at
home to you ') ; Capt. 187 Habe modo bonum animum ; Rud. 1255 Ego
tibi daturus nil sum ; Trin. 750 Sed ut ego nunc (with emphatic Pro-
noun) Men.
; 1001 ||quid ego ocuWs aspicio meis ? ; Pers. 147 propera, abi
domum ; 2gjpotin abeas ? ; Asin. 724 Quid ego #//ud exoptem ? ; Cas. 613
Abi et aliud cura ; Asin. 488 ||ita facito, age ambula ergo. Dull indeed
would be the editor who would lop off a short syllable where he could
and efface the Proceleusmatic, let us say, in Bacch. 1043 (Ego neque te
iubeo neque veto neque suadeo) by printing Ego nee. For even careless j
readers can see how Plautus revels in short syllables on occasion, and i
how he uses this device to mark the tone of the dialogue. In Menander I
(to put it forcibly, rather than quite accurately) there are only two tones,
the Iambic of ordinary dialogue and of soliloquy, the Trochaic of ex-
cited conversation (with quick music from the orchestra). But Plautus
can pass from the deliberate utterance of dignified talk or of soliloquy
(with few resolved feet ; e. g. Aul. Ill v) through various crescendo tones
of conversation (liveliest in slaves' banter, where Proceleusmatics are as
thick as blackberries) without quitting the Senarius. (Notice the effect-
of this foot in Most. 513, at the ghost-scare: Fuge, obsecro, hercle.
Quo fugiam? eiiam tu fuge.) If we remember that the Brevis Brevians
94 PLAUTUS AND MENANDER

is not a poetic licence, not a Procrustean plan of squeezing a square

word into a round hole, but echoes exactly the pronunciation of every-
day (educated) talk, we shall find no harshness in Proceleusmatics like
Trin. 595 Sed id si #//enatur ; Trin. 559 Meu' quidem hercle numquam
net j Rud. 1311W sine hoc a/iud fabulemur ; Bacch. 1050 Dolis ego
deprensus sum, ille, etc.; Capt. 133 Quis hie loquitur} \ Rud. ^^^
\\ quis hie loquitur? quern ego video?; Rud. 703 || metus has id ut
faa'ant subigit ; Rud. 506 Sce/u' te et sceleste parta qui vexit bona (' you
and yours') ; Men. 229 Quasi adveniens terram videas ; Asin. 699 Vehes
poi hodie me, si quidem hoc || argentum ferre speres.
But in this paragraph we are concerned with the division of the
Proceleusmatic between words in iambic lines. There are two rules like
the rules for an Anapaest : (1) Since an Anapaest is forbidden whose short
syllables form a pyrrhic word-ending (though a pyrrhic word is welcomed)
we must question (at any rate, outside Cantica) Proceleusmatics like :
Asin. 403 || erus in hara, haud zedibus habitat (IV 10, end) ;
485 Quid, verbero ? ain tu iurcifery [eruni] ||nosmet censes
fugitare
?;

Pers. 87 Commisce mulsum struthea ^/wteaque appara ;

Pseud. 1320 Heu heu heu ! Desine do/eo. Ni||doleres tu ego
dolerem (Heu heu des. A) (Cf. 59).
Ter. Ad. 60 Venit ad me saepe dzmitans 'quid agis, Micio ? '.
We must separate tenus from ea in, e. g. Most. 131 (Ea tenus abeunt
a fabris ||), pott from ut in e. g. Rud. 462 (Sati' nequam sum ut pott qui
hodie amare coeperim ; App. A, end).
But as with the Anapaest, so with the Proceleusmatic a polysyllabic
word with elided final is exempt from the ban. We may allow in Poen.
1042 :
Verum ego hie hospitium habeo : Antidamae filium (Some scan
with a Dactyl Verum ego hie).
(2) And since the final short syllable of a disyllabic (without Elision)
or longer word is the legitimate beginning of a Tribrach (e. g. Bacch.
254 Quid ita, obsecro, here/.?? Quia ^depol certo scio; to quote one
example of the commonest of all resolved feet), but not of an Anapaest,
we must question Proceleusmatics like :
Asin. 483 || de nobis detur. Atque eti&m (EmendedIV to 10, datur
end) ;

493 Merito meo neque me Athenis dXter est hodie quisquam

(Or hodiest ?) j
Cure. 401 Licetne infora/^ si ineoftotvaxt non licet? (Emended to
inc. haud licet) ;
 95
PLAUTUS AND MENANDER

Pers. 3T9 Enim metuo ut possiem in bubi||te reicere, ne (Cf.59);

vagentur

? Poen. 922 Ero uni potius intus ero odi\p quam hie sim vobis
omnibus (Rather trochaic ; see below, 56).
Ter. Ad. 475 Compressu gravida facta est, menszV hie decimus est.
So if the reading is to stand, the scansion must be altered in
Capt. 666 Confidentem esse snum apud erum potissumum (With
suym ? With ess' ?) ;
Epid. 364 ego||met solus, earn ego docebo (With eum ?; 33) ;
May ecquis hoc aperit ostium ? (Bacch. 582, etc.) escape by the same
plea as the Anapaest propter amorem (48) ?

(3) Like the Dactyl with trochaic caesura (see the next paragraph), so
the Proceleusmatic formed of a Tribrach and an initial short syllable is not
favoured. The first eleven plays supply these instances :
(In first foot)
Cas. 564 Hominem izmatorem ullum ad forum procedere (cf. 9) ;
Merc. pr. 29 Inerit *?tiam aviditas, desidia, iniuria (Or Inhaeret MSS ?) ;
(In other feet)
Epid. 332, 334 Alicunde ab aliqui aliqua tih'i spes [|
Quipp' tu mi aliquid rtliquo modo ||.
This scansion seems impossible (29) in Epid. 353 (Manibus his denu-
meravi, pater ||), where the traditional text must be emended. The true
pronunciation may be monosyllabic ill' in Cas. 432 (Ut ille trepidabat, ut
festinabat miser !) with the first foot an Iambus, not a Proceleusmatic.
No exception can be taken to a Proceleusmatic like Asin. 430 Erus in hara..
This is of Dactyl type and conforms to the rule for the Dactyl (53) that the
two short syllables, when they are divided between words, must be closely
connected in utterance. The words in hara would be actually written as
one word in ancient (and pre-Carolingian) times. Similar Proceleusmatics
are, e.g. : Asin. 699 Vehes poi hodie me ; Bacch. 508 Adeo ego //lam cogam ;
Men. pr. 70 Hodie in ispidamnum ; Mil. 1257 Quia me #mat; 1276 Egon ad
//lam earn ?
Rather scan quid ita as an Anapaest (III 4) with Syllaba Anceps 'in
pausa ', or as a Tribrach, than find a Proceleusmatic in the first foot of lines
like:
Cure. 48 Quid ita ? Quia proprium facio ; amo pariter simul ;
Mil. 1260 || quid ita? Quia stare nequeo ;
Poen. 691 Quid ita ? Quia (a.) muscis si mi hospitium quaererem
Pseud, jy Quid ita ; Genu* nostrum semper siccoculum fuit ;
as is enforced by Poen. 705 :
Quid ita ? Quia aurum poscunt praesentarium.
96 PLAUTUS AND MENANDER

53. (D) Dactyl. In Menander's Trimeters the trochaic caesura of

a Dactyl (- w, w) is not rare in the first foot (e. g. Hpa£iv ijSiai/ and the
like, almost twice in every hundred lines), but extremely rare elsewhere
(Peric. 272 *A£iov ISelv. | dAAa ri\ <f>4pw vvv cis /xeaov ;). So in Plautus :
e.g. Poen. pr. 73 and 88 Vendit tfum domino, Vendit *as omnes ; or
with the second short syllable due to a Brevis Brevians, Poen. 445 Illic
hinc iratus abiit. (But rather 111' qui than Ille qui ; rather 1st' qui than
Iste qui in lines like Pers. 520 1st' qui tabellas adfert ; cf. 36.) Plautus,
who does not tolerate an Iambic line-opening like Agit aim, with first
foot a Tribrach and the ictus on the final of the pyrrhic word, tolerates
this clash of accent and ictus when the line opens with a trochaic
(instead of a pyrrhic) word, Vendit (but not Agit). Still he does not
favour it. A portion of equal length to the Menander MSS. — the
Senarii of Pers. Poen., Stich. — offers only half as many instances. This
Dactyl-opening is therefore about as frequent (or rather infrequent) in
Plautus as the Tribrach-opening (of the type Kara \6\yov) in Menander.
But this trochaic caesura of the Dactyl must be formed by a trochaic
word. A trochaic word-ending is not allowed. In the whole of Plautus
the traditional text offers only :
Aul. 377 Abeo iratus //line, etc. (corrupt; see below) ;
When the last syllable is elided, we have such examples as (in the
Bacchides) :
939 Bacchidem hahei secum, etc. ;
960 ||detuli, *bi occidi Troilum ;
988 || Chrysale, ades diim ego has pellego,
all three in the first foot of the line or hemistich.
The two short syllables of the Dactyl, when they are divided between
words, are closely connected in utterance, e.g. Amph. pr. 41 ; pr. 56 i
pr. 109 ; 466 j 492 (all Senarii) :
Nam quid £go memorem, ut alios in tragoediis ;
Utrum sit an non vultis ? sed £go stultior ;
Et gravidam fecit is earn compressu suo ;
lam ille illuc ad *rum quom Amphitruonem advenerit ;
Amphitruo rem omnem, quid /gitur ? nemo id probro.
54. These statistics of the trochaic caesura of a Dactyl in the Iambic
lines of the first eleven plays will show on the one hand how infrequent it is,
and on the other that it cannot (like the pyrrhic caesura of a Tribrach) be
declared un-Plautine.
(1) Strong examples in first foot :
Amph. 943 Inter £OS, rursum si reventum in gratiam est ;
Asin. 773 Ne ilia minus aut plus quam tu sapiat. Sati' placet ;
 PLAUTUS AND MENANDER
97
Bacch. 808 Nullus homo ditit : hae tabellae te arguunt ;
Cas. 456 Ecquid «mas nunc me ? Immo edepol me quam te minus ;
Cure. 9 Tute tib\ puer es, lautu' luces cereum (cf. Capt. 371 : Cist. 563) ;
245 Aufer fttaec, quaeso, atque hoc responde quod rogo ;
Epid. 8 Cena tihx dabitur. Spondeo. ||;
488 Em ('extra metrum ' ? ?) Istic homo te articulatim concidit, senex
(cf. Amph. 676; Capt. 751 ; Men. 98) (III 52) ;
(2) Examples in other feet :
Aul. 40 Exi, inquam, age exi, exeundum hercle ttb\ hinc est foras ;
Asin. 100 Venari autem rete iazxAo in medio mari (Hardly * reti ' ; cf. 7) ;
Aul. 139 Nam optima nulla potest eligi (Dimeter) ;
446 Mihi vasa tubes pipulo || (Vers. Reizianus) ;
Epid. 179 || Herculi, quam ilia mihi obiectast (Or illaec?).
From these lists are excluded not merely lines with 111' (rather than Ille),
even (cf. Ill 46 end) Merc. 532 (Ille te homo : ita edepol deperit ||), but
also Quiv' of Amph. pr. 84 (Quive quo placeret alter fecisset minus), Siv'
(i.e. Seu) of Amph. pr. 69 (?), Post' (poste MSS.) of Bacch. 966 (Poste cum
magnufico milite ||), Fort' of Asin. 794 (Forte si tussire occepsit, ne sic
tussiat), ess' of Cas. 230 (Heia, mea Iuno, non decet || esse te tamtristem tuo
Iovi) (cf. 36). Further the alternative expression of two words as one,
e. g. Sic ut (Asin. 558 Sicut ego possum quae domi ||; cf. Cas. 566), Ne quid
(Aul. 305 Nequid animai forte amittat dormiens ; cf. Cas. 784 Nequis ; Men.
340 Siqua. But see III 42, s.v. siquis). Also Istaquidem Merc. 730 (cf.
Ill 9). These all appear in the first foot. Like cases in another foot are
Asin. 714 (|| abs te, atque ipse me adgredere. Read ips' med) ; Merc. 540
(|| puer est illequidem, stulta. Cf. Ill 9). Since this form of Dactyl is clearly
not a favourite of Plautus, we are probably justified in excluding the spellings
which necessitate it from the text also. The scansion of Epid. 541* is un-
certain (||primu' pudicitiam pepulit. Or primus pudicitiam pepulit ; cf. 27) ;
the reading of Aul. 377 (Abeo iratus illinc quoniam nil est qui emam).

55. (II) In Trochaic Lines.

(A) Tribrach and Anapaest. I. It is obvious that whether we
credit the Greek theory or not, by which a Trochaic Tetrameter Cata-
lectic (Septenarius) was merely an Iambic Trimeter (Senarius) with a
^Cretic prefixed, we must keep separate the Latin division of resolved
t between words in trochaic and in iambic lines. Klotz's attempt to
F^ike one investigation and one report for the two led him into errors.
The use in trochaic lines of a Tribrach like: omnia memini et scio
(Pers. 186), mi quoqu^ Zucridem confido (Pers. 627), seems to have made
him infer that in iambic lines such a Tribrach as : ' omnia z«£mini et
scio ', * mi quoque Zwcridem ' must be equally frequent. A great mistake !
/We have found reason to doubt whether the pyrrhic caesura of the
ITribrach is ever allowed by Plautus in his iambic verse (44). In his
2348 H
 98 PLAUTUS AND MENANDER

i trochaic verse it is common, e.g. || quia pentet praedico (Pers. 233),

when a pyrrhic word (not word-ending) forms the beginning of the
Tribrach.
V A pyrrhic word-ending cannot form the beginning of a Tribrach in
trochaic verse, nor yet of an Anapaest, so that e. g. these two lines of the
Pseudolus, cannot be trochaic with this scansion :
Pseud. 146 Ut ne peristrowa/a guidem aeque ||(cf. Stich. 378, with
1 ^Y^ Ace. PI. peristroma) ;
218 Ain excetra tu ? quae tibi amicos ||(tot habes) (cf. IV 3 C)
(but cf. Bacch. 615 and Stich. 736. The text of Men. 405 is doubtful. On
Cure. 192 see 9) ; and we must make two separate words of the Adverb
in Pseud. 357 (Pseudole, adsiste altrim secus, at||que), and one of the
Verb and Pronoun in Trin. 880 (Multa simul rogitas ; nescioquid || expe-
diam potissimum. Cf. Ill 38). And the same ban is laid on these types,
as we have seen, in iambic verse, where neither ' peristromata quidem '
satisfies the requirements of a Tribrach (47) nor ■ excetra tu ' of an
Anapaest (50). In trochaic (as in iambic) a polysyllable with elided
final is exempt ; e. g. zeeipere hunc is a legitimate Anapaest.
The divided Anapaest of iambic verse could not, we found (48), begin
with a final (short) syllable unless the two words were closely connected,
e. g. propter amorem (Prep, and Noun). Otherwise, we must recognize
a Tribrach or else emend the text. The same applies to the divided
Anapaest of trochaic verse. While, e. g., inter erum (Pseud. 648), inter
eos (Trin. 623) need not be questioned, we found (48) that most of the
so-called Anapaests in Klotz's lists were really Tribrachs, e. g. res zgitur
apud iudices (Pseud. 645), Die igxtur ubi z/last (Merc. 900). The others
are licences of Cantica (e. g. Stich. 85 ; 736) or suspect.
II. Ictus on final of pyrrhic-word in Trochaic Septenarii. We have men-
tioned (44) Plautus' dislike of a divided Tribrach like lAgit *?as' in Iambic
verse. Is this clash of ictus and accent in Pyrrhics (e. g. ' digit homines ')
avoided in the other conversational metre, the Trochaic Septenarius ? The
stronger examples are :
Asin. 184 Vult famulis, vult etiam anciliis, || et quoque catulo meo;
Aul. 208 Nimi' male timui. priu' quam intro || redii, exanimatus fui ;
Bacch. 83 Ubi tu lepide volgs esse tibi, ||' mea rosa ' mihi dicito ; laves ;
105 Cupio. Dabitur opera, aqua calet : || eamus hinc intro ut

? 472 Ubi ea mulier habitat ? Hie. || Unde earn 6sse aiunt ? Ex

Samo (Emended to Hie. (Hie ?)) ;
?75x Quia mi ita libet. potin ut cures || te atque ut ne parcas
mihi ? (Or mihi ?) ;
Capt. 358 || gratia ea gravidast bonis (Or ea gravi., Proceleusmatic ?) ;
 PLAUTUS AND MENANDER
Varro)99 ;
Merc. 620 Non tibi istuc magi' dividiaest || quam mihl hodie fuit (P,

Mil. 235 Eru' meus elephanti corio || (cf. 775 ; Pseud.

Most. 1063
660; ; True.
Poen. 579);
901 ;
727 || qui zsXprobus agoranomuS (Or probust ?) ;
Pers. 537 Mea quidem istuc nil refert : tu\\d ego hoc facio gratia
(Emended to qu. (hercle) ist.) ;
546 Nisi quia specie quidem edepol || (cf. Trin. 938 ; True. 786) ;
565 || te alter erit opulentior (Or er. opu., Proceleusmatic ?) ;
626 Mihi quoque Lucridem confido || (fore-te, etc.) ;
Poen. 903 || mebque ero eum hie vendidit ;
Pseud. 300 Ita miser et amore pereo ||;
317 Aut terra aut mart alicunde ||;
? Rud. 419 Sed quid ais, mea lepida, hilara ? Aha ! ||nimium familiariter
(Or with Hiatus at change of speaker ?) ;
1001 Quod scelus hodie hoc inveni ? ||;
1069 Quo modo habeas, id refert || (OrQuomodo, a Dactyl-word);
1 1 14 E6 tacent quia tacita bonast || mulier semper quam loquens
(Or quiast ?) ;
Stich. 696 [a]Mica titer utrubi accumbamus || (Or uterutrubi, agroup) word- ;

Trin. ? 684 Numquam erit alienis gravi' qui || suis se concinnat levem
(Or gravis ?) ;
? 714 || quod meum erit, id erit tuum (Or meiim er. id erit t.) ;
1046 || nam id genus hominum omnibus.
That is not a long list. It might conceivably be shortened by claiming for
Proceleusmatics (e.g. Mil. 618 || facinom! ^«*rilia), rather than for Pyrrhics,
the word-groups nimi'-male, mea-rosa, eru'-meus, nisi-quia. On the other
hand it might be greatly extended by the admission of lines where alternative
scansions are possible. For some prefer to avoid this ictus of a tribrach-foot
wherever they can, and would even ignore the prevalence of magi' etc. (before
a consonant ; III 16) in favour of, e. g. :
Mil. 615 Quis homo sit magis meu' quam tu's ? || ; confer) ;
782 Ecquem tu potis reperire ? (as in 781 Quam potis tarn verba

Similarly with erus, Aul. 619; tuus, Bacch. 994 ; modus, Merc. 652 : domus,
Merc. 653 ; opus, Most. 262 ; and meus, Pseud. 1235 :
Eru' meus tibi me salutem ||.
They prefer the Iambus at a pause (III 48-50) in the line, e. g. Iovls, Trin.
941 ; tuus Pseud. 11 52:
Sub solio Iovis ? Ita dico || (as in Trin. 971 Neque edepol tu is es
neque hodie is || umquam ens— auro huic quidem) ;
Eru' tuus — Ita dico. Miles ||.
Similarly they scan eris, Asin. 870; bibam, Aul. 623 ; agls, Aul. 636 ; erit,
Men. 956 ; lapis, Mil. 236 ; erus, Mil. 451 ; caput, Mil. 725. They have no
H 2
ioo PLAUTUS AND MENANDER
hesitation in disabling the Brevis Brevians, e. g. Amph. 778 tibl, rather than
'tibi':
Em tibi pateram. eccam. Cedo mi ||.
Similarly they scan sciat, Amph. 510 ; sibl, Asin. 945 ; abi, Aul. 455 ; vellt,
Aul. 599 ; tibi, Capt. 897; ego, Capt. 1000; homo, Cas. 303 ; utl, ego, Cas.
682; amo, Cas. 802; sino, Cist. 489; moror, Cist. 778; tibi, Cure. 195 ; amet,
Cure. 208 ; ego, Cure. 294; vale, Cure. 588 ; mihl, Epid. 50 ; iuvat, Epid. 113 ;
soror, Epid. 657 ; ego, Epid. 668 ; eg5, Epid. 677 ; mantis, Epid. 697 ; ibl,
Men. 187 ; cito, Men. 225 ; ego, Men. 463 ; tibi, Men. 680 ; tenet, Men. ion ;
enlm, Merc. 159 ; pater, Merc. 366 ; volo, Merc. 368 ; habet, Mil. 191 ; tibi,
Mil. 1419; mihl, Most. 386; ego, Most. 1096; tace tace, Pers. 591, Pseud.
6ooa ; volo, Pers. 612 ; soror, Poen. 263 ; ego, Poen. 274 ; mihl, Poen. 598 ;
mod5, Pseud. 724; utl, Poen. 840; libet, Poen. 841 ; ubl, Poen. 855 ; utl,
Poen. 1204 ; tibi, Pseud. 631 ; licet, Rud. 724 ; ego, Rud. 1077 ; tibi, Stich.
123 ; soror, Stich. 147 ; scio, Stich. 337 ; homo, Stich. 360 ; tibi, Stich. 547 ;
libet, Stich. 698 ; mihl, Stich. 753 ; eget, Trin. 330 ; moror, Trin. 337 ; volo,
Trin. 696 ; soror, Trin. 713 ; nonne ego, True. 257 ; dolet, True. 526 ; mihl,
True. 534 ; tibi, True. 540 : ego, True. 759 ; void, True. 899.
And they favour another scansion of Cist. 1 1 1 (opus erit, Proceleusmatic) ;
Cist. 116 (Quando ita, Dactyl) ; Cure. 208 (ego te hoc, Anapaest) ; Cure. 305
(cupis quam, Tribrach) ; Merc. 588 (Elide ego) ; Merc. 1013 (habeo, Ana-
paest) ;Rud. 1046 (med) ; Trin. 321 (Qui ipsu', Trochee) ; True. 564 (ea
aqua, Tribrach).
But are they right ? If their reason is merely the absurd belief that the
ictus of a line, the mere beat of the finger (or toe, * pollicis ictus ') to mark
the time, could alter the pronunciation of a word, we refuse them a hearing.
But if they appeal to Plautus' dislike of the clash of ictus and accent in a
Pyrrhic-word, this argument cannot be ignored. Yet it is quite conceivable
that the Tribrach agit ed& in an Iambic line offended Plautus' ear more than
the Tribrach digit homines in a Trochaic line. And even if Plautus disliked
the clash in Trochaic verse, two questions arise. Was the clash at all
lessened by the dwelling of the voice on the second (the unaccented) syl-
lable (e.g. Capt. 444 Tu hoc age ; tu mihi eru' nunc es ||) ? Was his dislike
greater than his dislike of the iambic pronunciation of modo (Adverb), ibi,
etc. (20) ? To both these questions we are inclined to give the answer : No.
But we feel a difficulty in settling with any certainty so nice a point of
Plautus' technique. Perhaps the following statistics may help to a decision.
In the Trochaics of the Persa, Poenulus and Stichus we find (excluding
alternative scansions) four strong instances of a Tribrach beginning with the
last syllable of a Dactyl-word, one with the last syllable of a Proceleusmatic-
word, five with the last syllable of a Trochee-word, four with the last
syllable of a Tribrach-word. Our list above offers from these plays five such
instances of a Tribrach beginning with the last syllable of a Pyrrhic- word
(one is Poen. 901). That does not look as if Plautus put this treatment of
a Pyrrhic-word in a separate category in Trochaic verse, as more objection-
able than any other incidence of ictus on a final short syllable.
 PLAUTUS AND MENANDER 101

56. (B) Dactyl. A great contrast between Plautus' and Aristophanes'^

Trochaics is the absence of the Dactyl from the Greek. White (Versel
of Greek Comedy, p. 10 1) finds only six occurrences of this foot in all
Aristophanes' (non-melic) Tetrameters (once in the first foot, three times
in the fourth, twice in the fifth). But whether it was banned also in the,
Greek New Comedy is open to doubt. There are only a hundred
Trochaic Tetrameters preserved in the Menander MSS., and in this
hundred the Dactyl appears once. Until the recovery of a sufficient
number of lines to approve or refute us, we cling to the belief that the
New Comedians, in adapting Trochaics to the better reproduction of
excited talk, interspersed them more freely with Dactyls. If iambic
verse had been made to echo every-day conversation by the substitution
of an Anapaest for an Iambus, *fi Ztv /3ao-i\ev (instead of the severely
correct type, *0 Zcv irarep), surely trochaic verse must in time have
come to substitute a Dactyl for a Trochee (e.g. Knights 319 N77 Ata
/cdjuc). That Plautus admitted the Dactyl more freely than Menander
lis quite likely. For although a mere hundred lines are a too unsafe
foundation for a guess at Menander's practice, it would be difficult to
find a page of Plautus' Trochaics without a fair sprinkling of Dactyls.
But that Plautus had to frame rules for the division of the trochaic
Dactyl between words, and had no clear guidance from Menander, is
quite unlikely. (The Anapaest seems to have been much commoner in
Menander's Trochaics than in those of Aristophanes.)
The trochaic caesura of a Dactyl (- w, w) is in trochaic verse (as in
iambic ; 53) mostly in the first foot (of line or hemistich), e. g. Asin.
II

272; 288; 509:

Illic homo aedes compilavit ||;
Illic homo socium ad malam rem || ;
|| matris zwperium minuere.
It seems to be commoner than in iambic verse, although its precise
percentage is difficult to determine, since alternative scansions are often
possible. Thus in Capt. 246 :
Perque conservitium commune quod || hostica evenit manu,
the first foot may be, not a Dactyl (Perque con-), but a Spondee (Perq'
con- j 36). The fourth foot (-mune quod) is a strange type of this
trochaic caesura of a Dactyl, not so much because away from its
commonest place, the first foot, but because a trochee word-ending
(not a trochee-word) appears. (Of course the shortening, after the
unaccented Relative, of the accented first syllable of hostica, the
emphatic word of the relative clause, is quite out of the question ; 22.
IQ2 PLAUTUS AND MENANDER

The Diaeresis comes after, not before, quod.) Another example of this
irregularity is Merc. 385 :
Eo ego ut quae mandata amicus «||micis tradam. Immo mane
(cf. Mil. 660),
where we may perhaps suppose amicus amicis to be uttered as a single
word-group (cf. Ill 17).
For as a rule the two short syllables of the Dactyl, when they are
divided between words, are closely connected in utterance (as in iambic
verse; 53), e. g. Amph. 292; 270; 597; 763; 825:
Sed quis hie est homo qui ante aedes ||;
Sed quid ?71uc est ? caelum aspectat || ;
Neque, ita me di ament, credebam || ;
Itaque nunc s(um) ut e& te patera || donem, sed quis frtuc tibi ? ;
Nescifl quid frtuc negoti || (Or a Tribrach ?).
Ennius' tragic fragments seem to supply examples of the same
irregularity, trag. 40 (Hecuba to Cassandra); 194:
Ubi illa(ec) paulo ante sapiens || virgin//' awdestia ? ;
Ego proiector quod tu peccas : || tu de/inquis, £go arguor.
Some examples of the trochaic caesura in the first foot (of line or hemi-
stich) may be added (from the first eleven plays) :
Amph. yy^ \\ credis id? qui in hac cistellula ;
Aul. 185 /(am) illic homo aurum scit me habere ||;
265 Illic hinc abiit. di immortales ! || (cf. 460) ;
636 Ecquid agis ? Quid agam ? Auferre |j;
Capt. 359 || dice, a^monstra, praecipe (but cf. Mil. 256);
901 Illic Mnc abiit, mihi rem summam || (cf. Epid. 81) ;
?Curc. 170 Ipsu* s(e) £rcruciat qui homo quodBrevis
amat Brevians
|| (an impossible
; 32 A) ;
Merc. 600 Tristis zncedh, pectus ardet ||.
Examples outside the first foot are :
Amph. 309 Quis homo ? Quisquis ^mo hue profecto ||;
816 Tute edictas facta tua, ex me ||;
? Aul. 45 1 Ite sane nunc [tarn] zntro omnes ||;
Capt. 437 Neque des operam pro me ut huius ^?^||reducem (Or facias filium ;
hui(u)s)
ecqui) ;
459 eadem percontabor ecquis hiinc |j adulescentem noverit (Or

? Cas. 351 Quid si abeamus [ac] decumbamus ? || (cf. Men. 152) ;

Epid. 30 || armane? Atque ^w/dem cito (Or ac. Cf. Trin. 935) ;
Mil. 311 Hercle quidquid est, mussitabo (Or quidquidst ?) ;
? 757 Fit pol illud ad illud £remplum (Or TUuc ad Tlluc) ;
997a £ra mea, quoius propter «morem ||;
 PLAUTUS AND MENANDER 103

Mil. 607 Sed speculabor nequis aiit hunc aut || (Or nequi) ;
? Most. 988 Heus vos ! ecquis Msce aperit ? Quid Tstas Cf. ? ||Trin.
(Or ecqui.
870) ;
612);
Pers. 42 Qui ipsu' siti aret. Sicin hdc te mi ? ||
833 Agite sultis hiinc ludificemus ||;
Poen. 922 [Ero] uni potius intus exo odio || (For the deletion cf. True.

Pseud. 704 Quaero quoi ter trina tripMcia. || (in Pseudolus' 'paratra-
740 || indidem. ecquid ^#bet ? Rogas ? (Or habet ?) ;goedia ') ;
1 1 82 In' malam crucem ? Ire /zcebit || (Or Ir' ? Cf. ilicet) ;
? 1 196 Quern ego hominem nul//[V] co\ox\s ||;
Trin. 386 Tute ad eum adeas, tute ^wzcilies || (Or tut' ?) ;
? 701 Effugias ex urbe \xiani' ,pro\\i\ig\zns patriam deseras (Emended
to profugus) ;
True. 593 Sed quisnam illic ^<?most qui ipsus se || (Or homSst).
Since so many are doubtful, we prefer to scan otherwise ; Amph. 267
(hui(u)s habere) ; 661 (ill' revertitur) ; 751 (vera volo) ; Capt. 350 (ess' scit);
887 (quoius erat) ; 940 (ei(u)s ut) ; Cas. 239 (cana culex) ; Mil. 192 (facta
domi) ; 628 (tamne tibi) ; 474 (huius. earn) ; Most. 605 (faenus mi). And
we accept Aulus Gellius' (7, 5) statement that piitus was the pronunciation,
not ' putus '. Ptirus putus is a phrase of Ballio's in the trochaic Septenarii :
Pseud. 989 Puru' putus est ipsus, novi, heus ! || Polymachaeroplagidi ;
1200 Puru' putus hie sycophantast. || Ego tibi argentum dedi.
We add these examples with Elision :—
(a) in first foot (from the first eleven plays) :
Aul. 728 || oppido £go interii. palamst res ;
Cas. 306 || specula in sortitust mihi ;
Cist. 62 Indidem unde oritur facito ut facias || (30, end) ;
526 || tertio omxies efflixero ;
(b) outside the first foot (from all the plays) :
? Cas. 362 Tace, Chaline. Comprime &tunc. Im||mo [istunc] cit
qui dare
didi-;

Trin. 1 163 Spondeo. Et ego spondeo z'dem hoc. O ||.

(We would avoid this unusual type by scanning : Mil. 761 neminem ebrum ;
Poen. 279 Milphio, ubi es ?)
(c) in polysyllabic word -ending :
Cure. 564 (if apud te has its usual scansion ; for te is not emphatic as
me is) :
Sed quid agit meum mexdmom(um) a\\pud te ? Nil apiid me quidem.

57. (C) Proceleusmatic. Some actually assert that Plautus ab-

stained altogether from this foot in his trochaic verse. And the
abstention would certainly be natural enough. The Proceleusmatic,
which accelerated the movement of the Senarius (52), would not be
io4 PLAUTUS AND MENANDER

needed for the ( running ' metre ; and a poet who excluded from iambic
lines the anapaestic types of Anapaest (48), would be quite capable of
the discrimination. Undoubtedly Plautus drew the sharpest possible
distinction between the two spheres of the Proceleusmatic. In iambic
verse it is as common as it is uncommon in trochaic.
It seems hard to get rid of these instances (among others) :
Amph. 442 (in-) || spexi, nimV similest mei ;
7 1 8 || tibi paritumm filium ;
Aul. 168 Clamores imperia, eburata. \\ vehicla, pallas, purpuram ;
595 Quasi pueri qui nare discunt ||;
655 Mane mane, quis illic est? quis hie intus ||;
Bacch. 78 Scio quid ago. Et pol ego scio quid || (32, end) ;
702 Nunc quid vis nos/acere? Emm nil est || ;
Capt. 288 Nam illi quidem 7'todoromedes || fmt germano nomine ;
493 Qui consilium iniere quo nos ||;
Cas. 262 || dare libetl Quia enim filio;
Cure. 311 Viden ut £*palluit ? datin isti ||;
Epid. 668 Tace sis, modo sine me hominem apisci ||;
Men. 461 Quoi tam credo datum voluisse ||;
592 Aut plus aut minu' quam opus erat dicto || ;
977-8 Id ego male malum metuo ; proptere||a bonum esse
certumst potius
Quam malum; nam magi' multo patior \\faciliuJ verba;
verbera ego odi ;
1069 || ea domus et patria est mihi ;
Merc. 465 Uxor tibi placida, et placatast ||;
1 01 6 Prki quam afeamus, qua se lege ||;
Mil. 451 Domicilium est, Athenis domus est ||;
707 Mea bona mea morti cognatis || ;
1437 Magi* metuxnX., minus has res studeant ||;
Most. 384 Cedo soleas mi, ut arma capiam ||;
1 1 16 || quia placeo, exemplum expetis;
Pers. 560 Ubi ea abexxmt, centumplex mums ||;
570 Proin tu tibi iubests concludi ||;
574 I sis malum crucisitum. I sane ||;
Poen. 282 || quibu1 tamen abstineo manus ;
1 2 13 || qmquidem inimicus non siet ;
Pseud. 314 Apiid novercam querere. Eho an umquam ||;
Rud. 1040 || tetulerit ibo ad arbitrum ;
Trin. 638 Nullum beneficium esse duco ||;
913 Vide ?nodo ut hominem noveris ! Tam||quam istuc
me. solet
fieri;
 PLAUTUS AND MENANDER 105

Trin. 923 Em istic exit., qui istum di perdant || (III 52) ;

1052 || beneficio invenias tuo ;
True. 914 || ubi mea arnica est gentium?
At the beginning of trochaic lines of the Miles Gloriosus we find a
Tribrach :
235 Eru' mews elephanti corio ||;
775 Eru1 mens ita magnus moechus || ;
at the beginning of an iambic, a Proceleusmatic in the same phrase:
in Qu(a)m en? mens amabat, etc.
58. Undivided Feet. When a foot is not divided between words,
when, e. g., an iambic word is also an iambic foot, or when a dactylic
word plays the part of a Dactyl in iambic or trochaic verse, what rules
are imposed by Plautus and Menander?
An Iambus-word at the end of the Senarius must not be preceded by
an Iambus unless a monosyllable is the penultimate word. How
strongly Plautus recognized this necessity is seen in his treatment of
such phrases as filio meo, matri meae, a. very common type of phrase,
both of them (one would have thought) equally available for a line-
ending. Plautus is quite ready to end a line with matri meae or matrem
meam or matri suae, etc. But when a son, not a mother, is spoken of, he
actually prefers to distort the natural order of the words, e. g. :
Aul. pr. 10 (Sen.) Numquam indicare id filio voluit suofilio (notsuo)
vo. ;

The Palimpsest corrects the impossible order of P at Cas. pr. 60 (sensit

filium suum P\ filium sensit suum A). Contrast, e.g. Asin. 76 (Sen.)
JEt id ego percupio obsequi gnato meo ; Trin. 570-1 Quid tibi libet tute
agito cum nato meo. Nunc tuam sororem filio posco meo. Even
Phaedrus follows this usage. It must have been congenial to Latin
speech. In Greek the restriction is quite unknown. Within five lines
of Menander's Epitrepontes (261-5) we find two examples: cyw totc
(261), 7rar/305 rtVos (265). However, a mere mention of this striking
difference between the Latin and Greek usage must suffice here, where
our concern is the rule for the undivided foot itself, not for its neigh-
bours. A full discussion will be found in chap. IV § 8. The same rule
applies to the close of a Trochaic Septenarius ; IV 16. (Therefore
e.g. subvenit, not 'subvenit', Pseud. 1146; surripui, not 'surpui',
Amph. 523; et malum et malum Amph. 723; tetigerunt, not ' teti-
gerunt', Aul. 198; divitias, not * ditias ', Capt. 299, etc.).
A Tribrach-word (or word-ending) must not be a foot of an Iambic v/
line. This rule too is a Roman innovation, quite unknown to Menander,

JL*
 106 PLAUTUS AND MENANDER

and hard to explain except by Roman aversion to the clash of ictus and
accent. For although Menander (44) does not greatly favour in his
Trimeters a Tribrach with pyrrhic caesura (e. g. Kara \6\yov), he shows
no hesitation to use a Tribrach-word (or word-ending) for any foot
(except the last) of the line. The only thing in Latin like this free use
of a Tribrach-word is to be found in trochaic verse, the verse where the
use brings not clash, but harmony of accent and ictus (e.g. Mil. 1003
Turn autem ilia ipsa est nimium lepida ||nimVque nitida femina).
•**'•,'• A Dactyl-word appears as a foot once in every hundred trimeters of
our Menander MSS., usually as the first foot, seldom as the fifth (e. g.
Epit. 164 irptor' a7rapi0fir)(rai kolO' Iv. «?x€ts I *°""i8a | rtva ;). In the first
eleven plays of Plautus there are hardly a score of iambic examples
(sometimes showing the Brevis Brevians Law, e. g. Capt. pr. 8 Alterum
quadrimum puerum servus surpuit, Epid. 179 Herculi, quam ilia mihi
obiectast) ; in all six plays of Terence not half a dozen (9 A) ; so that
y Latin iambic verse is not nearly so tolerant of this type. In the penul-
timate foot there is one instance (in a Canticum) -out of the number
(Cas. 229 || quid tu agis? abi atque abstirie manum) ; the rest are all in
the first foot. Whether a Dactyl-word can be the fourth foot of
a trochaic line (e.g. Ter. Hec. 453) is a much debated question.
/ Certainly the Plautine instance usually quoted (pertinet) seems to offer
" a Trochee rather than a Dactyl :
True. 810 Magi' pol haec malitia perti||net ad viros quam ad
mulieres (with emphasis on viros ; 28).
Trochaic verse seems no more tolerant than Iambic of the dactyl-foot, as
this selection from the first eleven plays will show :
Amph. 391 Dicito [si] quid vis ; non nocebo ||;
? Asin. 199 Cetera quae volumus (nos) uti || (Emended to ceterum) ;
Aul. 781 Noscere. Filiam ex te tu habes ||;
Bacch. 411 ||perdidit assentatio :
981 Optimif sum orator, ad lacrimas ||;
Capt. 280 Turn igitur ei quom in Aleis tanta ||;
321 Ne patri, tarn etsi unictf sum de||cere videatur magis;
? 330 Filiu' meus illic apud vos || (Or illi. Or apud) ;
408 ||gratiis emittat manu ;
558 Hegio\ fit quod tibi ego dixi ||;
840 \gaudiis. Noli irascier ||;
Cas. 242 (Anap. ?) Ecquid te pudet ? Omnia quae tu ||;
Epid. 232 Cumatile aut plumatile, carinum a.ut \\cerinfim. gerrae
maximae :
 PLAUTUS AND MENANDER 107

Epid. 706 Quomodo me ludos fecisti ? || (Or Quo mod6) ;

Men. 916 lam hercle occeptat insanire \\primuliim. Quin [tu] me
interrogas ? ;
Merc. 879 || aspicifC} ad sinisteram.
This list might be extended and might be curtailed. We should print
Haecin Amph. 362, Asin. 226 (not Haecine ; cf. 51) ; Hoccin Cure.
200, 695, True. 537 ; Sicin Merc. 158; Usquin Merc. 387. Terence's
six plays offer only : Heaut 199 Illene (?) ; Hec. 281 Nemini; 380 and
Ad. 971 Omnibu'.
A dactylic word-ending seems to be forbidden. We can hardly accept :
Cure. 189 Etiam dispertimini ? Nulli est ||homini perpetuum bonum,
fwhere editors make some such emendation as a transposition of est
((dispertimini nul||li hominist) ; nor Aul. 186 ain tu te valere ? Pol ego
'haud||a pecunia perbene (Emended to perb. a pec).
A_ Proceleustnatic-word we have already shown to be impossible as
a foot of Iambic verse, where it would require an ictus not allowed in
Latin (8). In Trochaic, only Men. 978 (|| faciliu') ; Rud. 1040
(|| tetulerit) ; hardly Stich. 524.
59. DIAERESIS. This is the last of the great differences between
Plautus and Menander which have to be treated here.
How are we to account for Plautus' free and unhesitating use of
Hiatus and Syllaba Anceps in the middle of long lines, licences
unknown to Ms Greek models ? We find them equally at a change of
speaker, e.g.:
Merc. 183 (Troch.) Qui potuit vide*?? Oculis. ||QuoHiantibus
pacto? ;

Cas. 738 (Bacch.) Servus sum tuus. Op||timest. Obsecro te.

Was then the new Latin departure, the use of these licences at the
Diaeresis, an extension of this practice ? Did the Romans think of the
Diaeresis as a break of the same kind as the break at a change of
speaker ?
Or was it that the Romans visualized the long Greek line as two short
lines ? For of course Syllaba Anceps is a feature of line-ending and
Elision between line and line is unknown (except in Synaphea).
The evidence which can help us to a verdict may be summarized as
follows (for full details see Appendix B) :
(1) Out of over 1300 Iambic Septenarii in Plautus about 165 show
Hiatus or Syllaba Anceps at the Diaeresis (occasionally both, e. g. Mil.
1226 :

Namque edepol vix fuit copia || adeundi atque impetrandi),

 io8 PLAUTUS AND MENANDER

Vihat is to say, one in every eight lines. The fourth foot is invariably an
llambus, as indeed it is normally in all his Iambic Septenarii.
(2) Less than half that number have not Diaeresis, in the traditional
text. But most of these retain the Iambus as fourth foot and stand
only a step removed from the normal type. They are of some such
pattern as this (Asin. 545, 469) :
Perfidiae laudes gratias ||q(ue) habemu' merito magnas ;
Nemo accipit : te aufer domum, abs||cede hinc, molestu' nete sis (au.
MSS.).
It is a small minority which effaces all trace of Diaeresis so thoroughly
as, e. g., Rud. 318 :
Tortis superciliis, contrac||ta fronte, fraudulentum,
and most (if not all) have to be emended. Altogether the sure examples
of Iambic Septenarii without Diaeresis are about sixty in number, and
in many of the sixty the lack of Diaeresis is hardly perceptible. If ac
(before vowel), estn', ips', etc., be allowed at Diaeresis, about a score
will have to be subtracted, reducing the total to 40 exceptional lines
(beside 1260 normal). And we may take it as a rule that the fourth foot
of any Iambic Septenarius is an Iambus (or Tribrach, Stich. 681) with
Plautus.
(3) Where there are two speakers the change of speaker usually (or
very often) coincides with the Diaeresis. There are some 140 examples,
e. g. Asin. 385 :
Nemo etiam tetigit. sanun es? ||At censebam attigisse.
(4) And in the great majority of the lines there is some pause, more
or less pronounced, at the Diaeresis, e.g. Asin. 384:
Quis nostras sic frangit fores ? ||ohe, inquam, siquid audis.
Some 60 are marked exceptions, of which most show a type like this
(Asin. 413) :
Cur non venisti, ut iusseram in ||tostrinam ? Hie me moratust,
a few like this (Asin. 695) :
Fac proserpentem bestiam || me, duplicem ut habeam linguam.
These 60 make a stepping-stone to the 60 already mentioned, which
lack Diaeresis.
In strong contrast to the Iambic Septenarii stand the Iambic Octonarii.
That is to say, usually. For one can almost discriminate two classes of
Octonarii in Plautus, one the class (favoured by Terence, etc.) which
disregards Diaeresis altogether and is at no pains to have an Iambus in
the fourth foot, the other (a smaller number) which follows the pattern
of the Septenarii.
 PLAUTUS AND MENANDER 109
We may take our examples of the first class from a Scene which is
almost wholly composed of Octonarii that defy Diaeresis, Asin. 830 sqq. :
Numquidnamst tibi molestumst, gna || te mi, si haec nunc mecurri
accubat ?
Pietas, pater, oculis dolo||rem prohibet. quamquam ego Istanc amo,
Possum equidem inducere animum ne ae||gre patiar quia tecum
accubat, and so on to the end of the Scene.
Some of the Octonarii at the beginning of the Persa may serve for the
second class (Pers. 9 sqq.) :
Quae ero placere censeat ||praesenti atque absenti suo.
Ego neque libenter servio || neque sati' sum ero ex sententia,
Sed quasi lippo oculo me eru' meus ||manum abstinere haudtamen quit
Quin mi imperet, quin me suis ||negotiis praefulciat.
These might be written as Dimeters, but were in fact written as Tetra-
meters inthe ancient editions (AP).
Plautus' Iambic Octonarii (some 400 in all) show Hiatus or Syllaba
Anceps at the Diaeresis about once in every ten lines, e. g. Poen. 818
(with both) :
Studeo hunc lenonem perdere, ||ut meum erum miserum macerat.
In such cases the fourth foot must be an Iambus, a restriction which
suggests that these licences — Hiatus and Syllaba Anceps — in this metre
were allowed on the ground that the long line was regarded as two short
lines, and not merely on the ground that there was a pause in the sense.
But if one asks what difference there is between Iambic Septenarii and
Octonarii which would account for Diaeresis being a characteristic of
Septenarii rather than Octonarii, one sees that the difference is that the
last foot of the Octonarius must be an Iambus, whereas the seventh of
the Septenarius, although it actually is in most lines an Iambus, may be
anything (Spondee, Dactyl, Anapaest), except on the few occasions
when the Septenarius ends in a (single) monosyllable. A monosyllabic
ending always requires a pure final foot, i. e. an Iambus in an iambic
line (Appendix D). So the Iambic character of the Octonarius is pro-
claimed at the end of the whole line, whereas in the Septenarius it is
proclaimed elsewhere, viz. in the middle of the whole line, i. e. in the
fourth foot. The Senarius announces itself to be iambic by invariably
ending in an Iambus j and a Diaeresis (on Aul. 511 see 42), or even
a rule that this or that interior foot must be an Iambus, is felt to be
unnecessary. This may have been the consideration which weighed
with Plautus. His Octonarius was to be one long line (not a mere
 no PLAUTUS AND MENANDER

combination of two short lines), whose Iambic nature was stamped un-
mistakeably on the line-ending and did not need to be stamped
elsewhere.
The Trochaic Septenarius, like the Iambic Octonarius, has its ending
restricted by rule, but not its middle. The whole line must end in
a Trochee (or Tribrach) 'plus' a syllable. What is its attitude to
Diaeresis ? Syllaba Anceps is of course out of the question. In, let us
say, the Amphitruo, Hiatus appears in about every 20 lines (much as in
all Plautus' Iambic Octonarii Hiatus or Syllaba Anceps appear, one or
other, in every 10 lines). Absence of Diaeresis is quite a common
thing, occurring about once in every ten or a dozen lines. Trochaic
Septenarii therefore go with Iambic Octonarii, not with Iambic
Septenarii.
The decision between the rival explanations of Plautus' licence at
Diaeresis has still to be made. Now that the reader has heard the
evidence we leave the verdict to him. For ourselves it is the facts, not
the why or the wherefore, that really matter. (On Trochaic Octonarii
o>^s ' see IV 18.) (On the traces, very doubtful indeed, of an older Diaeresis
*y> after the fourth foot of the Senarius see III 48.) (On the baseless
\ theory of other Diaereses in Trochaics see III 48.) (On Diaeresis in
Bacchiacs, Cretics, Anapaestics see chap. IV.)

60. Monosyllabic ending of Senarii. The only remaining contrast of

Plautus and Menander is slight enough for small print. Of the (complete)
Trimeters of the Menander MSS. more than two hundred (out of some 1300)
end in a monosyllable ; whereas out of an equivalent number of Senarii
(Pers., Poen., Stich.) there are only some thirty. Perhaps this does not
prove quite so great a difference of technique but is partly a mere accident,
due to the different character of the two languages. The Greek has its
monosyllabic Particles like ye, ydp, av, 8q, oZv, etc., which are in Latin di-
syllables, quidem, enim, etc. It would certainly be hard to find in Plautus'
Senarii a parallel to consecutive monosyllabic endings like Menander's
Epit. 580-582 Xopa>v a7ro(nra(r6el<rap — alaBduei ye ; Nat.
Nvi4 §' dvayvapia-pbs avrols yeyove /cat
"Kiravr dyadd. Tt <fir)(riv, iepoavXe ypav ;
Sam. 309-312 (Kat heCkorare j) yikoiov. r]7rei\r}(re p.01
Srt'fif"', Iva pd$f] trap, dia(f)ep(i b* ovfte ypv
yA.dUa)s iraOelv tovt' fj dtKatW. eari 8e
Udv rala-xpov ovk dcrrelov. Ovtos. Xalpe <rvt
even though one pressed into service lines ending in -st, -s as well as est, es,
and lines ending in -que (Greek re). (For a list of monosyllabic endings of
other lines and hemistichs of Plautus than Senarii see Appendix D). In
Terence's Senarii there are more examples of monosyllabic endings, though
 PLAUTUS AND MENANDER in

not nearly so many as in Menander's Trimeters. The three plays, Andria,

Phormio, Adelphi, muster some eighty (out of some 1650 Senarii).
Just as the Latin enclisis of est, es, que diminished the number of mono-
syllabic line-endings, so did an ancient editorial convention. Elsewhere
(IV 3 D) we mention a practice of that standard edition of Plautus whose
line-division is reproduced in the ' Ambrosian ' and ' Palatine ' texts, the
practice of writing at the beginning of a line a small word (construed with
that line) which properly stood at the end of the preceding line. Traces of
this practice in the (presumable) standard edition of Terence are not lacking :
Hec. 283 (Or hui ! 'extra metrum ' ; IV 3 C) ; Ad. 35 ; 55 ; Andr. 306 ; Eun.
349 ; Ad. 465. And so we cannot avoid a suspicion of the exact accuracy of
any comparative lists of monosyllabic line-endings in the two Latin Comedians.
Further an editor of Plautus need not hesitate to transfer a monosyllable (or
any small word) to the end of one line from the beginning of the next. Some-
times he is forced to do this by the metre (e.g. Cas. 827 ; IV 3 D) ; some-
times by other considerations. We have already seen (31) that eo, if left at
the beginning of Capt. 70, involves an unlikely operation of a Brevis Brevians.
Similarly the unfavoured shortening (27) of a long vowel (/ before ns) would
be avoided by reading in Capt. pr. 48-49 :
Itaque hi commenti de sua sententia ut
In servitute hie ad suum maneat patrem (Ut in serv. MSS.) ;
just as Menander often ends one line with a short monosyllable which is con-
strued with the next, e.g. Her. 40-41 :
yE\evdepios Koi Kocrpia. Ti olip ; crv tL
Updrreis vnep aravrov J etc.
or as Plautus often makes the end of a line the starting-point for the next,
e.g.:
Rud. pr. 24 id eo fit quia
Nil ei acceptumst, etc. ;
Cure. 5 1 Tarn a me pudica est quasi soror mea sit, nisi
Si est osculando quidpiam impudicior.
Other Senarii which might perhaps be improved by this touch are, e. g. :
Merc. 827-828 Si quis clam uxorem duxerit scortum suam, ut
Illae exiguntur quae in se culpam commerent ;
Pseud. 552-553 Libidost ludos tubs spectare, Pseudole, et
Si hunc videbo non dare argentum tibi ( . . . ego dabo) ;
Ter. Eun. 70-71 Ultro supplicium : O indignum facinus ! nunc ego et
1 11am scelestam esse et me miserum sentio.
61. Line and Sentence. Still, after allowing for all these obstacles to
precise numerical comparison, it is unmistakeable that Terence makes some
advance towards Menander's practice. And it is significant. For this
practice of Menander is not unconnected with his device for reflecting in the
disposition of his Trimeters the uninterrupted flow of talk. If each line were
to make a separate sentence, the movement of the dialogue would be too
ii2 PLAUTUS AND MENANDER

stilted ; and so Menander every now and then effaces the division between
line and line. He does not scruple to use Synaphea on occasion.
It is to Terence, the ' dimidiatus Menander ', rather than to Plautus, that
one turns for Latin parallels, e. g. (in the three plays cited) :
Adel. 375 Rationem. Est hercle inepta, ne dicam dolo, atq(ue)
Absurda. pisces ceteros purga, Dromo (cf. Adel. 465) ;
Andr. 560 Uxorem demus. spero consuetudine et
Coniugio liberali devinctum, etc. (cf. Phorm. 57) ;
804 Quid vos ? quo pacto hie ? satine recte ? Nosne ? sic
Ut quimus, aiunt, quando ut volumus non licet ;
Phorm. 130 Quae mater, qui cognata tibi sit, omnia haec
Confingam, etc. (cf. Phorm. 423) ;
Adel. 38. Aut uspiam ceciderit aut praefregerit
Aliquid. vah ! quemquamne hominem in animo instituere aut
Parare, etc.
To treat fully the question of how far the Roman Comedians make the line
commensurate with the sentence, and how far they allow overflow from line
to line, is a task that does not concern us here. We have confined ourselves
to one detail, the use of monosyllabic conjunctions (or the like) to minimize
the pause at the end of a Senarius.
62. Elision in last foot of Senarius; Synaphea in Dialogue Metres.
In the list of these monosyllabic endings of Senarii selected from three plays
of Terence it will be noticed how often there is Elision immediately before
the monosyllable. This is not unknown in Plautus, e. g. (to omit est preceded
by Elision) :
Amph. pr. 91-92 Etiam histriones anno quom in proscaenio hie
Iovem invocarunt, etc. (apparently a reference to the
Rudens. Cf. Rud. pr. 1-30) ;
True. 28-30 quot illic iracundiae
Sunt, quot super clamanda, di vestram fidem ! hui !
Quid peierandum est autem, etc.,
though more characteristic of Terence. (This Interjection hui, a whistle of
astonishment, our ' whew ! ', seems to be mostly used by the Comedians like
this, at the end of a line, the preceding word being elided.).
Terence's elision too of the final vowel of atque at the end of a line of
dialogue (producing Synaphea) is not favoured by Plautus. Atque is treated
differently in the Iambic Septenarius, Asin. 585 :
Manedum. Quid est : Philaenium est ||ne haec quae intus exit atque
Una Argyrippus ? (exit | Atque arg. u. MSS.).
(On Synaphea in systems, i. e. ' runs ', of Canticum-metres, see chap. IV.)
 Ill

EARLY LATIN PROSODY; HIATUS

i. In the previous chapter a few peculiarities of Prosody were men-

tioned, ebdem, cui(u)s, etc., conversationalisms rather than antiquarian-
isms. They represent the pronunciation in everyday (educated) talk,
not merely of the Plautine but of the Augustan period. Propertius scans
eosdem (4, 7, 7) :
Eosdem habuit secum quibus est elata capillis,
Eosdem oculos ;
Virgil, ebdem (Eel. 8, 81 ; Aen. 12, 847) :
Uno eodemque igni ;
Uno eodemque tulit ;
the author of Catalept. 9 (line 35), cui(u)s :
Non cuius ob raptum pulsi liquere penates.
So too with that peculiarity of Latin utterance already mentioned, the
shortening of unaccented syllables when they were preceded by a short
syllable. The Adverbs modo, ubi, etc., the Imperatives cave, puta, etc.,
were pronounced as Pyrrhics in conversation not merely by Plautus'
contemporaries but also by Cicero's and Virgil's. And even if it were
true (but it is not) that Lucan preferred to exclude ubi, ibi, etc., from his
Epic, it would not be (as Klotz imagined, p. 50) because these words
had resumed their original iambic form in conversation, but because he
had determined that his Calliope should shun each and every tone of the
1 profanum vulgus ' ; much as one writer of a historical novel may be more
punctilious than others in excluding from his pages such contemporary
parlance as ' don't go ' or even * do not go ' and admit only the phrase
redolent of antiquity ' go not '. Although the Dramatists' ad illam vallem,
per ilium saltum, etc., were not recognized by the Augustan and later
poets, we cannot believe that these pronunciations had fallen into disuse.
Italian ' alia valle ', etc., tell a different story, that they persisted even to
the Romance period. Probus' remark on calamus Nom., pennae Gen.
(Gram. Lat. iv, 215) suggests that any Iambic Imperative of the First
Conjugation might be pronounced as a Pyrrhic in the Empire (Amer.
Journ. Phil. 37, 33).
ii4 EARLY LATIN PROSODY
In this chapter we are chiefly concerned with antiquarianisms, the
early pronunciations which had become obsolete in Virgil's time. But
since it is to the chapter on Prosody that a reader will naturally turn for
all information on the subject, it will be convenient to recapitulate here
the results reached in chapter II.
2. The scientific study of the Latin language began, we may say, with
Ritschl's collection of the earliest inscriptions, the Priscae Latinitatis
Monumenta Epigraphica (now vol. I part i of the Corpus Inscriptionum
Latinarum ; not yet available in a second edition, although much
material has been discovered meanwhile). In his time Plautine and
Latin linguistic study went, as they should always go, hand-in-hand.
But when the two fields of research became larger and larger, Plautine
scholars spared themselves trouble by ignoring Linguistics. Even the
study of the old Italian dialects was scouted and its votaries (e. g.
Buecheler) sneered at, 'qui Osce et Volsce fabulantur'. After Brug-
mann and his school had put the scientific treatment of early Latin
within the reach of every one, this early Latin author was still treated in
books and monographs and articles of what we may call the ' Rip van
Winkle ' type, which wore the garb and spoke the language of a bygone
stage of knowledge. The Weidmann editor actually allowed a second
issue of his Forschungen to appear without correcting the absurd state-
ment of the first, that the Nom. Plur. of the Second Declension had lost
a final s (Leo, Plaut. Forsch.2, p. 291 ; cf. p. 323), although every
undergraduate ought to be aware that Latin agri has the same formation
as Greek aypoi
Most of the new theories of Plautine Prosody which Leo has stated,
unhappily with so much success, reflect the linguistic attainments of
Ritschl's time. Ritschl had found in the Senatus Consultum de
Bacchanalibus the language of Plautus' generation. Subsequent study
has found it to be the language of an earlier generation, the archaic
language of law (like our * this indenture witnesseth ' for ' witnesses ') ;
but Leo has (practically) revived Ritschl's theory that Plautus spoke
agrod, anginad, auspiciod, hodied, etc. We can only suppose that he
relied on the analogy of med, ted, etc., and did not know (what every
student of Linguistics knows) that monosyllables are retentive of a final
consonant long after it has been dropped by other words. Leo's treat-
ment of -s, -m seems an amateur's reasoning of this kind : ■ Since in
early inscriptions -s and -m are dropped, the two must have taken the
same course J so that (1) factum volo (Leo's scansion at Aul. 146) is as
justifiable as factu(s) volo, (2) orat(us) advenio as orat(um) advenio '.
AH experts, so far as we are aware, are agreed that the usage of early
 HIATUS

inscriptions is consistent with what we know to have been the facts x

viz. that -s was dropped (after a short vowel) before an initial consonant ;
-m before an initial vowel. Instead of accepting apud mensam as the
same operation of the Brevis Brevians Law as penes sese (emphatic
Pronoun), Leo seems to have argued thus : ' Since final d was dropped
in agrod, med, haud% etc., therefore it must have been dropped in apud,
id, quid, quod, etc.' (Hence his weird scansions td quod Amph. 793,
etc. ; quod tibi Stich. 21 ; nam quid mi Aul. 723.) A mere glance at
any book on Linguistics would have shown him that phonetic laws work
under strict conditions. That a final consonant disappears after a long 1
vowel (agrod, med) is no reason why it should disappear after a short!
vowel (apud, quod). The pronunciation factu' for /actus is no proof of
a pronunciation facto' for /adds or facti' for factis or factu' for /actus
(Gen. Sing., etc.). When caussa became causa, missus, iussus, /issus,
did not at the same time become ' misus ', c iusus ', I fisus '.
3. Final Vowels. In the pre-scientific days there was a vague
notion that any final short vowel (or syllable) might have been originally
long. 'You never can tell.' And so a Saturnian line was recklessly
scanned (by quantity) :
Runcus atque Purpureus || filii Terras,

and the length of the final ^n£jhf Trffinit"1'vp was 'prnv^H ' from lines of
Pjautusjike :
Rud. 244 (Cretic) Tu facis me quidem ut || vivere nunc velim
(Emended to vivere ut).
These days of liberty or rather licence are over. We have now the
reign of law. We know now that -us (Greek -os) of Nom. Sing. 2 Decl.,
-e of que (Greek re) and -e oi gignere (the same as in Abl. genere, Greek
-l of the Dative) were always short syllables. Final a of Nom. Sing.
1 Decl. (Greek -a, -77) was indeed originally long, but it had become
short before the age of Plautus. Of these things, which every one knows
by this time, a brief treatment may be substituted for all the pages they
occupy in C. F. Miiller's Plautinische Prosodie (published in 1869). At
the outset we may remind the reader, once for all, of the circumstances
under which Syllaba Anceps (i. e. the pause upon short syllables which
makes two ' morae ' out of one) and its companion, Hiatus, are per-
mitted in Plautine Verse :
(1) at the end of a line ;
(2) at the Diaeresis. (48) ;
(3) at a change of speaker (49) ;
(4) at any other marked pause in the line (50). •
u6 EARLY LATIN PROSODY

4. -A. If the scansion in Ennius :

Ann. 147 Et densis aquila pinnis obnixa volabat,

is not an archaism (like Virgil's olli, aulai), but the same imitation of
Homer's lengthening ' in arsi ' as (apparently) his
Ann. 87 Sic expectabat populus atque ore timebat (Or at aSyll.
pauseAnc.
?),

then there is no trace in Early Latin Verse of the original -a of the

Nom. Sing, of Latin A-stems. (On Greek Proper Names in -a?, etc.,
see 30.)
The origin of ita is unknown. That this Conjunction had a long
vowel is unlikely (cf. itidem) and not proved by lines like :
Amph. 635 (Bacchiac) Ita dis est placitum ||(EmendedItanto dis,
Ita etc.);
divis,

Capt. 372 (Sen.) Cum servitutem ita fers ut ferri decet (Emended
to ita fers ut earn f.).
(On the metre of Amph. 572 see IV 24.) Quid ita ?, as the first foot of
an Iambic line (II 52) was perhaps rather an Anapaest (with Syllaba
Anceps ' in pausa ') than a Tribrach, since the accentuation in talk was
quid ita (like our ' why so ? '), and the phrase seems not far removed
from a single word (but see Bursian's Jahresbericht 130, 164).
When Adverbs like contra, frustra appear as contra* frustra in the
older writers (even in Catullus 66, 66 iuxta Lycaoniam), that is no
change of quantity. They are rather different formations (perhaps A hi
Sing. Fern, and Ace. PI. NeutA just as e.g. amice and amiciter are
different ways of forming Adverbs. Examples of the short scansion are :
Pseud. 156 (Iamb.) Adsistite omnes contra me et || quae loquor
/ advertite animum ;
Ennius Ann. 463 Quis pater aut cognatu' volet vos contra tueri ?
(cf. Naev. praetext. 6) ;
Lucilius 864 Contra venis (but 1335 Contra defensorem hominum);
and a common line-ending in Plautus is (Capt. 854 ; Men. 692 j Merc.
528; Pers. 140; Rud. 969; 1255; True. 754): ne frustra sis. (Also
presumably frustra in Amph. 974 ; cf. IV 8.) The long scansion was
also known in early Latin, for the Senatus Consultum de Bacchanalibus
has extrad, suprad. But the Dramatists' metre offers only slight evidence
of contra :

Amph. 217 (Iamb.) Producit omnem exercitum, ||contra Teloboae

ex oppido (Emended to Tel. co.) ;
 HIATUS 117

So we may scan :
Rud. 242 (Cretic) Accede ad me atque adi || contra. Fit sedulo;
True. 124 (Anap.) Fer contra manum et ||pariter gradere.
Supra appears e.g. Pers. 819 (Troch.) :
||te ille qui supra nos habitat (cf. Cure. 477).
And since a Molossus is not uncommon in the first foot of a Cretic
Tetrameter (IV 25), we may scan :
Amph. 224 Extra turbam ordinum || colloquuntur simul.

5. -E. Linguists disdain to speak of the ' Third Declension '. They
make out of it two declensions, of Consonant Stems (with -um Gen. PL,
e. g. consulum) and of I-stems (with -ium Gen. PL, e. g. finium). The
Abl. Sing, of the first ended, they tell us, in \ (the Greek Dat. Sing,
ending), originally the Locative Suffix, and this i became -e ; while the
I-stem Abl. Sing, ended originally in -Id (like the -6d of O-stems, the
-ud of U-stems) which became -I. Since a Consonant Stem often took
an I-stem case-ending (e. g. ferentium) and an I-stem a Consonant-stem
ending (e. g. agrestum Virg., fine), we must, unless we dispute the
linguists' conclusions, print e.g. carni at Capt. 914, pectori at Bacch.
628, and so on (not ' carne ', ' pectore ') ; though we may allow ourselves
to remind our linguistic guides that they have not yet accounted for
some inscriptions (which will appear in the second edition of C.I.L. I i)
with cosoled 'consule', etc.
The Adverbial suffix -e (originally -ed) had become -e (by the Brevis
Brevians Law) in bene, male before Plautus' time. There is no reason
to believe that he ever scans 'bene'. In Cure. 517 (Iamb.) || cures,
bene sit isti (MSS.) the traditional reading should be emended to e. g.
bene (ei) sit isti (Adv.). In Asin. 137 bene shows Syllaba Anceps at
the Diaeresis. Nor yet ' male ' (see 42). (On the suppression of short
final e, of nempe, ille, etc., see II 36.)
6. -I. The -1 of /ibi, sibi, ibi, ubi, etc., was originally a diphthong
and appears in early inscriptions as -ei, e. g. in the elegiacs (c. 130 b. c.)
on Scipio :
ut sibei me esse creatum (Laetentur),
where the word shows the pyrrhic form which (thanks to the phonetic
law of the Brevis Brevians) it usually had in talk. How persistently ibi
was so pronounced we see from the shortening of even ibidem ; although
ibidem is not unknown in Plautus (42) and ublque is the only form of
the Adverb (while ubique means 'and where', Merc. 839, 840, etc.).
Ubiquomque appears in Pseud. 580 (Anap.).
ir8 EARLY LATIN PROSODY

7. -O. Every one knows that -6 was shortened after Virgil's time in
poetry j that is to say, that it had come to be so persistently pronounced
as 6 that even poets had to recognize the innovation. Philologists tell
us that the shortening probably began with Iambic words (scio, volo,
etc.) and Dactylic (nescio, dixer6, mentio, etc.), the Brevis Brevians Law
lending a hand, then became universal.
A passage of Lucilius (552) cited by Nonius (7, 1) and referring to a
law of the Twelve Tables was changed by those ' saecli incommoda ',
reckless emendators, into this form :
' Si non it, capito ', inquit, ' eum, et si calvitur, endo
Ferto manum,'
though common sense should have told them that ferto was impossible
in Lucilius' time. We welcome Marx's happy defence of the traditional
reading :
1 Si non it, capito ', inquit, ' eum, et si calvitur.' Ergo
Fur dominum ?
i. e. would this law entitle a servant to lay hands on his master ?
Indo-European short -O seems to have become -e in Latin; e.g.
sequeso became ' sequese ', sequere (Ind. or Imper.). The -6 of cedo
1 give here ' may be a shortening (after the * brevis brevians ') of an old
Imperative ' do \ Ennius' (and Lucilius') indo (older endo) came to be
spelled indu (e. g. indugredi) :
Epigram. 23 Si fas endo plagas caelestum ascendere cuiquam est ;
Ann. 70 Indotuetur ibi lupu' femina, conspicit omnes ;
238 Consilio indu foro lato sanctoque senatu.

8. Final Consonants. In Plautus' time the short consonant-

endings could almost be counted on the ringers of one hand : -us and
-is (to the same extent as in classical Latin), -it (only in 3 Sg. Pres. Ind.
of Third Conjugation, 3 Sg. Fut. and Fut. Perf. Ind.), -ut (as in classical
Latin) and so on. But not -at, -et, -it of any 3 Sg. of a Verb whose
2 Sg. shows in class. Lat. -as, -es, -Is {manddt like mandds, currdt like
currds, terret like terres, curret like curres, amaret like amares, audit like
audis, nollt like noils). (Cf. mandatur, curratur, etc.) Similarly -it in
3 Sg. Perf. Ind., Perf. Subj., and so on. The phonetic change by
which a long vowel became shortened before final / (much as we cannot
give to ' note ' the long sound of ' node ') swept over the language in
the next generation. The shortening of a long vowel before -r was also
subsequent to Plautus, who scans -or in Verbs (Passive, -0 Active) and
Nouns (with Gen. -Oris) and Comparatives (with Gen. -oris). It is true
that after a short syllable, acting as ■ brevis brevians ', these long finals
 HIATUS 119

were often shortened in talk, so that Plautus allows himself amor, amat,
etc. ; but only when there is a short syllable preceding (never ' clamor ',
1 fervor ', never ' mandat ', etc. ; 18). And of course there were short
monosyllables like the Conjunctions at, sed, and the Prepositions ad,
in, per, available for him. Still, when we glance over any few pages of
his plays we see that -us (of bonus, etc., quercus, etc., maius, etc., )
finibus, etc., amamus, etc.) is his chief means of a short (consonantal) )
ending before an initial vowel.
And so the weakness is revealed of the argument for Leo's wild
theory of the elision of -us (like -um) by Plautus, 'orat(us) advenio'
(like oratum advenio). He claims with truth that the majority of the
short final syllables (ending in a consonant) which break this or that
metrical law are us-syllables and that the elision of -ias would greatly
help to keep the laws unbroken. On the other hand, since -us is the
commonest of the short final syllables (ending in a consonant), it is
naturally in a majority of those which observe these laws as well as of
those which violate them. What he appeals to is a mere accident, due
to the great number of occurrences of final -tis.
9. -C. The Particle -ce (cf. Greek 1-kCi, K€ivo<s, Lat. ce-do 'give here ')
was attached more capriciously to Demonstratives in Plautus' time than
Virgil's. To Plautus illlc (as well as illi) is Dat. Sing., illi (as well as <J
illlc) is the Adverb, illic (as well as ille) Nom. Sing. Illuc, Neut. Sing.,
should really be illucc, since it represents illudce (cf. accurro for adcurro,
etc.) : the final syllable is long not by nature, but by position, and was
becoming short in Terence's time (see 20). So with istlc, istic, istuc(c).
Similarly hoc (Neut. Nom. Ace.) was really hocc, for 'hod-ce' (cf.
hoccine for ' hod-ce-ne ') and teachers in the Empire made their pupils
pronounce in Virgil's line : hoccerat, alma parens.
Whether the Adverbs hue, illuc, istuc are merely the Neut. Ace.
(with a final long by position) or add the Particle to the rival forms illo,
isto (cf. eo) ' thither ' (and so have a final long by nature) is matter of
dispute. Adhuc, the grammarians tell us, was always accented on the
final, and this is corroborated by the absence of a shortening ' adhiic '
(under the Brevis Brevians Law). In all the occurrences in Plautusy
Terence, etc., the word is an Iambus, never a Pyrrhic.
Hie Nom. Sing. Masc, earlier hec (C. I. L. I 32), a formation like
illic (ille + ce), istic (iste + ce), came somehow or other to be pronounced
as a long monosyllable (like hoc), e. g. in Virgil's line, Aen. n, 16 :
manibusque meis Mezentius hie est,
which the grammarians of the Empire direct to be pronounced ' hicc
est ' (a spelling actually found in an inscription a century later, C. I. L.
 120 EARLY LATIN PROSODY

IX 60). But we can hardly refer this mispronunciation to Plautus'

time and efface his distinction of hie ' he ', hie ■ here ', illic « he ', illlc
' there
in : ' or ' to him ', etc. The traditional text must rather be altered
Am ph. 300 (Troch.) || hie auscultet quae loquar auscultet
(Emendedhie)to;

Capt. 547 (Troch.) Hegio, hie homo rabiosus || (With Hiatus at

pause after Hegio ?).
(In Capt. 98 hie is Adverb, 'in this house'. In Pseud. 1018 quam hie
est. In Ter. Ph. 1028 the MSS. disagree. Ennius Ann. 124 is too
doubtful. So is Lucil. 117. In Lucil. 422 probably the Adverb.)
Apparently ' illic quidem ', ' istic quidem ' are not allowed by Plautus,
but only illequidem, istequidem ; which throws light on the scansion
hicquidem and suggests that it is no shortening of a long syllable, as in
siquidem, etc. (II 37), but merely hi-quidem (without the Particle ; like
illequidem). We shall print it here hi(c)quidem.
10. -D. That the -d (after a long vowel) of Old Latin Ablatives
(terrad, agrod, partid, etc.), Adverbs (exstrad, porod, facilumed, etc.),
Imperatives (facitod, etc.) had been dropped from pronunciation before
Plautus' time has already been mentioned (2). It still kept a footing in
the monosyllables med, ted (Plautus does not recognize sed), Ace, Abl.
forms whose origin is matter of dispute (31). That the Abl. 'red*
should be added is extremely doubtful (30 M). And hand not merely
held its ground beside hau (used before an initial consonant, e. g. hau
secus), but in time drove the new comer off.
The rashness of Leo's inference that every -d (even after a short
vowel) must have been faintly pronounced has been remarked (2). The
true explanation of apiid mensam, etc., is not any weak pronunciation
of the final -d but the Brevis Brevians Law (II 22). Even if Leo's
account were right, it would not justify his statement that id and quid
dropped their -d like apud. For, as every linguist knows, mono-
syllables retain a final far longer than other words. Besides, if i' for id,
why not a' for ad? Why does not ad templum become a Bacchius
on Leo's theory? His scansions id quod (Amph. 793; Rud. 1335;
Epid. 507), quid quod (Trin. 413), quod tibi (Stich. 21), etc., are im-
possible, unless a ' brevis brevians ' precedes id or quod. (On ' id qui-
dem ',' quidquid ' see 33 A j II 37.)
Not merely Leo, but many others cling to Ritschl's theory of -d (e. g.
1 anginad ' Trin. 540) in a modified form. They allow that -d had been
dropped from all Ablatives, etc., except monosyllables (which often
retain a final for hundreds of years after longer words have dropped it) ;
 HIATUS 121

but they believe that the tradition survived in Dramatic Verse of allow-
ing a vowel-initial to follow an Ablative like angina. Some even would
admit Ritschl's 'alternas1 Norn. PI. (Trin. 539), though -as Norn. PI. 1
Decl. is now known to be a dialectal suffix and not Latin (30 F). We
find it hard to accept the theory of an * anginad ', ' auspiciod ', ' hodied ',
etc. (in however modified a form it is presented) or an 'alternas ' Norn.
PI. in Comedy. For a Comedian must always use the language of his
time and country. An archaism, a provincialism would raise a laugh
at the wrong place and damn the play. If all or most or even a strik-
ingly large number of (otherwise inexplicable) cases of Hiatus could be
removed by these theories, we might reconsider our attitude to them.
But it is only when the instances of ' Hiatus after an Abl. Sing.' are
presented in a separate list, that their number seems imposing. Com-
pare them with other instances of Hiatus, and they are seen in true
perspective. There is no trace in our MSS. of an ' agrod ', ' terrad ',
etc., as there is of med, ted.
An official inscription earlier than the S. C. de Bacchanalibus of
186 b.c. spells its Ablatives without -d. When a final consonant is
dropped in pronunciation some time elapses before this is recognized in
spelling. That is a fact known to all students of language. We are
!driven to regard 'agrod ', etc., as pre-Plautine.
11. -L. The classical scansion is animal, capital, etc. ; the original
(and Plautine) -al (for -ale Neut. Adj.). At what time the shortening
of a long vowel before -1 asserted itself is hard to say : there is not
enough evidence. An example of -al in Plautus is Aul. 41 ia (Troch.) :
Aperit baccha||nal, adest.
Nil, nihilum (-li, -lo), the Plautine scansions, offer a difficult problem.
T\e^-»
One would expect 'nihilum' (wrongly foisted on Bacch. 518), since the
word is a compound of ne (cf. ne-fas, ne-scio) and hllum. Perhaps the
Brevis Brevians Law, at a time when the accent fell on the Negative
Prefix, changed nihilum into nihilum, and the dropping of the word's
tail left nihil, which everyday talk reduced to nil. There is no clear
Itrace of the disyllable nihil in Plautus, but Terence certainly twice uses
itjp end a line :
Haut. 896 (Troch.) || ne is quidem quicquam ? Nihil ;
Ph. 940 (Sen.) Etiam dotatis soleo. Quid id nostra ? Nihil.
(Also e. g. Andr. 449.) In no line of Plautus is it necessary ; on True.
696 see below, 49.
Proof of nihilum is :
Cure. 155 (Dactylic first hemistich) Respicio nihili mearri vos ||.
i2Z- EARLY LATIN PROSODY

Even Capt. 103 (Sen.), True. 333 (Sen.) add their testimony :
Nam ni Ilium recipit, nihilumst quo me recipiam (nihil est quo
MSS., which less suits the assonance) ;
Quid iam revocabas, improbe nihilique homo (see II 46).
Hardly (see Appendix D) Pseud. 937 (Anap. Septenarius) :
Nam si exoptem quantum dignu's || tantum dent, minu' nihilo sit.
12. -M. The chief thing that calls for discussion, Hiatus with -m,
belongs to another paragraph (43). But since Leo in the second edition
of his Forschungen (p. 330) has not made a whole-hearted recantation
of his heresy (a scansion like * factum volo ' Aul. 146), a word or two here
becomes necessary.
Leo, we have seen (2), seems to have argued from the dropping of -m
and -s on early inscriptions that these two finals were faintly pronounced
under precisely the same circumstances (or rather under all circum-
stances whatsoever). His first position apparently was that any -s 1
might be elided before an initial vowel (like any -m) and that -m might'
fail to make ' position ' before an initial consonant (like -s after a short
vowel). Before the first edition of his book he had come to see that -s
after a long vowel was differently treated from -s after a short vowel.
Before his second edition he came to doubt his scansions ' factum volo '
Aul. 146, 'cum machaera' Pseud. 593, etc.; though he clung to his
wrong explanation of eriim (vero\ etc., a shortening really due to the
Brevis Brevians Law.
But this heresy may spring up again at any moment and make even a
braver show than the heresy of ' ornat(us) advenio '. It may plead, e.g. :
(1) traditional readings like ' factum volo ' ; ' aequo(m) sit ' at end
of line, Mil. 515 ;

(2) Cato's spellings dic[a]e, etc. for dicam, etc. (now explained as
an obsolete 1 Sing. Fut. dice, whence dices, dicet, etc. ; cf.
Journ. Phil. 34, 263);
(3) Lucretius' donique for Plautus' donicum, classical donee (ex-
plained as an archaistic coinage on the type neque : nee),
noenu for Old Latin noenum (explained as noenu' Masc,
like nullus venit ' he did not come ');
(4) the loss of -um in donee, nihil, non, sed, etc. (not yet satisfac-
torily explained) j
(5) quasi and quam si (see II 37).
So we had better enter our affidavit that we find in Plautus absolutely
no justification for it. They may tell us that ' some ancient editor ' may
have gone carefully through the plays and effaced every trace of the
 HIATUS I23
practice from the text. We do not believe there ever was such a
person.
On the question whether a long vowel had already become shortened
before -m in Plautus' time see 49, note.
13. -N. If the theory were right that an had sprung from anne (as
dein from deinde), we should find a long monosyllable (' ann ') in
Plautus. But the word is invariably short (indeed acts as a ' brevis
brevians ', e. g. An ilia Merc. 962); so we had better equate the pair
an : anne to the pair non : nonne or nam : namque (II 36).
The shortening viden became so universal in conversation that, as
mentioned above (II 25), Virgil himself had to recognize it in viden ut.
In Plautus viden ut is universal, and viden is usually a Pyrrhic even
before a consonant initial (not Bacch. 834, 11 30, Capt. 304, Mil. 376?,
Rud. 157, 869), although viden appears (in the Ithyphallic colon of a
Cretic line) once before a vowel :
Rud. 253 Sed quid hoc, obsecro, est? || Quid? Viden, amabo?
It is of courseia result of the ' brevis brevians \ Final n did not (like
final /, r, I) shorten a preceding long vowel. Aiidin never became
1 audin '. The scansion sanun (pergin, etc.), e. g. Asin. 385 (Iamb.) :
Nemo etiam tetigit. sanun es ? ||,
merely means that sanus (pronounced sanu') of sanun (like vides of
viden) was treated as if the phrase were two words, sanus ne or sanu*
ne (like vides ne or vides ne). (Cf. II 18 C.) A close parallel is, e. g.,
moribundu'que est (at the end of a Senarius, Bacch. 193).
14. -R. The balance of evidence inclines decidedly one way. Plautus
did not recognize the shortening of a long vowel before a final -r. The
frequency of such scansions as amor, soror, pater, vocor in his Comedies 1
may indeed be taken to indicate that something more than the mereJ
Brevis Brevians Law was operating, some new law of phonetic changef
that was affecting these finals. But this new law he does not recognize ;
he does not shorten these syllables after a long syllable, but only after
a short syllable, i.e. only after a 'brevis brevians*. The contrary
evidence is mainly
Aul. 140 (Iamb.) Alia alia peior, /rater, est. Idem ego arbitror,
if we suppose the first part of the line to be an Iambic Dimeter and
mark Diaeresis at est. But the line is clearly an Iambic Senarius and
its somewhat unusual construction is quite natural in the Senarius of
a Canticum. So there can be no objection to the fourth foot as a
Spondee, (rater est.
i24 EARLY LATIN PROSODY
Against this illusory instance we place these examples :
soror, Poen. 895 (Troch.); 364 (Troch.) :
ebdem quo soror illius || altera Anterastilis (AP) ;
I, soror, abscede tu a me || (AP) (cf. Stich. 7) ;
pater, Aul. 779 (Troch.) :
Meu' fuit pater Antimachus, || ego vocor Lyconides ;
(cf. Cist. 101, if we read subigit Pres. Tense. On Men. pr. 34 see 34.
In Trin. 645 pater avusque A, paterque avusque P).
uxor, Asin. 927 (Troch.); Merc. 800 (Sen.) :
|| odium, non uxor eram (cf. Stich. 140) ;
Uxor, heus uxor, quamquam tu irata es mihi ;
imperator, Amph. 229 (Cret.) :
Imperator utrim||que, hinc et illinc, Iovi ;
amator, Bacch. n 63 (Anap.) :
Tun, homo putide, amator istac ||;
auctor, Pseud. 231 (Anap.):
Quid mihi's auctor huic ut mittam ? ||;
gubernator, Rud. 10 14 (Troch.) :
|| ego gubernator ero ;
exercitor, Trin. 226 (Bacch.) :
Magister mihi exer||citor animu' nunc est ;
amor, Trin. 260 (Vers. Reiz.) :
Amor amara dat tamen || sati' quod aegre sit.
-or of Comparative, e. g. Amph. 548 (Troch.) :
||longior hac proxima (cf. stultior Bacch. 123; auctior Capt. 782 ;
versutior Epid. 371);
-or of Verb, e. g. Capt. 791 (Troch.) :
Eminor interminorque || (cf. machinor Capt. 530 ; regredior 1023 ;
experior Mil. 633; perpetior Most. 621; rapior Rud. 868;
moror 1248);

-ar of Verb, e.g. Amph. 559 (Bacch.) :

Tamen quin loquar haec || uti facta sunt hie (cf. loquar Amph. 38 ;
opprimar 1056 ; utar Aul. 232).
Some scholars, while admitting the preservation of other long vowels
before -r in early Latin, make an exception of Nouns of Relationship.
They think that the frequent use of the Vocative Sing, may have com-
municated itsshort vowel (Greek irdrep, etc.) to the Nominative. And
yet the early Latin tendency was in precisely the opposite direction, to
conform the Vocative to the Nominative, not the Nominative to the
 HIATUS 125

Vocative. It resulted in leaving to the Vocative only one foothold, the

Second Declension, and even that not wholly secure (e. g. oculus meus
Pers. 765, meus ocellus Poen. 366, etc.). We do not like to allow this
argument even in the case of Jupiter, which some philologists indeed
refer to a Vocative (like Zev Trarep), while others explain it merely as
Jov? pater (like mavolo for magi' volo). In Amph. pr. 94 Jupiter would
make a false or doubtful Tribrach (II 46).
Mutter, of unknown derivation, is suggested by the combination of :
Mil. 1223 (Iamb.) O fortunata mulier es ! || Ut amari videor?
Dignu's (II 59) J
Pseud. 13 10 sq. Pessimus homo.
Mulier haec facit.
At any rate the neighbouring lines of this last metre (curtailed Cretic
Dimeters ; IV 26) seem not to show a Spondee in the first foot (1307 Sed
Simo, ut probe, 1308 Tacty' Ballioest). The -er was therefore originally
short, like the -er of per (Mil. 1391 proves nothing). (On cor, ter see 20.)
15. After Plautus the course of the shortening may be traced as follows :
Ennius follows Plautus' usage, e.g. :
Ann. 41 Postilla, germana soror, errare videbar ;
113 O pater, o genitor, o sanguen dis oriundum (Pater by Brevis
Brevians Law, like Iupiter 97) ;
117 . . Quirine pater, veneror Horamque Quirini ; 422) ;
442 Tollitur in caelum clamor exortus utrimque (cf. 531. But clamos

444 Imbricitor aquiloque suo cum flamine contra ;

557 Interea fugit albu' iubar Hyperioni' cursum ;
(So in line 406 read Totum corpus habet sudor.)
trag. 45 (Troch.) Virgines vereor aequales ||.
For Terence's practice evidence is lacking. The monosyllable at the end
of Hec. 258 (Iamb.) ought to be preceded by an Iambus (Appendix D), but
we should have to change illi of the MSS. to ei (or omit it) :
At ita me di ament, haud tibi hoc || concedo, et si ei pater es.
A Turpilius fragment can hardly be said to prove mat^r (see II 59) :
117 (Iamb. Octonar.) Em meas fortunas, mater; em || qua causa hue
opere maximo (Contenditur).
Lucilius shortens the syllable, like the classical poets (e.g. praetor 91;
quattuor 126; 160; longior 168). An instance of the older scansion seems
to be :
391 Langudr, obrepsitque pigror torporque quietis (Or Languor, three
syllables).
(But 1049 is very doubtful.)
In an Anapaestic fragment of Afranius some editors read clamos N
12 Hoc haeret, at haud clamor oritur, ||sed spiritiim firmiter ihstat.
 126 EARLY LATIN PROSODY

16. -S. The faint pronunciation of -s after a short vowel, in early

times and as late as Cicero's youth, has already been mentioned (I 3).
Can we elicit from the earliest poets, the pioneers in quantitative metre,
any rules for its prosody before an initial consonant ? Under what
circumstances did they determine that e.g. sanus sis should be (1) a
Molossus, (2) a Cretic ? Monosyllables are retentive of final letters :
therefore i' for is is quite unlikely (33). Plautus' usage clearly shows
jthat Pyrrhic words of the type of magis, nimis, satis, prius normally (but
jnot invariably ; see 42) remain Pyrrhics before a word beginning with a
/consonant; e.g. priu' quam, nimi' quam, magi' quam. These are pre-
cisely the words that would naturally yield to the Brevis Brevians Law,
subordinate or enclitic words like the Adverbs male, modo, cito (II 22).
The combined pressure of the two tendencies, (1) to drop their -s before
an initial consonant and (2) to shorten their final syllable, must have
been wellnigh irresistible j and it may be significant that Catullus' one
example is the phrase dabi'-supplicium. Less clear is the influence of
a following s- (seen in the Catullus line), though it stands to reason that
euphony would gain by the loss of one of the sibilant pair ; and we have
no difficulty in finding Iambic line-endings like (Most. 555) confessu'
sit. But on the other hand we find endings often enough like (Bacch.
313) occidisti* me. (And we have lines like Bacch. 956, Troch. : Paria
item tria is tribus sunt ||.)
What prevents the result of such investigation from being final and
convincing is the inadequacy of the material used by the investigators.
They practically confine themselves to line-endings, since there only can
we be quite sure that -s has been dropped by a Dramatist. And yet
when we see Ennius and Lucilius dropping -s of any and every kind,
for the mere convenience of the metre, it is surely more natural to believe
fthatPlautus did the same in the interior of his lines (see below, 17 C).
Although Plautus allowed a Spondee in the ■ even ' feet of the Senarius,
we may be sure that he preferred an Iambus ; and if this Iambus could
be got by the faint pronunciation of -s (the current pronunciation, no
mere poetical licence), why should he (a Comedian) have hesitated to
use it, especially in the dearth of short finals at his time (8) ? For our
own part, we would print e. g. munu' (not munus) in Trin. 1 (Sen.) :
Sequere hac me, gnata, ut munu' fungaris tuum,
without feeling any necessity to prove by statistics that a Spondee was
much rarer than an Iambus in this foot. Even the timorous scholar,
who will tread only on the firm footing of figures, need not hesitate to
pronounce e. g. molestus as molestu1 in Asin. 469 (Iamb.) :
Nemo accipit, te aufer domum, abs||cede hinc, molestu' nete sis. (au.
MSS.).
 HIATUS 127

For of the 1,300 Iambic Septenarii of Plautus only some 20 per cent. /
have a Spondee in the seventh foot, while the Iambus is the rule (IV 10).
Until the ictus- heresy (that ictus regulated quantity) has been finally
exorcized by bell, book, and candle, it will be well to remind the reader
that there is nothing unnatural in a scansion like :
Accius trag. 534 (Anap.) Divisus : eum || dic/«' Prometheus.
A short i never ends a Latin word. The Romans made it e (e. g.
here, for pyrrhic heri). The Neuter of a Greek I-stem Adjective ends
in -i (t8pi5, ISpts, IBpt), of a Latin in -e (similis, similis, simile). The
close connexion of Latin 1 and e is seen also in a word like pedes, older
pedess for ped-it-s ' the foot-goer ', ' foot-soldier ' (20) : -its became -ess.

II
Nonius Marcellus tells us that the enclisis of est produced similest out
of similis est (as bonust out of bonus est ; II 38), a spelling of which
abundant traces remain in our MSS. of Plautus (e. g. True. 170 similest).
And Leo rightly pointed out that this clears up the declension of possum.
Potest would be the slurred pronunciation of potis est (Masc, Fem.) as
well as of pote est (Neut). (See below, 40 K.)
In very early times the faint sound of final s and final m had brought
in a fashion of suppressing these finals in spelling. Very early inscrip-
tions show -e for final -is and -im or -em, -o for final -6s and -6m (class,
-us and -um). The material they provide is scanty, and complicated by
the old use of e for ei ; but we have no reason to believe that this fashion
of spelling accompanied any practice of pronunciation but that found in
the earliest poetry, viz. the suppression of -s after a short vowel before
an initial consonant and of -m before an initial vowel. Nor can we tell
how long precisely this practice of spelling lasted in literature, how long
a word like militaris continued to be spelt in books (normally or
occasionally) militare (see 17 B). To a linguist these are doublet forms,
the one (militaris) being the pre-vocalic form (e. g. militaris erat ille), the
other (militare) the pre-consonantal form (e.g. militare fuit ille). A
purist would cling to the distinction of potis erat ille from pote fuit ille,
until the distinction of Gender in this much-used phrase had been
effaced by the wear and tear of daily utterance (40 K). A Pyrrhic like
the Adjective potis is the Adverb satis. Confusion of the much-used
doublets satis erat and sate fuit must have been almost as inevitable and
rapid. Since potest, satest meant ' he is able', 'sufficient' (as well as 'it
is able ', ' sufficient ') it was but a step to poterat, saterat ' he was able ',
' sufficient '. We have no such clear trace of the spelling sate as we have
of pote. Sat is the form in which the pre-consonantal doublet appears ;
and we doubt whether it would be possible to find clear proof that
Plautus or Terence were conscious of the true, original distinction of.
 128 EARLY LATIN PROSODY

sat(e) from satis (or indeed of pote from pott's), and whether the editor
of Terence (when an edition at last appears) will print e. g, in Andr. 705
(Iamb.) :
Satis habeo. Quid facies ? cedo. || Dies hie mi ut sate sit vereor,
where Umpfenbach offers Sat habeo . . . satis sit. To judge from
mavolo (probably spelt by Plautus mavvolo, from magi'- volo), magis had
suffered as much from wear and tear as potis or satis. Still, so far as
we can gather from an unedited text of his plays, Terence did not use
mage-erat as he used pot(e)-erat, sat(e)-erat, while Plautus did ; though
since Plautus elides the -e of mage only a few times (mostly in mage-
amo) in all his twenty-one plays (42), the mere absence of this elision
from the six plays of Terence in our MSS. cannot be called convincing.
(In fact we seem to find one example of elision in Eun. 356, though the
MSS. all show the spelling magis. See 42 s. v. So Leo's theory is very
insecure.) Yet Terence may have deliberately stopped short of mag(e)-
erat, just as all writers seem to have stopped short of * nim(e)-erat '.
Indeed there is no evidence (see 42 s. v.) of a ' nimest ' (like potest, satest),
nor, so far as that goes, of a ' magest ' either. In studying these niceties
of early Latin speech one must remember the virtue c aliqua nescire ' and
take the facts as he finds them, without trying to give reasons for every-
thing. We all know the ludicrous mistakes of foreigners who apply
strict rules of reason and analogy to the wayward caprice of English
speech. All that we can safely infer is that it will not do to say ' Every
-is in Latin could be pronounced -e even before an initial vowel '. If
we attempt to reduce the facts to a formula, it must be a modest state-
ment, e. g. that ' a few verb-phrases of daily use like potest, satest, pote-
fuit, sat(e)-fuit, mag(e)-volo led by false analogy to such irregularities as
\pot(e)-erat, sat{e)-erat, sat(e)-habeo, mag{e)-amo '. Beyond that we cannot
safely go.1 And if it would be unreasonable to say ' Every -is could
drop -s before an initial vowel ', it would be sheer madness to say ' Also
-us could drop -s before an initial vowel '. Though priu1 quam goes
hand in hand with nim? quam in Plautus, and factust with potest, how

1 The occasional appearance of mag{e)amo in Plautus may be compared with the

unique occurrence of seu istuc (if the reading be right) in Merc. 306 (Sen.) :
Si canum seu istuc rutilumst sive atrumst, amo {AP),
The pre-consonantal doublet seu has audaciously usurped the place of the pre-
vocalic doublet sive. The frequency of sat ertt, sat habeo, etc., in all writers maybe
compared with the frequent use of nee (instead of neque) before a vowel.
If volupis really was an Adjective in Early Latin we may appeal also to the
transition from volupest to e. g. Men. 678, where Erotium ends a line with the words :
ut tibi ex me sit volup.
 HIATUS 129

can we dare to say that fadust led to * facterat ' as potest to poterat ?
What we can dare to say is precisely the opposite, that factust appears
never and under no circumstances whatsoever to have led to ■ facterat '.
17. (A) Leo's theory of Elision of -s What we have described as a mad
saying is— alas !— precisely what Leo has said, that Plautus scanned ' fact(us)
erat ' as three syllables, just as he scanned fact{um) erat as three syllables.
He goes further. He actually says that Plautus elided before a vowel any
and every final syllable containing a short vowel before -s, i. e. that Plautus
pronounced oratus advenio in the same way as oratum advenio. Once his
statement is so formulated, its improbability stands self-confessed. But since
this wild theory of Leo's has been cleverly presented and has gained many
adherents, it must not be dismissed so unceremoniously as by most Compara-
tive Philologists hitherto. Leo's arguments must be answered one by one.
(The page- references are to the second edition of his Plautinische Forschun-
gen, 1912.):
(pp. 248-252) Leo compares the loss of -s to the loss of -d, and seems to
think that the spelling aput proves a weak pronunciation of -d. For the true
account of the loss of -d see above (10). Plautus' and Terence's apiid men-
sam is not the dropping of -d after a short vowel. It is due to the Law of
Breves Breviantes (like penes sese, with emphatic Pronoun). Leo's scansions
id quod (Rud. 1335), etc«> are impossible.
(pp. 253-257) Leo points out that one Grammarian of the Empire scanned
a line of Virgil with elision of -us before a vowel (Aen. 3, 229) :
Rurs(us) in secessu longo sub rupe cavata,
through failing to see that the true reading was rursum ; also that another
may have scanned Plaut. Capt. 997 (Troch.) :
Sed eccum incedit hue ornat(us) haud ||ex suis virtutibus,
through failing to see that the true scansion of the second last word was suis
or else sins.
Granted. Grammarians of the Empire will scan ' through a brick wall '.
A comical instance is Consentius' scansion (Gram. Lat. V. 403,24) of Aen. 2,
457. His copy had (instead of Ad soceros, et avo) :
Ad soceros, atque avo puerum Astyanacta trahebat.
1 Scandimus enim sic, rostque a, ex quo apparet inter duas consonas A
vocalem perisse.'
(pp. 258 sqq.) Leo points out that the MSS. of Plautus present many lines
in such a form that the elision of -us before a vowel would reduce the
scansion to order. The weakness of that argument has already been
mentioned (8). Since -us was the commonest short syllable (ending in a
consonant) in Plautus' time, it follows that most of the examples whether for
or against any metrical law of this type must be words ending in -us. And
what of the many examples of similarly defective scansion where the line does
not happen to contain a word in -us ? His first examples are lines with
incedit) lines for which we have the evidence of the minuscule MSS. only.
2348 K
i3o EARLY LATIN PROSODY
When we get the evidence of the Palimpsest for the phrase, we find that it
seems to testify to cedit (the Old Latin equivalent of class. Lat. incedit) :
Poen. 577 (Troch.) Basilice exornatus cedit || (cedit A ?, incedit P).
And Nonius (251, 9) seems to have found incedunt as gloss on cedunl (Au\.
517) in his copy of Plautus (see Philologus 63, 284). So the minuscule arche-
type had written the suprascript gloss instead of the old Latin word in these
lines to which Leo appeals (probably also in Asin. 403, where the metre
allows either form, since the Simple Verb is required in 405).
Next, the phrase amicus amico produces an awkward division of a foot
between words sometimes, especially in Merc. 385 (Troch.) :
Eo ego ut quae mandata amicus a||micis tradam. Immo mane (cf. II 56).
Leo shows that the natural order of the phrase was amicus amico and that
Terence has departed from this order :
Ph. 562 (Troch.) Solus est homo amico amicus ||,
thereby making the metre smooth ; while Pomponius (time of Sulla) retains it :
145 (Sen.) Amicus amici, gaudet si cui quid boni,
thereby producing (like Plautus) an awkward division of a foot (here an
Anapaest ; II 48) between words. Surely the obvious inference is that
Plautus and Pomponius preferred the unsmooth metre to the unnatural order,
while Terence did not. (Leo, apparently, has been misled by Klotz and does
not know that Pomponius' Anapaest is abnormal.)
Of the remaining examples many have been treated above (II 48 sqq.).
More reasonable methods of emending the residue will be found in any good
edition.
(pp. 273 sqq.) Leo continues to give examples of unsound lines which this
remedy would heal. Yes ; but is not the remedy worse than the disease ?
Are there not better remedies ? For example, Leo would elide -us of fiaeda-
gogus in Bacch. 142 (Sen.):
Praesentibus illis paedagogwj una ut siet (with a false Anapaest in second
foot). May not the root of the malady lie in praesentibus illis (instead of
praesente illis, a Plautine phrase. Cf. II 50) ? He strangely finds evidence
of elision of -us of nullus in Bacch. 90 (Troch.) rather than the following
scansion :
111' quidem hanc abducet, tu nullus ||adfueris, si non libet
or rather than the substitution of nullum (as elsewhere before a vowel ; see
41).
(pp. 279-284) Leo, recognizing the weakness of the evidence of mere
corrupt lines in the traditional text of Plautus, now advances a positive proof
of the elision of -us before a vowel, viz. bonust, etc. These slurred forms
have been already discussed (II 38) and a rival explanation offered. But
even if Leo's explanation be accepted, is it not highly dangerous to infer
phonetic laws from the slurred pronunciation of everyday phrases like ' good,
bye ' (for ' good be with ye ') ? What English phonetic law would Leo infer
from ' gadzooks ' ?
 HIATUS 131

(pp. 285-287) Leo accepts Nonius' explanation of similest as simile est (see
II 38). But surely all that is of value in Nonius' statement is that Nonius
found similest, etc., in his copies of early authors. His own attempt at ex-
plaining these forms is worthless, as worthless as the eleventh-century scribe's
alteration of nundinalest into nundinale est in our unique MS. of Festus
(176, 29 ; a citation of Plaut. Aul. 324).
(pp. 288-291) Leo defies Comparative Philology by insisting that sequeris
is older than sequere (see below, 40 D).
(pp. 292-301) Leo's strongest evidence of Elision is Plautus' mag(e) amo,
a scansion which we have discussed in the preceding paragraph. Here we
may remark that his account of magis, mage in Plautus is incorrect (see
below, 42). Magis is clearly the pre- vocalic form in Plautus, mage the pre-
consonantal. In a note on p. 300 he adds fortasse, -is (see 42), ignoring the
explanation accepted by Comparative Philologists, thztfortasse is Imperative,
fortassis Indicative of an S-tense of an old verb fortare (like ad-Jirmare)
' affirm ', ' you (or one) will affirm '.
(pp. 301 sqq.) Leo, rightly seeing that, if sequeris became sequere, other
parts of the Verb should have followed suit, makes a brave but unsuccessful
effort to find other instances, to call up reserves to save the weak position.
They are all ' peditastelli '— mere tag-rag-and-bobtail soldiery — and one may
let them live. A single victim will suffice. The attigat in the Diomede
MSS. is a mere error of the scribe of the archetype whose eye was caught by
the attigat in the previous line (Gram. Lat. i, 382, 18 ; a Pacuvius citation:
nequis vim attulat neve attigat).
(pp. 304 sqq.) Leo in these pages partly cites examples of the spelling -e
for -is before an initial consonant (These we discuss below) ; partly before an
initial vowel; and, being now confident that he has proved his theory,
ventures to explain by it some anomalies in Plautine vocabulary, e. g. nullum
(beside nullus) for non (see 41).
The only example of Elision that seems worth discussing is (p. 307) pane
in Cure. 367. Nouns in -is appear often to have been originally adjectives.
Hence their various gender. Thus amnis (fluvius) Masc, amnis (aqua) Fern.,
e.g. Merc. 859. In rete Neut., beside retis Fern. (?), we have a similar pair
to pane and panis. Since panis would suit the Metre here without Elision
(II 56), why should Plautus have elided it ? Clearly (if the reading be right)
he chose the (vulgar?) by-form, a Neuter, for effect (cf. Class. Quart. 7, 118).
On Rud. 888 (p. 310) see II 49.
(pp. 313 sqq.) Leo now tries to answer the objection : Why is there no
mention of this Elision by Grammarians ? He has recourse to the same
theory as Ritschl, viz. that Ennius made a revolution in the language. After
Ennius there is a new world which forgot what had happened in the old. (On
Ritschl's explanation of ferunt, etc., see I 6.) But in denying that Ennius or
any subsequent poet elided -s before a vowel, why does not Leo mention a
phrase of Lucretius which seems as strong an example as any he has cited
from Plautus ?
6, 230 Et liquidum puncto facit aes in tempore et aurum.
K 2
i32 EARLY LATIN PROSODY

Surely ftuncto tempore is as likely to represent puncto temporis as pane to

represent pants (Class. Quart. 13, 19). If it is Lucretius' invention, then
Lucretius elided -s before a vowel ; if it is borrowed from Ennius, then Ennius
may (if we argue in Leo's style) have done the same. To dismiss briefly
some of the remaining witnesses, quisque is not the same word as quisquis.
Its -que '-ever' is the Particle in quandoque (38 D). Mil. 685 (uxor suave
ductust) is of course like Virgil's triste lupus stabulis. And from other parts
of this book it will be seen that Leo's explanation of illius (p. 320) is wrong
(see II 35) ; that he wrongly disallows (p. 266) a Dactyl as fifth foot of a
Trochaic Septenarius (see II 56 ; 58) ; a shortening (p. 271) like (Aul. 603)
meus amat (II 27) ; the absence of et (p. 263) in (Pseud. 805) qui optimus
carissimust (see II 51). (On ipse, a. different formation from ipsus, ipsu\ see
39-)
(B) The spelling -e for -is. What gave Leo's theory its lease of life was
partly his explanation of potest (like similest), partly his interesting collection
of spellings in MSS. and inscriptions like the very early specimens tribunos
militare, etc. For previously the only example (to disregard Lachmann's
note on Lucr. 1, 186) of the recognition in literary spelling of the faint pro-
nunciation of -s was Lucretius' noenii (for noenus the Masc. by-form of
noenum, as nullus of nullum ; all four words playing the part of a Negative
Conjunction).
In late Latin -is definitely passed into -e (see Bonnet, Latin de Gregoire
de Tours, p. 341), just as -urn passed into -o (cf. Ital. vino). By means of
these spellings unearthed by Leo we learn the persistence of this pronuncia-
tion of -is from the earliest to the latest periods, even although in Cicero's
lifetime (and after) literature disdained to recognize it. Leo's examples from
inscriptions are the safer guides, e. g. corpore custos (on an inscription of the
early Empire) ; those from MSS. lie under suspicion of reflecting the late
Latin of scribes, just as scribes in the Dark Ages often substitute -o (their
own pronunciation) for -um. And since the same abbreviation-symbol was
sometimes used for -e, sometimes for -is, many of Leo's examples may be due
to this confusion merely. However, we are wandering outside our province
here. A full discussion of these spellings and their value as evidence of pro-
nunciation iswork for a book on Latin Phonetics or Orthography. Herei
is sufficient to warn our readers to use Leo's examples (and examples added
by other scholars) with caution. Of course a spelling like noenic is really the
pre-consonantal form of noenus and gives no proof that noenus elided its -us
before a vowel. No more does the spelling corpore for corporis, etc. A
scientific student of language views them as doublet-forms, like atque
(prevocalic) and ac (preconsonantal), neque and nee, neve and neu, etc. The
strongest Plautine example is usually said to be splendore plenum (p. 308) :
Merc. 880 (Troch.) Caelum ut est splendore plenum (B: spl. est CD).
To us it seems very weak. For although Plautus may have twenty-six times
the Genitive with plenus and only here the Ablative, yet Plautine Syntax is
now seen to have been so elastic (cf. Burs. Jahresber. 167, 41) that the word
\ never ' has generally to be replaced by ' hardly ever ' in all these dicta of
 HIATUS 133

last century : ' Plautus never uses this construction ; therefore this line inust
be spurious ', etc., etc. (Similarly in Merc. 795, Suspicione implevit me, the
first word may be Abl. Leo has no right to claim it for a Genitive.) But
further, there is no certainty that what Plautus wrote here was not :
Caelum uti splendorist plenum,
since in the Mercator the Codex Vetus is carelessly written and since -st was
often muddled by the minuscule scribes (II 38).
(C) Rules for the suppression of -s. Can rules be framed for the help of
editors ? Editors of Ennius' Annals or Lucilius' Satires need no rules. The
metre shows them where they must print bonu', magnu', where bonus,
magnus. But rules of some kind or other are necessary to editors of Plautus,
for the present practice of suppressing -s only at the ends of lines hangs a
millstone about the neck of Plautus' tripping verses. Take these three
Trochaic Septenarii of the Asinaria (163-165) and see how the lines are
marred by substituting a_Spondee solus :
Solu' solitudine ego te ||atque ab egestate abstuli ;
Solu' si ductem, referre ||gratiam numquam potes.
Solu' ductato, si semper ||solu' quae poscam dabis.
It almost seems as if editors perverted the true statement, that Plautus allows
a Spondee in the odd feet of trochaic lines, into the false form that Plautus
demands a Spondee there. The (comparative) closeness of Plautus' Trochaics
to the Greek type is hidden from the reader who naturally scans the final
syllable long by position when he finds -us, -is printed before an initial con-
sonant. Editors w^nlf! g*fr ^nrh nAoi-Ar t-n thp; tmfh if they made it their
pract" *» to VPnt (prftrnnsnnantaH -u\ -i' always unless the metre postulated
a long syllable.
On the other hand, they would very often be wrong in so doing before a
final Cretic word in a Senarius (IV 8 B).
Theycan safely print -u', -i'to secure a Trochee in the odd feet of trochaic,
an Iambus in the even feet of iambic lines and (we may add) aJ3ajcchius_(or
Cretic) at the end of a Bacchiac (or Cretic) line pr hftmistirh (IV 33 ; 25).
Also to secure an Iambus in the seventh foot of an Iambic Septenarius, etc.
We have already stated the chances in favour of editors who do this (16) and
have taken for example Asin. 469 :

(abs-) ||-cede'hinc ; molestu' ne sis ' go away ; don't bother'.

Of this everyday phrase ' don't bother ' there is a score of occurrences in
Plautus. We print them according to the three rules just framed :
Asin. 469 (Iamb.) (just quoted) ;
Aul. 458 (Troch.) Lege agito mecum ; molestus ||ne sis ; i cenam coque ;
Epid. 63 (Iamb.) Potin ut molestus ne sies ?||
Men. 250 (Senar.) Molestu' ne sis. non tuo hoc fiet modo. Em ! ;
627 (Troch.) Potin ut mihi molestu' ne sis ? ||
Merc. 779 (Senar.) Potin ut molestus ne sis ? Agite, apponite ;
Most. 74 (Senar.) Molestu' ne sis nunciam ; i rus ; te amove ;
601 (Senar.) Molestu' ne sis. nemo dat. age quid libet ; '
 134 EARLY LATIN PROSODY

771 (Senar.) Molestu' ne sis. haec sunt sicut praedico ;

877 (Iamb.?) Non eo. moles||tu' ne sis;
886 (Troch.) Mihi molestus|| ne sies ;
955 (Troch.) Egone ? Tu. Tu ne molestus. || (scil. sis. Some
prefer ne [' verily '] molestu's) ;
Pers. 287 (Iamb.) Potin ut molestus ne sies ? || ;
Pseud. (118 (Senar.) Dabo. molestus nunciam ne sis mihi ;)
889 (Senar.) Molestu' ne sis. nimium iam tinnis. tace ;
Rud. 1031 (Troch.) ||mihi molestus ne sies;
1254 (Senar.) Abi intro; ne molestus ; linguae temperaon Most(see
. 955)
note ;

True. 897 (Troch.) Potin ut mihi molestu' ne si-|| (-es) ;

918 (Troch.) Hoc modo, ut molestu' ne sis ||.
Is there not great probability that the phrase was pronounced everywhere
with molestu' (not molestus), perhaps even at the Diaeresis (Aul. 458 ; Most.
886), and possibly even at a pause in the sense (Rud. 1254) and, with
Diaeresis, (Most. 955)? We would print subvenisti' as well as tempu' in
True. 187 (Iamb. Septenar.) :
Per tempu' subvenisti'. Sed ||quid ais, Astaphium ? Quid vis ? (IV 10).
Still, there is no example in the (extant) lines of Ennius' Annals (line 166?)
or Lucilius' Satires (line 359 ?) of -u', -i' at a very strong pause ; so that an
editor would be justified in printing -us, -is at such a pause. And at the
f
Diaeresis, at any rate where there is anything of a pause at the Diaeresis.
(Not in enumerations, e.g. Ann. 63 ; 245 ; Lucil. 358.)
Iambus-words have another way of becoming Pyrrhics, through the
operation of the Brevis Brevians. How can we be sure that opus is not as
correct as opu' in e. g. Capt. 163 (Senar.) ? :
Opu' Turdetanis, opust Ficedulensibus.
And some editors will prefer to present such disyllables at the Diaeresis (and,
more confidently, at a strong pause) as opus rather than opu'. Lists of
debateable -s at any pause (or at the Diaeresis) in Plautus would occupy too
much space ; but here is a list of lines divided between speakers, where the
word stands at the end of one speaker's remark :
Bacch. 220 (Senar.) Nam istoc fortasse aurost opus. Philippeo quidem ;
1 1 76 (Anap.) Abin a me, scelns. Sine, mea Pietas ||;
Cure, no (Iamb.) Sitit haec anus. Ouantillum sitit?||
Merc. 133 (Iamb. Octonar.) Acanthio, quern quaeris. Nus||quamst
disciplina ignavior (Or quaeris ; IV 11) ;
Most. 723 (Cret.) Intus. Quid id est? Scis iam ||quid loquar. sic decet
(Or intus ; IV 25) ;
modo ;
1 173 (Troch.) Tranio, quiesce si sapis.\\ Tu quiesce hanc rem

Pers. ? 514 (Troch.) Aut quid erus tints 1 Tace, stultiloque ||(Or eru') ;
814 (Bacch.) Atque hoc quod tibi sua.\\deo facts? Quid est id?;
 HIATUS

Rud. 1305 (Iamb.) Immo edepol una littera j|plus sum quam ??iedicus.
Turn tu (Mendicus es) (Or medicus ; Appendix D III).
1337 (Iamb.) ebdem die, <tui> vidull ||ubi sis potitus. Fiat (Or
potitus ; IV 10) ;
Stich. 710 (Troch.) Bibe, si bibis. Non mora erit apud me||.
We ourselves prefer opus, scelus, anus, sapis, tuiis, facis, bibis (i.e. shorten-
ings by the Brevis Brevians). But each editor must follow his own guidance,
never forgetting the virtue ' aliqua nescire '.
Another problem of hair-splitting is the correct presentation of these
disyllabic Adverbs in -is before an initial consonant. Is the opening Dactyl
of e.g. Mil. 1337 (Troch.) Si magis vis, earn omittam. Nolo|| to be printed
with magi' or mage ? It seems hard to believe that magis with final s dropped
or faintly pronounced (magi') was a different thing from mage and that we
should suppose a trio of (preconsonantal) forms (1) magis (scanned as an
Iambus or, under the Brevis Brevians Law, a Pyrrhic), (2) magi' (a Pyrrhic),
(3) mage (a Pyrrhic) ; although indeed we might place beside it another trio
of forms : satis, sati' (sate ?), sat (see 42, s. v. satis) ; and certainly poti'
(Masc. Fern.) can reasonably be kept apart (always or often) from pote
I(Neut.). Was mage a false coinage on the type of pote ? The ancient editors
of Plautus wrote mage only when the last syllable was elided before an initial
vowel (see 42, s. v. magis, for details) ; elsewhere magi's, for (pace Lachmann)
the ancient editorial practice was to write final s even when it failed to make
' position', e.g. occidistis me (at tne end ot a Senarius). A modern editor
whose guiding maxim is 'stick to the MSS.' will print Si magi' vis (Mil. 1337)
just as he prints occidisti' me, but Mage amo (an Anapaest). But if an
editor prefers to print Mage vis, Mage pulcher, etc. (like Mage amo) and
appeals to common-sense, we shall not quarrel with him. Until linguists
have cleared up the origin of these Adverbs, magis, satis, nimis, (and ot potts
too ; 40 K) certainty is unattainable. For other words the guidance of the
MSS. should be followed. Thus, while qualest, quale' s should be printed for
the slurred forms of qualis est, qualis es, print rather (Pseud. 275) quali' sis
(qualis sis MSS.), since qualest is no evidence of a (Masc. Fern.) by-form
'quale '(II 38).
18. -T. What has been said of -r applies apparently to -t, but not
quite so convincingly. The balance of evidence inclines to show that
Plautus did not recognize the shortening of a long vowel before a final
-t. The frequency of such scansions as amat, monet, perit, fult, vellt in
his Comedies may indeed be taken to indicate that something more
than the mere Brevis Brevians law was operating, some new law of
phonetic change which was affecting these finals. But this new law he
seems not to recognize. He seems not to shorten these_syllables after
a long syllable^but^pnly after a short syllable, i. e^a 'brevis brevians '.
Our traditional text, full of errors, can, as has been said already (I 6),
lend a modicum of support to any wrong theory whatsoever, and lends
136 EARLY LATIN PROSODY
this not very strong support to the theory of shortening (stronger, it
must be confessed, than for the shortening before -r) :
Asin. 524 (Troch.) Quid dedit? quid deportari ||iussit ad nos? an
tu tibi ? (Scan dedit, or alter the order) ;
Bacch. 665 (Cretic colon) \fecit ex patre (but see IV 26 A) ;
Capt. 198 (Iamb. Octonar.) Nunc servitus si evenit ei ||vos mori-
gerari mos bonust ;
Cist. 312 (Iamb.) Nimi' lepide exconcinnavit hasc' ||aedes Alcesi-
marchus (but see App. B) ;
Merc. 121 (Iamb. Octonar.) Curaest negoti quid sit aut ||quid
nuntiet. Nugas ago (cf. Capt. 921) ;
Pseud. 934 (Cretic) Iuppiter te mihi ||servet. Immo mihi (But in
vv. 931) ;sqq., unlike 926 sqq., the feet are
iv 25 not all pure Cretics ;

Rud. 212 (Cretic miuric colon) ||monstret \ ita nunc IV(but26 see
B).

(On the three Iambic Octonarii see II 59.)

Against this handful, all doubtful, we set these examples :
Asin. 874 (Troch.) Fundumalienum arat, incultum || familiarem
deserit (Not a strong example) ;
Cist. 583 (Sen.) Instare usque odio donee se adiurat anus ;
Merc. 648 (Troch.) || quia enim me affiictat amor ;
Cas. pr. 49 (Sen.) Amat efflictim, et item contra filius ;
Poen. 489 (Sen.) Faciat ut semper sacrificem nee umquam litem
(Strong emphasis on last MS S.);;
word)
Rud. 113 (Sen.) Quern ero praesente praetrahat oratio (praetereat

Most. 788 (Bacch.) Seni non erat o||tium ; id sum opperitus ;

Bacch. 229 (Sen.) Negotium hoc ad me attinet aurarium ;
Merc. 696 (Sen.) Solet hortator remiges hortarier (Emended to
Solet ageator ; II 44) ;
Trin. 206 (Sen.) Quod quisque in animo habet aut habiturust sciunt
(Emended to (aut) habet) ;
Rud. 1333 (Iamb.) Quod tibi libet, id mi impera ||;
Merc. 1022 (Troch.) Quin a met et scortum ducat || ;
Mil. 1244 (Iamb.) ||desideret, expectet; Pers. 327);
Pers. 68 (Sen.) Si legerupam qui damnet, det in publicum (cf.

(Sit is frequent, but may always be read as siet)

Capt. pr. 25 (Sen.) Ut fit in bello, capitur alter filius ;
Cure. 489 (Anap.) Nemo it infitias. At tamen ||(cf. Poen. 683) ;
 HIATUS 137

Pers. 762 (Anap.) ||scit accipere et reddere nescit;

Men. pr. 52 (Sen.) Vellt, audacter imperato et dicito ;
Merc. 924 (Troch.) Ob oculos adduxerlt in || aedes, dum ruriabest
ipsa;

Bacch. 642 (Bacch.) Vicit et domum lau||di' compos revenit ;

Asin. 501 (Iamb.) Adnumeravit et credidit ||;
Capt. pr. 9 Eumque hinc profugiens vendidlt in Alide ;
(-fert);
Men. 759 (Bacch.) Nam res plurimas pes||simas, quom advenitad-

Merc. 428 (Troch.) Mandavlt, ad illam faciem ||;

Mil. 832 (Sen.) Neque illic calidum (?) exhibit in prandium ;
Most. 894 (Vers. Reiz.) Novit erus me. Suam quidem ||culcitulam

Poen. 260 (Bacch.) Lingua huic excidit, ut ego opinor ; oportet

1059 (Sen.) Emit, et is me sibi adoptavit filium ;
Pseud. 311 (Troch.) Ilico vixit amator|| ;
Rud. 199 (Cret.) || perdidit in mari ;
Stich. 384 (Troch.) ||mi obtiglt hereditas.
Therefore monstret must not be foisted on Epid. 537 (after Glyconics
and before Choriambics), and the text altered into Glyconic metre :
\ Monstret eum mi hominem aut ubi hab[it]et.' The hemistich is
rather Choriambic, like Horace's ' Nullam, Vare, sacra vite prius ', but
with resolution (suitable to Philippa's agitation) of the first long syllable
of the foot :
||monstret | eum mi hominem aut | ubi habitet.
We must resist the temptation to shorten in Merc. 343 (Bacch.) :
Is rescivlt et vi||dit et perdidit me.
(On dat see 42 ; on percipit see 40 H.)
19. After Plautus the course of the shortening may be traced as follows :
Ennius (who must have been badly in need of short finals for his Dactylic
Verse ; 8) clearly treats these syllables as half-long, scanning them long in
one line, short in another. This is so unmistakeable that one passage will
suffice for illustration (Ann. 79 sqq.) : devovet (79) ; servat (80) ; esset (83) ;
recessit (89) ; dedit (90) ; volavit (92). This departure from Plautus' usage
by a younger contemporary of Plautus is puzzling, since the Comedian would
be expected to welcome the new pronunciation and the Epic poet to cling to
the old. It makes a strong argument that Plautus may have occasionally
shortened.
Caecilius' nescit may be mentioned (frag. 143).
Terence, curiously, does not often reveal his scansion. We may believe
him to have at least gone no further than Ennius on the strength of:
Andr. 682 (Iamb.) ||crepult a Glycerio ostium ;
 138 EARLY LATIN PROSODY
Ad. pr. 25 (Sen.) Poetae ad scribendurn augeat industriam ;
453 (Sen.) Utinam hie prope adesset alicubi atque audiret hanc ;
Eun. 701 (Sen.) Dicebat eum esse, is mihi dedit hanc. Occidi ;
Ph. pr. 9 (Sen.) Quod si intellegeret cum stetlt olim nova ;
(The ending of an Iambic Octonarius hemistich, Andr. 606 esset hie ||, may
be a Spondee ; IV 11. Adducet et sumptum, Ad. 913, is impossible ; II 48).
There is no doubt about Lucilius. The syllables are invariably short, as
I in classical poetry. So we must regard sit (470) as Syllaba Anceps ■ in
I pausa'), and read Praestringat splendore micanti (1094). Whether 330
I begins with a Perfect or a Future is doubtful. In either case the -It must be
a lengthening ' in arsi \ like Cetera contemnlt et (550).
20. DOUBLE FINAL CONSONANTS. The ' double consonants '
are those which in classical Latin were written and (in all but mono-
syllables) pronounced as single consonants. Possibly the spelling
ess has left some traces in the MSS. of Plautus, e. g. Merc. 489 (see
Class. Quart. 1, 107), but that Propertius wrote ess (cf. Homer's iaai)
in 2, 32, 61 :
Quod si tu Graias tuque es imitata Latinas
is quite unlikely. (The short quantity in classical Latin, invariable in
a disyllable like prodes, usually prevails with the monosyllable too.
Contrast Terence's is Ad. 696). These endings are generally the result
of Assimilation. Thus mil-its ' the goer in a thousand ' became miles(s),
equ-it-s ' the goer on a horse ' became eques(s), ped -it-s ■ the goer on
foot' became pedes(s). Whether Plautus spelled mi/ess, equess,
pedesst sospess, etc. is doubtful. That he so pronounced them is certain,
4 for the final syllable is never short :
Aul. 528 (Sen.) Miles impransus astat, aes censet dari ;
Asin. 330 (Troch.) Turn igitur tu dives es factus || (see 53) ;
Cas. 817 (Ionic) Sospes iter incipe hoc uti viro tub ;
Cas. 629 (Choriambic) || quae suist impos animi ;
Jll It was shortened by Ennius :
Ann. 269 Spernitur orator bonus, horridu' miles amatur ;
though his es of Ann. 578 (Ausus es hoc ex ore tuo; unless Ausu's be
the true reading) may be due to the ■ brevis brevians ' : ausus-es like
auded. A more certain example of es is Lucilius 1238 :
O Publi, o gurges Galloni, es homo miser, inquit (cf. 1 1 1 7).
With this shortening goes apparently the scansion istuc (older istucc,
from istud-ce; 9) in Ter. Andr. 941 (Iamb.) :
Cum tua religione, odium. no||dum in scirpo quaeris. Quid lstiic est ?,
where scribes are not likely to have substituted the unfamiliar istuc for
 HIATUS I39

the familiar istad. (Cf. Eun. 830 ; II 49, end.) And yet istuc, illuc
are fully attested for Terence by lines like Eun. 349 ; 782 ; Heaut. 238 ;
346 ; 348 ; Ph. 58 ; 294 ; Hec. 608.
To this category we must add such words as cor (for cord or corde),
lac (apparently spelt lact or lacte in Republican Latin), ter (cf. terrunci) :
Bacch. 1 127 (Bacchiac) Rerin ter in anno ||.
21. Some additional details may stand in small type :
cor, Mil. 1088 (Anap.) ; Poen. 390s (Troch.) ; Pers. 803 (Anap.) :
||cordate, ut cor ei saliat (cf. 35) ;
Huiu' cor, hum' studium, huius || savium, mastigia (a weak example) ;
Cor uritiir, caput ||ne ardescat.
Lucilius shortens :
488 Vera putant, credunt signis cor inesse in aenis.
lact{e) :
Amph. 601 (Troch.) Neque lact lactis magis est simile ||(lac MSS.) (cf.
Bacch. 6, where lacte is attested) ;
Bacch. 1 1 34 (Bacch.) Quae nee lact' nee lanam ul||lam
astenthabent. sic sine;
(lacte MSS.)
Men. 1089 (Troch.) Neque aqua aquae neque lacte est lactis ||, crede mi,
usquam similius ;
Mil. 240 (Troch.) Tarn similem quam lacte lactist (Or lactest lactis) ;
True. 903 (Troch.) Opu' nutrici, lact ut habeat, ||veteris vini largiter ;
Ennius Ann. 352 Et simul erubuit ceu lacte et purpura mixta ;
Caecilius 220 (Sen.) Praesertim quae non peperit lacte non habet ;
Presumably mell (printed mel), though evidence is lacking (cf. Poen. 388).
22. H, J, V.
H. An eminent scholar, who has done admirable work in other
departments of Latin Literature, tried in vain to prove that h- in
Plautus' time was so strongly pronounced as to resist the elision of a
preceding final vowel (Rhein. Mus. 54, 40 and 201). That theory (and
its weakness), mentioned below (44), need not be discussed here. But
an argument used to support it provokes a lover of Plautus and deserves
the pillory, the extraordinary argument that, because Catullus' Arrius
pronounced ■ hinsidias ', therefore Plautus must have pronounced, e. g.,
amico hinsipientiast ' in Poen. 1090. As a sample of lack of common-
sense in Plautine research (48) this may challenge for first place the
silly theory of a ' metrical ' Law of Breves Breviantes, that an actor
pronounced venire as ' venire ', duorum as ' duorum ', eamus as ' eamus '
(ii 21).
The treatment of h between vowels in Plautus' plays illustrates the
weak sound of this spirant at his time. Nemo is always a disyllabic and
 Mo EARLY LATIN PROSODY

has so lost trace of l ne-hemo ' that Plautus uses the phrase nemo homo.
Nil is apparently always a monosyllable in Plautus (beside nihili, -lo,
-lum', n). Prehendo and prendo are both Plautine, e.g. Asin. 668
|((Iamb.); 563 (Iamb.) ; 569 (Iamb.) :
Prehende auriculis, compara ||labella cum labellis ;
[| in furto ubi sis prehensus ;
Ubi prensus in furto sies || ;
but the trisyllable is much more frequent. By Terence's time the
disyllable has ousted the trisyllable (prehendit Andr. 353). (For fuller
details see 42 s. vv. Prehendo, Apprehendo, Comprehendo, Deprehendo,
Reprehendo.)
Perhaps cohibeo, prohibeo were pronounced as three syllables in
Plautus' time, like debeo (dehibuisti Trin. 426 A) praebeo (sometimes
spelt praehibeo in the MSS. : e. g. Cas. 537 ; Mil. 591 ; Pers. 429 ; 510;
Pseud. 182 ; Rud. 530 ; cf. Merc. 1023), since (unlike adhibeo, exhibeo)
quadrisyllabic scansion is never necessary and an ictus prohibeam or
cohibeam, etc., is never found. Cohonesto appears in the MSS. (at
least in the archetype) as conestat in a Trochaic Septenarius of Accius :
trag. 445 Pro se quisque cum corona || clarum conestat caput.

23. -J-
Compounds of iacio. The evidence shows that -iec- was the
Republican spelling, e. g. coniecit Mil. 112, Ter. Heaut. 277 ; iniecit Ad.
710 ; whereas -ic- (the result of the change of -jec- to -jic-) is later. The
scansion abicit, etc., belongs to Imperial Poetry. The Republican
scansion is abicit (abjecit), e. g.
Aul. 197 (Troch.) Ubi manum inicit benigne || ;
Ter. Ad. 710 (Iamb.) Itaque adeo magnam mi inicit ||.
A curious exception is Asin. 8r4 (Sen.) :
Praeripias scortum amanti atque argentum obicias (Emended to
obsupes) (cf. Naevius com. 94). And when a vowel preceded the / it
appears to have blended with the /-sound (as in Virgil's reice
capellas) sometimes, e. g. :
Rud. 769 (Troch ) lam hercle ego te continuo barba ar||ripiam, in
ignem coiciam (coniciam AP) (cf. Merc. 932.);
Asin. 161 (Troch.) || tractas quomque eicis domo,
if that is the right treatment of these examples. Examples of the usual
pronunciation are :
Asin. 127 (Cret.) Sicin hoc fit ? foras ||aedibus me eici ? ;
425 (Iamb.) Iussin columnis deici ||operas araneorum ?
 HIATUS 141

J for I in ' Synizesis '. In our discussion of Synizesis (e. g. debs)

we included such pronunciations as die (II 33). Conveniently, rather
than correctly. For the two phonetic processes are different. Debs
might be expressed by deos. The e is reduced to vanishing point ; and
indeed sorsum was actually written for seorsum. But die seems rathen
to represent dje (pronounced * dye ' as in our ' d'ye hear ? ' : for Latin /
had of course the sound of our y). The i was not reduced but rather
changed from a vowel to a consonant. While etiam is an Anapaest (or
Tribrach), etjam is a Spondee, (or Trochee). Nerienem, when the i
Isuffered this change, would become not ' Nerienem ' but Nerienem
(with first syllable long by 'position') in Ennius Ann. 104 (II 33).
Some departures by so fastidious poets as Virgil and Horace from the
usual scansion show us that this change of the vowel i to the consonant
j was quite a feature of Latin speech : e. g. Virgil's abjete costas, not to
mention Fluvjorum rex ; Horace's Ut Nasidjeni ; Vindemjator. Some-
how it seems nowadays more uncouth than the parallel change of u to
v (e. g. Virgil's Genva labant ; Tenvis ubi argilla), and one is averse to
allow it even in early poetry unless metrical necessity justifies the
abnormal form. Some scholars prefer to find an Anapaest at the
((beginning of Ennius Ann. 94 :
\ Avjum, praepetibus sese pulchrisque locis dant,
though they are surely straining at the gnat and swallowing the camel
(IV 37). We ourselves would rather doubt Cicero's accuracy in his
quotation of the line, since disyllabic avium is not expressly attested.
It seems better to read : MSS.),
Ann. 436 Insidiantes hie vigilant, partim requiescunt (Hie ins.

Tecti (cum) gladiis, sub scutis ore faventes,

where faventes is attested, but not ' insidjantes '. It is the Verb elido
not the word 'injurjatum ' which Nonius attests in Lucilius 57 :
Iniurfatum hunc in fauces invasse animamque
Elis[is]se illi (Emended to Impuratum hunc).
In Cist. 297 (Sen.) :
Praestigiator es, siquidem hie non es atque ades (cf. Aul. 630),
some editors scan praestigiator-es (like molestae-sunt ; II 29) ; some go
so far as to delete es. Praestigiator's (cf. Epid. in?) can hardly be
called impossible (II 38). But the quadrisyllable is most likely (42).
(Cf. Ennius Ann. 251.)
24. -V-. Between two identical long vowels (or the like) v was apt
to disappear in Roman talk (like our w in ' Hawarden '). -Provorsus
i42 EARLY LATIN PROSODY

became prorsus (prosus, whence prosa oratio ' forward speech ', not like
poetry that is ever turning back to begin a new line). The full form
was used by Plautus in the description of the wriggling ' leno ', Pseud.
955(Troch.):
Ut transvorsu', non provorsu', || cedit, quasi cancer solet !
fVarro has preserved the ' ipsa verba ' of Plautus here, for both A and P
show the ' Revival ' re-casting :
Non prorsus, verum ex transverso || cedit, quasi cancer solet.
And Varro, who, no doubt, had seen the play acted, describes the
stage-tradition : quod cum leno non faceret, sed secundum parietem
transversus iret, etc. (Ling. Lat. 7, 81). Other pairs are e.g. aliovorsum
and aliorsum (42), dextrovorsum and dextrorsum (42).
But ' controrsia ' is not a necessary inference. We may scan Men. 593
(Troch.) with controversia :
Aut plus aut minu' quam opus erat dicto || dixeram controversiam
And ' clator ' for clavator seems somehow unattractive : (n 58).
Rud. 804 (Sen.) Ehem ! optime edepol eccum clavator [adjvenit.
For -ivi- we may cite obliviscor and obliscor, the latter in e. g. :
Mil. 1359 (Troch.) Muliebres mores discendi ob||liscendi stratiotici ;
Accius 190 (Troch.) || memet possim obliscier;
188 (Troch.) An ego Ulixem obliscar umquam? ||.
But that divinus had in Plautus' verse a by-form 'dinus' is hardly
certain (Mil. 675 ; Epid. 314. In True. 307 duarum A P. In Amph.
672 omit quicquani). Ditiae (which perhaps belongs to 27), occasional
in Plautus, is regular in Terence. (The MSS. of both dramatists know
only the spelling divitiae.) Terence recognizes e. g. diviti (at end of
line, Ph. 276), though ditis, -tem, -tes are his usual forms, and on their
analogy he actually offers dis Nom. Sing, in Ad. 770 (Sen.) :
Tun si meus esses — Dis quidem esses, Demea (MSS., Gram.).
Plautus has both divitis, etc., and ditis , etc. ; never dis, only dives s
(printed dives ; 20). It would be rash to find in Terence's innovation a
justification for a one-syllabled ' nav(i)s ' in :
Men. 344 (Sen.) Nunc in istoc portu stat navis praedatoria (Aratis
P) ;;
Bacch. 797 (Sen.) Bene navis agitatur, pulchre haec confertur

Most. 737 (Sen.?) Sed, Simo, ita nunc ventus navem deseruit.
Quid est ? ;
Ennius trag. 74 (Anap.) Rapit ex alto || naves velivolas.
 HIATUS 143

And editors are suspicious of ' vi(ve)ndi ', ' vi(ve)re ', of the traditional
text in :

Ter. Andr. 52 (Sen.) Liberiu' vivendi fuit potestas,Epid.

nam 177;
antea 40);
(cf.

Ad. 445 (Sen.) Restare video, [v]ah ! vivere etiam nunc libet.
Also of such a scansion as prae(ve)rtar (wrongly proposed for Pseud.
602): ca(vi)sse Bacch. 10 17.
I Si vis became (in the rapid utterance of the familiar phrase) j/> (see
I42), the two differing as our ■ if you please ' from the more usual
r please '. Only the full form, as the less usual, need here be supported
I by a quotation :
Afranius 1 79 (Sen.) Mea nutrix, surge, si vis, profer purpuram.
Di, afc^have been explained (rightly or wrongly) as similar reductions of
divi (deivei), divis (deiveis), not of ' dei ', ' deis ', just as peri, perit
may represent perivi, perivlt. But while the reduction of -Ivit to -It is
normal, the supposed ' -at ' for -avlt would be quite abnormal and should
not be foisted on such lines as Asin. 501, Pers. 834, Mil. 1038, Bacch.
1097. If later poets ventured on it, they may have blindly followed the
analogy of -amus for -avimus, etc. This Syncope, e.g. Ennius trag. 138
nomus ambo Ulixem (for novimus), hardly justifies * vi(vi)mus ', the
reading of A P in Stich. 695 (Troch.) :

Tamen bibimus nos, tamen efflcimus || pro opibu' nostra munia

(cf. Poen. 1 187).
In Merc. 846 amicitiam is preferable to ci(vi)tatem. Dido (dis-do) was
not fused with divide?. (On aunculus, the everyday form of avunculus,
see 42.)

V for U in ' Synizesis \ We have seen (23) that debs was a^

different thing (to a phonetician) from die. So was it from dubs,
duellum, pronounced dw- (as in 'dwell'). To the examples of this
pronunciation of u as v (our w) given in II 33, may be added here two
that are far from certain :

Lucilius 1246 Pascuali pecore ac montano, hirto atque soloce

(Pastali MSS.) ;
Afranius 237 (Sen.?) Fluctuatim ire ad ilium, accipite hoc, tege tu
et sustine.

25. LONG VOWEL SHORTENED BEFORE VOWEL. The Latin

phonetic law ' vocalis ante vocalem corripitur ' takes a wayward course
 M4 EARLY LATIN PROSODY

in Plautus. We find Chius (Adj.) (cf. Anth. Pal. 182 %v XtW esse
ovx
LKavov) : ) ;

Cure. 78 (Sen.) Quasi tu lagoenam dicas, ubi vinum Chium (solet

Poen. 699 (Sen.) Ubi tu Leucadio, Lesbio, Thasio, Chio ;

Pelldeus (but long, Asin. 397), Dicaea :
Asin. 333 (Troch.) |[ mercatori Pellaeo ;
Mil. 808 (Troch.) Dicaeam. Nempe eandem quae dudum.
lit is clearly no mere metrical law (or licence) that Plautus follows. He
■allows only the pronunciations which were actually current. The Dat.
Sing, of is is ei or ei, never ' ei ' (see 35) ; apparently (but this is not
quite certain) illlus, istlus have no by-forms 'illnis', 'istius' (II 35).
Never ' f 10 ', ' f 1am ' apparently (42).
/ The old length (obsolete in classical Latin) appears, not merely in
<! Perfects like fui, metui (40 F), beside fui, metui, but also apparently in
L the Genitive siiis :
Cure. 323 (Troch.) Pernam, abdomen, sumen suis ||;
and one is tempted to find propnus (from pro and privus) in :
Capt. 862 (Troch.) Atque agnum adferri proprium || pinguem.
Cur? Ut sacrifices (cf. Merc. 338 ; Trin. 1130),
and plus (cf. Oscan Piihioi ' Pio '), pietas in :
Ennius Ann. no Pectora pfa tenet desiderium, simul inter (diu
corr. to dia MS.)

Sese sic memorant ' O Romule, Romule die ' ;

Trin. 280 (Vers. Reiz. ?) Patrem tuum si percoles || per pietatem
(Cure. 639; Pseud. 293. On Poen. 1137 see II 45).
(On a very doubtful prior see 42.)
The reproduction of preyocalic Greek ct by Latin e seems normal,
e. g. platea, balineae, gynaeceum :
Most. 759 (Sen.) Ad earn rem facere vult novum gynaeceum.
26. PARASITIC VOWEL AND SYNCOPE. It is convenient to
take together peric(u)lum, dext(e)ra, ar(i)dus, etc.
(1) Parasitic Vowel. The Parasitic Vowel (a less alarming name
than ' Svarabhaktic ') of Greek loan-words like drach^ma, m/na is men-
tioned below (28). Here we are concerned with its appearance in
native Latin words in Plautus' time, e. g. discipwlina, quadrup&lari, an
appearance that is not wholly capricious, but has not as yet been
successfully reduced to rule. One thing is clear, that Plautus draws a
distinction between the Diminutive Suffix -co-lo- (always disyllabic,
 M5

HIATUS

e. g. corculum, uxorcula) and the Instrumental Suffix -tlo- (e. g. vehi-

clum, cubi-clum, peri-clum). He confines himself to vehidum, and
allows periculum at the end of a line only (i. e. through metrical cogency),
e.g. Capt. 740 (Sen.):
Periclum vitae meae tuo stat periculo.
h(But always cubiculum apparently.) So that in his time the Parasitic
'Vowel between c and / had not quite asserted its claim to rank as a
^separate syllable. (For details see Class. Rev. 6, 87.) We must scan :
Asin. 666 (Iamb.) Die me igitur tubm passerculum ||(ig. me MSS.) ;
Cas. 837 (Bacch.) Meum corculum, mel||culum, verculum. Heus
tu (melliculum P) ;
MSS.) ;
917 (?) Amabo mea uxorcula cur ||;
Cure. 1 1 (Sen.) Ex dulci oriundum melculo dulci meo (melliculo

Mil. 18 (Sen.) Quasi ventu' folia aut paniculum tectorium j

1006 (Troch.) Turn haec celocla, ilia autem absente celoc- la)e.;
|| (i.

1060 (Anap.) || proculenam impertiturust (Not porcl-) ;Adito (?)

j
Pers. 310 (Iamb.) Ecquid quod mandavi tibi estne in ||te speculai ?

Rud. 1170 (Troch.) Sucula. Quin tu i dierecta || cum sucula et

cum porculis ;
Stich. 91 (Troch.) Osculum. Satest osculi mi || vestri. Quid,
amabo, pater ? ;
Accius 100 (Sen.) Quot luna circ[ul]os annuo in cursu institit.
The derivation of vidulus is not clear. But editors of the Rudens find
it difficult to accept the disyllable which the MSS. seem now and then
to suggest to them (see 42). If we may rely on the manuscript tradition
of the Carthaginian lines in the Poenulus, ligula (i.e. lig-la) was the echo
of Carthaginian lechla- (Poen. 1013-14; but cf. 1023 and 1025).
27 (2) Syncope. Linguists have not yet cleared up the movements of
this phonetic tendency of Latin, a tendency due to stress-accentuation.
(So far as present information goes, the syncope of the final syllable was
a feature of Oscan, etc. (e. g. Campans of Trin. 545, and perhaps True.
942 for Latin Campanus) rather than of Latin. So that Terence's disk
(Ad. 770), Class. Lat. sors, Ops, nostras do not represent any actuals
pronunciation of dives, sortis (Cas. 380), Opis (Cist. 515), nostratis (the!
only Plautine form), any more than deum Gen. Plur. is a syncope of
deorum. Rather the analogy of lis, litis, liti suggested dis as a suitable
Nominative for ditis, diti. It is quite wrong to say : Since dives was
2348 L
146 EARLY LATIN PROSODY
reduced to dis} therefore navis may have been reduced to ' naus ' or
1 nas ' (24).
And although the precise laws of Latin Syncope (if indeed there were
any) are doubtful, so much seems certain, that a form accented like
altrovorsum does not permit us to infer a form accented like ' altrum ',
1 altris '. It will be safer to retain the alterum, alteris of the MSS. and
scan :
Capt. pr. 8 (Sen.) Alterum quadrimum puerum servus surpuit ;
Naevius com. 23-24 (Sen.) Alteris inanem vulvulam madidam dari,
Alteris nuces in proclivi profundier ;

and not to change their alterius to { altrius ' in Capt. 306 (see II 35).
For a form cetr- there is no evidence at all : we must scan ceteris Poen.
1 183 (if Anapaestic Metre), ceteri True. 102. But there is warrant (see
42) for dextra beside dextera, sinistra (Terence) beside sinistera (Plautus).
1 Opra ' is unbelievable in Ennius' epitaph of Scipio. Vahlen reads
(Epigr. 20) : reddere opis pretium (operae MSS.). ' Intero ' (Cist. 704)
has no more claim than l supera ' (see 42). On aspera and aspra see

Though lardum (Lucilius 79) is a post-Plautine syncope of laridum,

42.
ardus (cf. Lucil. 733) for aridus appears in :
Aul. 297 (Sen.) Pumex non aeque est ardus atque hie est senex
(n 50) ;
Pers. 266 (Troch.) ||homines vetulos, avidos, ardos.
Surpui (beside surripui) is certain, but never ' surptus ' (see 42). It
hardly justifies 'erpe' in Stich. 718 (Troch.) :
|| eripe ex ore tibias.
furigo, obiurigo, purigo (beside obiurgo, purgo) in Plautus (42) hardly
justify ' iurigium ' in Pers. 797. Of the same type are, to outward
appearance, compounds of rego, but in them Syncope is capricious :
pergo, surgo, porgo on the one hand, erigo, corrigo on the other (see 42),
a warning to editors against rash inference from analogy.
Since lamina (lammina) occurs only once in Plautus (Asin. 548), it is
impossible to say whether Horace's lamna was known in Plautus' time.
But a by-form domnus (with dominus) is suggested by the pun in :
Cas. 722 (Anap.) ||duplici damno dominos multant;
Ter. Heaut. 268 (Troch.) || domina ego, eru' damno auctus est.
(On valide, ba/ineae, the Plautine forms of va/de, balneae, see 42.)
28. GREEK LOAN-WORDS. It was Accius who initiated the
reform in the spelling (and pronunciation) of loan-words from the Greek,
 HIATUS 147

a reform which ended in the addition of two letters, y and s, to the

alphabet. It was said of him (ap. Varr. L. L. 10, 70) :

Accius ' Hectorem ' nollet facere, ■ Hectora ' mallet.

Previously these Greek immigrants had been roughly handled. Indeed
the correct three-syllabled Tecmessa was not heard on the Roman stage
till the tragedian Julius Caesar Vopiscus, an older contemporary of
Cicero, ordered his actors to drop the fashion of pronouncing it Tecu-
messa Some Greek consonant-groups required on Roman lips a parasitic
vowel (much as Teutonic hnapp- became Late Latin hanappus^ Fr.
hanap ; Low German ' knif ' became * canif ' in French utterance) : fiv
of jjiva (Lat. mina\ yy oirkxyt) (techina), ^ of fyaXPV (drachuma), Kfiof
'AA-Ac/xatW, A\K/Arjvr) (Alciimeo, Alciimena), kv of Kv/cvos (Cucinus), rv
of Acrvrj (Aetma, Mil. 1065), etc. To discuss the question fully lies
outside our province ; but one point, the early Roman treatment of
Greek \ between vowels, must be mentioned here.
In all Latin Ppayiuv became bracchium. In Plautine Latin Axepwi/
became Accheruns. What other Greek words were so treated ? Not
words seldom used, like Cleomachus (Bacch. 589), Eutychus (Merc.
474), Mnesilochus, Philolaches. What of the often used Achilles'*
Here are the occurrences earlier than Accius :

Bacch. 938 (Iamb.) Relictus, ellum non in bus||to Achilli, sed In

lecto accubat ;
Merc. 488 (Troch.) Achillem orabo aurum [ut] mihi det || Hector
qui expensus fuit;
Mil. 61 (Sen.) Rogitabant : ' hicin Achilles est?' inquit mihi ;
1054 (Anap.) Age, mi Achilles, fiat quod te oro, || serva illam
pulchram pulchre ;
1289 (Sen.) Mitto iam ut occidi Achilles cives passus est;
Poen. pr. r (Sen.) Achillem Aristarchi mihi commentari libet ;
Ennius trag. 148 (Troch.) Qui cupiant dare arma Achilli, ut || ipsi
cunctent . . .

The only two that throw light on the quantity of the first syllable are
Merc. 488 (where a long syllable is required) and Mil. 1054. The
latter can hardly be scanned, as it stands, with either quantity ; but if
we put Age (an ejaculation ' extra metrum ' ; IV 3 C) in a separate line,
we get :
Mi Achilles, fiat quod te oro ||.
The natural inference is that the word was pronounced by Plautus (and
should be spelled in our editions) Acchilles. The first foot of Merc.
l 2
148 EARLY LATIN PROSODY
488 is a Spondee ; of Mil. 1054 possibly the same, but more probably
an Anapaest (Mi Acchill-, with Mi in prosodic Hiatus as 'brevis
brevians '). Of course Merc. 488 is the only line which proves the word
to be a Molossus (cf. II 18 B), since a Bacchius would suit Mil. 1054
equally.
What of another word in common use, machaera! To exhibit
all the numerous occurrences of the word (or its Compound Polyma-
chaeroplagides or Derivative Machaerio) would take too much space. As
with Achilles, only one line seems decisive, Pseud. 593, a line which
follows an Anapaestic Octonarius and is presumably the same (or
a Septenarius) :
Lubet scire quid hie venit cum machae||ra et huic quam rem agat
hinc dabo insidias.

Macchaera seems necessary. Of course Leo's ' ciim machaera ' may be
left out of the question).
The other occurrences before Accius' time are three in number. Two
prove nothing:
Ennius trag. 149 (Troch.) Quae mea comminus machaera at||que
hasta hostibit e manu ;
Caecilius 68 (Iamb.) (Quae) Narrare inepti est (ad) scutras ||
ferventes. Quin machaera
Licitari adversum ahenum coepisti
sciens ?
One is quite uncertain, a mere couple of words quoted by Servius, that
notoriously inaccurate quoter, from Ennius (whether from his Dramas
or his Epic is not stated): heia machaeras (Ann. 597). But another
quotation by Servius from Ennius is more readily accepted as a Hexa-
meter ending (Ann. 400) : succincti corda machaeris. The early Latin
scansion of machaera is thus even more doubtful than of Achilles.
Ennius seems not to have followed Plautus in Accheruns. His
Andromacha had an Anapaestic line :
trag. 70 Acherusia tem||pla alta Orci,
and we naturally find in Festus' quotation a complete Senarius (I 4) of
the Iphigenia :
trag. 202 Acherontem obibo, ubi morti' thesauri obiacent.
To the quantity of the first syllable of Achivi in early Latin there is
no clue. The scansion of the single line with Antiochus is doubtful :
Poen. 694 (Sen.) Quam regi Antiocho oculi curari solent (AP),
 HIATUS 149

but Antiocho is likely. Stomachus (orro/xaxo?, accented on first syllable

like 3>iAi7r7ros) would hardly be a popular word. It occurs in : tare ;
Asin. 423 (Iamb.) Clamore ac stomacho non queo ||labori suppedi-

Lucilius 155 Quam gladium in stomacho furia ac pulmonibu' sisto ;

and the derivative Verb in :

Ter. Eun. 323 (Sen.) Id equidem adveniens mecum stomachabar

modo.
(For other words see paragraph 42 s. vv.)

29. PREPOSITIONS IN COMPOUNDS. (On Compounds of iacio

see 23.)
ante elides -e in antehac (also antidhac in Plautus), a disyllable (see
42) ; anteo (nomally antideo in Plautus) ; antea (always antidea in
Plautus) :
Ph. 247);
Amph. 649 (Bacch.) Virtus omnibus re||bus anteit profecto (cf. Ter.

Ter. Andr. 52 (of doubtful text) . . . nam antea (Plaut. frag. 175 is
still more doubtful).
circum. On circu(m)eo see 45 and 42.
cum. The co of cogito, comptionalis, etc., we have already referred
to the accentuation c6agito, c6emptionalis (II 18). The fusion of 6 with
a or e produced 6. If the pronunciation of the disyllable (?) coegi
should be expressed by coegi (cf. debrum) in :

Bacch. 981 (Troch.) Optimu' sum orator; ad lacrimas || hominem

coegi castigando (co. horn. MSS.),
then Plautus' pronunciation of coepi may have been different from
Virgil's. Pacuvius makes cderce a disyllable :
47 (Sen.) Gradere atque atrocem coerce confidentiam (cf. 345).

The Perfects coegi always (in Terence's three occurrences at least), and
coepi sometimes, are trisyllables :
Cas. 651 (Bacch.) Tua ancilla hoc pacto ex||ordiri coepit;
701 (Bacch.) Nam cur non ego id per||petrem quod coepi? :
Cist. 687 (Bacch.) Sed pergam ut coepi || tamen, quaeritabo ;
Merc. 533 (Iamb.) || cum mecum rem coepit (cf. Aul. 626).
The relation of the trisyllable coactio to the Noun cocio is uncertain ;

Asin. 203 (Troch.) Vetus est ' nihili coactiost '— scis || quoius. non
dico amplius.
150 EARLY LATIN PROSODY
Nor can we say whether coaccedunt (Brev. Brev.) or coaccedunt (Syniz.)
was the pronunciation in :
Cure. 344 (Troch.) Triginta minis, vestem, aurum ; et || pro is
decern coaccedunt minae,
nor even whether it be the right reading (eo accedunt edd.). Synizesis
would be the only possibility for coadsolet Epid. 7, but this is a mere
editors' reading (eo adsolet A, eo assolet P). (On coagmenta Most.
829, coepulonus Pers. 100, see II 18.)
The spelling of the MSS. suggests that Accius pronounced conestat
(for cohonestat ; see 22), but we have no clue to Plautus' pronunciation
of coopertus (frag. 176), cooriuntur (Pers. 313).
de shortens_the vowel be_fore_another vowel in deamo, deuro (Lucilius
1037), etc. That the corruption in the MSS. conceals deagetur in Epid.
65 (Sen.) is by no means certain :
Deperit. Degetur corium de tergo meo (detegetur MSS.),
Presumably dehortor in Plautus (cf. Ter. Ph. 910):
Capt. 209 (Troch.) || si erit occasio, hau dehortor ;
Poen. 674 (Sen.) Neque nos hortari neque dehortari decet.
(Cf. Ennius Ann. 381 Hannibal audaci cum pectore de me hortatur.)
Presumably deargentassere in Lucilius 682 (Troch.) || me ac deargen-
tassere. Presumably deorsum in_Lutilius 703 (Troch.) :
Modo sursum, modo deorsum ||.
But the usual pronunciation of the last is with Synizesis deorsum (II 33);
and similarly, we may presume :
debsculor, Cas. 136 (Sen.) :
Sine tubs ocellos debsculer, voluptas mea (cf. 453-454; 467);
deartuo, Capt. 64 1 (Troch.) :
Turn igitur ego deruncinatus, || deartuatus sum miser (cf. 672) ;
deascio, Mil. 884 (Iamb.) :
Tibi dixi miles quern ad modum || potisset deasciari ;
deambulo, Ter. Heaut. 587 (Iamb.):
Abi deambulatum. Deambula||tum ? quo ? Vah ! quasi desit locus ;
(cf. 806)
deintegro, Caecilius 255 (Troch. ?) :
Nomen virginis, nisi mirum est, || delntegravit ....
dehinc, a monosyllable (see 42).
Shall we add deerro (or derro), Men. n 13 (Troch.) ? :
Inter homines me deerrare ||a patre atque inde avehi (With Hiatus
at Diaeresis).
 HIATUS 151

prae^ with shortening of diphthong (printed e in firehendd) before a

vowel, e. g. :
Mil. 41 (Sen.) Curamque adhibere ut praeolat mihi quod tu velis.
pro and pro, e. g. protinam and protinus ; protervus (see 42).
Notice procuro, Accius Ann. 4 :
Exercent epulis laeti famulosque procurant.
May we therefore scan in the seventh foot of an Iambic Septenarius
(see IV 10) similarly ?
Cure. 519 || tibi orTust qui te procures ;
525 || decern qui me procurem.
Or should we follow the one decisive Plautine example (out of
many) ?
Stich. 200 (Sen.) Quibus ipsis nullast res quam procurent
True.sua878).
(cf.

Notice profiteor, Men. 643 (Troch.), Ennius trag. 293 (Sen.) :

Neque vis tua voluntate ipse || profiteri, audi atque ades j
Te ipsum hoc oportet profiteri et proloqui.
beside profiteor (Capt. 480 ; Ter. Eun. pr. 3).
Since propino is proved by Stich. 425 (Sen.), Ter. Eun. 1087 (Troch.)
Cadum tibi veteris vini propino. Papae ! (cf. Cure. 124);
Hunc comedendum et deridendum || vobis propino. Placet,
editors are (rightly or wrongly) reluctant to find propino in the corrupt
Pseud. 1262. Compare prologus (but see 4a s.v.), propola (Lucilius
198).
Notice proficiscor, Trin. 149 (Sen.)
Quoniam hinc est profectu(ru)s peregre Charmides ;
but since proficiscor appears in all the other Plautine occurrences where
the scansion can be determined (Cure. 1 ; Merc. 939 ; 946 ; Mil. 1329 ;
Pseud. 425), the Trinummus variety is suspect.
re anfl red. While redduco is invariable and an occasional reddux
(beside the usual redux) possible (see 42), the following are isolated and
therefore, most of them, suspect :
reccidere, the spelling in A in Ter. Hec. pr. 47 (where the metre gives
no clue to the scansion) ;
rellatum, Ter. Ph. pr. 2 1 (Sen.) :
Quod ab Illo allatumst, sibi esse rellatum putet ;
rellevabis, Ter. Ad. 602 (Iamb.) :
Nam et illi animum iam relleva||bis, quae dolore ac miseria (Or
illic, with rele- ?) ;
152 EARLY LATIN PROSODY
rellicta, Lucilius 1012 :
Et sua percepere retro rellicta iacere ;
remmotum, Tec. Hec. pr. 22 (attested by Donatus : geminavit M
secundum antiquos) ;
reppedabam, Lucilius 676 (Troch.):
Sanctum ego a Metello Romam || reppedabam munere (Emended
to Ro. rem repe.) ;
repplebo, Poen. 701 (Sen.) :
Ibi ego te repplebo usque unguentum geumatis ;
repprime, Accius 381 (Sen.):
Repprime parumper vim citatum quadripedum (see II 9).
revvortimini, Amph. 689 (Troch.) :
Quid animi habeam ? sed quid hue vos || revvortimini tarn cito ?
30. PROSODY OF NOUN. (A) First Declension. Norn. Sing. As
Greek Proper Names in -77s, e. g. Anchises, naturally passed into the
Fifth Latin Declension (see below), so those in -as into the First. The
name Antidamas (Gen. -ae Poen. 1042, 1047) in the Poenulus offers a
problem. Hanno uses this form in his Carthaginian (line 934) as well
as his Latin utterances (955 and 1051, two Senarii with this ending:
hospes Antidamas fuit). The substitution of Antidama required by the
metre in the Senarius spoken by Agorastocles looks like one of those
true-to-nature inconsistencies which are so often effaced by medieval
scribes (or correctors) and modern editors. For the MSS. show the -as
form as in the other occurrences. And editors make this metrically
possible by altering Mine to illim. The line can be scanned without it :
Read:

Poen. 1058 (Sen.) Surruptu' sum illinc, hie me Antidama hospes tuus.
For the Gen. Sing. Antidamati (A), Antidamarchi (F) editors substitute
(Poen. 1045) Antidamai. Otherwise Plautus follows the rule mentioned
by the grammarians, that these Greek loan-names took -a in the Nom.
Sing, in Early Latin. This -a one would expect to be short like the -a
of Latin Nouns (4). But the evidence, such as it is, points to -a :
Amph. 439 (Troch.) Ubi ego Sosia nolim esse ||;
Asin. 740 (Iamb.) Leonida, curre, obsecro, hue ||;
Ennius Ann. 275 at non sic dubius fuit hostis
Aeacida Burrus ;
179 Aio te, Aeacida, Romanos vincere posse;
Lucilius 1 1 08 Perditu' Tiresia tussi grandaevu' gemebat.
(On the improbability that a Vocative form was used for a Nominative
see 14.)
 HIATUS 153

Do other loan-words from the Greek show long -a? In Asin. 762
(Sen.) epistula would give a false Tribrach (II 44) :
Ne epistula quidem ulla sit in aedibus,
but there is another possibility. In this contract, written (in part) and
read aloud on the stage, there is often a Hiatus to mark a pause (51).
So this may conceivably be Syllaba Anceps designed for the same effect.
The Tribrach produced by tessera (Poen. 1052) would not be un-
Plautine (in the fourth foot of a Senarius) :
Haec mi hospitalis tessera cum illo fuit (On cum Kilo see 53).
The line (or fragmentary couplet, whose second line begins with Longa)
cited by Isidore from Ennius is too doubtful evidence of aged (Ann.
492) : Multa foro ponet et agea longa repletur (Emended to lo. re. ag.)
(B) Gen. Sing. The relation of -ai to -ae (-ai) in Early Latin has not
yet been investigated along with that of -ei to -ei (or -1). Apparently,
the reduced form -ae (in Plautine spelling -ai) of the Gen. Sing. Suffix
was a new-comer in Plautus' time, restricted to a position before an
initial consonant. Before an initial vowel the full form of the suffix, -ai, 1
was used and its second part, -I, elided (see 57). So Plautus' age
recognized e.g. terrae fuit. pars, terra! erat pars. The result is that!
Plautus never (or hardly ever) has a scansion like Lucilius 1029 ; 25 :
Sicuti te, qui ea quae speciem vit(ae) esse putamus ;
Thestiados Led(ae) atque l&ovfys a\6xou>.
There are, however, plenty of examples of -ai before a consonant, and
while we might distinguish some as aiming at an effect of dignity or
ceremoniousness, e. g. :
Mil. pr. 103 (Sen.) Magnai rei publicai gratia ;
Aul. 121 (Bacch.) Meai fidei tuaique rei,
the distinction would be arbitrary (cf. Ter. Andr. 439). For in most
cases the choice of the full form of the suffix seems due to mere metrical
convenience or even caprice. Details (to be used with caution) of the
infrequency of the elision of Gen. Sing, -ae in subsequent poetry will be
found in Leo, Plaut. Forsch.2 ch. vi. In Plautus the alleged examples
of elision are :

Amph. 1 135 (Sen.) Primum omnium Alcumena* usuram corporis

(cepi) (Possibly Dative) ;
Bacch. 1083 (Anap.) Aequum esse puto sed nimi' nolo ||desidia* ei
dare ludum (Rather desidiai ei) ;
1 105 (Anap.) Hi(c)quidemst pater Mnesilochi. Eugae !
soci||um aerumna* et mei mail video (Or Octonarius) ;
154 EARLY LATIN PROSODY
Capt. 585 (Troch.) Atque ut perspicio profecto ||iam aliquid pugna^
edidit (Rather with elision of iam. See App. C, end) ;
Epid. 563 (Troch.) Domi meae eccam salvam et sanam || (Locative,
not Genitive ; see below) ;
Mil. 645 (Troch.) Commemini et meae orationis ||iustam partem
persequi. (The sense favours meam. And conceivably meae
might be in Prosodic Hiatus, a ' brevis brevians ');
883 (Iamb.) Postquam adbibere aures meae ||tuae oram
orationis (tuam moram MSS. Possibly tuam) ;
Most. 173 (Iamb.) Virtute format id evenit||ut deceat quidquid
habitas (Read formai ev.) ;
Poen. 802 (Sen.) (paululum) Fmedae intus feci : dum lenonis
familia (Emended to fee. int.) ;
Stich. 202 (Sen.) Adeunt, perquirunt quid siet causae ilico (Rather
sit, or siet, causai ; II 33) ;
erft).
Trin. 676 (Troch.) Turn igitur tibi aquae erit cupido ||(Rather aquai

(A Dative, not a Genitive, is to be found in Capt. 1036 ; Merc. 521 ;

Trin. 365; 628; True. 883.)
The old ending -as was quite obsolete in the speech of Plautus' time
(except in the legal phrases pater familias, mater familiasy etc.), though
Ennius seems to have used this archaism in his Epic :
Ann. 441 dux ipse vias.
(C) Dat. Sing. The suffix -ae (in Plautine spelling -ai) was never a
disyllable, as the Gen. suffix was. It is freely elided.
(D) Loc. Sing. This case was still distinct from the Gen. Sing, in
Plautus' time. Like the Dat. Sing, it has for suffix -ae (in Plautine
spelling -ai). Examples (not very strong, as it happens) of Elision are :
Epid. 563 (Troch.) Domi me(ae) eccam salvam et sanam || (Or
meae in Prosodic Hiatus) ;
Ennius Hedyph. 36 Mitylen(ae) est pecten caradrumque apud
Ambraciai (Finis) (Or Mitylenae'st).
(E) Abl. Sing. Always in -a. The older -ad was quite obsolete in
the talk of Plautus' and presumably Naevius' time, though the latter
poet uses this archaism in his Saturnian Epic, the Bellum Punicum :
Noctu Troiad exibant || capitibus opertis (Troia de MSS.).
(F) Nom. Plur. The original form of -ae is not clear, since a
comparison with Greek -at offers difficulties. But until Comparative
Philology provides justification for an early -ai (like the Geji. Sing, suffix)
the likelihood of -ae (-ai) being treated before a vowel like Gen. Sing.
 HIATUS 155
-ai seems remote. Till then the lines which show Hiatus must be
treated like any other examples of Hiatus. Thus the Hiatus in Asin.
759 (Sen.) :
Fores occluft** omnibus sint nisi tibi,
may be ascribed to the pause in writing and reading the formal contract
(50) ; in Most. 709 (Cretic) :
Atque pol nescio ut || moribus sient
Vostrae : haec sat scio || quam me habet male,
to the pause in the sense (50) ; in Poen. 11 36 (Sen.) :
Eho an hum' sunt illaec Rliae? Ita tit praedicas,
to the pause at the change of speaker (49). Others (see Leo's list in
Plaut. Forsch.2, pp. 346 sq.) must be emended, e. g. Trin. 539 (Sen.) :
Nam fulguritae sunt alternae arbores (Emended to alternas Adv.,
scil. vices).
RitschFs remedy was to substitute that Nom. PI. suffix -as which is found
(with -s dropped) on old inscriptions (in rude Latin) of Pisaurum and
(in full) in Oscan inscriptions. But the scientific study of the language
has shown that, although at some remote period it may have been a
Latin suffix, it had been replaced by -ai (cf. Greek -ai) before (perhaps
long before) Plautus' time and survived only in dialects. The Atellan
farces exhibited the life of the poorer classes in country-towns and
deliberately used some un-Latin forms (e. g. dicebo for dicam Fut.). So
it is natural to find this dialectal Nom. PI. employed by Pomponius
(time of Sulla) :
141 (Troch.) Quot laetitias insperatas || modo mi irrepsere in sinum !
(cf. 1 5 1 has, Nom. PI. ?). But it must be ruled out of the ' pura oratio '
of the other Dramatists.
(G) Second Declension. (On Abl. Sing, -o, not -od, see 10.) The
Gen. Sing, of IO-stems ended in -I, the Loc. Sing, in -ii (1 2). In Most.
213 vitii Gen. finds no supporter (rather || viti malesuada lena); and the
Hiatus in Ennuis' epitaph is a puzzle :
Aspicite, o cives, senis Enwi imagini' formam (Details in Leo,
Forschungen2, p. 338).
The Dat. Abl. Plur. of IO-stems ended normally in -iis. Did it end
occasionally in -is (II 33) ?
Ennius Ann. 163 nonis Iunis soli luna obstitit et nox ;
Turp. 162 (Iamb.) Detegere, despoliare opple||reque adeo fama ac
flagitis (Emended to famae ac flagiti).
(But Turp. 168 superciliis, not -lis.)
156 EARLY LATIN PROSODY
(H) A Nom. Plur. suffix -eis (is) found in some inscriptions of the
end of the second and beginning of the first century B.C. (apparently by
addition of a Plural suffix -s to the already formed Plural in -ei) is
unknown to Plautus, who however uses in the Pronoun Declension
/usee, illisce (31) before an initial vowel, e. g. Mil. 374 (Iamb.) :
Non possunt mihi minaciis || tuis hisce oculi effodiri.
Nor does it appeal anywhere in early literature. The theory of Ritschl's
day, that the suffix -I (really identical with Greek -01) represents an
earlier -Is, is by this time (or rather ought to be ; 2) as dead as Queen
Anne.
Gen. PI. The rivalry of -um (cf. Greek -wv) and -orum may be
illustrated from :
Most. 120 (Bacch.) Primumdum parentes || fabri liberum sunt:
I fundamentum sub||struunt liberorum.
(I) Third Declension. The Abl. Sing, of Consonant-stems ended
normally in -e, of I-stems in -1 (never -Id ; 10). But -1 seems often to
have encroached on the province of -e (see 5). Whether Abl. avt, etc.,
when scanned as a Pyrrhic should be printed with -i or with -e is an
open question.
(On the old declension of homo with Ace. Sing, homonem see 42.
It is confined to early Epic.)
(K) Fifth Declension. This Declension goes hand-in-hand with
the First, but is not so coherent a collection. For example, res did not
wholly conform to fides, etc. ; canes Fern., fames, sordes, etc., are on the
border-line between Third and Fifth. (On spes see 42.)
Gen. Sing. The full suffix -el became -ei (-1) just as the full suffix -al
became -ae ; e. g. fidei and fidei (fidi). But res has three Genitives, rel,
rel, ra (II 33), e. g. :
Mil. pr. 103 (Sen.) Magnai rei publicai gratia;
Men. 494 (Sen.) Adulescens, quaeso, quid tibi mecum
(cf. Men.est323);
rei?

Pers. 65 (Sen.) Nam publicae rei causa quicumque id facit.

(L) Dat. Sing. The forms ' fidei ', ' die! ' in our school-grammars are
unknown to early (and probably to classical) Latin (Class. Rev. 10, 424),
which recognizes only fide, die. An exception is rei (35), always a
monosyllable and often elided, e. g. in the common phrase ei r(ei)
operam dabo. (Hardly 're'; Amph. 674; Merc. 300; Poen. 815;
Trin. 757 ; Lucilius 237.) The doubtful tradition of Ennius Ann. 107
suggests reique, fide, regno vobisque.
 HIATUS 157

(M) Abl. Sing, in -e (never -ed; 10). Since monosyllables like hand,
medy ted managed to retain -d after a long vowel, it would not be
unreasonable to look for ' red ' Abl. in Plautus. But the evidence is far
too slight (Aul. 141 ; Merc. 629; Pseud. 19 ; Pacuvius frag. 237).

(N) Since Greek Nouns in -77s, like 7ron7r>js, became First Declension
Nouns when they were borrowed by the Romans, e.g. poeta, it was
always taken for granted that Greek Proper Names in -rj<s did the same
(cf. Quintilian 1, 5, 61 on Anchises). But the acute mind of Wacker-
nagel saw that these Proper Names, which were hardly used in the
Plural (an exception is Atridae Bacch. 925) must have suggested to the
Romans rather the Fifth Declension than the First. And the MSS.
confirmed this discovery, which unluckily was made too late for most
editions of Plautus. The ' Charmidal\ etc., of our printed texts must
be changed to Charmida, etc., the ( Charmidae* to Charmidl (-ei). The
shorter form of the suffix produced in classical Latin these anomalous
Genitives, Socrati, Achilla Aristidi, etc. Even Hercules was in Plautus'
time a Fifth Declension Noun, e.g. Pers. 2 (Iamb.), Cas. 398 (Troch.),
Rud. 822 (Sen.):
Superavit aerumnis suis || aerumnas Herculei ;
Utinam tua quidem ista sicut || Herculei praedicant (Quondam
prognatis) ;
lam hoc Herculei est, Veneri' fanum quod fuit.
So read Epid. 626 Apelles ac (-es atque MSS.), Poen. 1271 o Apelle
(with A\ not ' Apella '. The manuscript evidence strongly supports -e
(cf. Priscian Gram. Lat. ii, 288) as the Voc. Sing, in Terence (e. g.
Chreme) and perhaps also in Plautus. (See Hauler's note on Ter. Ph.
567.) The Nominative form seems to have been a rival. So that the
Vocative shows both -e and -es.

31. PROSODY OF PRONOUN. Conversational Latin of the

beginning of the second century b. c, echoed in Plautus' verse, is a
fascinating study for a linguist. It reveals to him the remote beginnings
of such things as the treatment of Pronouns in the Romance languages
(e. g. French toi; te, t' and so on), and some of these, although already
mentioned in chap. II (e. g. Plaut. ego ilium scio, French je l'sais), had
better stand here too. With Pronouns the question we have usually to
ask is : Which is the enclitic, which the emphatic form (or the form of
deliberate utterance) ? This question does not always find an unhesita-
ting answer, and one reader will differ from another. Luckily there is
often no room for doubt.
158 EARLY LATIN PROSODY
Personal Pronouns.
Ego (more often ego by the Brevis Brevians Law) is not rare, even in
Dialogue Metres ; e. g. :
Trin. 173 (Sen.) Sed nunc rogare ego vicissim te volo ;
Aul. 457 (Troch.) Coctum ego, non vapulatum ||;
True. 357 (Sen.) Vah ! vapulo hercle ego nunc, atque adeo male
(hercule A ; cf. 42) ;
Cist. 745 (Iamb.) || ego sum illiu' mater ;
Mil. pr. 142 (Sen.) In eo conclavi ego perfodi parietem;
Rud. 1 184 (Troch.) Sumne ego scelestu' qui illunc ||;
It can therefore stand in Prosodic Hiatus, e.g. Poen. 1042 (Sen.) ;
Cas. 724 (Anap.); Ennius Ann. 193 :
Verum ego hie hospitium habeo : Antidamae filium ;
|| tii amas, ego esurio et sitio ;
Hos ego in pugna vici, victusque sum ab isdem.
The pyrrhic form appears in egoquidem (cf. II 37) of which equidem
(origin unknown) must not be claimed as a slurred form on the mere
evidence of Epid. 202 (Troch.) :
Et ego Apoecides sum. Et egoquidem || [sum] Epidicus. sed,
ere, optima,
although it is likely that the accentuation corresponded with the ictus
here, Egoquidem. In Amph. 764 (Troch.) editors change ' ego equidem '
of the MSS. :
Dixit ? Egoquidem ex te audivi et || ex tua accepi manu.
But (like tu quidem beside tiiquidem, etc.) ego quidem appears in, e. g.,
Aul. 570 (Sen.) :
Non potem ego quidem hercle. At ego iussero (With Hiatus at
change of speaker).
Mis, the old rival form of mei, must be taken along with Us. Its
puzzling scansion (mis or mis ?) is discussed elsewhere (42). We add
a hint that it may have been supplanted by mei (and tis by tui) here
and there in the traditional text, through the practice of writing a gloss
above (e. g. Pseud. 6 Mei te rogandi et [tui] tis respondendi mihi ;
Trin. 343 tis A, tui P). Mei is really Gen. Sing, of the Possessive, ' of
mine' (32). An example of iambic mei is Vid. 67 (Troch.) :
Nisi quid ego mei simile aliquid ||contra consilium paro (Emphatic) ;
of monosyllabic, the line which caught Terence's fancy (Eun. 801),
Capt. 800 (Troch.) :
Faciam ut hui(u)s diei locique || meique semper meminerit (Or
meique, emphatic ; II 18 C).
 HIATUS 159

Mihi. The unaccented form mi (found also ' in pausa ', e. g. Asin.
614, Most. 175 ; but not when emphatic. Read Ace. me Most. 871) is
hardly distinguishable from mihi in the traditional text (cf. nihil and nil),
since an abbreviation-symbol (mi, etc.) was often used, and since a scribe
would be apt to regard * mi ' as a mere miswriting of ' mihi ' (e. g. Merc.
106, Pers. 739). And the scansion mihi is to us nowadays indistinguish-
able from mi. (It is assured for the emphatic Pronoun, e. g. Pers. 487.)
Presumably mihi and ml differed as huius and hui(u)s. A good example
of emphatic mihi in Prosodic Hiatus is Merc. 619 (see 54), a Trochaic
Septenarius :
Non tibi istuc magi' dividiaest || quam mihi hodie fuit.
The contrast of mi and mihi appears e. g. at Cas. 920 (Bacchiac) :
(Nimis tu quidem hercle immerito)
Meo mi haec facis, quia || mihi te expetivi;
and (in spite of Persius' Min tu istud ais ?) we incline to follow the MSS.
in Pseud. 472 (Sen.):
end).
Mihin domino servu' tu suscenses ? Tarn tibi (Mihi A) (See II 27,

At any rate an editor should print at Bacch. 125-126 (Senar.) :

Non hie placet w/ornatus. Nemo ergo tibi (Enclitic monosyllable)
Haec apparavit : mihi paratum est, cui placet (Emphatic Pyrrhic) ;
at Cas. 849 and 851 (Senar.) :
Quid est ? Pectu' mi icit non cubito, verum ariete.
At mihi) qui belle hanc tracto, non bellum facit ;
in Pseud. 945 mihi obtrudere (not ' mi obtr.').
Med (Ace, Abl.) and ted might be altered to me and te by any scribe
at any time (I 6). We can hardly elicit from the traditional text the
precise circumstances under which the fuller and under which the
classical form was used by Plautus. Since the word is often elided, we
see that med (ted) was not the mere prevocalic form (like haud aliter
beside hau secus). That more emphasis was attached to med, ted than
to me, te would seem likely (see below), but cannot be proved (53).
The expression of emphatic me, te by Prosodic Hiatus is discussed
below (53), e.g. Asin. 820 (Sen.) :
Ego sic faciendum censeo : me honestiust
(Quam te palam hanc rem facere, etc.),
where med cannot be substituted.
 160 EARLY LATIN PROSODY

Plautus seems1 not to have used sed (Mil. 1275 ; Men. 909; Merc. 379),
form found on the S. C. Bacch. etc., but only sese as by-form of se.
Med, ted seem not to be recognized by Terence : in Eun. 307 tete (tete,
like sese, according to Donatus on Adel. 33) is as likely as ted. In the
citation of a line of Caecilius (between Plautus' and Terence's time) the MS.
reads me oporteat, but editors :
9 (Troch.) Sed ego stolidus : gratulatum || med oportebal_p_rius,
probably a just emendation. The same cannot be said of Leo's reading in
Turpilius 209 ; and in Accius 372 (Troch.) tete is a rival conjecture :
Quae subsistat : modo tute ipse || ted offirma et compara.
It need not be questioned in Ennius1 Epicharmus 45 (the third or fourth line
of the poem which began with the acrostic Q. ennius fecit), a Trochaic
Septenarius :
Nam videbar somniare || me(d) ego esse mortuum.
Nos, vos might be specially emphasized by the addition of the Particle
-met of egomet (cf. nobismet, etc.) ; and one explanation of med, ted
sees in them an addition to me, te like the -dem of ibidem ' precisely
there', totidem 'just so many', idem (for is-dem). Cf. memet Men.
1 145; mepte Men. 1059. As example of the emphasized form take
Poen. 250-251 :
Soror, parce, amabo : sat est istuc alios
Dicere nobis, ne nosmet in nostra etiam vitia loquamur.
Ictus marks emphasis in, e. g. Poen. 630 (Sen.) :
Nunc vos mihi amnes estis ; vos certum est sequi.
Tu. No trace of the Greek quantity is to be found in tiiquidem
(tequidem, etc.) (II 37). The other cases of this Pronoun (as of the
Reflexive) have already been treated under the First Personal Pronoun.

1 This will be variously interpreted. One scholar, who has a propensity to what
may be called thrilling fiction (and thrills are more attractive with readers than
sober sense) will erect on it a theory that our text is a modernized text, a kind of
Chaucer in Dryden's setting, that the re -caster used sese for Plautus' sed, but re-
wrote lines containing med, ted (e. g. Pseud. 523 survives in two versions ; also
Merc. 555), and so on, and so on.
Another, who likes to be guided by facts and detests Horace's motto :
Et mihi res, non me rebus subiungere conor,
] will find from the evidence of inscriptions, etc. (1) that sese was current in Plautus'
\j^r. /time (sesed in the inscription of the Faliscan cooks) (2) that med, ted soon disap-
/ peared after Plautus' time. He will infer that Plautus had no particular liking for
med, ted and used them only because a handy, current by-form like sese was not
available. , Cautious editors will take the hint that they must not replace me, te by
med, ted where this is not quite necessary (e. g. at the Diaeresis, where Hiatus is
freely allowed by Plautus ; 48).
 HIATUS i6t

There remains only

tibi (like sibi), an Iambus or a Pyrrhic (6), without an enclitic by-
form like mi. In the pre-scientific days, when Ecthlipsis was in the air,
a scansion ' t'bi ' (' s'bi ') was talked of, and this weird pronunciation has
actually in our own time found favour with Leo.
On the marking of emphasis by Prosodic Hiatus, e. g. Men. 389
(Troch.) :
|| tibi et parasito tuo
(cf. Merc. 966), see below, 54. Examples of the emphatic Dative not
so marked are :
Men. 439 (Troch.) Mihi dolebit, non tibi si ||quid ego stulte fecero ;
True. 379 (Sen.) lam lauta es? lam pol mihi quidem atque oculis
meis.
Num tibi sordere videor ? Non pol mihi quidem ;
398 (Sen.) Tu nunc superstes solu' sermoni meo es,
Tibi mea consilia semper summa credidi (Not
marked by ictus ; II 52) ;
745 (Troch.) Nam invidere alii bene esse, || tibi miseriast
male esse ;

Men. 646-7 (Troch.) || huic surreptast, non tibi. foret ;

Nam profecto tibi surrepta || si esset, salva non

Ter. Heaut. 977 (Troch.) Neque tibi nee tibi: nee vos est ||.
We would find an Iambus in Stich. 123 (Troch.), where the father after
catechizing the one daughter, turns to the other :
Quae tibi mulier videtur || multo sapientissima ? (cf. Cure. 202).
We would scan Cure. 179 (Troch.) with sibi in Hiatus :
Sibi sua habeant regna reges, ||sibi divitias divites,
Sibi honores, sibi virtutes, ||sibi pugnas, sibi proelia :
Dum mihi abstineant invidere, ||sibi quique habeant quod suumst.
32. Possessive Pronouns. These have been treated in the section
on Synizesis (II 33), where an attempt was made to discriminate enclitic
mei (tin, sui), mebs, etc. from the accented (or deliberate) disyllable and
to show how the same operation of the Brevis Brevians Law as in
Terence's bonis (II 22 ; 28) turned the Iambus into a Pyrrhic to mark
special emphasis. This explanation runs counter to the theory of C. F.
Mulier and his pupil Skutsch, who deny Synizesis and recognize only the
Iambus and the Pyrrhic (the latter being at once the form of special
emphasis and of complete absence of emphasis !). To the chief argu-
ment on which their theory rests, that the monosyllabic scansion never
2348 m
162 EARLY LATIN PROSODY
appears at the end of a line, we reply that it could not appear at the end
of a line (nor at the end of a speaker's remarks nor at any other break
in the line), since the deliberate form of utterance is always the form
used 'in pausa ' (just as huius, etc., II 35; nonne, etc., II 36. See
below, 59).
Some illuminating examples may be added here to the others in II 33 :
Most. 37 (Sen.) Mel tergi facio haec, non tui fiducia (Emphatic) ;
49 (Sen.) Meum bonum me. te tuum maneat malum (Emphatic) ;
Mil. 1377 (Troch.) Ad amores meos. sed sensi || (' In pausa') ;
Trin. 1 165-6 (Troch.) Quid ego feci ? Meum corrumpi || quia perpessu's
filium. (Probably meum)
Si id met volimtate factumst, || est quod mihi
suscenseas (Emphatic) ;
True. 819-820 (Troch.) Res palam omnis est, meo illic |l nunc sunt
capiti comitia, (Emphatic ; meo ?)
Me(um) Illuc facinus, mea stultitiast ||;
Men. 291 (Sen.) lube te piari de mea pecunia;
Merc. 680 (Sen.) Meoque Ut parcas gnato pace propitius (In a prayer
to Apollo).
Contrast the Enclitic, e. g. Trin. 1156 (Troch.) :
Filiam meam tibi desponsatam es||se audio. Nisi tu nevis,
where an emphatic Possessive ('my daughter and nobody's else') would be
ridiculous. Still we must repeat the warning already given, that it is
impossible nowadays to determine each and every detail of early Roman
sentence-accentuation. A reader should follow his own interpretation of the
lines, believing that the variety of these Pronoun-forms had (always or often)
a significance for the audience.
We would scan, e.g. :
Pseud. 558 (Sen.) Agite amolimini hinc vos intro nunciam
Ac meis vicissim date locum fallaciis ;
Merc. pr. 8 (Sen.) Vobis narrabo potiu' meas nunc miserias (cf. 5) ;
Bacch. 752 (Troch.) Mea. fiducia opu' conduxi et || meo periclo rem gero ;
True. 137 (Iamb.) Quia tuo vestimento et cibo || alienis rebu' curas;
True. 259-260 (Troch.) Salve. Sat mi est tuae salutis. || nil moror.
non salveo.
Aegrotare malim quam esse || tua salute sanior j
372 (Sen.) Hoc tuis fortunis, Iuppiter, praestant meae;
741 (Troch.) De eb nunc bene sunt tua virtute ||;
Merc. 662 (Troch.) Si ille abierit, mea factum omnes || dicent esse
ignavia ;
True. 965 (Troch.) Meamque ut rem video bene gestam, || vestram rur-
sum bene geram ;
Men. 393-4 (Troch.) Detulisti quam ab uxore ||tua surripuisti. Quid est ?
Tibi pallam dedi quam uxori || meae surripui ?
sanan es ? (Menaechmus 11 was a bachelor) ;
 HIATUS 163

Trin. 684 (Troch.) Numquam erit alienis gravi' qui || suis se concinnat
foot) ; (With Trochee in first, Tribrach in second
levem

Cist. 678 (Troch.) Mei homines, mei spectatores || (With MeiHiainProsodic

tus ; 38) ;

Rud. 1390 (Troch.) (opera mea) Haec tibi sunt servata. Immo hercle ||
mea, ne tu dicas tua ;
1392 (Troch.) (nemp' pro meo) lure oras ? Mirum quin tuiim
ius || meo periclo abs te expetam ;
Trin. 81 (Senar.) Ne admittam culpam, ego meo sum promus pectori ;
Suspicio est in pectore alieno sita.
The Interrogative (Relative, Indefinite) Possessive quoius is not always
distinguishable from the Gen. Sing. Since it is always a Trochee (in True
604 read quoiu's), we must find the Genitive in e. g., Cas. 734 (Anap.) :
Eru' sum. Quis erus ? Quoi(u)s tu servu's ||.
We emphasize the Personal Pronoun in 'what's that to me?' Similarly
mea (tua, etc.) is not a monosyllable in quid mea refertf, quid id tneaf, etc.
Read in Rud. 746 (Troch.) Quid mea refert Athenis ||.
The Gen. Plur. of the Possessive Pronouns is normally (like debrum,
ebrum; II 33) a Spondee, but not in Poen. 766 (Senar.), where it is
emphatic :
Tuorum apud me nemost nee quicquam tui.
The Tragedians' ready recognition of these conversational forms may be
illustrated by: meas (Ennius 296), mea (Ennius 171), meum (Pacuvius 139),
meis (Accius 293),tuae (Ace. 313), tub (Ace. 171, 600), suas (Ace. 170), suis
(Ace. 52), suapte (Ace. 235).
The iambic and trochaic fragments of Lucilius offer : mei (730) ; meo
(965); tuam(6i6); ? tuorum (661).
33. Demonstrative Pronouns. (A) Nominative Singular. Hie,
never a long syllable in Plautus (9), probably lost its final in hi(c)quidem
(9), the companion of illequidem, istequidem (cf. egoquidem, 31).
Is (like quis) never drops its -s in poetry. We must reject any scan-
sion like (Lucil. 197) sol i' mihi, (Lucil. 474) saeva i' febris. Mono-
syllables always retain a final letter lonoiiexdjsyllables, trisyllables, etc.
discanLit. But it is of course subject to the Brevis Brevians Law, e. g.
Trin. 766 and 769 (Senar.) :
Quasi sit peregrinus. Quid is scit facere postea ?
Mendaciloquum aliquem. Quid is scit facere postea ?
Ilk (never Mile '), iste (never ' ste ') lost -e normally (or at least very
often) before an initial consonant (II 36), but retained it 'in pausa'
(e. g. Pseud. 1215, at the Diaeresis of a Trochaic Septenarius, ille ||Syru'
cor perfrigefacit). The loss is conceivable only when the Pronoun
(emphatic or unemphatic) is closely combined in utterance with the
m 2
164 EARLY LATIN PROSODY

following word. So scan Mil. 481 (Senar.) with a Proceleusmatic in the

first foot (II 44) :
Satin abiit ille ? neque erili negotio (Plus curat ?, etc.).
Considering the huge number of occurrences of the word in Plautus and
the faultiness of MSS., we are not surprised to find one or two (illusory)
examples of ill- (on Trin. 809, etc., see II 36, end) :
Mil. 1231 (Iamb.) Spero ita futurum, quamquam ilium ||multae sibi
expetessunt (Emended to eum ; cf. True. 559) ;
Pseud. 954 (Troch.) Illicinest? Jllic[e]st Mala mercist ||(AP).
A line of the Truculentus weakens the force of the second. For in
True. 599 Illicinest? Illic est ||is a Cretic half-line. (In Epid. 135
nunc iam ; see 42, s. v. nunciam.) The MSS. of Terence add another ;
an unknown Tragedian another : H 36);
Eun. 343 (Sen.) Ilia sese interea commodum hue adverterat (see

Trag. inc. 210 (Troch.) ||muliebrem, ilia virgo viri.

Of the by-forms tlfifc, istic (often corrupted by scribes to ' ille hie ', and
eventually sometimes to ' ille ' alone or ' hie ' alone) the precise nuance
escapes us nowadays (35). With Plautus illic homo is the rule, and
suspicion attaches to the exception ille homo of Rud. 147. We can see
that Plautus did not prefer to make the way smooth by using the form
without the Particle in such line openings as the Dactyls Illic hinc abiit
and Istic hinc abiit (Trochaic line-openings), Illic hinc abscessit (Iambic
line-opening), Nam illic est (Mil. 271 AP). Nor does he secure a
Trochee for the first foot of a Septenarius in Capt. 609 Istic qui vult
vinciatur. Some would pronounce ill'c (like ill! for ille), which suits
e. g. Trin. 998 (Sen.) Postquam illic hinc abiit, etc. (II 52) ; but istic is
less tractable. It would presumably have to be pronounced isc' (like ac).
If istic est, e. g. Poen. 625 (Sen.) :
Istic est thesauru' stultis in lingua situs,
Ut quaestui habeant male loqui melioribus,
is not a Dactyl (by the Brevis Brevians Law), a likelier possibility is
is tics t (like bonus t, etc.), although the reduction of the Substantive Verb
after -c is denied by some scholars, as well as after -n, e. g. Illicinst True.
599 (Cret.) (cf. II 38). Cf. Pseud. 954 (just quoted) illicst; True. 122
and Rud. pr. 79 (where ' illic est ' is impossible ; II 48).
The Nom. Sing. Feminines ilia, ista never lose (II 36) their final -a
(formerly long) as ille, isle lose their final -e (formerly 6). The by-forms
illaec, istaec (often corrupted by scribes to ' ilia haec ', and eventually to
 HIATUS 165

• ilia ' alone or ' haec ' alone) an editor need not hesitate to substitute for
ilia, ista when the metre demands. Skutsch's substitution of a (fictitious)
' illic ' Fern, in Mil. 361 (Iamb.) :
Respicedum ad laevam. quis lllaec est ||mulier ? Pro di immortales !
(Emended to illaec quis est),
is quite improbable.

In the Norn., Ace. Sing. Neuter Mud (never ■ ill'd ' ; II 36) normally,
and istud perhaps invariably, have the form increased by the Particle -ce,
Muc, istuc (20). The shortening of id in idquidem is probably illusory
(e.g. Aul. 637; Asin. 149. Rather Id quidem). When id is shortened,
it is after a ' brevis brevians', e.g. (Men. 141) quis Id coxit coquus?;
(Mil. 633, Poen. 29i)polId quidem. (On Leo's wrong-headed scansions
' id ', ' quod ', etc., before an initial consonant see 10.) Hocce (cf. hocci-
ne) Nom. and the like Abl. form (for hod-ce) seem not to have been used
in conversation in Plautus' time. Nor yet haece Plur. For while they
were admitted by Ennius into Epic (Ann. 234 Haece locutu' vocat. Cf.
O. Lat. hocedie, etc.), they never appear in the Comedies.
34. (B) Genitive. In II 35 we have attempted (but perhaps not
succeeded in the attempt) to class Mi(u)s, isti(u)s with the enclitic forms
/iui(u)s, ei(u)s, of which the disyllabic forms (the forms of classical
poetry) appear 'in pausa' and whenever they had an accent in the
utterance of the sentence.
A full list of the occurrences in Plautus of illius, istius (perhaps never
Dactyls in Plautus and Terence) was given there. Here we add some
examples of the others to the examples there offered :

Mil. 519 (Sen.) Itast ista huius simili' nostrai tua (Huius suits -ai) ;
634 (Troch.) Nam benignitas quidem huius || oppido adulescentul
est (Deliberate
in line) ; utterance ; not merely due to place

it 53 (Troch.) Miles, nil erTerri poterit || huius : nunc quom maxime

(In pausa) ;
Pers. 602 (Troch.) Dixit dare potestatem eius || (At Diaeresis) ;
787 (Anap.) Siquidem hue umquam em' redierit eius |j (Ditto) ;
Poen. 245* (Bacch.) (Item nos sumus,) Eius seminis mulieres sunt
(Emphatic) ;
Pseud. 733 (Troch.) || n(am) hum' mihi debet pater (Deictic) ;
857 (Sen.) Turn ut huius oculos in oculis habeas tuis
and (Emphatic
Deictic) ;
Trin. 7^7 (Sen.) Post adeas tute Philtonem et dotem dare
Te ei dicas, facere id ei(u)s ob amicitiam patris ;
741 (Sen.) Datam tibi dotem, ei quam dares, ei(u)s a p'atre;
166 EARLY LATIN PROSODY

Since the trochaic form of eius is natural at the Diaeresis, ' ei(u)s rei'MSis not
needed at True. 467 (Troch.) : S.);

Bene si facere inceplt, eius || nimi' cito odium percipit (eius enim scito
similarly huius (Deictic) is natural in True. 770 (see below).
There is no strong evidence for a (fictitious) trisyllable 'huius' in the
(apparently) corrupt lines :
Poen. pr. 83 (Sen.) Sed illi patruo huius qui vivit senex ;
pr. 120 (Sen.) Is illi Poeno huius patruo hospes fuit ;
True. 772 (Troch.) Alteram tonstricem huius, || alteram ancillam suam.
Capt. pr. 10 (Sen.) Patri huiusce. iam hoc tenetis ? optumest.
ffui(u)smodz, or rather hui{u)s modi, is naturally a trisyllable, e. g. Poen.
824 (Troch.) :
|| velut ego habeo hunc hui(u)s modi ;
though it may be a quadrisyllable when the Pronoun is emphatic (So scan
Capt. 1033 Huius modi or Huiu' modi) just as quoius{-u') ?nodi (Interrogative)
seems to differ from quoi{u)s modi (Relative); 38. Similarly with ei(u)s
modi. Thus the trisyllable would give an awkward rhythm to Rud. 127
(Sen.), where the emphatic quadrisyllable seems required by the context :
Malum, periurum, palpatorem. Plurimos,
Nam ego propter eius modi viros vivo miser (Or eiu' mod! ; II 48),
unless we are to take ei(u)s-modi-viros for a word-group.
We would scan :
Rud. 198 (Anap.) Sed erile scelus me sollicitat, || eius me impietas male
habet. (Is navem, etc.) ;
200 (Cret.) Haec bonorum eiu' sunt || reliquiae, etc. ;
Trin. 175-176 (Sen.) Utrum indicare me ei thesaurum aequum fuit,
Adversum quam ei(u)s me(d) obsecravisset pater ?;
illic est ;
True. 596 (Troch.) Militem, hie apud me qui erat. || huiu' pater pueri
Men. pr. 34 (Sen.) Pater ei(u)s autem postquam puerum perdidit (But
eius is not unsuitable for a Prologue) ;
Mil. 448 (Troch.) An ista non sit Philocomasiiim, || atque alia ei(u)s
similis siet (with Hiatus at Diaeresis) ;
Since huius is not likely to be emphatic in Poen. 1246 (Iamb.) and since
apud vos (with unemphatic Pronoun) seems natural, this emendation is
suggested :
Quoque modo hui(u)s filias apud || vos (vos) habeati' servas,
since the diiambic ending of the first hemistich may be allowed in a Canti-
cum-metre.

35. (C) Dative. The forms illic, istic were inevitably (if recognized
as Datives) altered by scribes and correctors to illi, isti (I 6). What
nuance, if any, was expressed by the addition of the Particle (like
 HIATUS 167

vulgar English ' that there ') we cannot tell. Mere metrical convenience
(to avoid Hiatus) seems the reason for, e. g. istlc (Dat.) in True. 202.
(contrast ista in Elision in the next line).
If a short syllable such as id, eg(p\ ut, precedes, the enclitic form of
the Demonstrative normally (or rather invariably) begins with a short
syllable. So scan True. 466 (Troch.), with Hiatus ' in pausa ' :
Id illi morbo, id illi seniost, || ea illi miserae miseriast.
This operation of the Brevis Brevians Law is (provided the necessary
condition, a preceding short syllable of this kind) the unmistakable
indication of the Enclitic Demonstrative in any of the Cases, Singular
or Plural. We abstain from collecting every possible exception — e. g.
Pseud. 502, 927 (Cretic) — illusory rather than real, for this practice
tends to focus the reader's attention on exceptions rather than rules and
puts wrong notions into his head. We rather ask him to read any
score of Plautus' pages with a view to this treatment of Demonstratives.
If he once does this he will never again doubt that Plautus' scansion
echoes the everyday (educated) conversation of that early time (and
later).
To the examples cited in II 31 may be added here : II18C);
Capt. 370 (Sen.) Vel ego hue vel Illuc vortar quo imperabitis (Emphatic ;

Mil. 1 1 49 (Troch.) Facile istuc quidemst, si et Ilia || volt et Ille autem

cupit (Emphatic) ;
Poen. 1220 (Troch.) lam hercle ego Illam uxorem ducam et || Iunonem
extrudam foras (A P. Emphatic) ;
Bacch. 164-5 (Sen.) Nimio es tu ad Istas res discipulus docilior
(Emphatic)
Quam ad Ilia quae te docui, ubi operam perdidi
(Emphatic) ;
Epid. 323 (Cret.) Scire cupio. Per Il||lam tibi copiam
Copiam parare aliam licet, etc. (Emphatic) ;
Most. 467 (Sen.) Et heus, iube Illos illinc ambo abscedere (Deictic) ;
Rud. 831 (Sen.) (num molestiaest) Me adire ad Illas propius? Nil
nobis quidem (Deictic).
Examples of the enclitic in such phrases, vel illuc, et ilia, ego illam, ad istas,
ad ilia, per illam, iube illos, ad illas need not be given here. They are the
commonest of common things in Dramatic Verse. So we get a hint that
the true reading in Pseud. 447 (Sen.) is (with P) hie illest, not (with A) hie
illist. For the latter would probably be an Anapaest :
Hie mihi corrumpit filium, scelerum caput;
Hie dux, hie illest paedagogus, hunc ego
Cupio excruciari.
(But see II 12.)
168 EARLY LATIN PROSODY

We would scan :
Capt. 716 (Sen.) Quia Illi fuisti quam mihi fidelior ;
Men. yjy (Troch.) Quid tu tristis es ? quid Tile au||tem abs destitit
te iratus
?;
Pers. 832 (Troch.) N(am) ego nil merui. At enim quod Tile ||meruit, tibi
id obsit volo ;
Since there is as yet no adequate edition of Terence, we forbear to add
suggestions for the true scansion of some of his lines. That Terence's usage
is precisely the usage of Plautus is unmistakable. Examples of the emphatic
Pronoun are : Andr. 535 tun an Illi insaniant; Eun. 965 ne neque Illi prosis
et tu pereas; Hec. 161 (postquam et ipse se) Et Illam et hanc quae domi
erat cognovit satis ; Hec. 599 Et me hac suspicione exsol||vam et Illis morem
gessero ; Ph. 332 Quia enim in Illis fructus est, in || Illis opera luditur.
In lines like Poen. 679-80 (Sen.) the addition of -ce seems intelligible :
Cum illoc te meliust tuam rem, adulescens, loqui :
Illic est ad Istas res probus quas quaeritas (AP) (Or Illicst).
It has correct deictic significance. Similarly in Most. 669 istunc (in an
1 aside' to the audience) is contrasted with istum (spoken to Theopropides)
and means ' the man you see before you '. And probably istlc (istuc MSS.)
in Asin. 702. Old Demipho would be startled when he heard (Merc. 270)
ego Illunc. We would read in Asin. 675 (Iamb.) :
|| illunc te orare meliust (Leonida points to Libanus),
and make the second foot an Iambus (not a Tribrach) in Bacch. 799.
Huic, a are Plautine, along with huic (rare), ei. The latter may be
called emphatic in such a line as :
Alius ;
Bacch. 484 (Troch.) Mihi discipulus, tibi sodalis || periit, huic

?Ter. Heaut. 455 (Sen.) Sensi : nam unam ei cenam atque eius
comitibus (Dedi).
But the use in prologues, e. g. (all Senarii) :
Aul. 13 Agri reliqult ei non magnum modum ;
Cas. 37 Est ei quidam servu' qui in morbo cubat;
? Cist. 138 Feci eius ei quod me oravit copiam ;
Rud. 39 Huic filiola virgo periit parvula,
rather proclaims them as the correct, unslurred forms, the forms of
deliberate utterance. With ei we may compare the Gen. Sing, (r) ret
(e.g. Aul. 121), (2) rei (e.g. Men. 323, 494), (3) rei (e.g. Men. 764a).
Is el Plautine ? The following lines have been cited in proof:
Most. 700 (Cretic) Nam et cenandum et cuban||dumst ei male (?) {A
opposes P) (see IV 26 A) ;
Stich. 653 (Sen.) Salutem ut nuntiaret atque ei ut diceret ;
Rud. 934a (Anap.) Ei ego urbi Gripo indam nomen ;
Pers. 785 (Anap.) Quia ei fidem non habui argenti, etc.
 HIATUS 169

None are convincing. The second and third are rather of the type (53) quo
in loco (Relative), with ei in Prosodic Hiatus ; the last is an instance of the
Brevis Brevians Law (as in Bacch. 45, etc.). So we may scan ei in Pers.
776a (Bacchiac ?), Bacch. 554 (Troch.) :
Ei qui invidet mi || et ei qui hoc gaudet (With Hiatus at Diaeresis) ;
|| ei quod posses mali,
and the monosyllable in Capt. 198 (cf. 18). Bacch. 265, Mil. 720, offer a
variety of choice. In Pseud. 1242 ei (P) seems better than el (A). Read
in Rud. pr. 49 el (II 45). The true version of Aul. 565 is doubtful ; but in
lines like Mil. 93 * el ' is a wrong conjecture. The Spondee suits Cure. 433,
Most. 287, 481, 947, etc.
The Dat. Sing. Fern, is eae1 in Mil. 348, a form which would suit Rud. 934*
(just quoted).
Monosyllabic ei is in most cases indistinguishable from a (fictitious)
Pyrrhic ' ei ', just as monosyllabic huic from a (fictitious) Pyrrhic ' huic '. Still
the common combination (also True. 713 ?, Trin. 1 123 ?) ei rei would, if each
were a Pyrrhic, give that disliked step-child of the Plautine family, a Proce-
leusmatic in Trochaic Verse (II 57), too free play. (That rei Dative is not
a mere scribe's error for ' re ' is fairly proved by this spelling on an inscrip-
tion of 160 B. c.). And other proof of the monosyllabic form is not wanting,
e. g. Trin. 906 quid est ei ; Bacch. 45 Ubi ei dediderit (Anapaest followed by
Tribrach), etc., etc. For Plautine huic we may appeal to the usage of the
classical poets, but not for Plautine ei.
Is it possible to find in ei and ei the same contrast as in eius and ei(u)s ?
And is it possible to class huic with el ? Certainly we must accept el even
where any special emphasis or deliberateness of utterance is not clearly
present : in such lines as :
Bacch. 525 (Sen.) Mendacium ei dixit, nunc me sequimini ;
Aul. 316 (Sen.) Pulmentum pridem eripult ei miluus ;
Epid. 36 (Troch.) Sine perdat ; alia apportabunt || ei Neri filiae.
And as certainly the monosyllable must be called emphatic in such lines as :
? Bacch. 666 (Cret.) Decimam partem ei dedit, || sibi novem abstuiit ;
Men. 646 (Troch.) || huic surreptast, non tibi ;
Capt. 1028 (Troch.) Tibi adimam, huic dem, etc. ;
Cure. 255 (Sen.) Fateor. Abi, deprome. Age tu interea huic somnium
Narra ; meliorem quam ego sum suppono tibi ;
Cas. 410 (Troch.) Tarn huic ioqui licere oportet || quam isti, etc. ;
1 Since this book may be read in countries where Latin verse writing is not
practised and therefore Latin quantities are imperfectly known, it may be well to
say that the e of ea (Indo-Eur. EYA), eae, earn, eum, eo, etc., is a short vowel. The
scansion et (Dat.) apparently represents a pronunciation eiei (eyyei), so that its
opening syllable is not the long vowel e but merely the diphthong ei (like maior,
pronounced ' mayyor ', for mag-yor ; cf. magis) and might easily succumb to a ' brevis
brevians'. Klotz seems to speak of earn, etc., as Spondees (Grundz. p. 130 'earn',
'eum'; p. 169 ' ea ' ; p. 179 'ea')!
170 EARLY LATIN PROSODY

(so huic may be a monosyllable in Pseud. 503, etc.). It appears 'in

pausa ', e. g. :
Rud. 389 (Iamb.) Quia leno ademit cistulam ei ||;
(cf. Bacch. 1098, Cure. 557, Rud. 1292).
So the first question gets a negative answer.
When the similar Interrogative pair, quoii and quoi, come to be treated
(38), we may attempt an answer to the second. Meanwhile we mention that
Terence shows the same usage of el and ei as Plautus. Thus in the Andria
we have ei in lines 106, 390, 484, 641, 878 ; ei in line 443 (Sen.) :
Laudo. Dum licitumst ei dumque aetas tulit (Amavit).
A trochaic line of Pomponius (time of Sulla) has el (Atellan. 177) Ei te
obiectes frustatim ? ||.
36. (D) Accusative. (On illunc, istunc see II 18 C.) Eum, earn, when
monosyllables (with Synizesis, II 33), must not be replaced (e.g. Mil. 560)
by em (im), an old form of the Twelve Tables (cited from them in a speech
of Cato frag, xl 7 si em percussi saepe; cf. Class. Quart. 7, 116). The
forms with Synizesis are very common. The disyllabic forms are, of course,
required not merely when they are emphatic, but also ' in pausa \ In Mil.
243 (Troch.) the disyllable is both emphatic and ' in pausa '.
Meu' conservus, earn vidisse hie || cum alieno osculari, earn
Arguam vidisse apud te || contra conservum meum.
In Pseud. 682 (Troch.) it seems emphatic :
Bene ubi quod scimus consilium || accidisse, hominem catum
Eum esse declaratur, stultum au||tem ilium quoi vertit male.
What of, e.g. Men. 211 (Troch.), 468 (Sen.) ?
Aut sincipitamenta porci||na aut aliquid ad eum modum ;
Non faxo earn esse dices : ita ignorabitur.
When we compare Merc. pr. 107 (where the context approves earn) with the
next line, mod{6) earn seems preferable to mod(o) earn :
Merc. 107-8 (Sen.) Earn me advexisse nolo resciscat pater.
Modo earn reliqui ad portum in navi et servulum.
(Contrast the disyllabic Abl. in Merc. 223 and 411.)
37. Of the other Cases of the Demonstratives only these need a separate
treatment :
Nom. Plur. Masc. Illisce, istisce, hisce are found when the next word
begins with a vowel. (Similarly Dat., Abl. Plur., e.g. Trin. pr. 12 in hisce
habitat aedibus.) They used to be accepted as evidence that -Is was the
Nom. Plur. Masc. suffix of Pronouns (as -I of 2 Decl. Nouns) but are now
ascribed to the addition of the usual Plural -s to the already formed Plural
in -i (like Greek els and « ' thou art '). Plautus never uses this double
suffix without the addition of the Particle -ce (never ' ill]s ' Nom. PI., but
only illi% illisce).
 HIATUS 171

In the Neuter Plural istaec seems to be as invariably used as in the Neut.

Sing. (33) istuc.
Ei (on early inscriptions iei), when monosyllabic, is printed i (e. g. Most.
859), and eis (Pat. Abl. Plur.) similarly is (e. g. Pseud. 1 109), though Plautus
wrote the diphthong (preserved in Pers. 684). The disyllabic forms are to
be expected ' in pausa ', e. g. :
Merc. 869 (Troch.) Non amittunt hi me comites || qui tenent,sunt eiQui ?;
Rud. pr. 73 (Sen.) Sedent eiecti. navis confractast eis ;
Stich. 81 (Troch.) || aetatis spatio c(um) eis (Or ciim is? Cf. 47).
Or when emphatic, e. g. :
Men. 585 (Iamb.) Eis ubi dicitur dies, || simul patronis dicitur ;
But they also seem necessary in Asin. 437 (Iamb.), Cure. 373 (Sen.), Men.
85 (Sen.) :
lam pro eis sati' fecit Sticho ? ||;
Dives sum, si non reddo eis quibu' debeo ;
Turn compediti ei anum lima praeterunt ;
though in Pseud. 819 (Sen.), Rud. 156 (Sen.) 1 homines (a word-group) is
quite possible (38 G) :
I homines cenas ubi coquunt, quom condiunt ;
Ubi sunt i homines, obsecro ? Hac ad dexteram ;
and in Men. 972 an Iambus in the fifth foot would not be irregular (IV 8).
On the other hand 'there is no clear example of i, is encroaching on the
province of ei, eis (unless Pseud. 1112 ?, Bacch. 1185 ?). So we had better
print eis at the Diaeresis in Asin. 269. But we have no right to exclude eis
from Stich. 17, Rud. 186, on the ground that emphasis seems lacking.
The disyllabic Fern. Nom. Plur. rightly appears 'in pausa', e.g. Men. 86
(Sen.) :
Aut lapide excutiunt clavum. nugae sunt eae,
but has less justification in Rud. 646 (Troch.), which should perhaps be
scanned with the monosyllable shortened by the Brevis Brevians Law :
|| audeat. sed eae mulieres (Quae sunt ?).
Ennius' is (eis) Dat. PI. may be mentioned :
Ann. 286 Is pernas succidit iniqua potentia Poeni. (Cf. Ann. 193
Abl.isdem
PI.)

The Gen. PI. of is is normally (like debrum, mebrum, mearum, etc.)

a Spondee, not however in e. g. Asin. 554 (Iamb.) :
(Qui saepe ante in nostras scaplas cicatrices indiderunt)
Eae nunc legiones, copiae || exercitusque eorum.
Dat. Abl. Plur. While the Third Declension (I-Stem) by-form for the
Relative is quibus, it is for the Demonstrative hibus, e.g. Cure. 506 (Iamb.) :
ebdem hercle vos pono et paro : || parissimi estis hibus ;
172 EARLY LATIN PROSODY

ibus e. g. Mil. 74 (Sen.) :

Latrones, ibus dinumerem stipendium.
All the case-forms augmented by the Particle -ce may reduce it to -c
before a consonant (II 36), e. g. True. 638 (Sen.) : .
Ut ego hisc' suffringam talos totis aedibus.
And all the case-forms of is which admit Synizesis follow the rules given
in II 33. Thus no exception can be taken to iambic ea (Abl.) of A in
True. 278-9 (Troch.) :
Cumque ea noctem in stramentis || pernoctare perpetem
Quam tuas centum cenatas || noctes mihi dono dari (Or mi),
since ea and tuas are contrasted and therefore emphatic. But if the same
form is accented in Trin. 594 (Sen.), it must get its accent from the
Enclitic re :
In ambiguo est etiam nunc quid ea re fuat.
In Trin. 623 (Troch.) iambic eos is natural at the Diaeresis :
Nescioquid non satis inter eos || convenit, etc.
So scan True. 193 (Iamb.) :
Heia! haud itast res. Ain t(u) earn || me amare? Immo unice unum.

38. Quis, qui, etc. (A) The discrimination of quis Pron. and qui
Adj. was later than Plautus' and Terence's time (Details in Berliner
Philologische Wochenschrift 13, 278), e.g. Ter. Phorm. 129 qui fuerit
pater. This allows us to evade the necessity of scanning ' siquis' (like
slquidem ; II 37) in Stich 182 (Senar.) :
Nulli negare soleo siqui [me] essum vocat (siquis essum A : siquis
essum meP.)

The other alleged examples of ' siquis ', ' nequis ' seem also illusory
(Amph. 391; Aul. 340; Capt. 791; Cist. 531; Epid. 339 and 729;
Men. 556; Stich. 67; Vid. 19; Ter. Andr. 258; Phorm. 643). In fact
Most. 86 (Bacch.) suggests that 'siquod' was unknown :
Ego atque in meo cor||de, si est quod mihi cor.

' N urn quid ' (Merc. 282 ; cf, II 51 ; delete me in Men. 548) is as im-
possible asLeo's ' nequld tibi ' (!) Epid. 339. (See also below, on quis-
que) There is stronger evidence (but hardly quite strong enough) of
1 e(c)quis ' (like equidem, e. g. Epid. 603 ?). Some editors alternate
ecqui and ecquis in Bacch. 581 sqq. (Sen.) :
Fores pultare nescis. ecquis in aedibust ?
Heus ! ecquis hie est ? ecquis hoc aperit ostium ?
Ecquis exit? (cf. Amph. 1020; Capt. 830; True. 664).
 HIATUS 173

If 'ecquid ' were better established, the evidence would be irresistible

(Bacch. 980 ; Pers. 107 ; Pseud. 740; Trin. 717). Or if ' ecqu(am) ' in
elision were found. (In Cure. 131 put Ah ! 'extra metrum '; Pers. 225
is iambic.)
(B) The old declension showed : Nom. Sing. Masc. quis or qui ;
Fern, quis (Interrog., Indef.) and quae (Rel.), e. g. Pers. 200 (Troch.) :
Illic hmc abiit intro hue. sed quis || haec est quae me adversum
, incedit ? ;
Abl. Sing, qui and quo ; Nom. Plur. ques (Interrog., Indef. ; e. g.
Pacuv. 221 ; apparently replaced by qui in the conversation of Plautus'
time) and qui (Rel.) ; Dat, Abl. Plur. quis (in old spelling queis) and
qulbus (contrast Ibus, hibus).
(On the improbability that the monosyllables quis, \s ever dropped -s,
see 16; 33. In the disyllable Ennius has (Ann. 536) siqui' ferat. On
Leo's wrong scansion ' quid quod ', etc., see 10.)
(C) While our attempt to discriminate the varieties of is, hie, etc.,
was hardly convincing, the distinction between this enclitic and accented
Pronoun is easy. The Interrogative was accented, the Relative (and
Indefinite) enclitic, (though it might assume something of an accent
before another Enclitic), just as in Greek t« Interrog., (oo-)tis Indef.,
or in English ' who is c6ming ? ', ( the man who is c6ming \ So elide
the Relative, but not the Interrogative in Trin. pr. 6-7 (Sen.) :
Nunc igitur primum quae ego sim et quae lllaec siet (Or Illaec
qu(ae)?)

Hue qu(ae) abiit intro dicam, s(i) animadvertitis.

(D) Quisgue, (1) 'whoever', (2) 'each', is sometimes altered by

scribes to guisguis. Read in

Trin. 218 (Sen.) Und' quidque auditum dicant; nisi id appareat

(quidquid AP) ;
Stich. 686 (Troch.) Quisq' praetereat, comissatum || volo vocari.
Convenit (quisquis AP).
It was of course a quite different word from quisquis, its second part
being the same -que as in quandeque (=quandocumque).
J (E) Nescioguis (cf. nescioquoia vox Merc. 864) is discriminated from
nescio guis in classical Latin (II 22). But not invariably by Plautus,
e.g.:
Bacch. 107 (Troch.) Simul huic nescio cui turbae || quae hue it
decedamus hinc (cf. IV 2) ;
Men. 406 (Troch.) Nescio quern, mulier, alium || hominem, non
me quaeritas ;
174 EARLY LATIN PROSODY

? Merc. 365 (Troch.) Sollicitus mi nescio qua || re videtur. Attatae

(cf. 31) ;!

Trin. 880 (Troch.) Multa simul rogitas : nescioquid || expediam

potissimum (cf. II 55) (cf. Poen. 856 ;
Amph. 354; Merc. 723). Nor by Terence,
e. g. Andr. 734 (Sen.) :
Nescioquid narres. Ego quoque hinc ab dextera. (Notice Aul. 71.)

(F) Gen. Sing. Cuius (the trochaic accented form and form of deliberate
utterance) and cui(u)s (an enclitic monosyllable, e.g. Lucilius 1039) have
been discussed in II 35, e. g. :
Pers. 386 (Sen.) Quoiu' modi hie cum mala fama facile nubitur ? (Or
Quoius),
the Interrogative, whereas the usual form (with the Relative and even, e. g.
Rud. 8^, Most. 817) is quoi(u)smodi, a trisyllable. Philologists dispute
whether quoivismodi (Pseud. 741, Bacch. 400) is a mere form of quoi(u)svis
or quoi(u)s modi (like divello for disvello, dimoveo for dismoveo) or points
to a rival Genitive with the -I of the Second Declension (or rather O-stems).
Cf. istimodi True. 930 (II 35). In Men. 577 the Interrogativejjhrase quoius
mod? clueat (quoiu' ? modi ?) is a Colon Reizianum (IV 15).
Notice the disyllabic Gen. of quisque in Ter. Heaut. 284 (Sen.) :
Quae cui(u)sque ingenium ut sit declarat maxime,
like the Gen. of the Relative (with que) in Ter. Hec. 478 (Troch.) :
(matri meae) Cui concedat cui(u)sque mores || toleret sua modestia.
(G) Pat. Sing. Like huic and huic, el and ei (35) is the Plautine pair
quoii (Spondee) and quoi (monosyllable).
We have seen that huic is emphatic in such lines as Bacch. 640 (Troch.) ;
True. 442 (Sen.) :
Hunc hominem decet auro expendl, || huic decet statuam statui ex auro ;
Me potiu' non amabo quam huic deslt amor.
Similarly the Interrogative (emphatic), equally with the Relative (enclitic),
appears as a monosyllable, e. g. :
Pseud. 1203 (Troch.) || servo? quoi servo? Syro. (Cf. the common
quoi rS ' why ? ', e. g. True. 394.)
Like huic homini is, e. g. Men. 473 (Sen.) :
Pro di immortales ! quoi homini umquam uno die (Boni dedistis plus ?)
(Cf. Capt. 973 ; Most 948 ; Trin. 604 ; Ter. Hec. 431), for an emphatic mono-
syl able isnot elided before a short syllable ; 53 ; and we may retain the
same phrase with the same scansion, even with the Relative or Indefinite, e.g. :
Cure. 557 (Troch.) Quoi homini di sunt propitii, ei || non esse iratos
puto (cf. Cas. 258) ;
 HIATUS

531 (Iamb.) Quoi homini di sunt propitii, || lucrum ei obiciunt75;

profecto

Rud. 1 1 93 (Sen.) Satin siquoi homini di esse bene factum volunt ;

since in the word-groups, qui homo, hie homo, tu homo, etc., the Noun seems
to have been an Enclitic, and the phrase was pronounced as a Fourth P^pnn .
'
(cf.But53).
the Enclitic quoi is elided in such a line as :
Trin. 106 1 (Troch.) Emere meliust quoi imperes. Pol e||go emi atque
argentum dedi (Imperes is emphatic)
(Cf. Aul. 75, 420 ; Accius praetext. 30) ; though it must remain in Prosodic
Hiatus before an iambic word (46), e.g. :
True. 501 (Troch.) Quoi adhuc ego tarn mala eram monetrix ||.
If both A and P have wrongly substituted qui in Pseud. 703 (Troch.), we
have a good contrast of the elided Relative with the Interrogative :
Io te, te, tyranne, te, te || ego, qu(o)i imperitas Pseudolo,
Quaero quoi ter trina triplicia, || tribu modis tria gaudia.
Lastly, another situation where quoi need not be altered is at the Diaeresis
in Hiatus, e.g. Cist. 695, Men. 623, since the monosyllable quoi (like the
monosyllables huic, ei) is common at the Diaeresis and since Hiatus at the
Diaeresis is a common licence.

(H) Neither quoii nor huic seems to be employed except in special cir-
cumstances. Afull list of the (real or supposed) occurrences of both will be
useful.
(a) Quoii.
Amph. 520 (Troch.) Quoii ego iam hoc scipione. Ah || noli. Muttito
modo (Spoken by Juppiter) ;
861 (Sen.) Ego sum ille Amphitruo quoii est servus Sosia (Ditto) ;
Asin. 459 (Iamb.) Suscenseat, quoii omnium || rerum ipsu* semper credit
(Or quoi is om.) ;
yyj sqq. (Sen.) Neque quom descendat inde det quoiquam manum,
(Or ind' det quoiiquam ?) ;
Spectandum nequoii anulum det neque roget,
Talos ne quoiiquam homini admoveat nisi tibi
(Articles of the contract) ;
Men. 498 (Sen.) Cur ausu's facere, quoii ego aeque heres eram ? (Or
quoi ego ae. Emphatic ego) ;
Merc. 615 (Troch.) || quoii est empta? Nescio (But the insertion of ea
seems necessary) ;
Trin. 558 (Sen.) Siquem reperire possit quoii os sublinat;
{b) Huic.
Amph. 702 (Troch.) Etiam tu quoque adsentaris || huic ? Quid vis fieri ?
Bacch. 484 (Troch.) Mihi discipulus, tibi sodalis || periit, huic filius
(Spoken by trie tutor) ;
176 EARLY LATIN PROSODY
764 (Sen.) Nam non conducit huic sycophantiae ;
Capt. 364 (Sen.) Nam ego te aestimatum huic dedi vigintiae.minis (Or;
te huic)

Men. pr. 13 (Sen.) Huic argumento antelogium hoc fuit (?) ;

pr. 40 (Sen.) Immutat nomen avus huic gemino alteri (Spondee-
word in fourth foot ! Cf. Mil. 1381) ;
Most. 870 (Bacch.) Si huic imperabo, probe tectum habebo (Or Si huic ;
47);
Poen. 395 (Troch.) Ut tu huic irata ne sis (AP) (Or Ut tti huic ; 47) j
1036 (Sen.) Maledicere huic tu temperabis, si sapis (A and Codex
Turnebi omit tu) ;
Pseud. 327 (Troch.) Victimas, lanios, ut ego huic || sacrificem summo
Iovi (Or ego huic ; cf. line 334) ;
775 (Sen.) Nunc huic lenoni hodiest natalis dies (Or huic lenonist
hodie ; A n. /.) ;
Rud. pr. 39 (Sen.) Huic filiola virgo periit parvuja ;
750 (Troch.) Nam huic alterae quae patria || sit profecto nescio
(Emended to Namque huic) ;
1342-3 (Sen.) Turn ego huic Gripo, inquito et me tangito.
Turn ego huic Gripo (dico, Venus, ut tu audias)
(In administering an oath) ;
While el is common, this pair is rare, almost to vanishing point. (So scan
Asin. pr. 10 with Hiatus 'in pausa\) What verdict should be given upon
this evidence ? Opinions will differ. We ourselves feel the most credible
witnesses to be (for_quoii^ Asin. 777 sqq., (for huic) Rud. 1342 sq. (and
perhaps Bacch. 484, Rud. pr. 39), and pronounce the two forms to belong to
legal, old-fashioned, precise speech. So far as the unedited state of Terence's
plays allows a pronouncement, this old pair had become obsolete by his time
(cf. Heaut. 540 ; 81 ; 685 ; Ph. 363), but Caecilius, who belongs to the genera-
tion between Plautus and Terence, offers a possible example (in a line about
the ' deus summus ') :
com. 261 (Sen.) Quoii in manu sit, quern esse dementem velit, (Or
Quoi ?)
Quem sapere, quern sanari, quern in mortem inici.
A writer of Atellanae in the time of Sulla, who uses many dialectal forms, is
thought to have used cuii_\
Pomponius 146 (Sen.) Evenit, cuii amicus est germanitus.
If he did, we may suppose the old form to have survived longer in the
country than in the capital.
The supposed O-stem Dative ' quo ' seems to be merely due to scribes'
misunderstanding of quoi (e.g. Amph. 520).

(I) The ' Third Declension ' (or rather I-stem) Ablative qui (e. g. qui fit
ut ? * how comes it ? ') would be accented when an Interrogative, but un-
accented when a Relative. It became a mere Particle in phrases like hercle
 HIATUS 177

fm\ utinam qui (cf. classical atqui, in which sense Plautus rather uses clique),
and is elided, e.g. in Pseud. 473 (Sen.):
Mirum id videtur ? Hercle q(ui), ut tu praedicas.
Whether the Relative is elided in, e. g. Pseud. 487 (Sen.) :
Ita, quas meb gnato des, qui amicam liberet,
is impossible to say (53). But this line shows how the word had become a
mere stereotyped ' wherewith ', since qui here plays the part of quibus.
Certainly the Interrogative is not effaced, e.g. Stich. 91 (Troch.) :
|| vostri. QuT, amabo, pater ?
39. Ipse, Idem. Alius. While ipse in classical Latin declines only
the second part of the compound is-pse (is-pe-se?), the first part was
often declined in the older language, eapse, (and eapsa ?) eumpse, (and
eumpsum?), eampse (and eampsam?), and so on. Synizesis (II 33)
frequently reduced them to disyllables eapse (even reapse for re-eapse,
True. 815), eiimpse, etc., e.g. :
Pseud. 833 (Sen.) eaepsae se patinae fervefaciunt ilico (Or eaeps'
sese ; see below) ;
Rud. 1 1 78 (Sen.) Nam haec litteratast, eapse cantat quoia sit.
Contrast True. 133 (Iamb.) :
Quia te adducturam hue dixeras ||eumpse, non eampse.

I
The old Nom. Sing, by-form ipsus (normal in Plautus with se. sicbi) is
said by linguists to differ from the other as Homer's os from 6, the -sus
being I.-Eur. sos (whence the archaic Ace. Sing, sum-fise in True. 160;
a quotation from an older writer ? Cf. sapsa Pacuv. 324) and the -se
I.-Eur. so. Even when its final s was faintly pronounced, ipsu} was thus
a different form from ipse, e.g. Pseud. 929 Ipsu' sese (for a Molossus is
never seen in this Cretic passage). Ips' (like ill', ist' ; II 36) was the
pronunciation of ipse when closely joined in pronunciation with a word
beginning with a consonant.
In Latin, it should be noticed, this pronoun was often a Sentence-
Enclitic and liable to the Brevis Brevians Law, e. g. :
Poen. 664 (Sen.) Ut quidem ipse nobis dixit, apiid regem Attalum ;
669 (Sen.) Immo, ut Ipse nobis dixit, quo accures magis ;
Amph. 415 (Troch.) Et Ipsus Amphitruo obtruncavit || regem
Pterelam in proelio (Emphasis on Amphitruo apparently).
But in Pseud. 641 (Troch.) we should read quam (cf. Ellis on Catull. 9,
32) with P, not quasi (qu&m si A), for ipsi is contrasted with mi/ii &nd
emphatic :
||verum si dare vis mihi,
Magis erit solutum qu(am) Ipsi ||dederis, etc.
2348 N
 178 EARLY LATIN PROSODY
It is emphatic too a few lines below (644, Troch.) :
Ego nis(i) ipsi Ballioni || nummum credam nemini (AP).
Also Rud. 1260 (Sen.) :
Et ipsum sese et ilium furti astringeret.
Idem. In the paragraph on Synizesis (II 33) the common scansions
ebdem, eundem, etc., have been sufficiently illustrated. Contrast the
trisyllabic, e.g. in one of the Cistellaria's prologues, line 178 (Sen.) :
Eandem quam olim virginem compresserat.
Since the Adverb is normally eadem (scil. opera) we should scan with
Hiatus at the pause :
Stich. 438 Mi hanc occupatum noctem : eadem symbolum.
(In Bacch. 60 the addition of et is suspicious. Cf. Niemoeller De ipse
et idem, p. 53 ; Class. Rev. 17, 298.)
If ei and ei are the only Dative forms (35), we must presumably find
eidem in Pseud. 668 (Sen.) :
Puero, atque eidem si addunt turpitudinem (Hardly Nom. Plur.) ;
and eidem in Asin. 120 (Sen.) :
1231).
Eidem homini, si quid recte curatum velis, (Mandes) (cf. Lucilius

And if we print i for ei (Nom. Plur.), is for 8s (Dat., AbLPlur.) we

must present Asin. 342 (Troch.), Amph. pr. 55 (Sen.) in this form :
Teneo, atque idem te hinc vexerunt ||vinctum rus. Memor es probe
(Emphatic te) ;
Comoedia ut sit omni[bu]s isdem versibus ;
and Mil. 758 (Troch.) in this :
Sed eidem homines numquam dicunt ||.
Alius* It is difficult to believe (with Skutsch in Glotta 2, 154) that
the by-forms a/is, a/id were not older than the time of Catullus (66, 28)
and Lucretius (1, 263). But there is no clear trace of them. So editors
challenge the traditional reading in Lucilius 619 (Troch.) :
Curet aegrotum, sumptum homini ||praebeat, genium suum
Defrudet, all parcat.
Plautus has a/ii Dative (e. g. Mil. 1076, 1357), a/iae Dat. Fem. (Mil. 802).
40. PROSODY OF VERB.
(A) Shortened Forms. The shortening is usually due to one of
three causes :
tr- (t) Suppression of a syllable similar to a neighbouring syllable, as in
por(ti)torium, ido(lo)latria. Hence dix(is)ti, etc. Perhaps this is the
explanation of amatust, etc. (II 38).
 HIATUS i79

(2) Suppression of v between vowels, as in di(vi)tior, si(vi)s, pro(vo)rsus

(24). Not all vowels; e.g. adnumeravlt could never produce 'ad-
numerat' or 'adnumeraut' (as Leo scans Asin. 501; 24). Hence
audi(vi)stis, etc.
(3) Syncope, as in ar(i)dus (26). When the short vowel is preceded
by v, this type is not always distinguishable from n° 2. Hence por(ri)go,
sur(ri)pui, etc.
These shortenings are the slurred pronunciations of everyday life.
They are far more prevalent in Dramatic Verse (especially Comedy)
which echoes talk than in e. g. Epic. The first type we may illustrate
by the Plautine examples given in Class. Quart. 1, 48 and 106 :
abduxti Cure. 614 (cf. Rud. 862); adduxtin Capt. 1016 ; adduxe
Rud. 1047 ; admisse Mil. 1287 ; advexti Merc. 390 ; advexe Merc. 333 ;
conscripsti Asin. 746 ; depinxti Poen. 11 14 ; despexe Mil. 553 ; detraxe
Trin. 743; detrusti Aul. 335; devinxti Asin. 849; discesti Asin. 251 ;
dixti (passim) ; dixe Amph. fr. vii, Poen. 961 ; emunxti Most. 1 109-10 ;
immersti Bacch. 677 ; inlexe Merc. pr. 53 ; instruxti Mil. 981 ; intellexti
Rud. 1 103; intinxti True. 294; intromisti Aul. 553; iusti Men. 1146;
occlusti Trin. 188; promisti(n) Cure. 705, 709; scripsti Asin. 802;
traduxti Cas. 579.
Mediaeval scribes (and correctors) were apt to find in e. g. dixti the
common error of omitting a syllable and corrected the supposed faulty
transcription to dixisti, thereby effacing all trace of the ancient spelling.
But sometimes they misread it as dixit a mistake which reveals to us
that dixti stood in their original. Indeed it is quite likely that Roman
scribes themselves were not always careful to present a word in the form
of its pronunciation and might write dixisti for dixti, as surrupui (-rip-)
for surpui, and possibly siet for sit (see below). Editors therefore need
not hesitate to save the metre by substituting the shortened form. A
good example is found in the minuscule transcription of the Trinummus,
where the full form dixisti appears at 556, 567, 602, etc., lines in which
dixti is necessary. In the last of the three the Palimpsest enters the
witness-box and testifies to dixti :
602 (Troch.) Quomodo tu istuc, Stasime, dixti} || nostrum erilem
filium.
The full and the shortened form appear in one and the same line :
Merc. 658 (Troch.) lam dixisti ? Dixi. Frustra ||dixti. hoc mi
certissimumst.
Of the second type full details belong to a work on Accidence rather
than Prosody. (They will be found in Brock, Quaest. Gram, ii.)
N 2
180 EARLY LATIN PROSODY

Roughly speaking, the shortened forms have sole supremacy in Terence's

dialogue (though the v-forms sometimes appear through metrical
necessity at the end of a line), whereas there is still a rivalry in Plautus
between -averim and -arim, -ivissem and -issem, -averunt and -arunt, and
so on. Thus Plautus uses novisti, rarely nosti ; Terence nosti, rarely
novisti. All this is so perfectly in keeping with a language's change in
a single generation as to strengthen our confidence in the text-tradition
of both Comedians. Some scholars deny -1 (for -ivi), -it (for -ivit) to
Plautus, but probably unreasonably, e.g. :
Bacch. 1 7 (Troch.) ||peri, harundo alas verberat ;
Cist. 286 (Sen.) I curre, equum adfer. Peri hercle, hie insanit
miser.
Rud. 325 (Iamb.) Data verba ero sunt, leno abit ||scelestus exulatum.
These forms abit, etc. never become Pyrrhics, because their final is
accented. Indeed their ictus does not seem to clash with their
accentuation, e. g. :
Asin. 395 (Iamb.) Quom venisset, post non redit ? || Non edepol.
quid volebas ? ;
Bacch. 950 (Iamb.) Dolis ego [dejprensus sum, ill' mendi||cans
paene inventus ducere
interit ;;
Men. 450 (Troch.) Atque abit ad amicam, credo, ||neque me voluit

Mil. 416 (Iamb.) Haec mulier, quae hinc exit modo, concubina
est||ne erili'
?;
Rud. 325 (Iamb.) Data verba ero sunt, leno abit exulatum
||scelestus ;

and (if this be the right reading) Mil. 376 (Iamb.) (unde) Exit haec
[hue] ? Unde nisi domo? || (cf. Asin. 742 ; Bacch. n 15 ; Merc. 705 ;
True. 884).
The metre does not always enable us to discriminate -it and -iit. In
Mil. 251 (Troch.):
Non domist, abiit ambulatum, ||dormit, ornatur, lavat,
a pyrrhic abit is unnecessary. In Rud. nil (Troch.) Quibu'cum periit
parva Athenis ||(pa. pe. MSS.).
A few of the more bizarre forms are worth mention :
devoroy Accius praetext. 15 (Troch.) :
Patrio exemplo et me dicabo at||que animam devoro hostibus.
(Not Nonius, but rather mediaeval scribes should be charged with the absurd
item which accompanies this citation in our MSS. of Nonius 98, 10 Devorare :
absumere, eripere.)

\
 HIATUS 181

decrero, e.g. Cure. 703 (Troch.) :

Siquidem vultis quod decrero || facere. Tibi permittimus.
insueram, Capt. 306 (Troch.) :
Qui imperare insueram, nunc al||terius imperio obsequor.
siritis , e.g. Poen. 953 (Sen.) :
Reperire me siritis, di vostram fidem. ^
nomus, Ennius trag. 138 (Troch.) :
nos quiescere
Aequum est ? nomus ambo Ulixem ||.
commorat) Turpilius 30 (Sen.) :
Commorat hominem lacrimis, etc.
Hardly ' adiiiro * (rather adiuero; cf. Catull. 66, 18), Ennius Ann. 335
Ter. Ph. 537 (Troch.) :
O Tite, siquid ego adiuero curamve levasso (Donatus attests adiuto);
Fieri miserum, qui me dudum, ut || dixti, adiuerit comiter.
The Persons.
(B) Third Person Singular Active. Since the shortening of a long
vowel before -t was never (or rarely) recognized by Plautus (unlike
Ennius, 18) the 3 Sing, of any Tense whose 2 Sing, ended in -as, -es, -Is
must in his plays show the ending -at, -et, -it. We must therefore scan
curat (like curas), etc., possibly even dat (like das ; see 42) ; also -it in
3 Sg. Perf. (18). Since -it of Fut. Perf. Ind. was (like -it of Fut. Ind.)
originally short, editors regard the second of the two following doublets
as the ■ Revival ' version (i. e. substituted at a later revival of the play),
since Virgil's error (egerlmus Aen. 6, 514) can hardly be ascribed to
Plautus :
Trin. 788-9 (Sen.) Sed epistulas quando obsignatas adferet,
[Sed quom obsignatas attulerit epistulas] (cf.
II 9; HI 48).
(In True. 344 obtigerit Perf. Subj.? See also Pseud. 100; Poen. 213,
both with -is and presumably Perf. Subj.)
(C) Third Plural Perfect Active. The relation of -erunt to -ere is
obscure. The Roman Grammarians declared -ere to have been originally
Dual. The theory that holds the field at present is that the two original
forms were -erunt (from -Isont) and -ere, and that -erunt was a subsequent
blend. The form -erunt seems to be found only at the end of the line
(or hemistich) in Dramatic Verse (Poen. pr. 2 1 ?), and therefore was
used only under metrical necessity, e. g. True. 468-9 (Troch. Septenar.) :
Nimi' quam paucae sunt defessae ||male quae facere occeperunt, 16).
Nimi'que paucae efficiunt si quid ||facere occeperunt bene (cf. IV

(D) Second Singular Passive or Deponent. Comparative Philology

182 EARLY LATIN PROSODY

has established that -so the Indo-European ending of Greek i<ficpe(<T)o

Imperf., <£e/oe(<r)o Imperat. became in Latin -re. For example sequeso,
the Greek ?7reo, €7rov, became in Latin sequere, used as Indie, or Imperat.
of sequor. In course of time, in order to distinguish the Indie, from the
Imper., the common Active 2 Pers. Sing, suffix -s came to be added
(just as it was added in Greek to cT, for &rt, ' thou art ', and produced
cts), and sequeres (naturally) became sequeris. In dialects where the -6
of -so had not sunk to -e this fuller ending was not -rls but -rus (e. g.
sj>atiarus, Beneventum ; utarus, Venusia ; figarus^ Pompei). And this
pronouncement of Comparative Philology still stands in the last authori-
tative book (the second edition of Sommer's Handbuch, 19 14). It is
confirmed by the evidence of the Dramatists, which shows that the
addition of -s, which changed -re into -ris, was an innovation of their
time. In Plautus -ris is the exception (required by the metre only in 6
lines out of 234 examples of 2 Sing., according to Class. Quart. 1, 42) ;
Terence hardly recognizes it (at most, in 2 out of 54 examples Phorm.,
ed. Hauler4, p. 71); even Cicero favours it only in the Present Indica-
tive (not the Subjunctive), where the necessity for distinction from the
Imperative was most pressing. To declare -re to be a later form of -ris
would be contrary to history. And Comparative Philologists know that
if -\s had become -e in one part of the verb, it would have become -e in
all ; e. g. legitis would have become legite, legis would have become
lege, amabis, ■ amabe ', and so on. It was reserved for Leo, who ignored
Comparative Philology, to take this rash step and declare seqmre to be
a mere development of segueris. His theory is discussed in paragraph 16.
Tenses.

(E) Perfect. The relation of Plautus' Latin to Terence's is illustrated

by the Perfect of fero. The older form tetuli gradually became tuli
under the influence of Compounds like ret(e)tuli, de(te)tuli, con(te)tuli.
Plautus has always tetuli (also attetuli Aul. 433 ?), though tuli appears at
the end of the line, in Poen. 1067 (Sen.) :
An mortui sunt ? Factum, quod ego aegre tuli (cf. Cure. 644).
In Terence's earliest play (a score of years after Plautus' death) we find
tetuli :
Andr. 808 (Sen.) Nam pol, si id scissem, numquam tetulissem
pedem ;
832 (Troch.) Impetrasti : incepi dum res ||tetulit. nunc non
fert, feras ;

but elsewhere in Terence always tuli. Catullus' tetulisset in the Coma

Berenices (66, 35) is a deliberate archaism.
 HIATUS 183

(F) The Perfect ending -ui was originally -ui (-uvi). The older form
is not rare in Plautus, e. g. Capt. 555 (Troch.), Most. 86 (Bacch.) •
Quibus insputari saluti ||fuit ; atque is profiiit ;
Argumentaque in pec||tu' multa institui ;
but is naturally commoner in Ennuis' Epic3 e. g. :
Ann. 133 Adnuit sese mecum decernere ferro;
377 Nos sumu' Romani qui fuimus ante Rudini.
Moods.
(G) Imperative. A Second Person Singular might drop its final e in
rapid utterance (see 42 s. v. dico). Plautus' rule seems to be that dice,
etc. are used before a vowel or ' in pausa ', die, etc., when closely joined
with a word which began with a consonant (see 42), e.g. Rud. 124
(Sen.) :
Tu, siquid opus est, dice. Die quod te rogo.
So we must read in Pseud. 488 (Sen.) :
etc.).
Fatere, die kcu tovto vai. Kat tovto vcu (Not ' Fatere, die. Kcu ',

Who can refuse to find in all this an echo of the everyday pronunciation
of Plautus' time and precisely that anticipation which we should expect
to find in the second century b. c. of the classical usage ? Cicero tells
us that cape si vis was pronounced capsis (Or. 154; cf. Quint. 1, 5, 66),
and his story of Crassus shows that cave ne eas sounded like Cauneas
(Div. 2, 84). Catullus' inger (27, 2) clearly is table-talk. Fortasset/
linguists tell us, is Imperative of an S-Tense (cf. amasso) of a lost verb (
* fortare ' (cf. ad-firmare) ' to affirm ' and meant literally ■ do you affirm ',/
while fortassis was the Indicative ' you will affirm ', ' one will affirm '. If
this explanation be right, we can set in this paragraph the (apparent)
dropping of -e in :
Merc. 482 (Sen.) Sequimini. Fortass' te ilium mirari coquum
(Hardly Sequimini, a Proceleusmatic ; II 9 ; 27) ;
Ter. Heaut. 715 (Troch.) Tu fortass' quid me fuat parvi || pendis
dum illi consulas (fiat MSS.).
Some editors (cf. Rhein. Mus. 49, 286) substitute fors (as in Amph.
726).
(On the reduction of Iambic Imperatives to Pyrrhics, cave, abl, etc.
see II 27 D [1].).
Conjugations.
(H) The io-Verbs of the Third Conjugation are a puzzle to linguists.
Some think it was the Brevis Brevians Law which (at that early time
when the accent fell always on the first syllable) pushed cupio, facio,
i84 EARLY LATIN PROSODY
etc., out of the Fourth. Certainly Plautus often treats them like flnio,
etc., e. g. :
Amph. 555 (Bacch.) Facis ut tuis nul||la apud te fides sit;
True. 555 (Bacch.) Domist qui faclt im||probe facta amator ;
Cure. 364 (Troch.) Laudo. Laudato quando illud || quod cupls ;
effecero

Men. 921 (Troch.) ||perciplt insania ;

and Ennius, e. g. :
Ann. 391-2 Nunc est ille dies cum gloria maxima sese
Nobis ostendat, si vivimu' sive morlmur.
(I) Sum. Second Singular Present. In the Indicative Mood
originally es-sl (Homer's eo-<n), which appears in Plautus as ess (printed
es). Similarly in the Imperative, ess (printed es), another long syllable
(20). (On the forms bonust, etc. for bonus est, etc. see II 38.)
Perfect. Fui and fui (e.g. Pers. 168) (see above, 40 F).
Subjunctive. The Indo-European Optative had -ie- in the Singular,
but in the Plural -1-, which would give in early Latin siem, sies, siet,
simus, sitis, sient (this last with short e). In course of time the -1- of the
Plural found its way into the Singular, sis, etc. When sies, etc. appear
in the MSS. where a monosyllable (or two morae) is required (details
in Brock Quaest. Gram. 84), editors print sis, etc. This is convenient
for readers ; but is it correct ? An enclitic Verb would be especially
liable to Synizesis, and factu'-siet is perhaps the more likely interpreta-
tion of the traditional spelling than factu'-sit (see II 33). Even in
Cicero's time siet was still in use (Or. 157 'siet' plenum est, 'sit ' immi-
nutum. licet utare utroque). To alter siet in Ter. Hec. 637, where three
morae are needed, merely because it stands inside (and not at the end
of) the verse, seems unreasonable. On the other hand, in the final word
of a line (e. g. Ter. Hec. 334 auctu siet) Synizesis is open to challenge,
though even in this position it may be defended in an appended enclitic
(59, end).
(K) Possum. It potis is an Adj. (cf. potior Comparative) we should
expect the word-group to have originally varied in Gender, potis est
Masc, Fern., pote £tf Neut, until one of the rivals gained predominance,
all sense of the verb's Gender having been lost. That was the old
explanation, and we doubt whether it has, after all, been overthrown
by Leo's theory that potis was pronounced (like the Neuter) pote (not
poti') before a word beginning with a Consonant or with a Vowel ;
though certainly the old spelling similest, etc., for similis est, etc., shows
that potest may represent potis-est Masc, Fern, as well as pote-est Neut.
 HIATUS 185

However the difference between the two theories is slight, in fact

negligible to every one but a mere linguist (16). Here we may content
ourselves with insisting on the parallel between potest, poterat, potesse,
etc., and sat est, sat erat, sat esse, etc., and with calling attention to the
ignoring of Gender and Case in Terence's (Ph. 379) Si tibi placere potis
est, mi ut respondeas ; Plautus' (Ps. 1302) Credo equidem potis esse
te, scelus ; (Pers. 30-3 1 ) si tu tibi bene esse Pote pati, venl : vives
mecum.
To give statistics of the occurrences of the by-forms potis and pote
would be a difficult task, but fortunately it lies outside the province of
this book. The vagaries of MSS. are very baffling. Since pote some-
times istreated as an abbreviation and expanded to potestas (as volup to
voluptas), we get a clue to the emendation of Ter. Andr. 52 (cf. 24).
If we could be sure that potisit of the Senatus Consultum de Baccha-
nalibus (two years before Plautus' death) represented what we printj
potV sit rather than what we print potis sit, we should have no hesitation!
in refusing to substitute pote for a pyrrhic (preconsonantal) potis of the!
MSS. The Subject is Feminine (tabula) : uteique earn figier ioubeatis|
ubei facilumed gnoscier potisit.
(L) Malo, nolo. Magi' (mage) volo produced mag-volo which be-
came mavvolo (printed mavolo). In Plautus both mavolo and malo,
mavelim and malim are used (see Solmsen, Stud. Lautgesch. 55).
Terence affects the later forms, though mavolo appears in Hec. 540
(at the end of the line) || quam ipsam veram mavolo (but see 42 s.v.
Magis).
Ne-volo produced novolo which became nolo. For class. Lat. nonvis,
nonvult we find in early Latin nevis, nevult; but Plautus apparently
turns to these early forms under metrical necessity, e. g. at the end of
a line or hemistich. (Details are given by Seyffert in Bursian's Jahres-
bericht, 1894, p. 320.)
(On aio, coepi, see 42.)
In this section we have tried to confine ourselves rigorously within
the limits of the subject, the Prosody of the Verb, though tempted to
show how the Plautine Verb echoes the talk of that day almost as faith-
fully as the Plautine Pronoun. If e.g. ego ilium scio is the remote
ancestor of French je l'sais, so is e. g. ego ilium habeo-cognitum of French
~ ~~~
je l'ai connu.
41. PROSODY OF ADVERB. Adverbs in -a. Their various for-
mation,—(1)in -a (Accusatives Plural Neuter?), e.g. frustra, contra,
(2) in -a (Abl. Sing. Fern.), e. g. contra, supra— has been already men-
tioned (4).
186 EARLY LATIN PROSODY

Adverbs in -e. Second Declension Adjectives normally form Adverbs

in -e in classical Latin, Third Declension Adjectives in -iter. Although
this is not a rule for Plautus, who uses e. g. amiciter as well as amice,
it is so far recognized by him that the amiciter type is resorted to under
metrical necessity (at the end of a line, etc.). The -e had definitely
become -e in bene (see 5), male (see 42) by his time, through the opera-
tion of the ' brevis brevians '.
Adverbs in -urn, -us. Another early variety in Adverb-formation
(for there is more strict uniformity in Cicero's time) was the alternation
of the suffixes -us and -um. The former (linguists tell us) was a Nom.
Sing. Masc. suffix e.g. rursus venit (re versus venit) ' returning he
came ' ; the latter an Ace. Sing. (Neut. ?) e. g. rursum venit ' he came a
return ' (cf. circum ire ' to go a round '). Scribes (and correctors) were
at all times prone to substitute the familiar for the unfamiliar by-form,
and even scribes of Virgil substituted rursus for rursum to the detriment
of the metre in Aen. 3, 229 :
Rursus in secessu longo sub rupe cavata.
Scribes of Plautus and Terence play similar pranks with demus and
demum* protinus and protinam (42), nullus venit and nullum venit (like
noenu of Lucretius and noenum, whence non ; 17). Editors need not
hesitate to save the metre by substituting the unfamiliar for the familiar
form.

42. LIST OF WORDS. One class of words in this alphabetical list

requires some preliminary remarks, Conjunctions like velut, sicut, etsif
tametsi, quamobrem, quemadmodum. In the earliest Roman literature
we should expect to find the several elements of these stereotyped
word-groups (like our ' albeit ', ' howsoever ') written separately ; and
although an editor may print, e. g., quamvis as one word for the con-
venience ofreaders, he will be slow to believe that Plautus and Ennius
wrote it as one word, until he finds a clear case where quam volo (cf.
Plaut. frag. 6), quam vultis or the like would be appropriate and not
quam vis ' as you please ', ' as one pleases' (cf. quam veils, Pseud. 11 75.)
Apart from these a priori considerations we have definite proof that
some, at least, of these word-groups were not stereotyped in Plautus'
time. Velut would give a false tribrach in two lines (II 45). Plautus
must have written velut, Merc. 227, Rud. 596 (Senarii) :
Vel ut ego nocte hac quae praeteriit proxima j
Vel ut ego hac nocte quae processit proxima.
And if vel ut, presumably also sic ut (cf. Ter. Andr. 804, cited above,
 HIATUS 187

II 61). Sicut would give not a false, but a disliked form of Dactyl in
Iambic lines (II 54), e. g. Asin. 558 ; Cas. 566 :
Sic ut ego possum quae domi ||;
Sic ut ego feci stultus : contrivi diem.
For et si some slight evidence is afforded by e. g. Trin. 527 (Sen.) :
Consuadet homini, credo, et si scelestus est,
where etsi would be a Spondee (in the fourth foot of a Senarius) with
clash of ictus and accent (II 7).
For the others their variety of scansion is proof that they had not yet
been stereotyped. Since emphatic monosyllables seem not to be effaced
by Elision (53), we should probably scan, e.g. in Merc. 252 (Sen.) :
Hoc quam ad rem credam pertinere somnium ? (cf. True. pr. 70).
How can we treat quam ob rem differently when it is Interrogative
(cf. quamnam ob rem Mil. 360) ? Scan, e. g.
Mil. 319 (Troch.) Quam 6b rem iubeam ? Philocomasium ||.
And yet the word makes a Bacchius in Amph. 552. Similarly quo
usque ?, e. g. Asin. 42 (Sen.) :
Etiam amplius. Nam quo usque ? Usque ad mortem volo.
And if quo in loco? be the true scansion, why not, e.g.Merc. 352 (Bacch.):
Emisse indico, quem ad || modum existimet me ?,
although it must be confessed that 'qu(em) ad modum' would suit
Bacchiac Metre (IV 23).
Does the Brevis Brevians Law require prae ut, e. g. in Men. 935
(Troch.) ?
Immo Nestor nunc quidemst de || verbis prae iit dudum fuit.
Or may we compare Rud. 1335 (Prae! verbis quid vis) ? While
tametsi may be printed in Pseud. 244 (Bacch.), where it is a Bacchius :
Redi et respice ad nos, || tametsi occupatu's,
we must print tarn etsi or tarn et si in Aul. 768 (Troch.), where it is an
Anapaest or a Spondee :
Tarn et si fur mihi es, molestus || non ero. i vero, refer.
And this last is its normal scansion in Plautus and Terence (Amph.
pr. 21; 977; Aul. 768; Capt. 321; Cure. 259; 504; Men. 92; Mil.
744; Poen. 342; 1201; Pseud. 471; Stich. 41; 205; Ter. Andr. 348;
Eun. 216; 316; Ph. 428. Cf. Pers. 362; Stich. 27; Trin. 679;
Ennius trag. 165).
Compare Lucilius' scansion :
181 Quo me habeam pacto, tarn et si non quaeri', docebo ;
916 (Sen.) Colligere auxilium, tarn et si est indigna ut feram.
188 EARLY LATIN PROSODY
abeo {abiit and ablt ; see 40 A).
abhinc. In none of its four occurrences a Pyrrhic (by Brevis Brevians Law),
presumably because the second syllable was accented (cf. Adhuc).
abicio (see 23).
ac, the preconsonantal doublet (atqu') ; atque, the prevocalic (see II 36). Is
atque written for ac in Trin. 935 (Troch.) :
Sed ubi absinthium fit atque cu||nila gallinacea,
so that the divided Dactyl is dispensed with (II 56) ? Terence allows atque
at the end of a line in Synaphea, e.g. Ad. 217, 375, 465.
Accheruns ('Axepav) in Plautus, etc., Acheront- in Ennius (see 28). In Poen .
831 (Troch.) if quasi is right, this ' brevis brevians ' overcomes the secondary
accent of the first syllable of Accheruntem (II 29) :
|| quasi Accheruntem veneris.
Acchilles (?) (see 28).
Achivi (see 28).
adhibeo: adhtbuisti Poen. 131 7 (II 33).
adhuc. Never a Pyrrhic by Brevis Brevians Law (because accented on final ;
see 9).
adicio (see 23).
adiuvo : Fut. Perf. adiiiero (see 40 A).
adsuesco, trisyllable (see Suesco).
aes (see Ahenus).
Aetina (see 28).
Agathocles. So in Men. 411; Most. 532. The form ' Agathocules ' would
suit the remaining occurrence :
Pseud. 532 (Sen.) Virtute regi Agathocli antecesseris (Vir.(tu) Ritschl ;
re. (ipsi) Mueller),
ahenus and perhaps -neus. The first form of this Adjective (derived from
ahes, an older form of aes, which perhaps appears at the end of line 23 of
the Casina prologue) is. the usual (Cas. 122; Pseud. 157 ; 656; True. 54;
274) ; the second occurs in one version of True. 272 (Troch.) :
An eo bella es quia accepistl || armillas aheneas ?,
where A seems to insert tibi before armillas, and some scholars read aeneas
(cf. Herm. 34, 323). ■ Aereus ' is a much later Adjective.
aio. This verb is the strongest available piece of evidence for the theory
(generally discredited) that Latin exhibited a variation of quantity like
Greek 7rota> (Spondee or Iambus), toiovtos (with first syllable long or short).
But until its etymology is cleared up, its evidence is suspect. The quantity
of its a is doubtful. Assuming the quantity to be short and the original
spelling to be alio (two long syllables, the first of which is a diphthong), we
find a spondee in lines (questioned by some scholars) like :
Amph. 284 (Troch.) Aiin vero, verbero? debs || ess' tui similes
(Or putas
deos) ?;
 HIATUS 189

344 (Troch.) Aiin vero ? Aiio enfrn vero ||;

Cure. 323 (Troch.) Pernam, abdomen, sumen suis, || glandium. Aiin tu
omnia haec ? (Or with Hiatus and Syllaba Anceps
at change of Speaker ; cf. Most. 383 ; Asin. 896).
After a ' brevis brevians ' this would become an Iambus, for example in the
common phrases quid ais ?, quid ait ?, etc., e. g. :
Men. 1043 (Troch.) Is ait se mihi adlaturum ||;
Naevius com. 125 (Sen.) An nata est sponsa praegnans ? vel ai velnega.
But even without a 'brevis brevians' ais, ait seem to be the normal scansion
in dialogue metres, e.g.:
Asin. 285 (Troch.) || vinctos nescioquos ait ;
Rud. 1072 (Troch.) || ut ego dico. Quid tu ais ? ;
(Cf. Poen. 1013, 1017, Senarii, both ending so : quid nunc ait ?)
Monosyllabic ain is not wholly proved by such lines as :
Pers. 491 (Anap.) Ubi nunc tua libertast ? Apud te. Ain || apud mest.
Aiio, inquam, apud test, inquam ;
184 (Troch.) Meliu' quam tu qui docuisti. Ain || vero, verbereum
caput ? ;
since it might be ain (like vtden).
Nor ait by lines like:
Mil. 430 (Troch.) Nam nee te neque me novisse ait ||haec. Persectari
hie volo.
Nor ais by Epid. 29 (supposing the line to be a Cretic Dimeter) :
Sed quid ais ? || Quid rogas ?,
since the first foot might be a Choriambus (as in 98-98* Quid faciam ? || Men
rogas ?).
At the beginning of iambic lines, curiously, we find not Ait se but Ait
sese, e. g. :
Mil. pr. 91 (Sen.) Ait sese ultro omnes mulieres sectarier;
pr. 126 (Sen.) Ait sese Athenas fugere cupere ex hac domu;
Rud. 60 (Sen.) Ait sese Veneri velle votum solvere.
Lucilius' dactylic verse (like Virgil's) shows ait :
202 Laeviu' pauperem ait se ingentia munera fungi ;
1 193 Hymnis, cantando quae me adseruisse ait ad se ;
(unless line 11 17 be an exception).
The scansion of the Imperfect Tense is less dark. Apparently we should
find only two rival forms :
aiebam, e. g. :
mihi;
Amph. 383 (Troch.) Amphitruonis te esse aiebas ||;
387 (Troch.) Ego sum Sosia ille quern tu || dudum esse aiebas

Ter. Heaut. 924 (Sen.) Quid faciam ? Id quod me fecisse aiebas parum ;
aibam, e. g. :
Amph. 661 (Troch.) Qui dudum properare se[se] aibat ||;
i9o EARLY LATIN PROSODY
807 (Troch.) Te dormitare aibas ; mensa ab||lata est, cubitum
hinc abiimus ;
Ter. Andr. 534 (Iamb.) Aliquot me adierunt, ex te audi||tum qui aibant
hodie nubere.
Alcumena (see 28).
AlcumZus ('AA/tyieW). Only occurrence is Capt. 562 (Troch.) :
Et quidem Alcumeus atque Orestes || et Lycurgus postea.
Aleus ('HXeZor, 'AXflos), never -eus. The Hiatus in Capt. 24 (Sen.) ; 31 (Sen.).
Postquam belligerant Aetoli cum Aleis ;
Summoque genere captum esse equitem Aleum,
would indeed go if we read ' Valeis ' (cf. fakeloi, the native form).
But this form would not suit Capt. 59 (Sen.) ; 880 (Troch.) :
Quia bellum Aetolis esse dixi cum Aleis ;
Et captivum ilium Alidensem ||,
and the MSS. give no hint of it.
Alexander. Once with second syllable shortened by Brevis Brevians Law
(Most. 775)-
aliorsum (True. 403) and aliovorsum (Aul. 287).
aliquis and aliqui (True. 102).
alius (see 39).
alter, -tera, -terum. Never -tra, -trum ; but altrim secus (q. v.), altrovorsum
(see 27).
Gen. Sing, alterius (Capt. 306), alterlus (Ter. Andr. 628).
alterim (and altrim) secus. The instances in Plautus are : secus ;
Merc. 977 (Troch.) Optime hercle, perge ; eg6 ad||sistam hinc alterin-
Mil. 446 (Troch.) || quin retines altrim secus ? ;
Pseud. 357 (Troch.) Pseudole, adsiste altrim secus atqu(e) || onera hunc
maledictis. Licet ;
Rud. 1 158 (Troch.) Post altrim secust securicul(a) || ancipes, itidem
aurea.
In the first of these alterinsecus, if written as two words, would make a
false final Dipody (IV 16) ; in the third altrim secus, if written as one word,
would make a false division of an Anapaest (II 55).
altrovorsum. The word only occurs once, Cas. 555 (Troch.):
vorsum
Verum autem altrovorsum quom earn || mecum rationem puto A),
(altero-

where indeed we might conceivably scan alterflvorsum, or even thus : altero-

vorsum qu(om) earn || (46).
ambo. Ace. ambo (often altered by scribes to ambos) and ambos.
tin (see 13).
ancipes (i. e. -ess) Rud. 1158. The class. Lat. Norn, (anceps) is not found.
 HIATUS 191

angina proved by Lucil. 1093 (on Scipio's death) :

Insperato abiit, quem una angina sustulit hora.
antehac (2 syll.), antidhac (3 syll.).
Antidamas (see 30 A).
Antiochus (see 28).
apagesis, one word (not two), as is proved by Poen. 225 (9).
Apelles (rather than -a), Voc. -e (see 30 N).
apprendo ('apprehendo ' does not occur, unless Ter. Andr. 353): Poen.
1226; Caecilius 80 ; Ter. Ph. 863 ?
apud (see II 22). The Philoxenus Glossary has (Corp. Gloss. Lat. II 21,
40--41) :
Ape : 7rapd.
Ape : kq)\v<tov.
The second is clearly the same gloss as Paul. Fest. 21, 4 (Ape apud antiquos
dicebatur prohibe, compesce). The first may be a scribe's error for Apud :
napd, his eye having been caught by the Ape in the next line. (Or Ape :
kg>\v(tov Trapa (Nai/3ia)) ?).

Apulus (App- ?), a Dactyl. There are two occurrences, Cas. pr. 77 (Sen.),
Mil. 648 (Troch.) :
Vel Graecus adeo vel mea causa Apulus ;
Post Ephesi sum natus, non enim in || Apulis ; non sum Animula.
The district-name too, has presumably the first syllable long, though this
cannot be ascertained from :
Cas. pr. 72 (Sen.) Et hie in nostra terra in (terra) Apulia ;
Lucilius 824 (Sen.) Hoc turn ille habebat et fere omnem Apuliam.
aqua. No clear evidence of a trisyllable in Amph. 1058 ; Cure. 161 ; True.
564 (In Mil. 552 Aqua aquai, with Anapaest as first foot ; in Trin. 676 Turn
igitur tibi aquai erlt cupido ||, with Anapaest as second foot). But abundant
evidence of the class. Lat. form (e. g. at end of line, Aul. 564, Bacch. 247,
etc.). (On the trisyllable of Lucretius and a Tragic fragment, n°. in, of
unknown date see the Latin Thesaurus The reading of Enn. Ann. 168 is
doubtful.)
architectus, usual ; architectonem (Ace), at end of line (i.e. through metrical
exigency) twice :
Most. 760 (Sen.) Nam sibi laudavisse ait hasce architectonem ;
Poen. 1 1 10 (Sen.) Me quoque dolis iam superat architectonem.
aridus and ardus (27).
arvum. Only as an Adjective in Plautus and only once (True. 149), where
it may be scanned either as two or three syllables. In Naevius (trag. 19),
\ Pacuvius (trag. 396) as a Fern. Noun, and with the same uncertainty of
scansion.
asper. The two occurrences of oblique cases in Plautus are at ends of lines
and show the full trisyllable. Ennius (Hedyph. line 2) has probably aspra
(see I 2).
192 EARLY LATIN PROSODY
at que (see Ac).
attdt (e.g. Capt. 664), but attatae (always a Cretic ; never a Dactyl under
the Brevis Brevians Law. So presumably the final syllable had an accent
as in the Greek drroTai).
audacia. No clear trace of a five-syllabled 'avidacia' (cf. Capt. 287 avari-
tiam ipsius atque audaciam).
audax. No clear trace of a three-syllabled ' avidax ' (Amph. 985 ?).
audeo. No clear trace of a four-syllabled ' avideo '.
augurium. The trisyllabic Plural used by Accius, trag. 624 (Troch.) is pre-
sumably from a Neuter augur (or augus ?) :
Pro certo arbitrabor sortes, || oracla, adytus, augura.
avidus. No clear trace of a syncopated ' audus ', though there is a pun in
Bacch. 276 (Senar.) :
Quin tu audi. Immo ingenium avidi haud pernoram hospitis,
avis, Gen. Plur. avium. This Gen. Plur. at the beginning of a dactylic line
(Spondee or Anapaest) is puzzling :
Ennius Ann 93-94 Cedunt de caelo ter quattubr corpora sancta
Avium, praepetibus sese pulchrisque locis dant.
Has Cicero quoted the line correctly ? (See 23.)
aureus. Always trisyllabic in Plautus ; so scan in Stich. 25 (Anap.) :
Montes qui esse au||rel perhibentur (Not aurei).
avunculus. The Plautine form is aunculus (cf. French oncle). The quadri-
syllable appears only in Aul. 35 (Sen.):
Is adulescentis illi(u)s est avunculus.
By a mere transposition (est illius) the normal form of everyday speech
could be put here too, but editors usually keep to the traditional reading, on
the plea that this line occurs in the prologue spoken by the Lar Familiaris.
The word is not found in any other line of early poetry and is used by
Plautus himself only in this play. It is a lucky chance which has preserved
this echo of Roman talk.
bacchanal (see 11).
balineae. balin-, preserved by the MSS. of Plautus and by the Bembinus of
Terence in Phorm. 339 (the only occurrence of the word), is said by Isidore
(Diff. 1, 75 ; taken from Donatus ?) to be the current, bain- the literary
form. There is no counter-evidence. For Nonius in citing a line of
Caecilius (frag. 98), attests merely the feminine gender, not the three-
syllabled form. It was the accentuation (II 18) of the fourth last syllable of
Proceleusmatic (or Fourth Paeon) words (or word-endings) that produced
the balnea (and the puertia) of Horace.
balineator. balin-, preserved by the Palimpsest in Rud. 527, suits the metre
in the other two occurrences also (Poen. 703 ; True. 325).
battuo occurs once, Cas. 496 (Sen.), to be scanned probably :
Quibu' battuatur tibi os, senex nequissime.
 HIATUS 193

bellum (see Duellum).

ben?, never ' bene ' (5).
beneficiwn. Is it necessary to say that RitschPs ' benf- ' is a myth ?
cacula. Trin. 721 (Troch.) ; Accius Ann. frag. (Hexam.) :
Video caculam militarem || me futurum haud longius;
Calones famulique metellique caculaeque.
The writers of the acrostic (in P) and of the non-acrostic (in A) Arguments
of the Pseudolus get the word from the scene-headings of that play, with
harpax cacula. That both writers should make the same mistake in its
scansion ('cacula') is interesting, and hardly a mere coincidence. It was
associated with calo and calator by antiquarians, and is derived by Festus
(also by Verrius Flaccus ?) from Greek KaXov.
Calchas, Abl. Calcha (Men. 748, -chante MSS.).
calefacio, -cto (see II 27).
calidus. No clear trace of ' caldus '.
Callicles, Gen. -clei (not -' cletis ') or -cli, Dat. -cli, Voc. -cle,-cles (cf. II 30 N).
Campans, the local form of Campanus (for Oscan, unlike Latin, syncopated
a final syllable ; 57), is used (derisively) in Trin. 545 and probably True. 942.
cants Masc, canes Fern., probably.
caput. Abl. Plur. capitibus plays the part of a Dactyl (if the reading be
sound) in Ennius Ann. 490 :
Capitibu' nutantes pinos rectosque cupressos,
where only the Gender of cupressus is attested by Gellius (and Nonius who
excerpts Gellius). It could conceivably play the same part in the two
occurrences in Plautus (Bacch. 305 ; 1208). (See below, Columen.)
Carthagin(i)ensis. Both forms suit the metre in Plautus, and the MSS.
generally vary between the two. Probably Carthaginensis, the correct form,
should be printed in Plautus, etc. ; Carthaginiensis (a coinage of Ennius ?)
in Ennius' Annals, etc.
caveo. Imperat. cave" (normally).
ceteri. No hint of ' cetr- ' (27).
Charmides, Gen. -dei and -di, Voc. -de and -des (cf. II 30 N).
Chius. Adj. (Cure. 78 ; Poen. 699), with the same quantity as the name of
the island (cf. 25).
cicatrix. Plur. cicatrices Asin. 552 (by Brevis Brevians Law).
circu(m)eo and circ{um) eo (see 45) :
Asin. 742 (Iamb.) Iliac per hortum circ(um) iit || clam, nequi' se videret
(Or circuit ?) ;
Cure. 451 (Sen.) Ita non potuere uno anno circuirier;
Men. 231 (Sen.) An, quasi mare, omnes circuimus insulas ? ;
Pseud. 899 (Sen.) Ne fidem ei haberem. nam eum circ(um) ire in hunc
diem (Or circuire ?) ;
Rud. 140 (Sen.) Heus ! tu qui fana ventri' causa circuis ;
True. 407 (Sen.) Novi. Haec data opera circuit per familias.
2348 O
194 EARLY LATIN PROSODY
cito (see II 20).
clam. Ever elided (Cas. pr. 51, 54) ? (See App. C.)
clavator, hardly a disyllable (Rud. 805). (See 24.)
Cleostrata (Cleus-), always a trisyllable.
cochlea (see 26).
Coculites (see 26).
coepio (see 29).
cogo. Perf. coegi (disyll. and trisyll.). (See 29.)
cohibeo (cob- ?) (see 22).
columen Nom., culminis (from colum-) Gen., etc., perhaps the old declension.
comptionalis (see II 18).
compos (-oss) (see 20).
comprehendo and comprendo. Of the five Plautine occurrences (Aul. 346 ;
Capt. 594 ; 599 ; Mil. 579 ; Rud. 475) all admit the quadrisyllable and two
require it (Aul. 346 ; Capt. 594). The trisyllable is Terence's form (Eun.
836; 993).
consuadeo, quadrisyllable (see Suadeo).
consuescoj trisyllable (and quadrisyllable ? See Suesco).
contra (see 41).
controversia, hardly -trorsia (see 24).
cor, for cord(e) (see 20).
corrigo (Trin. 118; 653; Ter. Ad. 593; 741 ; 994; Hec. 254; Andr. 569).
A disyllable ' corgo ' is not proved by :
Ter. Andr. 596 (Iamb. Octonar.) Ego vero solus. Cor[ri]gere || mini
gnatum porro enitere (IV 11).
cotidie. No clue to quantity of first syllable. Presumably cot- (quot-), as in
Martial (10, 65, 8 ; II, I, 2), though the spelling cottidie offers a difficulty.
(In Catull. 68, 139 cottidjana is now abandoned by editors.)
coturnix (Asin. 666 ; Capt. 1003).
crocus ■; and corcotarii (II 40), Aul. 521 (Sen.) :
Quom incedunt infectores corcotarii.
cupio, partly Fourth Conjugation (see 40 H).
Dareusy presumably -e- (like platea). Occurs only Aul. 86 (see II 7 a).
deartuo (see 29).
deascio (see 29).
debeo or dehibeo (see 22).
debilis (Merc. 630) and debil (Enn. Ann. 324), like mugilis and mugil.
dehinc, a monosyllable. (For * dehinc ' is unlikely ; see above, Abhinc.)
dehortor, a. trisyllable (see 29).
deicio, a quadrisyllable (see 23).
deinde (a Trochee), before vowels ; dein, before consonants (see II 36).
deorsum, a disyllable in the Dramatists (like sorsum). There are four
occurrences of the word in Plautus :
Amph.i 108 (Troch.) Devolantangues iubatae [|deorsum in impluvium duo;
Aul. 708 (Sen.) Ubi Tile abiit, ego me deorsum duco de arbore ;
Rud. 178 (Sen.) Si ad saxum quo capessit ea deorsum cadet.
 HIATUS

So scan : Aul. 367 (Sen.) ST autem deorsum comedent siquid coxerint (47).
Three in Terence :
Eun. 278 (Iamb.) Ne sursum deorsum cursites|| ;
Ad. 573 (Troch.) Nostin porticum apiid macellum hanc ||deorsum ? Quid? ;
ni noverim

575 (Troch.) Clivu' deorsum vorsum est: hac te ||praecipitato :

But it is a trisyllable in Lucil. 703 (Troch.), where it is emphatic ('down'

postea.
contrasted with ' up ') :
Modo sursum, modo deorsum ||tamquam collus cernui.
Similarly seorsum (so-) is a disyllable in the Dramatists. In the MSS. of
Plautus it is spelled always sorsum, and that it was so pronounced by Plautus
appears from the alliteration in Asin. 362. The occurrences are :
Asin. 362 (Troch.) Nam me hodie senex seduxit ||solum sorsum aedibusab;

Capt. 710 (Sen.) At ego aio recte, qui abs te sorsum sentio ;
Epid. 402 (Sen.) In aediculam istanc sorsum concludi volo ;
Rud. 1314 (Iamb.) || Philippea in pasceolo sorsus ;
Demea ;
Ter. Ad. 971 (Troch.) Omnibu' gratiam habeo et seorsum ||tibi praeterea,
Afranius 85 (Sen.) Officiis cogi ut abs te sebrsus sentiam.
Is it a trisyllable in Accius 117 (Sen.)?
Nosque ut seorsum dividos leto offeres (Emended to Nosque uni
seorsum).

These two words taken together make a strong argument for Synizesis. It
is unreasonable to explain sorsum as anything but a spelling-device to ex-
press the pronunciation seorsum. And why may not Ennius have used the
same device in his sis (for suis), sas (for suas), etc. ?
debsculor (see 29).
depeculatus (see Peculatus).
deprehendo and deprendo. Two (doubtful) occurrences in Plautus :
Bacch. 950 (Iamb.) Dolis ego deprensus inventus
sum, ill'interit
mendi||cans
(prensuspaene
A) ;
Mil. 1276 (Iamb.) Egon ad lllam earn quae nupta sit ? ||vir ei(u)s me
deprehendat (Reading doubtful);
deus. Synizesis (II 33) is as frequent in the cases of this Noun as of Greek
6(6s, e. g. debs (normally ; cf. Class. Quart. 3, 9), di deaeque (normally),
debrum (normally). Notice the full form deos in :
Cas. 670 (Bacch.) Per omnes deos et||deas deieravit (cf. Rud. 191) ;
Poen. 950 (Sen.) Deos deasque veneror qui hanc urbem coluut of a (Opening
prayer) ;
Cist. 242 (Troch.) Neque deos neque homines aequumst ||(Emphatic).
On the Plural di (never ' dei '), dis (never ' deis '), see 24.
O 2
i96 EARLY LATIN PROSODY
dextera and dextra. The trisyllable (often spelt in the MSS. as a disyllable)
appears in all parts of the line (not merely at the end). Notice :
Merc. 965 cette dextras nunciam ;
Cure. 307 cedo tuam mi dexteram (cf. Merc. 149 ; Poen. 315).
dextrovorsum, a quadrisyllable in all three occurrences in Plautus (Cure. 70 ;
Rud. 176 and 368), but dextrorsicm in Accius. praetext. 27 (Sen.) :
Mirificum facinus : dextrorsum orbem flammeum.
Diana (e.g. Bacch. 312; Ennius Ann. 62; trag. 31).
dico. Imperat. dice (Plaut., not Ter.), but die when closely joined with a
following word which begins with a consonant (e.g. die mihi. With Rud.
1 156 dice dum cf. hici-ne, etc.). Some examples of the full form are :
Rud. 124 (Sen.) Tu, siquid opus est, dice. Die quod te rogo ;
Mil. 256 (Troch.) ||dice, monstra, praecipe (cf. Capt. 359) ;
Cas. 346 (Sen.) Bene dice, dis sum fretus, deos sperabimus ;
Cure. 132 (Troch.) ||me periisse. Age dice. Anus, audi ;
Merc. 159 (Troch.) || dice. At enim placide volo ;
Rud. 951 (Cret.) Quid negoti est modo ||dice. Dicam, tace ;
Bacch. 716 (Troch.) Quid nunc es facturus ? id mi ||dice. Coctumst
prandium ? ;
Naevius trag. 63 (Iamb.) Odi summussos ; proinde aper||te dice quid sit
quod times.
Similarly face and fac (e. g. fac sciam, not * face sciam '). Some of the
numerous examples of the former (at end of line in Terence, e. g. Ph. 397) :
Asin. pr. 4 (Sen.) Face nunciam tu, praeco, omnem auritum poplum ;
90 (Sen.) Face id ut paratum iam sit. Unde gentium ? ;
605 (Iamb.) Sermoni iam finem face ||;
726 (Iamb.) Dabuntur ; animo sis bono ||face; exoptata obtingent;
Aul. 153 (Vers. Reiz.) Heia ! hoc face quod te iubet soror|| ;
Cas. 353 (Troch.) Face, Chaline, certiorem||;
714 (uu-u-) Face ut impetres ;
Ter. Andr. 712 (Iamb.) Ut ducam. Ridiculum. Hue face ad||me ut
venias, siquid poteris.
Lucilius 890 (Troch.) Perge, amabo, ac, si pote, face||dignam me ut
vobis putem.
Similarly duce (Plaut., not Ter.) and due (but duce me Most. 324 ; 794 ;
Rud. 386). Some examples of the former :
Aul. 452 (Troch.) Etiam — intro duce, si vis, ||vel gregem venalium ;
Epid. 399 (Sen.) Exite hue aliquis. duce istam intro mulierem ;
Poen. 1229 (Iamb.) Ite in ius, ne moramini, an||testare me atque duce ;
Trin. 384 (Troch.) Tibi permitto ; posce, duce. || Di te servassint mihi !
dierectus always (but Trin. 457 ?).
dies. (On forms with Synizesis see II 33.)
Diespiter and Diesp- (II 33).
 HIATUS 197

discip{u)lina (see 26).

dissuadeo, quadrisyllable (see Suadeo).
diu (see II 33).
dives (-ess), divitiae (see 24).
dim mis (see 24).
dius (see 25).
diiitinus, -ne. The occurrences are :
Mil. 503 (Sen.) Longum diutinumque a mani ad vesperum ;
Rud. 1241 (Sen.) Diutine uti bene licet partum bene ;
Turpilius 38 (Sen.) Subm sibi amicum esse indulgentem et diutinum
(Or Suom . . . diutinum) ;
Ter. Ph. ion (Troch.) Haecin erant itiones crebrae et || mansiones
diutinae.
(Some scan in the last two diut-.)
diutius. The Comparative of diu is in Phaedrus (epilog. 4) diiitius, a Proce-
leusmatic (cf. diuturnus of Augustan poets), but in Rud. 93 diutius :
E6 vos, amici, detinui diutius.
Apart from the Rudens line, the word is always capable of being scanned
as a Dactyl diutius or as a Proceleusmatic ; see Class. Quart. 12, 47 and
above, Diu). The reading of Lucilius 929 is doubtful.
do. Possibly dat (like das Poen. 868 ; det Pers. 327) ? The passages that
throw light on the scansion hardly outweigh the evidence of datur :
Most. 601 (Sen.) Molestu' ne sis. nemo dat ; age quid libet (AP) ;
Men. 10 1 (Sen.) Ceriales cenas dat ; ita mensas extruit (P Gram.) ;
Rud. 1072 (Troch.) Verba dat ; hoc modo res gesta est ||;
Cas. pr. 44 (Sen.) Dat erae suae, orat ut earn curet, educet ;
Cure. 161 (Troch.) ||foribu' dat aquam quam bibant.
dom{i)nus (see 27).
donee and donicum, never ' donique ' (12).
drachuma (see 26).
ducenti. Lucilius' due- (but due- 481) is a puzzle :
555 Milia ducentum frumenti tolli' medimnum,
Vini mille cadum (Emended to Milia tu c.) :
1 05 1 Quid vero est centum ac ducentum possideas si
Milia (Emended to atque).
dueo. Imperat. duce and due (see above, Dico).
duelhim, a disyllable in True. 483 (Troch.) :
Manibu' duella praedicare ||.
A trisyllable in Amph. 189 (Iamb.) ; Ennius Ann. 559 :
Duello exstincto maximo|| (Emended to Ex. du.) ;
Hos pestis nocuit, pars occidit ilia duellis.
dum, the Particle (e. g.age dum, nondum, dudum) was an enclitic appendage.
For if it had been a separate word, we should find ' die dum ', not dicedum
(see Dico).
198 EARLY LATIN PROSODY
duo. (On the forms with Synizesis see II 33 end.) This numeral was (with
ambd) the last foothold of the Dual in Latin. What stand duo Ace. makes
in Plautus' plays against the encroachment of the new form duos cannot be
detected with precision, since any scribe (or rather corrector) who recognized
the word to be in the Accusative case might take duo to be a mere mis-
writing and alter it to duos. Studemund's attempt at discrimination (Arch.
Lat. Lexikogr. 3, 550) seems highly doubtful. It is quite as likely that mere
metrical convenience was what moved Plautus' choice. In Amph. 1108
(where Nonius attests the feminine gender) duo seems to be Dual Fem.
Devolant angues iubatae ||debrsum in impluvium duo.
ea tenus, not one word (Most. 131 ; II 52). Therefore print istac tenus as two
words (Bacch. 168) :
Istac tenus tibi, Lyde, libertas datast,
with benefit to the Alliteration (cf. II 47).
ecastor. Since the first syllable succumbs to a ' brevis brevians' it is less
likely to be long by nature (II 27). Presumably -or (Poen. 1176) (see 14).
(Mil. 1041 is not good evidence.)
eccere. The proof of -e is not quite conclusive :
Amph. 554 (Bacch.) Mihi praedicas. Ec||cere ! iam tuatim.
tecum (see 52 ; II 18 D).
ecquis probably always a Trochee (38), except after a ' brevis brevians ', e. g. :
Pers. 108 (Sen.) Sapi' multum ad genium. Sed ecquid meministin heri ?
ego (see 31). Dat. mi (enclitic) and mihi (see 31). Ace. Abl. med and me
(see 31).
egoquidem and (two words) ego quidem (see 31).
ehem, Pseud. 873 (Sen.) :
Immo edepol vero hominum servator. Ehem {A, not P).
Scan Ter. Andr. 417 (Sen.) :
Quasi de improviso respice ad eiim. Ehem, pater (Or ehem ?),
with Hiatus at change of Speaker and with the fifth foot a Tribrach (II 48;
IV 8 A).
eheu, e. g. e- :
Capt. 152 (Sen.) Nunc habe bonum animiim. Eheu ! huic illud dolet
(with Hiatus at change of speaker and after Interjection).
995 (Troch.) Eheu ! quom ego plus minusque ||;
Ter. Heaut. 1043 (Troch.) Facere puduit. Eheu! quam nunc ||totu' dis-
pliceo mihi ;
&-:
Pseud. 81-82 (Sen.) Eheu ! Neque intus nummus ullus est. Eheu!
Ille abducturust mulierem eras. (Cras?) eheu!\
Ter. Heaut. 83 (Sen.) Ouaeso, quid de te tantum meruistl? Eheu ! (with
Hiatus at change of speaker).
el (Interjection), whence the Verb eiluo (eiiulo) ; cf. Aul. 796 (ei mihi ! . . .
Cur eiulas r).
 HIATUS 199

eia (Greek eta ?), heia (the same word?). (Details in Studemund's Studien.
Ii 538-)
Merc. 998 is not conclusive proof of -a : still less Ennius Ann. 597 of -a,
since Servius' quotations are notoriously untrustworthy, and, besides, Servius
does not say that the phrase comes from the Annals.
ejicio (see 23).
em (see 52).
enim. Normally a Pyrrhic before an initial consonant (by Brevis Brevians
Law, 12), e. g. :
Epid. 94 (Cretic) At enim tu ;
but occasionally an Iambus (1) in Bacchiacs or Cretics, (2) at a pause,
etc., e. g. :
Pseud. 1267 (Bacch.) Dari dapsiles, non ||enim parcepromi ;
Most. 551 (Sen.) Quid tute tecum ? Nil enim. sed die mihi.
In Plautus non enim is a Dactyl, e. g. :
Asin. 808 (Sen.) Haec sunt non nugae, non enim mortualia;
Aul. 594 (Troch.) Retinere ad salutem, non enim ||quo incumbat eo im-
pellere ;
Epid. 162 (Troch.) Non enim nunc tibi dormitandi ||neque cunctandi
copia est ;
Trin. 705 (Troch.) Non enim possum quin exclamem : eu||gae, eugae !
Lysiteles, nakiv !
as in Ennius Ann. 371 :
Non enim rumores ponebat ante salutem.
eo (Verb) (see II 33). eamus and eamus (see II 33). Imperat. i in Prosodic
Hiatus, e. g. i Intro (frequently). Pluperf. leram is attested by Donatus (ad
Ad. I, 1, 2 ierant) : producte I pronuntiando, quod nos addita V 'iverant'
dicimus : tale est illud Vergilii ' Nos abiisse rati et vento petiisse Mycenas'.
eo (Adverb) and eo and eo (see II 33).
eques (-ess in Plautus presumably ; 20). •
et si (two words) (see beginning of this paragraph).
eugae (Greek cuye), e.g.:
Bacch. 1 105 (Anap.) Hi(c)quidemst pater Mnesilochi. Eugae ! soci||um
aerumnai et mei mail video ;
Ter. Andr. 345 (Troch.) Te ipsum quaero. eugae! Charine||.
No evidence of -ge. The ictus normally falls on eu- (Studemund's
Studien 1, 522). On eugae-eugae see Appendix A.
eugepae (not used by Terence).
exin and exinde (see II 36).
cxporgo (see Porgo).
txpurgo (see Purgo).
 2oo EARLY LATIN PROSODY
extemplo and -ftulo, the latter only in Plautus (not Terence) and only at the
end of a line (Aul. 93; Bacch. 968; Cist. 96; 572; Mil. 461; Poen. 183),
except possibly :
Bacch. 304 (Sen.) Quom extempulo a portu ire nos cum auro vident (Or
Quom extemplo ; 47) ;
True. 666 (Sen.) Qui non extemplo (intro) ieris. Anne oportuit ? ;
Naevius com. 98 (Sen.) Extemplo illo te ducam ubi non despuas.
facto. Sometimes of Fourth Conjugation (40 H), e. g. :
Amph. 555 (Bacch.) Fads, ut tuis nul||la apud te fides sit.
lmperat./ace and fac (see above, Dico). Passive yfo (see below).
fe{ne)stra (II 18). Perhaps always a disyllable in Plautus, for we have
express testimony tinaXfestra was the older form (Paul. Fest. 80, 27 Festram
antiqui dicebant quam nos fenestram) and the metre never requires anything
else :
Cas. 132 (Sen.) Quid facies ? Concludere in festram firmiter (fenestram

Mil. 379 (Iamb.) Neque festra nisi clatrata ; nam ||certe ego te hieAP)
vidi (fenestra intu';
Rud. 88 (Sen.) Illustriores fecit festrasque indidit (fenestras P : A n. L).
But Terence (at any rate the traditional text of Terence) has the trisyllable :
Heaut. 481 (Sen.) Quantam fenestram ad nequitiam patefeceris.
fero. Perfect tetuli (Plaut.), tuli (Ter.) (see 40 E).
finitor. If we follow the MSS. (the Palimpsest is not in evidence) we must
scan ' finitor ' (?) :
Poen. pr. 49 (Sen.) Determinabo : ei rei ego sum factus finitor (Emended
to fi. factu' sum).
fio, perhaps never 'fio' ; fiam, perhaps (Men. 352 ?) never ' fiam' (in Mil.
1054 put age in a separate line ; in Aul. 405, Ter. Heaut. 715, etc., read fu-) ;
fieri, fzer em, etc., but fi- is used at end of line, i.e. through metrical
exigency, both by Plautus and Terence (only f 1- in Lucilius' fragments) :
Some examples :
Amph. 593 (Troch.) Fieri, nunc uti tu hie sis ||;
Poen. 1056 (Sen.) Quid ais ? qui potuit fieri uti Carthagini ;
Ter. Ph. 760 (Iamb.) || dabamus operam ut fieret (The whole passage
consists of Iamb. Septenarii) ;
Ter. Ph. 593 (Sen.) Argentum opus esse et id quo pacto fieret ;
Plautus has fieri in the middle of a Bacchiac line :
Amph. 567 Vidit nee potest fi||eri, tempore uno.
The Active form of the Infinitive is attested for Ennius' Annals, presumably
jlere ; and we have either this form or a shortening (by the ' brevis brevians ')
of the final of the Deponent in :
Ann. 15 . . . memini me fieri pavum (Emended to fiere).
But fi- appears in Ann. 345 and 617 :
Pugnandi fieret aut duri (fini') laboris ;
Qua murum fieri volult, urgemur in unum.
 HIATUS 201

flagitium (see 45).

fors sit an (sit Ter. ?). ' ForsTtan ', as one word, would give a false Anapaest
(n 50) in :
Ter. Eun. 197 (Sen.) Me miseram ! fors (s)it an hie mi parvum habeat
fidem (cf. Ph. 717).
In Plautus the Palimpsest has fors fuat an (cf. fors fuat pol Ter. Hec.
610) but P has forsitan (i. e. fors sit an) in :
Pseud. 432 (Sen.) Fors fuat an istaec dicta sint mendacia (A) ;
Fors (s)it an ea tibi dicta sint mendacia (P ; not for-
sitan, which would give a false Tribrach ; II 44).
fortasse and fortassis (in MSS. at Asin. 493, Bacch. 671 ; never in MSS. of
Terence). On fortass' see 40 G.
forte. On forf see II 54. Some prefer* fors, a compendium-phrase for fors
sit an, fors fuat an, etc., instead of the traditional/^/* in Asin. 794 and the
iX3.d\i\0Ti2i\ fortasse in Ter. Heaut. 715.
frater (see 14).
frustra (see 4).
gnaruris (e. g. Poen. pr. 47).
gratia. Is gratiis ever a disyllable ? (See II 33.) The metre often proves
the trisyllabic scansion, and everywhere this scansion is possible. Ingratiis
is printed by editors of Cicero (e. g. Tull. 5) as four syllables and must have
been so scanned by Plautus. The likelihood of the classical form gratis
having been anticipated by Plautus is therefore not very great.
gynaecewn (see 25).
hahae and hahahae (-ha). The MSS. vary and obscure the exact scansion
of these imitations of a burst of laughter (our ' haha '). Thus hahahae might
be hahae in Pseud. 946 (cf. Mil. 1073), hahahae might be hahahahae in Ter.
Eun. 426.
hercle. A form hercule (substituted by MSS. occasionally ; e. g., against the
metre, in Pers. 591) would save the metre in :
Asin. 275 (Troch.) Mea quidem hercle opera liber \\ numquam fies ocius
(Emended to li. op.) ;
Aul. 392 (Sen.) Perii hercle ! aurum rapitur, aula quaeritur (With Hiatus ;
at pause)

Bacch. 902 (Sen.) Abeo ad forum igitur. Vel hercle in malam crucem
(Emended to he. i in) ;
Men. 316 (Sen.) Eu hercle ! hominem multum et odiosum mihi (With
Hiatus after Interjection ?)
Merc. 436 (Troch.) Hercle illunc di infelicent, || quisquis est. Ibidem
mihi (Emended to divi) ;
Pers. 140 (Sen.) Numquam hercle hodie hie prius edis, ne frustra sis
(Emended to hie ho.) ;
Poen. 173 (Sen.) Non scis ? Non hercle. At ego iam faxo scies (With
Hiatus at change of speaker) ;
566 (Troch.) Vix quidem hercle, ita pauxilla est, || digitulis pri-
moribus (Emended to itast pau.) ;
202 EARLY LATIN PROSODY

1302 (Troch.) lam hercle ego ilium excruciandum || totum carnifici

dabo (Emended to illunc ; cf. I 6) ;
Pseud. 508 (Sen.) Tu vives, tu mihi hercle argentum dabis (Emended to
(vivus) vives) ;
Rud. 656 (Troch.) At malo cum magno sub fecit || hercle. ite istinc
foras (With Hiatus at pause) ;
cule A)
True. 174 (Iamb.) Non hercle occidi, sunt mihi etiam fundi et aedes (?);;
357 (Sen.) Vah ! vapulo hercle ego nunc atque adeo male (her-

814 (Troch.) Et tibi quidem hercle idem || attulit magnum malum

(Emended to id. (vir)).
There are some 640 examples of the word in Plautus ; so these fourteen are
an insignificant minority, a negligible quantity.
Hercules, a Fifth Declension Noun, Gen. Herculel, etc. (30 N).
Mc, Norn. Sing., never ' hie ' (see 33).
Hiluricus, Hilurius l Illyrian ' :
Men. 235 (Sen.) Histros, Hispanos, Massilienses, Hilurios;
Trin. 852 (Troch.) Hilurica facies videtur ||.
hodie (some 400 instances) ; never ' hocedie ' in the Drama (33). And prob-
ably never ' hodie ' (cf. II. 37). Even hoc die would be preferable, with such
a host of examples of hodie. Scan Pseud. 1071 (Senar.) : Si ille hodie ilia
sit potitus muliere. (With emphatic ille ? Not S(i) ille hodie.)

Homeronida (presumably -da ; 30 A) ' Homer ' ? * Ennius ' ? :

True. 485 (Troch.) Et Homeronidam et postilla || mille memorari potest
(Emended to -da)
Qui et convict i et condemnati ||falsis de pugnis sient.
homo. Ace. homonem appears in Ennius (attested by Priscian) :
Ann. 138 Vulturus in spinis miserum mandebat homonem,
but probably as an archaism (like his Gen. Sing, vias ; 30) or a novel coin-
age. For there is no clear trace of this spelling in the MSS. of Plautus and
the number of occurrences of homin- is so large that it overwhelms the insig-
nificant number where homon- would suit the metre ; e. g. Men. 82 Nam
homini, rather than N(am) homoni (cf. 38 G), just as in Mil. 563 nam
hominem, where ' n(am) homonem ' would not scan.
honos. Normally mei honoris gratia, tui honoris gratia, etc :
Aul. 463 (Troch.) Qui simulavit mei honoris ||mittere hue causa coquos ;
Cure. 549 (Troch.) Quid nunc ? Quod mandasti feci || tui honoris gratia ;
Mil. 620 (Troch.) Ea te expetere ex opibu' summis || mei honoris gratia ;
So perhaps scan Poen. 638 (Sen.) :
Quia nos honoris tui causa ad te venimus.
But Stich. 338 (Troch.) offers a difficulty unless we allow ecquid (38 A) ;
Propere a portu tui honoris causa. Ecquid adportas boni ? {AP).
hospes (-ess) (see 20).
 HIATUS 203

ibidem, usually ; but also ibidem, e. g. Rud. 396 (see Sonnenschein's note for
more examples) :
Rud. 396 (Iamb.) Et aurum et argentum fuit || lenonis omne ibidem.
In fact Terence, who uses the word only once, gives it the long quantity :
Andr. yjy Provolvam teque ibidem pervolvam in luto.
ille (see 33).
Illyria (see Hilur-).
immo, Spondee and often Pyrrhic (see 58).
imposs (printed -os) (see 20).
ingratiis (see Gratia).
instituo. Perf. institui (Most. 86 -ui).
insnesco, trisyllable (see Suesco).
istac tenus (see Ea tenus).
ita (see 4).
Iuppiter, probably (see 14).
iurigo (see Obiurigo).
iurgium. No clear trace of ' lurigium ' (see 27).
iuxta, presumably (see 41), although the occurrences in Plautus give no clue
to the quantity. So scan :
Trin. 197 (Sen.) Iuxtaque earn euro cum mea. Recte facis.
lact{e) (see 20).
languor, two syllables (Asin. 574 ; Pseud. 216; Ter. Heaut. 807). A trisyl-
labiclanguor
' ' is doubtful :
Lucilius 391 Languor obrepsitque pigror torporque quietis (see 14).
laridum, three syllables in Plautus (Capt. 847 ; 903 ; 907 ; Men. 210) ; two
in Lucilius 79 :
Ut lurcaretur lardum et carnaria fartim.
Lartius, Bacch. 946 (Troch.) ; Trag. inc. 90 (Sen.):
|| idem Ulixes Lartius.
Nisi siqua Ulixes lintre evasit Lartius.
latro. Sisenna seems to be wrong when he says (apparently referring to
some line of the Rudens) : ' latronem ' producit metri causa.
lien. Since e is known to be long at a later time (cf. Mart. Cap. 3, 279 ;
Priscian Gram. Lat. 2, 149, 7), we may suppose it long in Plautus and scan
with Synizesis (II 33) :
Cure. 220 (Sen.) Nam iam quasi zona liene cinctus ambulo ;
244 (Sen.) Lien dierectust. Ambula, id lieni optimumst (Less
probably Lien by Brevis Brevians Law. Cf. Cure.
236; Merc. 124);
Cas. 414 (Troch.) Perii ! cor lienosum, opinor ||.
Compare rien, a disyllable in Plaut. frag. 1 10 (Troch.) || glaber erat tam-
quam rien ; Plur. renes (presumably rtenes), Cure. 236 (Sen.) :
Sed quid tibi est ? Lien enicat, renes dolent.
Priscian (1. c.) says : In -en producta Latina generis sunt masculini lien,
rien vel ren, et splen, splenis.
2o4 EARLY LATIN PROSODY

magis normal before a vowel, mage (magi') normal before a consonant.

There are some 230 occurrences of the word in Plautus, of which some 40
are at the end of a line or sentence (Pseud. 873) or hemistich (Stich. 298 ;
773). For examples of the normal usage we may take :
Mil. 1263 (Iamb.) Non edepol tu ilium magis amas ||;
1437 (Troch.) Mage metuant, minus has res studeant ||;
Rud. 1 1 80-1 (Troch.) Quae ex te poterit argumentis || hanc rem magis
exquirere,
Quae te mage tractavit mageque || signa pernovit
tua.
The abnormal occurrences are :
(a) Mage before vowel :
P: An.(mage
Asin. 66 (Sen.) Quipp' qui mage amico utantur gnato et benevolo I.) ;

394 (Iamb.) Nihilo mage intus est. Ubi est ? || Ad dixit (Ditto)ire;
tonsorem

Men. 386 (Troch.) Accipedum hoc iam scibo utrum haec me || mage
amet an marsuppium (Ditto) ;
Poen. 276 (Troch.) Quid habetis qui mage immortales ||vos credam esse
quam ego siem ? (Ditto) ;
461 (Sen.) Contentiores mage erunt atque avidi minus (mage-
runt A : mage erunt P) ;
Trin. 1053 (Troch.) Si mage exigere occipias, duarum || rerum exoritur
optio (mage AP) ;
True. 177 (Iamb.) Mage amat corde atque animo suo, || si quidem habes
fundum atque aedes (mage P : An. I.) ;
662 (Sen.) Argentum ad hanc, quam mage amo quam matrem
meam (Ditto) ;
887 (Troch.) Quern ego ecastor mage amo quam me, || dum id
quod cupio inde aufero (Ditto) ;
918 (Troch.) Quern ego ecastor mage amo quam te. || Quam me?
(Quam te.) Quo modo ? (Ditto).
(b) Magis (Iambus) before consonant :
Asin. 573 (Iamb.) Ubi amicae quam amico tuo || fueris magis fidelis ;
Bacch. 620 (Bacch.) Malos quam bonos par || magis me iuvare ;
Cas. 215 (Cret.) Mox magis quom otium || (et) mihi et tibi erit ;
Men. 576 (Bacch.) Magis quaeritur quam || clientum fides ;
Mil. 539 (Sen.) Magis facete vidi et mage miris modis;
Most. 702 (Cret.) Quom magis cogito || cum meo animo ;
Poen. 212 (Bacch.) Nam nullae magis res || duae plus nfigoti ;
Pseud. 1 21 4 (Troch.) Edepol ne istuc mage magisque || metuo quom
verba audio ;
True. 457 (Bacch.) Mater dicta quod sum, eo || magis studeo vitae;
720 (Bacch.) Molestusne sum ? Nunc || magis quam fuisti.
 HIATUS 205
So scan :
Amph. 254 (Iamb.) Hoc adeo hoc commemini magis || quia illo die im-
pransus fui (Rather than Trochaic with a Proce-
1 leusmatic) ;
Capt. 466 (Troch.) Neque ieiuniosiorem || neque magis effertum fame
(Or nee m. MSS. ; II 57) I
Cure. 305 (Troch.) Haud mage cupi' quam ego te cupio. || O mea tunitas
oppor- ;

463 (Troch.) || magis ess' dicam nescio ;

Men. 594 (Troch.) Nee mage manifestum ego hominem || umquam ullum
fuit;
teneri vidi ;
Merc. 619 (Troch.) Non tibi istuc mage dividiaest || quam mihi hodie

Mil. 613 (Troch.) Gerimu' rem mage non potest esse ||;
615 (Troch.) Ouis homo sit mage meu' quam tu's ? ||;
Trin. 169 (Sen.) Adesurivit magis [et] inhiavit acrius.
(On Mil. 615 see II 55 ii.)
The MSS. spell mage only when there is Elision ; otherwise magis. (On
the question whether an editor should print magi1 as well as mage see 17 C.)
In Terence there are^some 80 occurrences, of which 14 are at the end of
a line or hemistich (Eun. 561 ; Ph. 328 ; 1024 ; Ad. 680). For an example
of the normal usage we may take :
Eun. 935 (Sen.) Nee mage compositum quicquam nee magis elegans.
The abnormal occurrences are :
(a) Mage before vowel :
Eun. 356 (Troch.) Turn mage id dicas. Ouidnam quaeso herc||le ?
Eunuchum. Illumne obsecro ? (magis MSS.).
(Hardly Ad. 56.)
(b) Magis (Iambus) before consonant :
Eun. 507 (Sen.) Profecto quanto mage magisque cogito ;
Ad. 179 (Iamb.) Qui tibi magis licet meam habe||re, pro qua ego argen-
tum dedi ? (II 44).
The MSS. spell always magis.
It is noteworthy that the magis vis (i. e. mage vis) of the Codex Bembinus
has been ' modernized ' to mavis by one group of MSS. (to malis by another)
in Heaut. 396. The etymological spelling (was it magvis ?) of this verb seems
to have been a feature of the ancient standard edition of Terence (cf. Hec.
259 ; 473). (Cf. 40 L).
macchaera, probably (see 28).
male. There is no trace of the older ' male '. Scan :
Amph. 572 (Cola Reiziana) Merito male dicas || mi si id ita factum est ;
and possibly :
Cure. 508 (Iamb.) Vos faenori, hi male suaden||do et lustris lacerant
homines (with Hiatus at pause ? But cf. II 59) ;
206 EARLY LATIN PROSODY
Read :
Aul. 658 (Troch.) || haud male egit gratias (agit MSS.)
Pseud. 133 is presumably trochaic, not anapaestic :
Exite, agite exite, ignavi || male habiti et male conciliati.
In Cure. 622 the emendation male te (te male MSS.) is impossible. The
second foot of Aul. 208 (Troch.) is of course a Tribrach (II 32) :
Nimi' male timui. priu1 quam intro ||redii, exanimatus fui.
malesuada, quadrisyllable (see Suadeo).
malum * plague on it ' was apparently subordinated in accentuation to a pre-
cedingasmonosyllable
such a line : (as was em'm, or modo, or dum). It is a Pyrrhic in
Bacch. 696 (Troch.) (Facile effeceris.) Quam, malum, facile, quern
mendaci || prendit manifesto modo ? (AP) ;
and when the final syllable takes the ictus, it is merely under the same cir-
cumstances as would any Cretic word or word-group, e. g. :
Amph. 403 (Troch.) Quid, malum, non sum ego servus || Amphitruonis
Sosia?
Elision is frequent, e. g. :
Amph. 626 (Troch.) Qui, malum, intellegere quisquam || potis est ? ita
At the Diaeresis the ictus malum is natural : nugas blatis.
Rud. 945 Cave sis malo. quid tu, malum, || nam me retrahis audi.
But scan Most. 34 (Sen.) :
Quid tibi, malum, me(d), aut quid ego agam curatiost ? (Emphatic me).
manipulus (Cure. 585, end of line \ Ter. Eun. 776) ; maniftularis (True. 491)
and maniplaris (Mil. 815 ; Most. 312 ; 1048) ; maniplatim (Pseud. 181).
mansues, disyllable (see Suesco).
mater (see 14).
mediocris, suggested (not proved) by :
Accius 618 (Sen.) Profecto hauquaquam est ortus mediocri IV satu8 A).
(cf.

mehercle (3 syll.) :
Pseud. 1 175 (Troch.) Strenue mehercle iisti||(?) ;
Rud. 1365 (Troch.) Bene mehercle factum est quom istaec ||sententia
res tibi ex;

Ter. Eun. 67 (Sen.) Haec verba una mehercle falsa lacrimula ;

416 (Sen.) Pulchre mehercle dictum et sapienter. papae ! .
So scan :
Stich. 250 (Sen.) Ego illo(c) mehercle vero eo quantum potest (or 1II0).
mell (printed met) (see 20). Both this word and cor seem to have post-Plau-
tine scansion in :
Poen. 388 (Troch.) Huiu' mel, huiu' cor, huiu' labellum, hui||us lingua,
huius savium,
unless the emphatic Pronoun is made a monosyllable (cf. 34).
 HIATUS 207

melculum (not ' melliclum ') (see 26).

metuo. Perf. metui (Capt. 911 -ui).
miless (printed miles) (see 20).
viiluus, three syllables (e.g. Aul. 316,319); miluinus, four syllables (e.g.
Men. 212).
Minerva. No trace of a quadrisyllable, since Lato is shown by Varro (L. L
7, 16) to be the true reading in Racch. 893 (Sen.) :
Minerva, Lato[na], Spes, Opis, Virtus, Venus.
ministerium. No trace of ' minst- ' in Pseud. 772 (Sen.), where the word is
a mere conjecture :
Parvis magnisque miseriis praefulcior.
mis, tis (see 31).
Mil. 1033 (Anap.) Quia tis egeat, quia te careat ||;
Poen. 1 189 (Anap.) Rebus mis agundis, etc. (meis A).
How is this contradiction to be explained ? In Ennius Ann. 132 (where
Priscian attests mis Gen.).
Ingens cura mis cum concordibus aequiparare,
the suppression of -s in a monosyllable seems unlikely (16). Should we read
Cura ingens mis ?
modo, Adverb in Plautus (see II 20). In Terence always modo.
muli2r (see 14).
mundulus (26) ; hardly ' mundlus ' : —
True. 658 (Sen.) Nunc ego istos mundulos urbanos amasios (?) (Emended
to mundos).
namque before a vowel, nam before a consonant or a vowel. So namque is
never a trochee in the Dramatists, except :
Ennius trag. 370 (Troch.) Namque regnum suppetebat ||.
nauta and navita : —
Mil. 1430 (Troch.) || lanam nauta non erat ;
Men. 226 (Sen.) Voluptas nullast navitis, Messenio.
nescioquis (see 38 E).
neuter, a Trochee.
n(e)utiquam, i. e. ne utiquam. From Donatus' remark ' una pars est ora-
tionis ' we may infer that it was spelled as two words but pronounced as a
word-group, although it seems hard to separate it from numquam (i.e.
n(e)-umquam).
nil and nihtlum (-li, etc.) (see 11) ; sometimes nihil in Terence.
nimis normally a Pyrrhic before a consonant (as before a vowel). An
Iambus appears in :
Amph. 903 (Sen.) Nimis verecunda es. Potin ut abstineas manum ?
(II 45), (The usual phrase is Nimis iracunde) ;
Men. 760 (Bacch.) Fert, quas si autumem omnes, || nimis sermostlongu'
(?) ;
2o8 EARLY LATIN PROSODY

Pseud. 598 (Troch. ?) || argentum. nimis velim (IV 16) ;

1019 (Sen.) Nimisque ego ilium hominem metuo et formido male
(Or Nimi'que ego lllunc) ;
(cf. Pseud. 205 a).
The word occurs some 120 times in Plautus, some 10 times in Terence.
There is no trace of a spelling ' nime ' except the corruption in the minus-
cule MSS. at :
A n. I.)
Pseud. I274a (Ionic) Nimis ex discipulina, || quippe ego qui (nime P :

Probe Ionica perdidici. || sed palliolatim amictus.

Editors print nimist for the nimis est of the MSS. in Amph. 828 (end of line).
So scan :
Capt. 913 (Iamb.) Nimi'que hercle ego ilium male formi||dabam : ita
frendebat dentibus ;
Cas. 919 (Iamb.) Nimi' tu quidem herc||le immerito (?) ;
nisi (older spelling nesi), never ' nesi ' (II 37). The passages for which the
latter scansion is (wrongly) claimed will be found in Brock ' Quaestionum
Grammaticarum Capita Duo', pp. 170 sqq. (e.g. Most. 1006).
non and (Aul. 67) noenum (like nil and nihilum). Non enim (normally a
Dactyl) should not be altered to ' noenum '. Before a consonant non takes
the place of nonne, e. g. non vides ?, etc. (cf. Rhein. Mus. 52, 624, for
details).
non enim (see Enim).
nullus and nullum (like noenu' of Lucretius and noenum) are used for non,
e.g.:
Trin. 606 (Troch.) Non credibile dices. At tu || nullus edepol creduas ;
Cas. 795 (Sen.) Qui amat, tamen hercle si esurit, nullum esurit.
nunciam (3 syll.) and, more rare, nunc iam (cf. iam nunc), e. g. :
Epid. 135 (Troch.) Illam amabam olim ; nunc iam alia || cura impendit
pectori ;
Amph. pr. 38 (Sen.) Nunc iam hue animum omnes quae loquar ad-
vert ite.
nuculeus, Capt. 655 (Troch.) Nuculeum amisi, retinui || pigneri putamina.
Presumably a quadrisyllable in its other occurrence :
Cure. 55 (Sen.) Qui<s) e nuce nuculeum esse vult frangit nucem.
obex. If the Ace. end? Pers. 203 (obieci MSS.) the word is objex in Plautus.
objicio, once obicio (see 23).
obiurigo and obiurgo. The Plautine instances are :
domum ;
Amph. 706 (Troch.) Hanc est obiurgare quae me || hodie advenientem

Bacch. 1020 (Sen.) Me obiurigavit plurimis verbis (Or Med ;

malis obiurgavit)

Cure. 171 (Troch.) Recte obiurgat. sane hau quidquamst || mage quod
cupiam iam diu ;
 k HIATUS 209
Merc. pr. 46 (Sen.) Obiurigare pater haec noctes et dies ;
321-2 (Sen.) Ne, sis, me obiurga : hoc non voluntas me impulit.
Quin non obiurgo. At ne deteriorem tamen ;
Trin. 68 (Sen.) Malis te ut verbis multis multum obiurigem (obiurgem
P : obiurgitem A)
70 (Sen.) Nemost. Quid tu igitur rogitas ten obiurigem ? (A
obiurgitem P)
96 (Sen. ?) Si id non me accusas, tute ipse obiurgandus es ;
680 (Troch.) Sed tu obiurgans me a peccatis || rapi' deteriorem in
viam.
The simple verb occurs once :
Merc. 1 19 (Iamb.) Et currendum et pugnandum et au||tem iurigandum
est in via.
In Terence only obiurgo, iurgo.
obsecro, a Cretic in all the numerous occurrences in Plautus, except Bacch.
1 130 (Bacchiac), Stich. 325 (Anap.), where it becomes a Dactyl in Prosodic
Hiatus :
Viden limulis, ob||secro, ut intuentur ? ;
Quisnam, obsecro, has || frangit fores ? ubist ? (?).
The Dactyl ' obsecro ' (II 27, n°. 4) is extremely doubtful in :
Aul. 715 (Anap.) Nequeo cum animo certum investi||gare. obsecro vos
ego mi auxilio (Rather ego vos, the usual order).
(Cist. 453 is as likely to be Trochaic as Iambic.)
omnis. The first syllable is never short except after a ' brevis brevians '
(II 29), e. g. Ter. Hec. 867 Omnia omnes. (Scan Trin. 236, 933 omnium
primum ; Poen. 834 omnia genera ; Ter. Ad. 971 omnibu' gratiam.)
opilio. Are we to scan opilio (with Hiatus) or opilio in Asin. 540 (Troch.) ?
Etiam opilio qui pascit, mater, alienas oves (Emended to ovipilio).
oraclum and oraculum (see 26) :
Men. 840 (Troch.) Vae capiti tuo ! Ecce Apollo || mi ex oraculo imperat
(Emended to ex oraclo mi) ;
Accius 624 (Troch.) Pro certo arbitrabor sortes, ||oracla, adytus, augura ;
Trag. incert. (Ennius ?) 13 (Sen.) Ibi ex oraclo voce divina edidit.
fiarum (originally Neut. of parvus). On the traces in MSS. of a phrase
'parvum est fides ', see Archiv. Lat. Lexicogr. 13, 133. In Most. 11 52 the
Archetype had (for panunper) parvomper.
pater (see 14).
Pellaeus (see 25). Asin. 333 (Troch.), 397 (Iamb.) :
|| mercatori Pellaeo ;
|| Pellaeo mercatori.
penes^ presumably peness (for -et-s), since otherwise it would be as often
pene' as nimis is nimi' (cf. True. 858). When the final is shortened, it is
under the Brevis Brevians Law, e. g. penes-sese (with emphatic Pronoun).
2348 P
 2io EARLY LATIN PROSODY

per- 'very' seems to have been a separate word (like perquani). So the
accentuation was presumably peropus (not 'peropus'), pervelim (not
' pe'rvelim '). Therefore Scipio should not have been bantered for his
preference oipertaesus to fertisus by Lucilius (see Latin Language, p. 195).
perduellis, a trisyllable in Amph. 250 ; Cist. pr. 201 ; Mil. 222 ; Pseud. 583,
and presumably 589 ; Ennius Trag. 385 ; Accius praet. 12. Conceivably
four syllables in Amph. 642 (Bacch.).
periclum and periculum (see 26).
pirinde (like PhilTppus. The accentuation of the first syllable is attested by
Priscian Gramm. Lat. iii, 67 ; cf. Serv. ad Aen. vi, 743). It is replaced by
proinde occasionally in our MSS., e. g. :
Stich. 520 (Troch.) Ut quoique homini res paratast, || perinde amicis
utitur (proinde A) ;
100 (Troch.) Perinde habetis quasi praesentes || (Proinde A).
The remaining Plautine example (Pseud. 577) is also a tribrach if the metre
is Iambic Monometer :
Res perinde sunt.
But Terence scans perinde in Heaut. 195 (Iamb.) :
Atque haec perinde sunt ut il||li(u)s animus est qui ea possidet.
perpluo (see Pluo).
persuadeo, quadrisyllable (see Suadeo).
Philipfius (Noun), Philippens (Adj.), in speaking of the coin (II 40), e. g. :
Poen. 714 (Sen.) Trecenti nummi qui vocantur Philippei ;
781 (Sen.) Qui ad te trecentos Philippeos modo detulit (-pos P).
In True. 928 (Troch.) we seem to have a derivative Verb :
Philippeari satiust, miles, || si te amari postulas (Nihiliphiari and Ni-
hilpphiari MSS.).
Philoxenus (never -ox- : see II 31).
Phrygio and possibly Phyrgio. The latter in Aul. 508 (Sen.) :
Stat fullo, phyrgio, aurifex, lanarius (Or phrygio, with Hiatus at pause).
The former in all the other (five) occurrences in Plautus, all in the Me-
naechmi, e. g. Men. 426 (Troch.) :
|| ad phrygionem ut deferas.
pius (see 25).
platea (see 25).
pluo, Perf.plui (Men. 63 -uerat). Perpluo, Perf. -ui (Most. 164 -uit).
poculum and poclum (see 26), e. g. Cure. 368 (Troch.), 359 (Troch.) :
Poculum grande, aula magna ||;
|| magnum poclum : ille ebibit.
poples. Has it a Parasitic Vowel in Accius 456 (Sen.) ?
Cave lassitudo populitum cursum levet (popKt MSS.).
populus andpoplus (see II 39).
porcus, etc. ; once proculena (II 40), Mil. ic6o (Anap.) :
|| proculenam impertiturust.
 HIATUS 211

porgo and porrigo (?), exporgo :

Epid. 733 (Troch.) Plaudite et valete, lumbos || porgite atque exsurgite ;
Merc. 883 (Troch.) Porrige bracchium, prehende || (Emended to Porge) ;
Pseud. 708 (Troch.) Contra pariter, porge audacter || ad salutem brac-
chium.
Pseud, pr. 1 (Sen.) Exporgi meliust lumbos atque exsurgier :
Plautina longa fabula in scaenam venit.
post and poste. To the words whose -e is * squeezed out ' before a following
initial consonant, this Adverb-Preposition may be added. For there are
abundant traces of the fuller form in the MSS. of Plautus and Terence, but
(like nempe) the word never seems to make a Trochee, except in :
? Ter. Andr. 483 (Bacch.) || lavet post(e) deinde (Hardly deinde. Read-
ing doubtful) ;
Most. 290 (Troch.) Poste[a] nequiquam exornata est || (Or post ea ?).
In Ennius' Annals we have this instruction of rowers :
230 Poste recumbite vestraque pectora pellite tonsis.
potts and pote (see 40 K).
praeco; once, perhaps, praedico (the original form), Liv. trag. 41 (Sen.) :
Quinquertiones prae(di)co in medium vocat.
praestigiator (Poen. 1125), praestigiatrix (Amph. 782; True. 134). Appa-
rently also praestigjator (like Horace's vindemjator; 23) : modis ;
Aul. 630 (Troch.) Ego edepol te, praestigiator, || miseris iam accipiam

Cist. 297 (Senar.) Praestigiator es, siquidem hie non es atque ades (23).
(Amph. 830 admits both scansions : see above, Nescioquis.)
firaeut (or rather prae ut ; see the beginning of this paragraph).
prehendo and prendo. Prehendo usual in Plautus (Amph. 716; 11 16; Asin.
563 ; 668 ; Aul. 749 ; Merc. 213 ; 883 ; Mil. 1426 ; Most. 219 ; Pers. 294 :
Poen. 375 ; 1232; Pseud. 725 ; 1260; Rud. 91 ; 107 1 ; 1291) ; prendo,
which is normal in Terence, is occasional in Plautus (Asin. 569 ; Bacch.
696 ; 950 A ; Cure. 339 ; Epid. 326).
prior, prius. The oldest quantity was probably (not certainly) ' prior ' ; and
traces are not lacking :
Cas. 571 (Sen.) Rogitare oportet prius et contarier (AP) ;
839 (Bacch.) Meast haec. Scio sed || meus fructus[es]t prior
(Emended to fructu' prior est).
Pseud. 578 (Anap.) || nam ego in meb pectore prius (cf. Amph. 545 ;
Bacch. 932 ; Cas. 378 ; Rud. 455 ; Pacuvius 198).
On the other hand prior, prius are extremely ^common, and the pyrrhic
nature of the Adverb is shown by the normal suppression of -s (see 16) ;
for priu' goes with nimi', etc. It is an Iambus before a consonant only in :
Men. 276 (Sen.) Prius iam convivae ambulant ante ostium ;
Most. 867 (Bacch. ?) Prius quam ut || meum (?) ;
Poen. pr. 67 (Sen.) Sexennio prius quidem quam moritur pater (II 45) ;
P 2
212 EARLY LATIN PROSODY
also, of course, at the end of a line (or hemistich : Amph. 240 ; Mil. 404.
Cf. Most. 326). The MSS. haveyfc in Ennins trag. 239 (Troch.) :
Inspice hoc facinus priusquam || fiat : prohibessis scelus.
Even in his Annals Ennius makes the word a Pyrrhic :
219 In somnis vidit priu' quam sam discere coepit;
480 Hortatore bono priu' quam iam finibu' termo.
And apparently Lucilius in a Trochaic Septenarius :
642 Nequam priu' quam venas hominis || tetigit ac praecordia.
Therefore in the second century B.C. (at least) the pronunciation was
Priu' quam (not priusqua7ii).
pro euro (see 29).
profecio and sometimes profecto (II 18 D).
proficiscor (see 29).
projiteor and prof- (see 29).
prohibeo (prob- ?) (see 22).
proinde and proin (see II 36).
prologus. And prologus (Ter. Phorm. pr. 14) ?
propello (Rud. 672 ; Lucilius 259).
propino (see 29).
proprius (see 25).
protervus, Adv. -ve (apparently never prot- in Plautus nor Terence), e. g. :
Bacch. 612 (Troch.) Petulans, protervo, iracundo ||;
Ter. Hec. 503 (Sen.) Ecce autem tu quoque proterve iracundus es (II 47).
Similarly Pacuvius 147 (Troch.) :
|| spectu protervo ferox.
protinus and protinam. Notice that in Ter. Ph. 190 (Troch.) :
Aliquid convasassem atque hinc me || conicerem protinam in pedes,
the unfamiliar form is attested by Paul. Fest. 253, 17 and Donatus (ad loc.
'sic Nigidius legit'), but the familiar form has ousted it not merely in all
the MSS. of Terence but in those of Nonius and Arusianus Messius.
puella. And possibly puella, Cist. pr. 124 (Sen.), Poen. 1301 (Troch.) :
Puellam proiectam ex angiportu sustuli ;
Non pudet puellam amplexari || baiolum in media via ?
(Cf. Ter. Heaut. 1060 ?).
purigo andpurgo. The occurrences in Plautus are :
Amph. 909 (Sen.) Et id hue reverti uti me purgarem tibi ;
945 (Sen.) Verum eadem si idem purgas mi, patienda sunt ;
Aul. 753 (Troch.) Non mi homines placent qui quando || male fecerunt
purigant (purgitant MSS.);
791 (Troch.) Quin pudeat, quin purget sese ||;
Cas. 944 (?) ... purgem scio me meae uxori ;
Cist. 384 (Sen.) Quae quasi carnificis angi porta purigans (purgitans MSS.);
Merc. 739 (Sen.) Nil hercle istius quidquam est. Numero purigas ;
Poen. 1410 (Troch.) || tibi me purgatum volo.
In Terence only purgo (and purgatio Heaut. 625).
 HIATUS 213

Similarly expurigo and expurgo. In Plautus :

Amph. 965 (Troch.) Habui expurigationem ||;
Capt. 620 (Troch.) Sed h6c primum, me expurigare || (or med expur-
Cist. 302 (Sen.) Expurges, iures, ores blande per precem ; gare)
304 (Sen.) Expurigabo hercle omnia ad raucam ravim ;
453 (Troch. ?) || nil moror. Expurgare me;
Merc. 960 (Troch.) At ego expurigationem ha||bebo at ne suscenseat ;
Mil. 497 (Sen.) [Ex]purgare void me. Tun ted expurges mihi ?
517 (Sen.) Med expurgare haec tibi videtur aequius (Or Me ex-
In Terence only expurgo. purigare).
The Compound with per" occurs once :
Mil. 774 (Troch.) Quam institui. Perpurigatis || damu' tibi ambo operam
auribus (Or -tui perpurg.).
piitus (Pseud. 1200) (see II 56).
quadrigenti (see II 40). Here are all the occurrences of the word:
aureis ;
Bacch. 934 (Iamb.) Qui misere male mulcabere || quadrigentis PhilTppis

974 (Iamb.) Quadrigentos filios habet at||que equidem omnes

lectos sine probro ;
1 183 (Anap.) Quadrigentis Philippis filiu' me circumduxerunt
et || Chrysalu' ;

Rud. 1324 (Iamb.) Quadrigentos. Tramas putidas. || Quingentos. Cas-

sam glandem.
The spelling of the MSS. is -ing- (but cf. Bacch. 934 ; 1 183).
quadrupulum, quadrupulator^ quadrupulari (see 26). Never -pi- in
Plautus :
Pers. 62 (Sen.) Neque quadrupulari me volo, neque enim decet (II 44) ;
70 (Sen.) Ubi quadrupulator quempiam iniexit manum (II 44) ;
True. 762 (Troch.) Postid ego te manum iniciam || quadrupuli, venefica.
But quadruplico (Stich. 405) (Cf. Cure. 619).
quandoquidem. Rarely -do- in Plautus :
Stich. 483 (Sen.) Quando quidem tu ad me non vis promittere ;
559 (Troch.) || ill* senex, quando quidem.
quasi and quam si (see II 37).
quid ita (see 4).
quiesco, quietus. Did Plautus allow ' quiesco ' ' quietus ' ?, e. g. :
Merc. 448 (Troch.) Ouiesce, inquam, istanc rem ego recte || videro
Quid ais ? Quid est ? ;
Epid. 338 (Iamb.) Per hanc curam quieto tibi licet || esse, hoc quidem
iam periit.
qui quidem. Also quiquidem ? (see II 37).
quis and qui (Interrog.) (see 38).
2i4 EARLY LATIN PROSODY
quoad, presumably a Pyrrhic (even before a consonant, under the Brevis
Brevians Law) :
Men. 769 (Bacch.) Verum est modu' tamen quoad || pati uxorem oportet ;
Pseud. 623 (Troch.) Praestitutast, quoad referret || nobis, neque dum ;
rettulit

Ter. Ph. 148 (Sen.) Quoad expectati' vestrum ? Non certum scio ;
462 (Sen.) Percontatum ibo ad portum, quoad se recipiat.
Unless it is a monosyllable, as in Lucretius (2, 850 Quoad licet ac possis ;
cf. Lachmann on Lucr. 5, 1033) and Horace (Sat. 2, 3, 91) ; which would
suit the Bacchiac Metre of the^first line. But when a Preposition is put
after a Relative (or Interrogative), Prosodic Hiatus seems to be the rule in
Plautus : e. g. Bacch. 176 quern ad (53). This favours quoad, a Pyrrhic.
quotidie (see cotidie).
(ii 7) ;
rdvis. Aul. 336 (Sen.) Ubi si quid poscam, usque ad ravim poscam prius

Cist. 304 (Sen.) Expurigabo hercle omnia ad raucam ravim.

Should we read ' ravio ' or 'arvio' in Poen. 778 (Sen.) ?
Nego ; et negando, si quid refert, ravio (ra-i3; ra- or ar- A. Alliteration
favours ra-). Marx (in a note on Lucilius 1289) argues for the identity of
ravus ' grey ' with ravus ' hoarse ' on the plea that Greek (paios had these
two senses.
redducoy always. What of red{d)ux1 Here are the occurrences :
Capt. pr. 43(Sen.) Reducemque faciet liberum in patriam ad patrem ;
437(Troch.) || reducem facias filium ;
686(Sen.) Reducem fecisse liberum in patriam ad patrem ;
923(Bacch.) Quom te redducem tub || patri reddiderunt
(Emended to Quom reducem tuo te) ;
931 (Troch.) Tibique hunc reducem in libertatem ||.
Rud. 909 (Bacch.) Templis redducem, plu||rima praeda onustum
(Emended to Templisque reducem).
(Add Trin. 823, an anapaestic line whose reading is uncertain.)
Ter. Heaut. 398 (Iamb.) Ergo, mea Antiphila, tu nunc so||la reducem me
in patriam facis ;
Hec. 852 (Troch.) Egon qui ab Oreo mortuum me || reducem infeceris lucem ;

Naevius praet. 1 (Sen.) Vita insepulta laetus in patriam redux.

rZlicuus, always.
ten (see Lien).
reprehendo and reprendo. Of the seven Plautine occurrences (Bacch. 364;
Epid. 1 ; Mil. 60 ; Most. 862 ; Pseud. 223 ; 249 ; Trin. 624) all admit the
quadrisyllable and four require it (Mil. 60; Most. 862 ; Pseud. 223 ; 249).
The trisyllable is required by the metre in Ter. Ph. 863 ; Lucilius 573.
res. Gen. rei, rei and rei ; Dat. rei ; Abl. re, hardly • red ' (see 30 K. L. M.).
rubrica. True. 294 (Troch.) :
Buccas rubrica, creta omne || corpus intinxti tibi.
 HIATUS

rudensy Rud. 1015 (Troch.) :

Mitte rudentem, sceleste ||.
Presumably also in Rud. 938 (Iamb.), 103 1 (Troch.) :
|| quam trahi' rudentem complico ;
Ut abeas, rudentem amittas ||.
And in Pacuvius 336 (Troch.) || et rudentum sibilus.
sacri- (3 Decl. Adj.), Rud. 1208 (Troch.) :
Sunt domi agni et porci sacres || (Cf. Men. 290).
saeclum and (rarely, or never) saeculum (see 26) :
Aul. 126 (Bacch.) (Aut) hodie dicunt muli||erem (aut) ullo in saeclo;
Mil. 1079 (Anap.) || vivunt ab saeclo in saeclum ;
Trin. 283 (Cret.) Novi ego hoc saeculum || moribus quibu' siet ;
True. pr. 13 (Sen.) Haec hum' saecli mores in se possidet;
Ter. Eun. 246 (Troch.) || quaestus apud saeclum prius ;
Ad. 304 (Troch.) Hoccin saeclum ! o scelera, o genera ||;
Caecilius 210 (Bacch. ?) . . arbores quae al||teri saeclo prosint.
sagitta (II 40). The occurrences of the word are :
Aul. 393 (Sen.) Confige sagittis fures thesaurarios (II 48) ;
Pers. 25 (Iamb.) Sagitta Cupido cor meum || transfixit. lam servi
amant hie
?;
Trin. 725 (Troch.) || pharetram et sagittas sumpsero.
And Nonius attests nivit (for ninguif) in a line of Pacuvius, in which the
archetype of our MSS. seems to have omitted this Verb. Perhaps we
should make the line trochaic and read :
Pacuv. praetext. 4 . . . sagittis || plumbo et saxis grandinat.
salvus. Never a trisyllable.
sarrapis (see II 7).
satelless (printed -es) (see 20; II 40). Or- is the shortening in the Trin.
line like simillimae-sunt (II 29) ?
Mil. 78 (Sen.) Age demus ergo, sequimini, satellites ;
Trin. 833 (Anap.) Distraxissent disque tulissent || satellites tui me
miserum foede.
satis and sat. Since these words occur some 190 times in Plautus (not to
mention the numerous occurrences of the Interrogative satin) we can
determine his usage with some confidence. The (presumable) original dis-
tinction between sat (i. e. ' sate ') preconsonantal and satis prevocalic has
(if it ever existed) been completely lost, e. g. :
Poen. 458 (Sen.) Ouando id quod sat erat satis habere noluit (AP) ;
Pers. 839 (Troch.) Nee sati* liber sibi videtur || nee sati' frugihonestus nee sat;

The disyllable is usually a Pyrrhic sati1 (sate ?) before a consonant. It is

an Iambus in :
Amph. 168 (Ionic) Noctesque diesque assidu||o satis superque est ;
648 (Bacch.) Satis mi es||se ducam ;
2i6 EARLY LATIN PROSODY
Asin. 446 (Iamb.) || suo odio. Heus iam satis tu ;
Bacch. 627* (' Glyconic ' Dactyls) Sanu' satis non es. Perii ;
Epid. 346 (Iamb.) Quantum hie inest ? Quantum satest || et plus satis ;
superfit (At a pause) ;
Mil. 584 (Sen.) Nam uni satis populo impio merui mali (II 47) ;
Poen. 215 (Bacch.) Neque is ulla ornandi || satis satietas est ;
227 (Bacch.) Poplo quoilubet plus )| satis dare potis sunt ;
Pseud. 914 (Troch. ?) Istuc ego satis scio. Cur || ergo quod scis me
rogas ? ;
Trin. 227 (Bacch.) Sed hoc non liquet nee || satis cogitatumst (cf. 260) ;
321 (Troch.) Qui ipsu' sibi satis placet nee ||;
Ter. Eun. 577 (Iamb.) Commendat virginem. Cui? tibi||ne? Mihi.
Satis tuto tamen ;
Heaut. 198 (Iamb.) Vereor quam ne quid in ilium ira||tus plus satis le d.) \
faxit,

Ph. 915 (Sen.) Satis superbe illuditis me. Qui? Rogas? (Satinpater
var. ;

(Sati'Tribrach
Ad. 315 (Iamb.) Satis mi id habeam supplici a||false mihi woul give;
d 47)
; II
(Not Ad. 309.)
Lucilius 1 1 47 Cui, si coniuret, populus vix totu' satis sit.
There is no trace of a spelling 'sate'. (In Stich. 620 id mi sat e rest loci
' enough room under the circumstances '.)
The monosyllabic form sat appears in MSS. before vowels as well as
consonants, e. g. :
Aul. 560-1 (Sen.) Turn obsonium autem. Pol vel legioni sat est.
Etiam agnum misi. Quo quidem agno sat scio.
Notice that the (Anapaestic) metre requires satis habeo (Mil. 1063), sat
amabo est (Mil. 1084).
Also that both A and P offer satis (where sat would make smoother
metre) in the Cretic line :
Cas. 189 Hem ! quid est ? die idem. || nam pol hau sati' meo;
but sat where the disyllable would intrude a Proceleusmatic into a Trochaic
line (e. g. Most. 927 Sat sapio AP). At the end of a line (or first hemistich,
e. g. Ter. Andr. 692 ; Eun. 547) the disyllable always appears. Perhaps
editors should print satis and suppose the final syllable shortened under the
Brevis Brevians Law (17 C) in Men. 603 (Anap.), where it is emphatic and
makes a sentence by itself (17) :
|| satin audis quae illic loquitur ?
Satis. Si sapiam, hinc intro abeam ubi mi || bene sit. Mane : male erit

scapulae', and (possibly) scaplae in Asin. 552 (Iamb.) :

Qui saepe ante in nostras scaplas || cicatrices indiderunt (IV 10).potius.
scelestus ; never ' scelestus ' (see II 18 B).
schema, -ae (Amph. 117 ; Pers. 463 ; Caecilius 57 ; 76).
 HIATUS 217

secus (see Alterim secus).

seorsum (see Deorsum).
sicut% sicuti (see Ut).
sinister. Perhaps only -teri in Plautus ; -tri in Terence and (allowed at end
of line) -teri. Here are all the examples :
Merc. 880 (Troch.) || aspice ad sinisteram ;
Ter. Eun. 835 (Sen.) Habemus hominem ipsum. Ubi is est? sinisteram
Em ! ad;

775 (Troch.) Simalio, in sinistrum cornum ; || tu, Syrisce,

dexterum in;

Ad. 582 (Troch.) Ad sinistram hac recta platea, u||bi ad Dianae veneris.
We would scan with the full form :
Epid. 183 (Sen.) Liquido exeo foras auspicio ave sinistera (-tra AP) ;
Pseud. 762 (Troch.) Ave sinistera, auspicio || liquido atque ex sententia
(-tra P Gram.),
siquidem (cf. Sonnenschein's note on Rud. 1061 for details).
slqtcis. Perhaps never ' siquis ' (38 A). One word ; since Ennius scans
(Ann. 536) siqui' ferat. For a monosyllable would not lose (or weaken)
a final consonant. But also two words ; see II 54 ; 7 b.
sis ' please ' (see 24), joined into a word-group with a preceding Imperative,
e. g. properd-sis. Cicero tells us that cape-sis became capsis. An iambic
Imperative was pronounced as a Pyrrhic in this word-group, e. g. tene-sis,
tace-sis. So scan with Hiatus at the pause :
Pers. 413 (Sen.) Tene sis argentum ; etiam tu argentum tenes ? (AP).
sive. When scribes write this full form where a monosyllable is required by
the metre, should we print seu (cf. Amph. 69-71), viewing it in the light of
such spellings as atque for ac, siet for sit ? Or siv\ like quiv' Amph. 84 ?
soror (see 14).
sorsum (see Deorsum).
Sosia (see 30).
sospess (printed -es) (see 20).
spes. Gen. spei in Terence always :
Ter. Andr. pr. 25 (Sen.) Ut pernoscatis ecquid spei sit relicuum ;
Eun. 1053 (Troch.) Mihi tllam laudas ? Perii ! quanto || minu' speist,
tanto magis amo;
Heaut. 659 (Troch.) Si potis est reperiri. Interii ! || plus spei video
quam volo ;
Ph. 474 (Iamb.) || ecquid spei porrost ? Nescio. Ah !
Uat. spei presumably in Caecilius 25 (Sen.) :
Nil spei ego credo : omnes res spissas facit (ego spei or rei ego MSS.).
st! has the worth of a long syllable (cf. Hauler's note on Ter. Ph. 743 for
details).
 2l8 EARLY LATIN PROSODY

statim. Nonius (393, 5) attests statim for Ter. Ph. 790 (Iamb.) only:
Capiebam statim, etc.
He is apparently using a marginal note in his copy of Plautus. Elsewhere
(254, 5) he cites the line from his copy of Terence and gives it the form it
has in our MSS. :
Statim capiebat, etc.
Therefore ' statim ' is very doubtful.
statura. Lucilius 794 :
Quare pro facie, pro statura Accius . . .
status, Participle (Cure. 5 ; presumably Ennius trag. 253s) ; status, Noun
(e.g. Amph. 266 ; Mil. 206; Pseud. 1288; presumably Ter. Eun. 598), but
apparently status in Mil. 1389 (Sen.) :
Paratae insidiae sunt : in statu stat senex.
stomachus (see 28).
strix, -Igis, Pseud. 820 (Sen.) :
Non condimentis condiunt, sed strigibus
Vivis convivis intestina quae exedint (cf. Lachmann on Lucr. 1, 360).
Did Plautus write stringibus (cf. Festus 414, 23) ?
suadeo, a trisyllable. There are some forty occurrences of this verb and its
compounds in Plautus, and the few traces of a quadrisyllable are illusory :
Cure. 508 (Iamb.) Nos faenori, hi male suadendo et lustris lacerant
homines (male su- would be a false Tribrach ; II 44) ;
Rud. 875 (Sen.) Sequere, obsecro, m(e>. Pariter suades qualis es;
879 (Sen.) Manete dum ego hie redeo. Equidem suadeo ;
Most. 213 (Iamb.) || malesuada viti lena (-ada vi- would be a false
Dactyl ; II 53. Emended to viti malesuada lena).
suavis, a disyllabic
sticula, Rud. 1170 (Troch.) :
Sucula. Quin tu i dierecta || cum sucula et cum porculis (26).
suesco (only found in compounds), a disyllabic There are traces of a tri-
syl able :
Asin. 79 (Sen.) Patres ut consuerunt. ego mitto omnia haec (Emended
to consueverunt) ;
727 (Iamb.) Ut consuere, homines Salus || frustratur et Fortuna
(Emended to consuevere).
Since the occurrences (including mansues) are only fifteen in Plautus, the
choice is open to allow the trisyllable in these two lines or to find in them
a very natural miswriting (ue for ueue). The nine occurrences in Terence
all shew ue.
sum (see 40 I).
supra (see 4). If supera ' is a pseudo-archaism of Lucretius, it must not be
foisted on Plautus (Cas. 815?). It would indeed save the metre (IV 16) in
Cure. 477 (a line of that entr'acte which some call post-Plautine).
 HIATUS 219

surrtpio. The shortening of this Verb (surpuerat Horace Carm. 4, 13, 20;
surpere Lucr. 2, 314) is as early as Plautus. He, however, confines himself
to surpui and that only at the end of a line (three times in all : Capt. 8 ;
760; 1 on), i.e. through metrical necessity (elsewhere surrupui, e.g. Aul.
822). There is no authority for ' surptus \ We should scan :
Pers. 150 (Sen.) Qui sibi parentes fuerint und' surrupta sit (unde Rel. =
de quibus ; II 36)
380 (Sen.) Et iit vi surrupta fueris. Docte calleo ;
Poen. 902 (Troch.) Ibidem gnatust, ind' surruptus ||;
1058 (Sen.) Surruptu' sum illinc. hie me Antidama, hospes tuus
tabemaclttm and tabernaculum :
(30 A).
Amph. 428 (Troch.) Quid in tabernaclo fecisti ? || (cf. 426) ;
Trin. 726 (Troch.) || placide in tabernaculo.
talentum. The word so often occurs with the second syllable long that,
although the scansion of Philippus (<&L\urno$) as Philippus would justify the
scansion of talentwn (rakavTov) as talentum, the following hardly inspire
confidence :
Mil. 106 1 (Anap.) Dabitur quantum ipsuJ preti poscet. || Talentum
Philippi huic opus auri est (Emended to tantum,
in deictic use, as in Merc. 7) ;
Ter. Ph. 791 (Iamb.) Ac rebu' vilioribus || multo tamen talenta bina.
Hui ! {var. led. tamen duo talenta).
Scan Cist. 561 (Sen.) :
Und' tibi talenta magna viginti pater.
tametsi (see the beginning of this paragraph).
tarpessita (II 40), proved by :
Cure. 341 (Troch.) || tarpessitam noverim ;
406 (Sen.) Lyconem quaero tarpessitam. Die mihi ;
712 (Troch.) || tarpessita. Non taces ?
So read in Epid. 143 (Troch.) :
Die modo : unde auferre me vis ? || quo a tarpessita peto ? (a quo MSS.).
No trace of trap-.
techina (28), required by metre in three of the four occurrences in Plautus,
may be presumed for the fourth (Poen. 817) :
Bacch. 392 (Troch.) Condigne is quam techinam de auro ||;
Capt. 642 (Troch.) Hui(u)s scelesti techinis, qui me ||;
Most. 550 (Sen.) Metuo ne techinae meae perpetuo perierint.
Similarly in Terence, required in one occurrence and presumable for the
other (Heaut. 471) :
Eun. 718 (Troch.) Parmenonis tarn scio esse hanc || techinam quam vivereme ;

Whether the Derivative Verb is of three or four syllables cannot be deter-

mined from its single occurrence :
Pseud. 1096 (Sen.) Vide modo ne illic sit contech(i)natus quidpiam.
220 EARLY LATIN PROSODY
templum (see Extemplo).
tenus (see Ea tenus).
tis (see Mis).
1 trafiezita * (see Tarp-).
valide, apparently the Plautine spelling, not ' valde ' ; although the metre can
give no clue, unless we disallow Syll. Anceps (at change of speaker) in the
rapid replies of Ballio (see 49) in Pseud. 364 (Troch.) :
Legerupa. Valide. Permities || adulescentum. Acerrime.
ubi (see 6 ; II 20).
vehiclum. Here are all the occurrences :
Aul. 168 (Troch.) Clamores, imperia, eburata || vehicla, pallas, pur-
puram ;
502 (Sen.) Salutigerulos pueros, vehicla qui vehar;
Pers. 782 (Anap.) Vehiclum argenti miser eieci ||.
veluf, veluti (see Ut).
videlicet^ in all the occurrences (Capt. 286; Mil. 1283; Most. 980 ; Stich.
555 5 557 5 Ter. Heaut. 263 j 514 ; Ad. 450) except :
Asin. 599. (Iamb.) Negotiosum interdius || videlicet Solonem (Emended
to videl. int.).
vidulus in all the numerous occurrences. In two some editors prefer ' vidlum '
to a Dactyl-word forming a foot of a trochaic line (II 58) :
Rud. 1 106 Quid id ad vidulum pertinet ser||vae sint istae an liberae ? ;
1 130 Estne hie vidulus, ubi cistellam || tuam inesse aiebas? Is est.
Add:
Rud. 1 1 27 Cedo modo mihi istum vidulum, || Gripe. Concredam tibi
(Emended to vid. ist).
vinclum in all the numerous occurrences. So scan :
Capt. 204 (Cret.) Nostrum erum si vos || eximat vinclis (IV 26 C).
ut and uti. The relation of the one-syllabled to the two-syllabled forms is
not certain. If we suppose uti (e.g. Ennius Ann. 178; 360) to be the
earliest and the full form, then uti is due to the Brevis Brevians Law (cf.
calefacio) and ut has advanced a further stage of shortening (cf. calfacio).
At the end of a line Plautus is therefore correct in using the disyllable, the
form suitable to a pause. And Terence's utin omnes (Hec. 199) preserves
the final before -ne as is the Latin rule (cf. sicin, hoccin, hicin). Since the
unfamiliar uti would be apt to be replaced by the familiar ut, an editor need
not hesitate to save the metre by substituting uti for ut. But he should not
sacrifice any uti which the MSS. have preserved, e.g. :
Pseud. 164 (Troch.) || lautaque coctaque omnia uti sint ;
Cist. 596 (Sen.) Debs teque spero. Eosdem ego — uti abeas domum.
(Cf. Capt. 115 ; Poen. 1184 ; 141 2 ; Pers. 575, and perhaps 259 ; Stich. 444.
 HIATUS 221

Possibly even in Pers. 290; IV 10 B). And {pace Lachmann) sicuti, veluti
are to be kept wherever the MSS. show these unfamiliar forms, e.g. :
Bacch. 1068 (Sen.) Hoc est incepta efficere pulchre veluti mi (Evenit) ;
Ennius Ann. 536 Sicuti siqui' ferat vas vini dimidiatum ;
Lucilius 198 Sicuti cum primos ficos propola recentes (cf. 1029 ; 1298).
The strictly correct spelling is vel uti (II 32 A), vel ut (II 45), and pre-
sumably sic uti and sic ut (II 54).
uvidus ; never udus. The occurrences are at the end of a line or in Cretic
Verse (Rud. 251 ; 263; 409; 573; 585; 942).
43. HIATUS. (A) This is the thorniest thicket in our investigation.
Since Cicero was a friend of Varro, the leading authority on the early
Drama, and had himself written verse on the model of early Epic and
Tragedy, we had better take him for guide : (Or. 152 ; on orators who
allow in a speech a word ending in a vowel to precede one so beginning)
Sed Graeci viderint; nobis, ne si cupiamus quidem, distrahere (i.e.
hiopi&w, 'leave an interval between') voces conceditur. indicant orationes
illae ipsae horridulae Catonis, indicant omnes poetae praeter eos qui, ut
versum facerent, saepe hiabant, ut Naevius ' Vos qui accolitis Histrum
fluvium atque algidum ', et ibidem ' Quam numquam vobis Graii, atque
barbari ', at Ennius saepe (var. led. semel) * Scipio invicte ', et quidem
nos '.Hoc motu radiantis Etesiae in vada ponti '. hoc idem nostri saepius
non tulissent quod Graeci laudare etiam solent. (On the text see
Kroll in Rhein. Mus. 60, 554.) Cicero's remark on the older practice,
1 saepe hiabant ', is presumably not merely his own verdict, but rather
the common verdict of well-educated men of the time, or even of experts
like Varro.
There was then a great amount of Hiatus in early Latin verse. Can
we reduce it (or some of it) to rule ?
Before the discovery of the Ambrosian Palimpsest (A) the problem
seemed easy. All extant MSS. of Plautus agreed (since all were copies
of the same ninth century archetype) in such a reading as Trin. 18 (Sen.) :
Huic nomen Gmece est Thensauro fabulae.
This line was accordingly accepted as one of the instances that Plautus
1 saepe hiabat '. The Palimpsest changed all that. It exhibited the
line in this form :
Huic Graece nomen est Thensauro fabulae.
And Ritschl's publication of the early inscriptions and investigation of
the early language showed that the Hiatus was illusory, not real, in a
line like Merc. 692 (Sen ) :
Parumne est malae rei quod amat Demipho? (' Hiatus.' between
Parumne and est).
222 EARLY LATIN PROSODY
Rather had Plautus used a form of the Gen. Sing, that was current at
that early time :
Parumne est malai rei quod amat Demipho ? (No Hiatus).
Small wonder that Ritschl and his contemporaries sprang to the hope
that Hiatus would disappear from Plautus and all early poetry if the
genuine, old forms were restored for the modernized. But the scientific
study of the language was only beginning, and many mistakes were
made. The d-ending of Ablatives, Adverbs, etc., which we now know
to have been obsolete in Plautus' time (10), was used to remove Hiatus
in such lines as Trin. 540 (Sen.), Stich. 459 (Sen.):
Sues moriuntur angina acerrime ;
Auspicio hodie optimo exivi foras.
The as-suffix of Nom. Plur. 1 Decl., which we now know to be dialectal
and not Latin (30), was used for Trin. 539 (Sen.) :
Nam fulguritae sunt alternae arbores.

And Cicero's illustrations were got rid of by theories which subsequent

investigation has overthrown, a ' Nom. Plur. 2 Decl. in -Is ' (making out
Naevius to have written ' Graiis ' or rather ' Graieis ' ; but see 30 H), a
Nom. Plur. of the Relative ' ques ' (only the Interrogative and Indefinite
form;^8B).p. H3
(B) The refutation of these theories of Ritschl led to a reaction.
Scholars began to ask : ' Is it not possible that the evidence of the
minuscule MSS. is true after all ? May not the Palimpsest's version be
1 Revival ' re-casting, or even a scholar's adaptation of primitive lines
o please the ear of Augustan or Silver Age or still later readers ? ' A
closer inspection shows that this suspicion is unfounded. Where the
Palimpsest's version removes Hiatus it is often clearly the Plautine
version ; it gives the Plautine phrase, follows the Plautine usage, suits
the context better. The same is true of the lines where the Palimpsest
shows Hiatus, but not the minuscule archetype. The Hiatus is not
Plautine. (For details see a Breslau dissertation, Krawczynski De Hiatu
Plautino. Breslau, 1906.) And no one can read many pages of the plays
without coming upon a line where Plautus patently avoids Hiatus, e. g.
de for the usual e in :
Cure. 232 (Sen.) De forma novi, de colore non queo (Novisse).
When one considers how easily small words (especially otiose words)
drop out of the MSS. of Plautus and other writers, one sees the danger
of implicit faith in mediaeval scribes. The discovery of the collation of
 HIATUS 223

the codex Turnebi knocked away a prop of the ' Hiatus before h- '
theory :
Rud. ii (Sen.) Qui facta liominum, mores, pietatem et fidem
(moresque cod. Turn.).
The citation of True. 562 in a late writer (pseudo-Aurel. Victor 6, 7)
dealt a blow to ' Hiatus after -m '. (The line is a Trochaic Septenarius) :
Quinque nummos : mihi detraxi || in partem Herculaneam (detr.
par. MSS.).
Further (we press this on the reader's notice) one example where A
corrects P shows us the danger of acquiescing in Hiatus on the plea that
' the words as they stand give excellent sense : no alteration seems
possible : to meddle would be to mar \ P reads at Pseud. 449 a
Senarius which seems ' totus teres atque rotundus ' :
Iram in promptu gerere. quanto satius est ? (* Hiatus ' after Iram).
A gives the true version : «.
Iram in propromptu gerere. quanto satius est ?
Who would have guessed propromptu} The vocabulary of Plautus'
time is not wholly known to us. In Asin. 290 (where A is not in
evidence), if all the minuscule MSS. (as some do) had omitted the con-
of concesso, who could have found anything to emend in this trochaic
line?

Sed quid ego hie properans cesso ||pedibu', lingua largior ?before
({ Hiatus
hie.)'
If an editor had ventured to propose concesso, who would have accepted
the emendation ?
No; we cannot cut the knot as Ritschl did, by excluding all (or
nearly all) Hiatus ; nor yet as the reactionaries, by admitting all (or
nearly all) Hiatus.
(C) Leo adopts a policy of despair and prints each and every * versus
hians ' as it stands in the traditional text, even refusing to print illic
(Dat.) for Mi in Bacch. 799 (Sen.) ; Amph. 263 (Troch.) :
Constringe tu illic, Artamo, actutum manus (Palimpsest evidence)
not in;

||ibo ego illic obviam (Ditto) ;

though he has no hesitation in changing illic (Adv.) to Mi in lines where
there is one syllable too many (not too few), e.g. Capt. 278 (Troch.) :
Quod genus illi est unum pollens ||(Ditto).
But Leo's pessimism is part of his pessimist theory (one of those .schemes
1 on paper ' which have such attraction for some minds and yet seem so
224 EARLY LATIN PROSODY
unlikely to most), that all copies of Plautus had been scrapped after
Horace's attack on early poetry and that Valerius Probus had not been
able to rescue anything but the most inadequate material from the
scrap-heap, a material tinkered into one shape by the 'Ambrosian'
editor, into another by the ' Palatine '. These editors, he imagines (and
his imaginations are for him as strong as realities), would believe Hiatus
to be a feature of Plautine verse and would therefore make no attempt
to tinker any ' versus hians '.
This pessimism will not do for us. We hope to reduce a great part
of Hiatus to rule and reason by a close study of the available material.
But not all. That many ' versus hiantes ' are abnormal is perhaps to be
inferred from Cicero's words. Hiatus like that after qui in Naevius' line
is not (apparently) a type. It looks more like exception than rule. (If
atque after Graii means ' all at once ', the Hiatus in the second example
may belong to paragraph 50.)
44. We have been spared the preliminary labour of collecting all and
sundry, real or seeming ' versus hiantes ' in the traditional text, whether
in those portions which have the evidence of A as well as P or in those
which have only P to appeal to, and occasionally (True, Merc, and
second half of Mil.) only P at its worst. That has been done by other
investigators, by one who seeks to prove that Hiatus was permissible
after certain final vowels (e. g. -ae ; 30 B and F), by others who believe
in ' Hiatus after Ablatives ' (10) or * Hiatus after -m ' (12) or ' Hiatus
before h- ' (22). The material they have collected supplies of itself the
refutation of their theories. No distinction can be found- in Plautus'
treatment of, e. g., qui abit, quae abit, quo abit, a quo abit, quam amat,
qui habet. Of course the array of instances of this or that usage seems
convincing at first sight, but not when we look at the contrary instances.
And there is one consideration which deters us from inflicting on our
readers a huge ' omnium gatherum ' list. Namely this ; that most lines
of Plautus contain at least one word with initial vowel (or h-) following
a word with vowel- or m- ending. Most lines therefore which contain a
syllable too few can be added by the hunter of ' versus hiantes ' to his
bag. And so his bag is swollen to an enormous size. Most unfairly.
Suppose that the minuscule MSS. offered this Senarius :
Sed ubi illast ? Pasiphilae quaero filiam ;
this would be added to the bag, because ubi ends in a vowel and illast
begins with a vowel. But a rival hunter, a hunter for examples of
modernizing, would claim it for his bag, pleading that Pasiphilai had
been modernized to Pasiphilae :
Sed ubi illast ? Pasiphilai quaero filiam.
 HIATUS 225

And the rival hunter would gain the award. Ubi Mast! was pronounced
as one word-group (or rather with one accent) in Plautus' time (and
later), ub(i)-Illast ? (with shortening of unaccented ill- after the emphatic
1 brevis brevians ' of the Interrogative ; II 48). The change of -ai to -ae
is one that would be made by any scribe at any time (I 6).
The lack of an adequate edition of Terence (I 6) is felt in this section
more than any other. For if the use of his plays for school-teaching led
to a removal from his text of some of those early usages against which
the teacher was always warning the pupil, Hiatus would be the first thing
to go. The safest procedure for us would be to omit mention of
Terence here. But that would leave a wrong impression. In our own
judgement Terence uses Hiatus much as Plautus uses it, but to a less
extent. To what extent precisely cannot be settled until the history of
the text has been elucidated. We prefer therefore to cite Terence as a
witness in the paragraphs that follow, and to put our readers on their
guard by telling them now that Leo (Plaut. Forsch.2, p. 2 n.) declares
that Hiatus was a licence never used by the younger poet.
45. Both from Cicero's words and from a priori considerations (50),
we are inclined to regard early Hiatus as more or less independent of
Greek usage.1 But we had better begin our investigation of Plautus'
Hiatus by a glance at his Greek models. What Hiatus do we find in
Menander ? Only (1) ri ow, ovSk h, etc., e. g. :
Her. 40 'EA.cv0epios kcu Kooyx/a. Tt ovv ; o~v ri;
Epit. 99-100 "Hkw Bk kcu vvv ovk ifxavrov a ovBk tv
"iBiov airaiTwv. — xolvos 'Ep/Aiys. — fjirjhe ev.
(2) With Interjection ; e. g. w 'Hpa/cAeis begins the Trimeters, Epit. 146,
155. And (3) — but the Teubner editor emends the traditional text — :
Peric. 404-6 'Y-rrcpev Acycis, /3d8i£' e'yw <r' i\tv6epav (eyw 8' i\. Koerte)
Avpiov a(f>rjcr<i), Awpi' dXA.' o Set iroav (Aw/at, o-'" ak.
Koerte ; Awpis alii)
"Kkovctov, etc. (cf. Epit. 99, 118).
The first type clearly echoes everyday talk. The phrases n ovv and
ovBk h> were so pronounced, just as from Plautus' verse we see that
flagitium hominis (a phrase not used by Terence) was pronounced as
seven (not six) syllables. Here are all the occurrences of the phrase :
Asin. 473 (Iamb.) Flagitium hominis. da, obsecro ||;

1 Hardie says of Ennius (Res Metrica, p. 44) : ' Ennius in his Epic, both in regard
to Elision and Hiatus, had adopted very strict principles of versification. In this
matter he does not follow Homer. He has fewer elisions than Virgil, and Hiatus
is not certainly found in the extant lines except once or twice, . \ . perhaps inimi-
citiam agttantes,' etc.
?348 Q
226 EARLY LATIN PROSODY

Cas. 552 (Troch.) Quasi catillatum. flagitium || hominis, qui dixit

mihi (Proves nothing) ;
Men. 489 (Sen.) Flagitium hominis, subdole ac minimi preti ;
709 (Sen.) Flagitium hominis, cum istoc ornatu. Quid est ?
And this phrase is sufficient refutation of the statement that ' Roman
actors could not pronounce a line with Hiatus, since such an utterance
of a Latin sentence is inconceivable \ It is true that animadverto shows
the pronunciation to have been anim(um) adverto ; that magnopere,
tantopere attest magn(o) opere, tant(o) opere ; that the compound circitor
and the pun on Domitius (Ad Herenn. 3, 21) prove circium) itor and
dom(um) itio, as the name (or the popular etymology of the name) of the
insipid berry unedo proves un(um) edo. But another possibility of
pronunciation is indicated by circueo, circuis, circuit, e.g. Men. 231
(Sen.) :
An quasi mare omnes circumimus insulas? (cf. Ter. Ph. 614);
by introire, e. g. Men. 662 (Troch.) :
Nam domum numquam introibis || nisi feres pallam simul (cf. Ter.
Andr. 850, contrasted with Eun. 842).
It is plain. Both Hiatus and Elision were (on appropriate occasion)
tolerated in Roman talk.

46. (A) Qui amant, etc. Alongside of flagitium hominis we may

set another phrase, qui amant. This makes three syllables normally,
e.g. Cist. 280 (Sen.):
Nam qui amant stulte atque immodeste atque improbe.
We may take it that qui amant was the everyday pronunciation. Even
Virgil recognizes this (along with viden, modo, etc.), Eel. 8, 108 :
Credimus ? an qui amant ipsi sibi somnia fingunt ?
Similarly ita me di anient (e.g. Trin. 1023), a scansion recognized by
Catullus (97, 1) :
Non (ita me di ament) quicquam referre putavi.
(So scan the opening feet of Most. 1 70 as Anapaests.)
And si me amas of Horace (Sat. 1, 9, 38). Horace has also niim
adest (Sat. 1, 2, 28); Lucretius (3, 1080) dum abest.
Examples of qui amdt, etc., with Prosodic Hiatus :
Amph. 473 (Sen.) Illius quam amat. igitur demum omnes scient ;
Asin. 515 (Troch.) || illo quern amo prohibeor;
Cas. 225 (Anap.) Qui cum amo Casinam, magi' niteo ||;
Cist. 97 (Troch.) Melius 1111 multo quern ames ||;
280 (Sen.) Nam qui amant stulte atque immodeste atque improbe ;
Merc. 744 (Sen.) Nam qui amat quod amat si habet, id habet pro cibo ;
 HIATUS 227

Poen. 860 (Troch.) Neque erum meum adeo. Ouem ament igitur ? )|
(Emphatic quern) ;
Trin. 242 (Anap.) Nam qui amat quod amat cum extemplo ||;
True. 930 (Troch.) Qui, malum, bella aut faceta's [| quae ames hominem
istimodi ? (?)
Elision appears only in :
Mil. 998 (Troch.) Qu(ae) amat hunc hominem nimium lepidum ||
(and in Pers. 179, if it be an Anapaestic Dimeter :
Miser est qui amat. Certo is quidem nihili est).
May we venture to assume Prosodic Hiatus for Pseud. 203 (Troch.) ?
|| integrast, qui amant a lenone (But cf. II 27).
At any rate we may give it the preference in lines like Asin. 616 ; Aul. 619 ;
Cas. 795; Cist. 119 (?) ; Cure. 142 (cf. Merc. 744, above); Mil. 1232;
Pers. 1 ; Pseud. 774.
Examples of me amat, etc., with Prosodic Hiatus :
Amph. 542 (Troch.) Numquid vis ? Ut cum absim me ames ||J
Bacch. 1 165 (Anap). || si amant, sapienter faciunt ;
Cas. 724 (Anap.) j| tu amas, ego esurio et sitio (Emphatic tu) ;
Cure. 213 (Troch.) Si amas, erne, ne rogites. facito || ;
Mil. 1257 (Iamb.) Quia me amat, propterea Venus (II 45) ;
Most. 182 (Iamb.) Ita tu me ames, ita Philolaches || tuu' te amet, ut
venusta es (Emphatic te) ;
231 (Iamb.) || qui te amant. Magis amabunt ;
305 (Troch.) Tu me amas, ego te amo ; merito id || (Emphatic te
and perhaps me) ;
Poen. 289 (Troch.) Ita me di ament, ut ilia me amet || malim quam (?)
Milphio di,;

Rud. 438 (Troch.) || me ames. cedo mi urnam. Cape ;

Elision appears in :
Pers. 227 (Troch.) || subigitatrix. Sin t(e) amo? ;
Pseud. 1294 (Cretic) Di t(e) ament, Pseudole ||;
Most. 717 (Cretic) Accedam. di t(e) ament || plurimum, Simo;
Mil. 985 (Troch.) Venu' m(e) amat. St! tace. aperiuntur || (II 57).
So we may ascribe Prosodic Hiatus to Cist. 96, to the corrupt lines, Cist.
273, True. 929 (ne amem), and even to Trin. 244 (si me amas). It should
be added that in the Prosodic Hiatus list ' med amas ', etc. is often conceiva-
ble ;and that in the Elision list a Choriambus (di te ament) is not incapable
of playing the part of a Cretic (IV 25).
Examples of qui habet, etc., with Prosodic Hiatus :
Aul. 781 (Troch.) Noscere. Filiam ex te tu habes ||;
Cure. 548 (Troch.) || qui habent et eos deserunt ;
Epid. 410 (Sen.) Ne tu habes servum graphicum et quanti vis preti ;
683 (Troch.) Tu habes lora ; ego te emere vidi ||;

Q 2
228 EARLY LATIN PROSODY

Merc. 535 (Iamb.) || quom tecum rem habet. Certo ;

Mil. 1245 (Iamb.) || vis quam habes ? cave sis faxis;
habes)? ;
Most. 653 (Sen.) Adulescens, mecum rem habe. Nempe abs te petam
Pers. 341 (Sen.) Utrum pro ancilla me habes an pro filia? (Or med

Poen. 320 (Troch.) Quae habent nocturna ora, noctu ||;

833 (Troch.) II qui habet quod det, utut homo est ;
Rud. 222 (Anap.) Ita res se habent : vitae haud parco ||;
1297 (Iamb.) || scit qui habet, ut ego opinor (Emphatic qui, the;
Interrogative)
1369 (Troch.) || heus tu ! iam habes vidulum ;
True. 235 (Troch.) || aliis qui habent det locum ;
and apparently in :
True. 713 (Bacch.) dum habet.
Elision appears in :
Mil. 1041 (Anap.) || hau mirum si t(e) habes carum;
Stich. 712 (Troch.) || cantharum. vinum t(u) habes ;
Trin. 749 (Sen.) Ipsum adeam Lesbonicum, edoceam ut res
(Cf. s(e)
Aui. habet
47) ;

True. 745 (Troch.) Qui invident, egent ; illis quibus || invidetur, habent
i r(em) ;

and (apparently) in:

Cas. 200 (Cretic) Clam virum et quae habet || (Cf. Most. 709) ;
189 (Cretic) Vir me habet pessimis ||.
So we may give Prosodic Hiatus the preference in Aul. 542; Mil. 580;
and suspect intrusion of atque (as in Aul. 784; Bacch. 11 15 ; Capt. 658;
Cure. 280) in Asin. 885 (Troch.) :
Surripiam in deliciis pallam || quam habet, [atque] ad te deferam.
From the list in Appendix C it will be seen that tu homo, etc., and all
similar phrases were normally three syllables (tu homo).
So Lucilius 2 (of the letter R) :
Irritata canes quam homo quam planiu' dicit.
Examples from Terence, who conforms very closely to Plautus' usage in
these scansions, are :
(a) qui am at :
Andr. 191 (Iamb.) Hoc quid sit? omnes qui amant, gravi||ter sibi dari
uxorem ferunt;
Hec. 343 (Iamb.) Nam qui amat cui odio ipsus est ||;
(b) ita me di ament :
Andr. 947 (Iamb.) Te credo credere. Ita me di a||ment, credo. Quid
restat, pater ? ;
Eun. 615 (Troch.) Ita me di ament, quantum ego ilium ||;
1037 (Iamb.) Bene, ita me di ament, factum. Audin ||;
Heaut. 308 (Sen.) Prae gaudio, ita me di ament, ubi sim nescio ;
 HIATUS 229

383 (Troch.) Minimeque, ita me di ament, miror ||;

569 (Troch.) Ut equidem, ita me di ament, metui ! ||;
686 (Iamb.) Atque, ita me di ament, ut ego nunc ||;
Ph. 883 (Iamb.) Vale. Vale, Antipho. bene, ita || me di ament, factum.
gaudeo ;
954 (Sen.) Monstri, ita me di ament, simile. Inieci scrupulum. Hem ! ;
Hec. 233 (Troch.) Gaudeo, ita me di ament, gnati ||;
258 (lamb.) At, ita me di ament, haud tibi hoc ||;
276 (Troch.) Nam, ita me di ament, quod me accusat ;
579 (Iamb.) Verum, ita me di ament itaque obtin||gant ex te quae
exopto mihi ;
642 (Sen.) Bene, ita me di ament, nuntias, et gaudeo ;
864 (Iamb.) || verum. Ita me di ament, Pamphile ;
(V) me amat, etc. : Eun. 448 ; 1080; Heaut. 360 ; Ad. 680 ; 903 ;
{d) qui habet, etc.: Eun. 695; Ph. 1041. But m(e) habens, Eun. 634
(Sen.) :
Cum sensi, redeo rursum, male vero me habens.
Here then we seem able to reduce Hiatus to rule. Monosyllables
ending in a long vowel (or -m) were left in 'Prosodic' Hiatus (i. e. with
shortening of the final) before iambic words which began with a vowel
(or h-). This scansion, which is only rarely departed from and that
through metrical exigencies (e. g. at the end of a line), was an echo of
talk. If Virgil's ' qui amant ' stood alone, it might be called a Graecism
(like Homer's ot e^owi, etc.) ; but the Plautine (and Terentian) evidence
is conclusive. The Romans of Plautus' and Virgil's time (and later)
pronounced these phrases as three syllables just as the Greeks pro-
nounced ti ovv as two syllables. Of course, when they made a Pyrrhic
of the Iambus, the rule does not hold. The monosyllable might then
be elided, e. g. Rud. 1137 (Troch.) :
(Si falsa dicam, frustra dixero,) Sed s(i) ertint vera, turn obsecro te ||
(Emphasis on vera).
47. Cum hac, etc. Have we in Latin any exact parallel to ri ovv ;
any disyllabic phrase (or the like) with Hiatus ? The traditional text
offers such (well-authenticated) examples as :
Pseud. 1332 ('Cret.') Nil profecto. / hac. || te sequor. Quin
vocas? (AP);

Cas. 612 (Sen.) Cum hac, cum istac, cumque arnica etiam tua
(AP);
Most. 392 (Troch.) Ubi ego ero? Ubi maxime esse || vis. cum
hac, cum istac eris ;
Capt. 395 (Troch.) || mihi cum hoc convenerit ;
23o EARLY LATIN PROSODY
Poen. 1054 (Sen.) Nam hau repudio hospitium neque Carthaginem
(AP);
1235 (Iamb.) Dato mi pro offa savium, \ pro obicito
osse linguam
{A P) ;
Asin. 130 (Cret.) || nam iam ex hoc loco (IV 26).
Also, e. g. :
Asin. 706 (Iamb.) Demam hercle iam de hordeo || (P Gram. Like
dehortor, etc.)
And that such pronunciation was not impossible for Roman lips is
perhaps suggested by word-groups (42 beginning) like tametsi, quamob-
rem, quemadmodum, quousque.
From the (unedited) text of Terence we take these examples :
Andr. 182 (Iamb.) Ne esset spatium cogitan||di ad disturbandas
nuptias j

Eun. 662 (Iamb.) Quo ille abire ignavu' pos||sit longius,domum

nisi sij
modo j
Ad 313 (Iamb.) Satis mi id habeam supplici || dum illos ulciscar

514 (Sen.) Si est, facturus ut sit officium suum ('if it is the

From Ennius : case that'; cf. Poen. 1005?).

Sat. 8 (Sen.) Nam is non bene vult tibi qui falso criminat.

48. (B) At Diaeresis. It' is unlucky that Menander's third type of

Hiatus is doubtful. At least, it may serve as peg for a new topic, Latin
Hiatus at a break in the line or sentence :

Avpiov a<f>rjcroi, Awpt* dAA' o Set irozlv (Hiatus before sentence).

the new

One clear case has already been discussed (II 59), Hiatus (and
Syllaba Anceps) at the Diaeresis (at end of hemistich), a Latin usage,
not Greek, and so readily allowed by Plautus that one could almost
say that Tetrameters were in his eyes two Dimeters. Hiatus, not so
readily by Terence, who however makes so free use of Syllaba Anceps,
e.g. Heaut. 724 (Iamb. Septen.); Andr. 957 (Iamb. Octon.) :
Decern minas quas luihi dare || pollicitust. quod si is nunc me ;
Proviso quid agat PamphilQs : || atque eccum. Aliquis me
forsitan,

that the lack of undisputed examples may be due to mere editing

(e.g. Hec. 876; Ad. 947), especially since there is no lack of them in
the fragments of subsequent Dramatists (App. B, end). From the host
 HIATUS 231

of occurrences in Plautus we need set here only one example from each
of the leading metres for Hiatus and one for Syllaba Anceps (mostly
taken from Klotz's fuller lists on pp. 142 sqq. of his Grundziige. Of
course in Trochaics there is no room for Syllaba Anceps) :
(Iamb. Septenar.) Mil. 1269 Induxi in animum ne oderim || item
ut alias, quando orasti ;
Rud. 390 Qui subs parentes noscere || posset :
earn veretur ;
(Iamb. Octonar.) Amph. 190 Quod multa Thebano poplo ||acerba
obiecit funera ;
- 215 Propere suis de finibus || exercitus
leducerent (de su. MSS.) ;
te huic meae machaerae ||
obicio, mastigia ;
protervo, iracundo ||animo,
indomito, incogitato ;
I eri diligenter ;
I ut ornata incedo ;
adversetur suo ;
.ures tetigit meas ;
ictusque atque procellae ||
infensae frangere malum ;
go huic bene et hie mihi
is || et amicitia est antiqua ;
mnes sollicito, ||ubicumque
spidum unguentum, ungor ;
operam seni surripere ; ||
lo, ne senex me opprimeret.
as at the end of the line to

907 (Troch.) :
i - vae misero mihi,
editors add Hit,)

possibility of Plautus' partial

;he fourth foot, and of the
ipinion||
Liest imitators as a Dimeter
I be suggested for Hiatus
iatus atKt^
tus which is, in our opinion
7 with success (pp. 165 sqq.)
1 is now pronounced to be
re-casting of Plautus' plays
232 EARLY LATIN PROSODY

introduced this ' Hiatus at the Caesura ' into a number of Senarii, e. g.
the two versions of Aul. 399 :
Congrum, muraenam exdorsua quantum potest j
Congrum, mumenam exossata fac sient.
Since both the theory of Hiatus at the Caesura and Klotz's amend-
ment are nowadays discredited, we can dismiss them summarily. But
the theory of a Diaeresis after the fourth foot of the Senarius is more
recent and has been used by Leo in support of his Quantitative scheme
of Saturnian metre. We must allow it a fuller treatment (but in small
type, since its claims are very weak indeed).
The warning already given (I 6) cannot be given too often, that the state
of our MSS. of Plautus makes it possible to muster a fair show of evidence
for any theory under the sun. Each new pretender must face three test-
questions. Is the theory itself probable a priori ? Are the readings on which
it is based reliable readings (attested by AP or PGram., etc.) ? Are they
capable of a better explanation ?
On a priori considerations this theory is certainly entitled to a hearing.
There do seem to be some indications that the Dimeter was with Plautus
a favourite unit of Iambic Verse (1 11), though, on the other hand, the more
jclosely we examine Plautus' technique, the more we find in him no rude
/versifier but an expert with a good comprehension of the verse of his Greek
(models.
What then is the evidence ? Are the readings sound ? And, if sound, do
they admit of other, more likely explanations ? Here is the list.
(a) Hiatus :
Amph. 141 Et servu' quoi(u)s ego hanc iz.ro imaginem (II 35) ;
978 Fac iam Amphitruonem advenien/^w ab aedibus (II 7) ;
Asin. 85 Dotalem servum Saurea/// uxor tua ;
98 Non offuturum, si id hodi^ effeceris ;
775 Neque illaec ulli pede pedem homini premit (Or pedem ; II 46) ;
Bacch. 134 Ibidem ego meam operam perdiafr, ubi tu tuam (II 46) ;
235 I bo in Piraeum, visam ecqitae(n) advenerit ;
Capt. 373 Sequere. em tibi hominem. Gr&tiam habeo tibi (Or gratiam ;
55);
Cas. pr. 2 Fidem qui facitis maxiwz : et vos Fides (Or Hiatus at pause) ;
pr. 58 Senis uxor sensit virum amor/ operam dare (Or virum; 54) ;
343 Tibi et Chalino. ita rem nazVz//z intellego (Rather Hiatus at
pause after Chalino) ;
Cist. 753 Istic quidem edepol mei v\ri habitat gener (Or viri ; II 46) ;
Cure. 15 Huic proximum illud ostium(st) oculissimum ;
Men. 258 natitf Epidamnia (P ? : natio in Epidamniis A) ;
480 Ait hanc dedisse me s\bii atque earn meae (Or Hiatus at pause);
558 Ut haec quae bona dant di mi^z ex me sciat ;
739 Quae mea flagitia ? Pallam a\que aurum meum (??) ;
 HIATUS 233
nisi) ;
Merc. 706 Vidisse credo muWerem in aedibus ;
712 Quid nunc ego faciam nisi ut adeam atque adloquar ? (Rather

Mil. 45 Sunt homines, quos tu occidis/z uno die (Emended to tu quos) ;

59 Qui sis tarn pulcher ; vel lllae quae her! pallio (Rather quae) ;
1 120 Itan tu censes ? Quid ego ni ita censeam (Rather ni) ;
1286 Me amori' causa hoc orna/w incedere (Emended to {hue) hoc) ;
Most. 999 Numquid processlt ad for«w hodie novi ? (Or forum; II 46) ;
1037 Quid vis ? I mecum, obsem?, una simul (56 [c]) ;
Pers. 433 Mirum quin tibi ego crederaw ut idem mihi (Or crederem ; 55) ;
Poen. 176 Se velle amare atque obsegui animo suo (Or obsequi ; 55) ;
454-5 Propitiam Venerem facere u/z esset mihi.
Quoniam litare nequeo, abiz illim ilico (In a passage with
frequent Hiatus ; 51) ;
497 Certum. Turn tu igiturdie bono, Aphrodisiis (Or bono; II 46) ;
7969 Cretast profecto horum hommum oratio (Emended to Cr.
pr. (creta)) ;
988 Pro di immortales pluriwz ad illunc modum (Or plurimi ; 55
Emended to plurimin ') ;
1019 Ad messim credo, nisi quid tu aliud sapis (Rather tu) ;
1076 Mi patrue, salve. Et tu sak/<?, Agorastocles (Rather Hiatus
at the change of speaker) ;
1 108 Malum crudumque et c&Midum et subdolum (56 [b]) ;
1 144 Matrem hie salutat suam, haec autem hunc fllium (Rather
Hiatus at the pause after suam) ;
Pseud. 415 Si de damnosis aut si de amatoribus (Rather de ama.) ;
776 Interminatust a .minimi ad maximum ;
Rud. pr. 7 Inter mortales ambu/i interdius ;
pr. 70 Nam Arcturu' signum sum omnium acerrimum (Emended
to om.(unum) ac.) ;
183 Si apud med esuru's, mihi darz'operam volo (Or dart ; II 46) ;
Stich. 216 Consenui : paene sum fame emortuus ;
Trin. 540 Sues moriuntur angina acerrime (Emended to macerrimae) ;
True. 33 Aut aera aut vinum aut olez7//z aut triticum (56 [b]) ;
Vid. 87 Quom mihi qui vivam copiam inopi facis (Or copiam ; II 46).
(0) Syllaba Anceps :
Cure. 438 Quia nudiu' quartus venimus in Cariam ;
sum tolib.)
Epid. 498 Potuit : plus iam sum libera quinquennium (Emended qu.;

Men. 327 Proin tu nequo abeas longius ab aedibus ;

506 Sanum est; adulescens, sinciput, intellego (Emended
sin.(ut)Tnt);to
877 Qui me vi cogunt ut validus insaniam (ToTribrach
avoid the
; IIfalse
9) ;
Mil. 481 Satin abiit ille? neque erile negotium (Plus curat) (Emended
to erili negotio) ;
234 EARLY LATIN PROSODY
Pers. 398 Vel tu me vende vel face quid tibi libet ;
Poen. pr. 85 Altera quinquennis, altera quadrimula (Emended to
alt. (erat) quad.) ;
474 Volaticorum hominum ? Ita dico quidem (?) ;
Pseud. 100 Nisj tu illi drachumis fleverls
Subj.argenteis
like True.(May
344 ;beIIIPert".
8) ;
563 Me idcirco haec tanta facinora promittere ;
1003 Nullam salutem mittere scriptam solet ;
Stich. 221 Logos ridiculos vendo. age, licemini (Rather Hiatus at the
pause before age) ;
Trin. 788—788* Sed epistulas quando obsignatas adferet.
Sed quom obsignatas attulerlt epistulas (To avoid the
false Tribrach ; II 9. Which is the Plautine line ?) ;
True. 425 Non audes aliquid mihi dare munusculum (Emended to dare,
mihl ; but cf. 31).
We wish that we ourselves could, like any credulous reader, accept these
lists without question ; the theory is in many ways so attractive. But, as
judges, we must regard them dispassionately as mere lists of irregular lines
ending in a quadrisyllable (or the like), precisely as Senarii like :
Bacch. 218 Edepol, Mnesiloche, ut hanc rem natam [esse] intellego;
354 Senex in Ephesum (hinc) ibit aurum arcessere ;
1071 Domum redduco (iam) integrum omnem exercitum.
We must ask how the irregularity can best be removed : by this theory of
a Diaeresis or in some other way ? The one possibility has to be weighed
against the other.
We cannot allow the lists to be increased by examples of Hiatus (or
Syllaba Anceps) at a change of speaker, e. g. :
Mil. 49 Edepol memoria es opti;;/«. Offae monent ;
or at a pause in the sentence, e. g. :
Capt. pr. 1 1 Negat hercle illic ultimus. accedito.
Such examples can be added to the lists of Hiatus (and Syllaba Anceps) at
the Diaeresis of Iambic long lines, or any other equally well established
Diaeresis ; but not in this doubtful case. (Indeed some items, e.g. Bacch.
134; Cas. pr. 2; Men. 480 should rather be excluded on this score.) And
while we may allow an appeal to the Senarii with abnormal Tribrach in the
fourth foot (II 46), since this theory would remove the irregularity from
them, we cannot allow them to be entered on the second list.
What then is the worth of the evidence ? Alas ! how weak it seems when
we contrast it with the evidence for Diaeresis in other Iambic lines. In
Iambic Septenarii and Octonarii Hiatus or Syllaba Anceps, one or other,
appears at the Diaeresis in every eighth or tenth line (II 59). This theory
can only muster about a score of sufficiently strong examples of the two
licences out of all the 9,000 Senarii (and more) of Plautus. Such evidence
cannot possibly avail to prove that this was an actual Diaeresis for Plautus.
 HIATUS 235

All that it could ever hope to effect would be to establish at most a possibility
that these occasional scansions are due to Plautus' reminiscence of an earlier
stage, when such a Diaeresis had been actually recognized.
(c) Two attempts have been made to strengthen it. Firstly, a similar
Diaeresis has been claimed for the corresponding part of the Trochaic
Septenarius ; and, according to the Greek metricians, the long Trochaic line
is, to all intents and purposes, identical with an Iambic Senarius preceded
by a Cretic. We have already entered a caveat against this identification
of Trochaic with Iambic lines in the case of Plautus (II 55). But even if
we allow it for the sake of argument, what are the Trochaic examples but
mere rotten wood added to the flimsy scaffolding ? Out of the 8,000 Trochaic
Septenarii of Plautus all that seems worth collecting (cf. 3) is :

Asin. 250 Atque argento comparando || fingere" fallaciam (Emended to

fingeres ' you should have framed ', with atque ' forthwith ') ;
Gas. 550 Propter operam illius hircl || improbi, edentuli {AP) (see 56) ;
Cure. 364 Laudo. Laudato quando illud || quod capls effecero (see
40 H);
602 Pater istum meu' gestitavit. || At mea matertera ;
Men. 601 Quam hodie uxori abstuli atque huic || detull Erotio (Emended
to det. huic) ;

841 Ut ego illic oculos exuram || lampadi[bu]s ardentibus ;

900 Quae me clam ratus sum facere e||a omnia fecit palam
(Emended to om. ea) ;
921 Potionis aliquid priu' quam ||percipit insania (see 40 H) ;
Pers. 835 At tamen non — tamen — Cave ergo || sis malo et sequere me ;
Pseud. 346 Militi Macedonio, et iam || quindecTm habeo minas (see (AP)
55) ;

355 Ego scelestus nunc argenti || promere po(ti)s sum domo

(possum AP) ;
True. 757 Unum aiebas, tria iam dixti || verba, atque (ea) mendacia.
(d) Secondly, a curious argument has been used. While a word like danunt,
a by-form of dan t, occurs generally (10 times in all) at the end of a line, we
find it once before the final Dipody of a Senarius (Pseud. 767) :
Ouoi servitutem di danunt lenoniam.
A list of such words (including some indispensable and universal !) has been
compiled (in a Berlin dissertation) in support of the theory of this Diaeresis
at the fourth foot of Iambic Senarii.
Now if we ask why Plautus uses danunt mostly to close the line, common
sense answers that danunt must have been something out of the way, a form
to which the Comedian has recourse only through metrical exigencies or for
some special effect. (He uses it at the beginning of a line once, Pers. 256.)
When he has to find a disyllable for the fourth foot he prefers an Iambus-
word (II 7) ; and if the required sense is ' they give ', since dant and tribuunt
and donant and praebent and offerunt, etc., are not Iambi, he naturally turns
to danunt. The compiler of this list (whose industry deserves all praise)
236 EARLY LATIN PROSODY
would almost seem to believe that Plautus had ear-marked a portion of his
vocabulary ' To be used at the ends of lines, but nowhere else \ Is not
a poet sometimes at a difficulty at other parts of the line too ? Really, we
cannot refrain from the remark that while the study of Latin Comedy might
be expected to sharpen qualities like imagination, humour, common sense,
there is in the literature of Plautine research quite enough material for
a rival sketch to Shakespeare's Holofernes.
(e) The theory that there are traces of a Diaeresis before the last dipody
of a Senarius could not be dismissed unceremoniously ; and we had to
mention the dozen (so-called) traces of one at a similar part of the Trochaic
Septenarius, since they were used in support of the Senarius theory. The
score of witnesses (out of some 8,oco lines!) for a (supposed) Diaeresis at
the second 'rise' in a trochaic line would not deserve mention, had not Leo
used them for his theory of Saturnian Verse. Here they are :
Amph. 826 Amphitrutf alius qui forte || ted hinc absenti tamen ;
968 (evoca.) Blepharo;zm,.ut(i) re divina ||facta mecum prandeat
(Emended to Bl. qui) ;
Asin. 199 Cetera quae volumus uti || Graeca mercamur fideto (Emended
ceterum) ;
Aul. 455 Intro ab*. opera hue conducta est [| vestra, non oratio (Rather
Hiatus at pause) ;
Capt. 820 Quo homine (homo) adaeque nemo || vivit fortunatior ;
Cas. 258 Cui homiw/ hodie peculi || nummu' non est plumbeus ;
Cist. 510 Non edeflo/ istaec tua dicta || nunc in aures recipio (Emended
to ed. (ego) is.) ;
Cure. 192 E brio/a persolla, nugae. || Tun meam Venerem vituperas ?
(Emended to ebriola's ; see II 9 A) ;
463 Halophantam(ne) an sycophantam hunc || magis ess' dicam
z nescio ?
Men. 1087 Illic homo aut sycophanta aut || geminus est frater tuus
(Emended to homost) ;
1 160 Veni^zY uxor quoque etiam, || si quis emptor venerit ;
Merc. 920 Omni&us hie ludificatur |[ me modis. ego stultior (Qui isti
credam) (Emended to Om. istic ; 33 A) ;
947 lam redii (ex) exilio. salve, || mi sodalis Eutyche ;
Most. 259 Una opera ebur atramento || candefacere postules (Emended
to op. (era) eb.) ;

376 Quaeso edepol, exsurge.pater ad||venit. Tuu' venit pater ?

(Emended to (te) exs. ; II 9 A) ;
Poen. 1295 Propemodum hoc opsonare || prandium potero mihi (AP) ;
Stich. 344 Iamdu^ww(n') ego istum patior ||dicere iniuste mihi ? (AP) ;
374 Argen/z aurique advexit |j nimium. Nimi' factum (AP);
bene !
True. 701 Di magni, ut ego (laete) laetus || sum et laetitia differor ! ;
957 Qu'd dedtl ut discinxi hominem ! || Immo ego vero qui dedi
(Rather Hiatus at pause).
 HIATUS 237

If any one cares to take the trouble, he could probably scrape together no
less imposing evidence for 'Diaeresis' at other parts of Trochaic (and
Iambic) lines, say after the third ' rise' of the Trochaic Septenarius, e. g. :
Amph. 507 Observato/*? quam blande || mulieri palpabitur (Emended to
(ut) quam).
49. (C) At Change of Speaker. We have the same companion,
Syllaba Anceps, for another clear case of Hiatus, at a change of
speaker,1 in Plautus over and over again (Klotz gives a long list,
pp. in sqq.) j e. g. in these two lines of the Mercator :
183 (Troch.) Qui potuit videre? Oculis. || Quo p&cto ? Hiantibus;
749 (Sen.) Abi. Quid abeam ? St ! abi. Abeam ? Abi.
Whether there is Syllaba Anceps in the second, is difficult to say ; for
though philologists tell us that all long vowels became shortened before
-m, they cannot assure us that this shortening 2 was as early as Plautus.
For Syllaba Anceps along with Hiatus we had better cite another line
of the same play :
788 (Sen.) Ut veniat ad me iam simul tecum. Eo.
And for Syllaba Anceps alone, a third :
934 (Troch.) Stultus es ; noli istuc, quaeso, || dicere. Certum
exsequi est ;
or the Bacchiac line, Cas. 738 :
Servus sum tufis. Op||timest. Obsecro te ;
or the Trochaic line, Rud. 1086 :
Et crepunditf. Quid si ea sunt || aurea ? Quid Istuc tua ?
In the last example (since quid is a ' brevis brevians ' in such phrases ;
II 31) we are left in doubt whether to scan aurea or aurea. The former
may be approved on the ground that a pause before the sharp retort
' what 's that to you ? ' seems less natural. But not on the ground that
the Syllaba Anceps is abnormal. As at the Diaeresis (48), so at a
change of speaker it seems that Plautus had no hesitation in using
Syllaba Anceps or Hiatus. We need not hesitate to scan with Hiatus,
e.g. Cure. 88 (Sen.) :
Ita faciam. Agite bibite, festivae fores,
1 At the risk of offending any of our readers who see in every Greek practice
the acme of perfection, we have hinted (II 32 B) at a ridiculous aspect of invariable
Elision in dialogue. Surely no one can deny that Plautus, in substituting elasticity
for rigidity of practice, acted by deliberate choice, ' arte non inscitia '. How can
one suppose that he found any difficulty in invariable Elision (if he had approved of it)?
2 Oscan paam attests Early Latin quam. And demum points to -dem, e.g. ibi
demum and ibidem. So Ennuis' dumquidem may have been a Cretic word-group,
and rnilitum (Greek -oov) a Cretic word.
238 EARLY LATIN PROSODY
and so give to the phrase ita faciam its normal ictus and (presumably)
accent (cf. ' I'll d6 so ') and avoid the false Tribrach (II 44).
The raison d'etre is the pause in the enunciation of the line. And it
may not be fanciful to find in Elision a proof that the second speaker
interrupts the first or at least snaps out a reply. At any rate the
excitement of old Periplecomenus in Mil. II ii, while Palaestrio is
thinking out the plot, seems to suit his interruption of Palaestrio's
pronouncement (230-1) :
Pa. dico et recipio
Ad m(e). Pe4 Et ego impetrare dico id quod petis, etc.

We would express Pyrgopolinices' hasty correction and Plesidippus' slow

reply by scanning Mil. 1307 (Sen.) :
Pl. Habeo equidem hercle ocul(um). Py. At laevum dico.
Pl. Eloquar,

thus allowing the ictus to fall on the accented syllable of the emphasized
/aevum. The pretended slave-girl continues her remarks without a
break and Toxilus' exclamation is simultaneous in Pers. 557 (Troch.) :
Virgo. Septimum periuri(um) — To. Eug(ae) ! Vi. Octava
indiligentia.
Ballio's retorts to Calidorus' reproaches cost him not one moment ot
hesitation (Pseud. 360 sqq.). And so on. We would read in Trin.
1 108 (Sen.), where however there is no change of speaker :
Nihil est mora(i). i, I ambul(a), actutum redi.
A good example of rapid interchange of remarks is Ter. Ph. 198
(Troch.) :
Faci(am). Eloquere. Modo apud portum — || Meumn(e)? In-
tellext(i). Occid(i) ! Hem ! ,
where Elision is clearly used for effect. (Cf. Andr. 533.) That Terence
uses the alternative of Hiatus, though perhaps not so freely as Plautus,
is proved by Senarii like :
Eun. 409 Perpaucorum hominum. Immo nullorum arbitror \
697 Fratern(e)? Ita. Quanafo? Hodie. Quamdudum? Modo;
Andr. 665 Factum hoc est, Dave ? Factum. scelus ais,
Hem ! quid ?;
see 11) ;
Ph. 146 Quod det fortasrc? Immo nil nisi spem meram (On nil

Ad. 767 Exemplum d\so.\\Ainae. Ecce autem hie adest ciplinai)

(Or dis-;
 HIATUS 239

and by the Trochaic Septenarius :

Ph. 542 It&ne? Ita. Sane hercle pulchre || suades : etiam tu
hinc abis ?
Examples of Syllaba Anceps in Terence are : Nihil) ;
Andr. 437 (Sen.) Potin es mihi verum dicere ? Nil facilius (Hardly

Heaut. 611 (Troch.) Non emo : quid agls ? Optata || (Hardly with
ictus on ag-) ;

978 (Troch.) Abiit ? vah ! rogasse vellem. || Quid ? Und'

mi peterem cibum (Interrogative unde
would hardly succumb to a ' brevis
brevians '. Besides there is noH room
29) ;
here for the Law of Breves Breviantes ;

Ph. 528 (Troch.) Sic hunc decipls ? Immo enlm ve||ro, Antipho,
hie me decipit.
(Cf. Andr. 267.)
An example of Hiatus from Pomponius (time of Sulla) is :
Atellan. 49 (Sen.) Mi frater, salve. O soror, salve, mea.
Syllaba Anceps removes the unique ' nihil ' (11) in True. 696 (Sen.) and
brings the scene more home to us. At the repeated iamne the coquette
holds up a warning finger to the reformed woman-hater, who, after a
pause, submits :
Truc. Eu edepol hominem nihili ! Ast. Iamne autem ut soles ?
Iamne ? Truc. Nil dico, etc. (Truc. Iamne MSS.).
50. (D) In pausa When Lysiteles at long last overcomes his
father's reluctance to act as intermediary, we could not imagine Hiatus
in his hurried instructions, Trin. 390-1 (Troch.):
Ph. Dabitur opera. Lv. Lepidu' vivis. || haec sunt aedes. hie
habet.
Lesbonico est nomen. age, rem || cur(a). ego te opperiam
domi.
We can picture him hurrying off as soon as he could get the words out
of his mouth, in case Philto should change his mind. Hiatus before
the last sentence would be out of place. The elision of cura is natural.
The Teubner editor of Menander does not tolerate the hiatus in
Peric. 405 (45). And although Virgil's text is above suspicion at
Geo. 1, 4 (qui cultus habendo)
Sit pecori, apibus quanta experientia parvis ;
Eel. 8, 41 Ut vidi, ut perii, ut me malus abstulit error;
24o EARLY LATIN PROSODY
Geo. i, 341 Turn pingues agni, et turn mollissima vina;
Eel. 3, 6 Et sucus pecori, et lac subducitur agnis ;
io, 13 Ilium etiam lauri, etiam flevere myricae,
many scholars refuse to call it ' Hiatus in pausa '. They prefer to speak
of ' Hiatus at the Caesura ' (presumably a relic of the remote time when
the Greek Hexameter was two separate lines) or some other imitation of
Greek usage. Hiatus like (Eel. 2, 24) :
Amphion Dircaeus in Actaeo Aracyntho
is undoubtedly Greek. Virgil is unmistakably putting a Greek rhythm
in a Latin line, as Catullus puts a Greek Prosody in (4, 18) :
Et inde tot per impotentia freta,
where Catullus may be, and Virgil certainly is, reminding his readers of
some particular Greek line. But in our judgement the hiatus in the Virgil
quintette is native Roman. It is all of a piece with the Plautine Hiatus
at the Diaeresis of Septenarii, etc., and at a change of speaker. It too
is common in Plautus, e. g. :
Most. 484 (Sen.) Ego dicam ; ausculfo. ut foris cenaverat (Pause
before narrative. Ego is unlikely phrase)
in this;

devehor
Merc. 259 (Sen.) Inscendo in \emdum, atque ('forthwith ') ad (A P) j
navem

Cure. 46 (Sen.) Earn vult meretricem face^. ea me deperit ;

Pseud. 890 (Sen.) Em illic ego habi/<?. intro abi et cenam (AP);
coque

Most. 498 (Sen.) Hie habita. haec mihi deditast habitatio ;

Pseud. 673 (Troch.) Hie argentum, hie amanti a||mica erili filio
(AP);
Men. 1038 (Troch.) Salvum tibi, ita ut mihi dedisti ||reddifo. hie
me mane (P Gram.) ;
Poen. 1009 (Sen.) Quid in hanc venistis urfiem ? aut quid quaeritis ?
(AP);
685 (Sen.) Blande hominem compellafo. hospes hospitem
(AP);

Pers. 550 (Troch.) Urbi' speciem vidi, hominum || mores perspexi

parum ;
Stich. 270 (Sen.) Sed eccum Pinacium eius puerum. hoc(AP);
vide

2 2 1 (Sen.) Logos ridiculos vendo. age, licemini (A P) ;

Cist. 619-20 (Sen.) Ego earn proie«. alia mulier sustulit.
Ego inspectaz'/. erus hanc duxit postea ;
 HIATUS 241

Rud. 851 (Sen.) Due me ad lenonem recta, ubi illic est homo?
(U48);
Mil. 24 Nisi unum (' but one thing I will say ') : epityra estur
insanum bene (A, Varro ; and the ' Palatine ' edition
probably).
We have Syllaba Anceps at such a break in Capt. 444 (Troch.) :
Tii hoc age. first foot ; II 57)
tu mihi eru' nunc es || (Hardly a Proceleusmatic for;

Indeed we are inclined to believe that ' Hiatus in pausa ' gives the key
to most of the difficulties. And only those scholars who imagine all
Latin poetry to be bound to the Greek chariot-wheels will find anything
remarkable in this free use of Hiatus in Dramatic poetry, where there
is an echo of actual Roman talk and not a mere slavish imitation of
Menander's echo of Greek talk.
We cannot agree to its banishment from the text of Terence, e. g. :
Hec. pr. 1 (Sen.) Hecyra est huic nomen fabu/ae : haec cum datast
(Nova) ;
Heaut. 890 (Troch.) || mane, hoc priu' scire expeto ;
Ph. 656 (Sen.) Quae debe^ : etiam nunc, si vult Demipho ;
Ad. 574 (Troch.) Praeterlto hac recta platea || sursum: ubi eo
veneris.
(Cf. Andr. 605.)
From the Dramatic Fragments we cite :
Accius. trag. 451 (Troch.) (fore) Melea^w, ubi torrus esset || inter-
fectus flammeus ;
meum;
Titinius com. 21 (Sen.) Aspecta formam, atque os contempla

Novius com. 80 (Troch.) Quid ego facerem ? otiosi || rodebam

rutabulum.
To the instances cited from Virgil may be added, e. g. :
Eel. 2, 53 Addam cerea pruna : honos erit huic quoque porno ;
Aen. 1, 405 Et vera incessu patuit dea. ille ubi matrem ;
1, 16 Posthabita coluisse Samo; hie illius arma.
Is it not impossible to believe that anything but the pause in the sense
was what justified this Hiatus to Virgil ?

51. (E) Freak-Lines. We call 'freak-lines' those which were

deliberately presented in abnormal form by Plautus for effect, such as
the chattering lines of the drenched Charmides (comic companion to
Labrax) in the Rudens. They are introduced by his intimation (Rud.
526) : omnia corusca prae tremore fabulor. And they are confined to
242 EARLY LATIN PROSODY
his utterances. It is wrong to find them in line 528. And line 535,
with the second part of 536, belongs to him, not to Labrax. The
1 corusca verba ' leave an appearance of Hiatus, but only an appearance.
For of course the syllable was repeated (cf. Cartwright The Siege V vii :
My teeth do ch-ch-chatter).
Rud. 529 Ne thermopolium quidem ullum in-instruit
533-4 Utinam fortuna nunc anetina ut-uterer,
Ut quom exiissem ex-ex aqua arerem tamen.
(A drunkard's articulation ma-ma-madere appears at Most. 319 and
33i-)-
May not the Hiatus in Asin. 921 (Miseram-odio-enecavit) represent
the sobbing reply of the frightened Philaenium to the angry question of
Artemona ?
Art. Quid tibi hunc receptio ad te est || meurn virum ? Ph. Pol
me quidem
Miseram odio enecavit ||.
And does not the Hiatus at another passage of the same play, where
the Parasite reads the contract aloud to Diabolus, exactly tally with
stage-convention ? It appears when a new clause is written in sight of
the audience (Contrast the procedure l in Bacch. IV iv.) :
Asin. 755 sqq. Par. Addone ? Di. Adde et scribas vide plane
et probe.
Par. Alienum — hominem — intro mittat neminem.
Quod ilia aut amicum — aut patronum nominet,
Fores occlusae — omnibus sint nisi tibi.
In foribus scribat occupatam — esse se.
(In 4 we suggested, but diffidently, that Syllaba Anceps might accom-
pany this Hiatus and show itself at 762, epistula.) That the Hiatus
here is not merely a possible but a necessary consequence of writing
down the additional words in the presence of the audience will, we
fancy, be denied by those only who have never seen a play acted.
Is it possible that the Hiatus in Stich. 459-461, if it is genuine, may
be an indication that Gelasimus is reading aloud from his ' libri ' (line
454) of jests ?
Auspicio — hodie — optimo exivi foras.
Mustek murem—abstulit praeter pedes ;
Quom strena— obscaevavit, spectatum hoc mihist.
1 Plautus is not unmindful of future actors (and readers) and drops a hint, when
necessary. Thus Alcmena's words at Amph. 954 are meant to show that the two
preceding lines are spoken ' aside ' by Jupiter.
 HIATUS 243

This is one of the two passages (the other is Poen. 453 sqq.) by which
Leo tries to prove that AP have a common archetype (viz. a sole sur-
viving copy of the play, found by Valerius Probus at Bey rout), with
patent corruptions shared by the two texts (43 end ; I 6). When one
thinks of the traditional way of reading aloud a letter on the modern
stage, the pauses occasioned by the pretended difficulty of deciphering
the writing, or by the affectation of incredulity or of astonishment at
what the letter is found to contain, or by the meaning glances directed
at the audience, one persuades oneself that the recurrent Hiatus in the
Stichus passage possibly, in. the Asinaria passage certainly, is really
Plautine. Nay even that the New Comedy may have sanctioned it.
Do we not hear Euclio speaking to his old serving-woman at Aul. 55,
with Hiatus and perhaps Syllaba Anceps (49 n.) before the last word ?
Abscede etiam nunc, etiam nunc, etiam — ohe !
Istic astato, etc.
If Caesura was occasionally discarded for effect e. g. in the reproach-
ful utterance of Charmides :
Trin. 1094 O Callicles, o Callicles, o Callicles !
(Qualine amico mea commendavi bona?),
why not Elision ?

52. (F) With Interjections. A few examples of this well-known

usage, a usage of the Greek Drama too, must suffice :
Capt. 823 (Troch.) Euge/ai ! edictiones ||aedilicias habet ;
hi(c)quidem

148 (Sen.) Alienus ? ego alienus illi ! ah! Hegio ;

True. 162 (Iamb.) O Astaphium, haud istoc modo|| ;
Ter. Andr. 817 (Sen.) O optime hospes, pol, Crito, antiquum
obtines ;
Ph. 754 (Iamb.) Quid? duasne uxores habet? Aul ob||secro
unam illequidem hanc solam (cf. 803).
There seems no reason to put em in a separate class on the ground
that it was originally Imperative erne ' take'. (In Mil. 687 erne means
'buy'; cf. Men. 121, etc.) In fact em ilium produced ellum, which
suggests a rapid pronunciation em-Ilium with enclitic Pronoun, or even
a Crasis like Terence's :
Andr. 270 (Sen.) Ne deseras se. Hem, cgone istuc conari queam ? ;
Ad. 407 (Sen.) Homo de improviso coepit clamare ' O ^schine '
(cf. Ad. 449).
(Cf. Ter. Eun. 459 and 472 ; also Plaut. Bacch. 809.) Therefore em
R 2
 244 EARLY LATIN PROSODY

need not be put ' extra metrum ' (IV 3 C) in a line like Epid. 488 Em !
istic homo, &c.
Eccum (i. e. ecce hunc ; Italian ecco) is followed by Hiatus in such
lines as :
Cas. 536 (Troch.) Sed eccum egreditur senati ||columen, praesidium
popli(^);
Men. 567 (Sen.) Atque edepol eccum optime revertitur (AP) ;
Most. 686 (Sen.) Eugae ! optime eccum aedium dominus foras
(Simo progreditur ipsus) (AP) ;
similarly eccillum (i. e. ecce ilium, Ital. quello ; cf. French celui) in :
Pers. 392 (Sen.) Librorum eccillum habeo plenum soracum (cf.
Class. Quart. 7, 118).
Hiatus precedes apparently in some less well-attested examples, e. g. :
Aul. 781 (Troch.) ||immo eccillam domi;
Most. 560 (Sen.) Sed Philolachetis servum eccum Tranium ;
similarly ellum in :
Cure. 278 (Sen.) Video currentem ellum usque in platea ultima.
Perhaps the Hiatus is due not to their interjectional but to their
parenthetic character. They interrupt the flow of the sentence. The
Hiatus is rather ' Hiatus in pausa '. Elision is frequent, e. g. :
Mil. 470 (Troch.) Domi eccam erilem concubinam ;
545 (Sen.) Nam Philocomasium eccam intus. Quid nunc,
furcifer ?;

but in a line like Merc. 330 (Sen.) where there is a break in the sentence :
Nunc adeo ibo illuc. sed optime gnatum meum
Video eccum. opperiar hominem. hoc nunc mi viso opust,
it seems unlikely. Read opperiam, with Hiatus after eccum.

53. (G) Emphatic Monosyllable. The raison d'etre of Hiatus

with Interjections seems to be the dwelling of the voice on e.g., O, a
pronunciation indicated by the spelling Oh (cf. proh, which is merely
the Preposition- Adverb pro ' forth ' ; ah or aha ; vah or vaha, a disyllable
in Ter. Ad. 439 ; see Kauer's note). The raison d'etre of Menander's
tl ovv ; (45) is rather the dislike of effacing an important monosyllable.
\ Virgil avoids the elision of monosyllables before short vowels (though
there are examples of it) : Plautus, who echoes talk, cannot be expected
to do this. An unemphatic Pronoun, say me, would, we may be sure,
be suppressed in Latin talk, as in French (il m'a dit voleur), e. g. :
Most. 1035 (Sen.) Deludificatust m(e) hodie in perpetuum modum ;
 HIATUS 245
habes ;
Trin. 654 (Troch.) Cives obiectare possint ||tibi quos t(u) inimicos

Rud. 382 (Iamb.) ||scin t(u) ? etiam qui it lavatum ;

Ter. Eun. 470 (Sen.) Exire, quos iussi, ocius. Procede t(u) hue.

But Virgil's te, amice, nequivi (Aen. 6, 507) finds a parallel in Plautus'
treatment of such words when emphatic, e. g. Interrogatives ; though we
find elision before a long vowel at Stich. 97 (Troch.), Capt. 479 (Troch.),
etc. :
Qu(em) aequiust nos potiorem habere ||quam te ? postidea, pater
(Or aequust, i. e. aequum est) ;
Salvete, inquam. qu(o) imus una ? || inquam. — atque illi tacent.
Before a short vowel, or even a syllable long by position, it is difficult
to find a sure example of elision. In Rud. 249 (Cretic) || nos. Quo,
amabo, ibimus, there is perhaps a Choriambus for a Cretic (IV 25), in
Pseud. 1297 (Cretic) Quae istaec audaciast ||, an Ionic a Minore, as also
in Rud. 676 (Cretic) || quae lllaec oratiost? (II 31). The reading of A
is to be preferred in Rud. 733 (Troch.) :
Vi agis [mecum]. Etiam vim proportas ? ||.
(Read te quidem, not 'te equidem ' in Pseud. 509.) '
1 But while Prosodic Hiatus is the rule with a long emphatic mono-j
syllable before a short vowel (or h-), (e. g. Trin. 693 Te honestet, me^
conlutulentet), we cannot quite say that it is an unquestionable proof
of emphasis. Virgil's reluctance to elide monosyllables is reflected by
such scansions (the echo of everyday talk) as Most. 11 60 qui ('where-
with') arnica (empta) est; Pseud. 280 pro arnica (AP). We may put
beside them Lucretius' (2, 68 1) ctim odore, (6, 716) qui Etesiae. If we
render sitta sunt by ' if indeed it is so ' in Trin. 1098 (Sen.) :
Credo, omnia istaec si ita sunt ut praedicas,

si amabat'm Rud. 379, * if he really loved Tier ', we cannot be quite sure
that this is correct, since we cannot claim the sense ' for indeed ' for,
e. g., nam in Pseud. 699 (Troch.) :
Nam et amicum et benevolentem ||ducis, etc. {AP).
So we had better state our rule so : An emphatic monosyllable always,
and an unemphatic often, is left in Prosodic Hiatus before an initial.
short vowel (or h- with short vowel following). And yet we must
remember that in Latin (as in Greek, English, etc.) an Enclitic became
accented before another Enclitic (e. g. quae before isii'm. Mil. 7314, 1338
omnia quae Isti dedi).
 246 EARLY LATIN PROSODY
And even before a long syllable Prosodic Hiatus is possible. It is
proved by the Canticum-metre in :
Cas. 708 (Anap. + Iamb.) SI effexis hoc ;
Bacch. 989 (Dactyls and Trochees) Ut scias quae hie scripta sient.
So we may save the emphatic monosyllable in Mil. n 24 (Sen.) and
scan :
Quin si voluntate nolet, vi extrudam foras.
Similarly perhaps Asin. 258 (Troch.) :
Unde sumam ? quern intervortam ? ||quo hanc celocem conferam ?
and the frequent Quo argumento?, e.g. Asin. 302. (On Bacch. 274
em accipitrina see 52.) More certainly, e. g., Bacch. 583 (Sen.) :
Ecqui[s] exit ? Quid Istuc ? quae Istaec est pulsatio ?
(The complete effacement in talk of an emphatic word seems unnatural.
A full (?) list of so-called instances in the first eight plays of Plautus, in
Appendix B, shows that the words which follow the monosyllable are
usually subordinate words whose first syllable is long by position (not by
nature), words like ille, atque, esse, etc., and that we have often close
parallels where the Brevis Brevians Law certainly operated. Thus we
may infer the pronunciation se et suos (Amph. 214) from e.g., Amph.
1 131 tibi et tuis ; quo hi loco (Amph. 699 ; Cure. 711) from, e. g., Cure.
354 sibi In manum ; i intro (Aul. 800, etc.) from, e.g., Cas. 881 ubi
Intro, and so on. The practice therefore of the Teubner editors (large
edition) is wrong. They put the slanting stroke by which linguistic
writers denote the accent, but they the ictus, over the wrong syllable in
such lines. Of course the context is not always of a kind to convince
every one that the monosyllable had emphasis, and in such cases each
reader must follow his own fancy. Even the shortening of a naturally
long vowel has often (e. g. autem Bacch. 155, Aul. 333 ; eveniat Bacch.
144) great probability. No one will hesitate over em istuc, em tlloc, etc.
(cf. 52). Our own opinion is that in nearly every line on the list Plautus
made the emphatic monosyllable (in Prosodic Hiatus) a ' brevis
brevians' ; Amph. pr. 17 quam 6b rem; 191 vi et virtute, and so on.
At any rate, in Dialogue Verse. How far the same should be postulated
for verse which does not echo talk is another question.
Another unreasonable practice of the Teubner editors may be men-
tioned here (cf. 38). In the trochaic line-opening, Cure. 557, they scan
rightly Quol homini. But in the iambic line-opening, Cure. 531, they
print Quoii homini. They seem to think that the ictus of the verse, not
everyday pronunciation, determined the quantity. Surely if the pro-
nunciation quol homini was normal, the phrase suits an iambic line-
 HIATUS 247

opening equally well. And any restriction (e. g. through consideration

of the metrical ictus) which treats these scansions as peculiarities of
verse and not as echoes of actual speech must be ruled out. It will not
do to restrict Prosodical Hiatus to the first (not the second) syllable of
a Resolution and therefore to deny the genuineness of a line like Mil.
1067 (Anap.) :
Sed, amabo, mitte me actutum. || Quin tu huic respondes aliquid ?
(cf. Cas. 134; Poen. 1052; Aul. 251).
To return to monosyllables, emphatic or unemphatic, before a short
syllable. Examples, a few out of very many, of Prosodic Hiatus are :
Bacch. 176 (Sen.) Mnesilochi Pistoclerum quern ad epistulam
Mnesilochu' misit super arnica Bacchide ; AP)i
Pseud. 318 (Troch.) || qua opera credam tibi (Una opera alligem, etc.;

72 (Sen.) Haec quae ego scivi ut scires curavi omnia (AP) ;

Epid. 34 (Troch.) Muiciber, credo, arma fecit || quae habuit Stratip-

Merc. 398 (Troch.) Quae habeat quotidianum ||familiae coctum pocles

cibum ;;
Men. 695 (Troch.) Aliam posthac invenito || quam habeas frustratui ;
True. 130 (Iamb.) Die quo iter inceptas ? quis est ? || (AP) ;
Trin. pr. 12 (Sen.) Adulescens quidam est qui in hisce habitat aedibus
(AP);

True. 150 (Iamb.) Habitiiris (' you are fain to have '), qui arari solent|| ;
Pseud. 880 Quin tu illo[s] inimicos potiu' quam amicos vocas ? (AP) ;
Stich. 91 (Troch.) [| vostri. Qui, amabo, pater ? (AP) ;
?Pers. 716 (Sen.) Argentum acceplt, abiit. qui ego nunc scio ? (AP) ;
Merc. 451 (Troch.) Post autem communest ilia || mihi cum alio, qui
scio ? (Emphatic alio) ;
Mil. 1259 (Iamb.) Naso pol haec quidem plus videt || quam oculis.
Caeca amore est (Emphatic oculis) ;
Rud. 937 (Anap.) Sed hie rex cum aceto pransurust ||.
Not the scansion but the sense necessitates Hiatus in Men. 653-4 (Troch.) :
Egon dedi ? Tu, tu istic, inquam. || Vin afferri noctuam
Quae ' tu tu ' usque dicat tibi ? nam || nos iam defessi sumus.
For contrast, we add some examples of Elision of unemphatic mono-
syllables before a short syllable :
Pseud. 153 (Iamb.) Hue adhibete aures qu(ae) ego loquor || (AP) ;
Bacch. 983 (Troch.) || auscultabat qu(ae) ego loquebar (Hardly a Pro- ;
celeusmatic)
ego);
Pseud. 184 (Anap.) ||madefactatis, qu(om) ego sim hie siccus (Emphatic

Poen. 704 (Sen.) Sed haec latrocinantur qu(ae) ego dixi omnia ;
Cist. 100 (Troch.) ||qu(ae) habitat hie in proximo;
 248 EARLY LATIN PROSODY
True. 855 (Troch.) S(i) alia membra vino madeant, || cor sit saltern
sobrium (Emphatic alia) ;
Pseud. 24ia (Anap.) || at priu' qu(am) abeat ;
592 (Anap.) Sed hiinc quern video quis hie est qu(i) oculis ||;
Poen. pr. 124 (Sen.) Hie qu(i) hodie veniet reperiet suas filias ;
animus) ;
Trin. 308 (Troch.) S(i) animus hominem pepullt, actumst (Emphatic

Bacch. H7ia (Anap.) N(i) abeas, quamquam || tu bella es.

To formulate precise rules would be dangerous. Can we safely discriminate
between Most. 1033, io35> io4° (an" Senarii) ?
Deludificatust me hodie indignis modis ;
Deludificatust m(e) hodie in perpetuum modum ;
Quis med exemplis hodie eludificatus est.
(On the treatment of cut see 38 G.)
T(nm) igitur appears in :
Capt. 641 (Troch.) Turn igitur ego deruncinatus ||;
Cas. 374 (Troch.) Turn igitur ego sortes utrumque j|;
Trin. 676 (Troch.) Turn igitur tibi aquai erit cupido ||.
So scan :
Asin. 330 (Troch.) Turn igitur tu dives es factus? |j Mitte ridicularia.
Sim has been declared immune from Elision, but hardly with reason. To
us it seems that the Elision of sim is more likely than the Elision of both
Interrogatives in Trin. pr. 6 (Sen.) :
Nunc igitur primum quae ego s(im) et quae lllaec siet.
(Cf. Pers. 276; Most. 199; Men. 644 ; also Cure. 164.)
Plautus seems to put a Preposition after an Interrogative in phrases like
Mil. 1047 (Qua ab illarum ?), and the emphatic Interrogative is (naturally)
not elided before the short syllable of the (enclitic) ' Postposition '. When
llput after a Relative, the first of the two enclitics takes (naturally) an accent,
Ijand Prosodic Hiatus again appears, e.g. Bacch. 176 quern ad. (See 42
beginning.)
We would scan :
Cure. 467 (Troch.) Commonstrabo quo in quemque hominem ||tass
facile
is) ;;
inveniatis loco
Vid. 90 (Senar.) Quam ad redditurum te mihi dicis diem (Cave demu-

Asin. 119 (Senar.) Nee mage versutu' nee quo ab caveas aegrius.
54. (H) Emphatic Iambic Words. The consensus of A, P and
Nonius Marcellus is strong evidence for Pseud. 319 (Troch.):
Una opera alligem fugitivam ||cariem agninis lactibus (Ace. Sing, of
canes Fern., cf. 49 n.).

We have already compared with Terence's emphatic bonis (II 23 ; 28)

this method of indicating an emphatic iambic Pronoun, e. g. :
 HIATUS 249
Personal : cf. 31) ;

Cas. 724 (Anap.) || tii amas, ego esurio et sitio (On Plautine ego

Men. 389 (Troch.) || tibi et parasito tuo ;

Pers. 61 (Sen.) Unde ego hunc quaestum locum;
obtineo et maiorum

Cist. 499 (Troch.) Et me, si umquam tibi uxorem ||meam

filiam (AP)
dedero ;

Poen. 1308 (Sen.) Quid tibi hanc digito tactio est? Quia mihi
Possessive : libet (AP) (cf. ' what right have you ? ')

Pers. 537 (Troch.) Mea quidem istuc nil refert tu\\d ego hoc facio
gratia {AP) ;
Aul. 463 (Troch.) Qui simulavit met honoris || mittere hue causa
coquos (cf. 42, s.v. Honos) ;
Cas. 915 (Troch.) Tin amoris causa ego istuc || feci, Mil
etc.. 626(Cf.
) ;

Examples of other emphatic words are :

Bacch. 51 (Troch.) Dude unum expetitis palumbem || (P Gram.);
Mil. 1 9 1-2 (Troch.) Domi habet animum falsiloquum ||
Domi dolos, domi delenifica || facta, domi
fallacias (Contrast 194) ;
Merc. 845 (Troch.) Domi erat quod quaeritabam ||;
and perhaps :
Pseud. 317 (Troch.) Aut terra aut mari alicunde || (AP).
But no special emphasis seems present in :
Gram.)
Merc. 257 (Sen.) Navem ex Rhodo quast fieri advectus Alius (AP ;

Stich. 152 (Troch.) Siquae forte ex Asia navis || heri aut hodie
venerit (AP) (cf. Amph. 714);
Poen. 873 (Troch.) Volucres tibi erunt tuae hirquinae || (AP))
Cas. pr. 50 (Sen.) Nunc sibi uterque contra legiones parat (AP) ;
pr. 58 (Sen.) Senis uxor sensit viriim amori operam dare (AP) ;
Poen. 497 (Sen.) Certum. Turn tu igitur die bono Aphrodisiis
(AP);
though we may perhaps ascribe deliberate utterance to a curiously
similar pair of trochaic couplets :

Merc. 1 80-1 Eloquar, quandoquidem me oras. ||tuu' pater — Quid

meu' pater?
Tudm amicam — Quid earn ? Vidit ||
 25© EARLY LATIN PROSODY

478-9 Quid id est quod scis ? Tuu' pater vult || vendere —

Omnem rem tenes.
Tuam amicam — Nimium multum || scis. Tuis in-

(Cf. Merc. 888.) And the word-group seems the efficient cause in
Cure. 497 manu — emittitis ; True. 762 maniim — iniciam. gratiis.
So that the
rule for Iambic words may have to be shaped like the rule for mono-
syllables (53) : An emphatic iambic word always (or normally ?), and an
\\ unemphatic sometimes, is left in Prosodic Hiatus. Virgil avoids elision
\\ of iambic words, as of monosyllables.
We would scan :
Cure. 368 (Troch.) hie ministrabit, ego edam ;
Merc. 619 (Troch.) Non tibi istuc magi' dividiaest ||. quam mi hi hodie
fuit (This line was borrowed by Plautus from
Naevius. ; Since it is quoted by Varro its form is
certain)

977 (Troch.) Optime hercle. perge : ego ad||sistam hinc alterinsecus

(With Hiatus ' in pausa') ;
True. 113-114 (Anap.) Me lllis quid (em) haec verberat verbis,
N(am) ego hue bona mea degessi ;
Pers. 495 (Anap.) Vivas. Benedicts tuis benefact(a) || aures meac
auxilium exposcunt ;
Cure. 519 (Iamb.) || tibi opust, qui te procures ;
Pers. 512 (Sen.) cura qu(ae) is volet ;
N(am) is mihi honores sitae &om\ habuit maximos.
Examples in Terence, like Heaut. 890 (quoted in 50}, Ph. 958, are very
doubtful. And Virgil's artistic imitation, whether of a second call more
faintly heard than the first, as the caller goes further away, or of an echo :
vale, vale, inquit, Iolla (Eel. 3, 79) ; Hyla, Hyla omne sonaret (Eel. 6, 44)
had better be left out of our discussion. But Caecilius, the Comedian who
held the stage between Plautus' and Terence's time, supplies an instance :
186 (Sen.) Opu'entitate nostra sibi iniuriam (Factam) (Cf. Accius 10? ;
Also Ennius in his Annals (193) : 85?).
Hos ego in pugna vici, victusque sum ab isdem.

55. (I) Cretic Words. The last of Cicero's examples (43), from a
hexameter of his own (Etesiae in vada ponti), is clearly a Graecism, if
not also the one from Ennius (frag. 3 of his Panegyric on Scipio) :
Scipio invicte.
Another device for getting this awkward name into a dactylic Hexa-
meter is clearly a Graecism, Lucilius' Scipiadas (cf. Munro on Lucr.
1, 26). But Priscian's example of Hiatus in early poetry :
 HIATUS 251

Ennius Ann. 332 Insignita fere turn milia militum octo

Duxit delectos bellum tolerare potentes,
is not of the Greek type. A parallel (see 49 n.) is :
Ann. 494 Dumquidem unus homo Romanu' toga superescit.
And beside this pair of Ennius we may place these of Plautus :
Mil. 1425 (Troch.) || gratiam habeo tibi;
Capt. 373 (Sen.) Sequere. em tibi hominem. Gratiam habeo
tibi.
These two examples point to pronunciation of this phrase as gratiam-
habeo, like the pronunciation of flagitium-hominis as seven syllables
(45), and suggest the scansion of the four trochaic lines, Epid. 266,
Pers. 540, Rud. 1397 and 141 2, with a Dactyl for first foot (but see
II 58), not a Trochee. (Terence elides, e.g. Ad. 887, 971; not to
mention Eun. 109 1.) Certainly the imitation of a Greek Hiatus in
dialogue-verse seems preposterous. It is Roman talk that is echoed
there. Even in Anapaestic Cantica where dactylic words are most
frequent and dactylic word-endings are most freely used in divided feet
(II 50), and where therefore this Hiatus has most opportunity to occur
and does sometimes occur, — even there we cannot imagine Plautus'
independent mind condescending to a Graecism. The less, because in
the fragments of the Annals of Ennius, where imitation of Homeric
usage would be natural, the Greek type does not show itself. It is
apparently a later fashion in the Roman Epic. Virgil's dislike to elide
a Cretic-word suggests that Plautus' occasional Hiatus with these words
reflects the actual F^mP pron MP^VJQ" Another strong example is :
Pseud. 346 (Troch.) || quindecim habeo minas (AP).
Not strong, but of the same type :
Most. 1 165 (Troch.) Si hoc pudet, fecisse sumptum, || supplici
habeo satis (Emended to su. id hab.).
(See also II 46.)
From Plautus' Anapaestics we take this well-attested example :
Pseud. 1 121 || atque aliquem evocem hinc intus (AP).

56. (J) Other types. Thus far we may have taken our readers with
us. The footing has been fairly sound (except in 51). But in this
paragraph we quit safe ground and tread boldly (or rather hesitatingly)
on surmise. We shall try to keep to Hiatus where the text is well sup-
ported or which seems justified by a priori considerations.
(a) Hiatus * in pausa ' we found to he. a Roman-tape. There might
 252 EARLY LATIN PROSODY

be a pause before a bizarre word, an unexpected turn of the sentence,

etc. (We nowadays use a hyphen in such cases), e. g. :
Most. 1032 (Sen.) Turbavit. Immo — exturbavlt omnia
Most. (But cf.;
1 1 12)

Cure. 334 (Troch.) Quod tibi est item sibi esse, || magnam argenti
— inopiam ('E£ airpoo-SoKrJTov, for copiam) ;
True. 913 (Troch.) Plus decern pondo — amoris || pauxillisper per-
didi (P Gram.) ;
Poen. 443 (Sen.) Nam isti quidem hercle orationi — OedipoGram.) ;
Opust coniectore, qui Sphingi interpres fuit (AP

pr. 89 (Sen.) Praesenti argento — homini, si leno estPoen.

homo 474).
(cf.

» The lusty soldier of fortune in the Miles is a punster (4 ; 8 ; 1022),

even in his darkest hour (1424). The first time this trait is revealed to
the audience there is a pause to mark it, Mil. 4 (Sen.) :
Praestringat oculorum aciem — in acie hostibus.
May we compare Virgil's Hiatus ?
Aen. 3, 606 Si pereo, hominum manibus periisse iuvabit.
/1 (b) A pause is sometimes natural in the enumeration of a list, etc.
Perhaps Virgil might be quoted in support (Geo. 2, 86) :
Orchades et radii et amara pausia baca.
So in Plautus, e. g. :
Men. 1 158 Venibunt servi, supellex, || fundi, aedes, omnia (AP) ;
Merc. 745 (Sen.) Videre, amplec//, auscular/, adloqui ; Gram.) ;
Poen. 1 1 13 (Sen.) Specie venusfo, ore atque oculis pernigris (AP

Aul. 511 (Sen.) Aut manuleariz, aut murobatharii ; (II 42) ;

Men. 476 (Sen.) Prandi, potavi, scortum accubuz, abstuli (AP).
Whether Prosodic Hiatus (amplecti, ausculari, adloqui) may be
inferred from a Cretic line, Most. 152 ;
Disco, hastis, pila, || cursu, armis, equo,
is very doubtful, since Virgil's line, quoted above, makes the vowel in
this Hiatus long. Cf. Cas. 550 (Troch.) :
Propter operam illius hirci || impro^/, edentuli (AP).
(c) That a pause might precede such a parenthesis as haud secus is
suggested by a comparison of Poen. 835 (Troch.) :
Tenebrae, latebrae. bibitur, estur || quasi in popi#<2 — haud secus,
 HIATUS 253

with Liv. Andr. com. 2, if the right setting is (Sen.) :

CorruTt quasi ictus scena — haud multo secus.
What of amabo ? E. g. :
Merc. 538 (Iamb.) || amafo, an maritust ? (AP) ;
Or inquam ? E. g. :

Rud. 1342 (Sen.) 'Turn ego hulc Gri/0' inquito' et me tangito ;

Although the Interjection itself might sanction Hiatus (52), we sug-
gest that Gelasimus puts out his tongue in Stich. 108 (Sen.) :
Nullan tibi lingua est ? Quae quidem dicat ' dabo ' ;
Ventri reWqui — eccam ! quae dicat ' cedo '.
(So At but Coffers Veterem reliqui eccillam, etc.)
But perhaps the types of Hiatus suggested in this paragraph will find
in readers the same incredulity as Klotz's ' Hiatus with Proper Names '
finds in us. Klotz's list (pp. 108- no) provides indeed sufficient refu-
tation of his own theory. In Cure. 358 (Troch.) :
Talos arripio, invoco almam || meam nutricem — Herculem to(Emended
nu. me.),

the Hiatus (if there is Hiatus) seems to us due not to the mere Proper
Name, but to the unexpectedness of the Proper Name (just as the!
Hiatus, already mentioned, before Oedipo, Poen. 443). In Amph. 145
(Sen.) :
Turn meb patri autem torulus inerit aureus
Sub petaso. id signum Amphitruoni non erit,
we put the Hiatus at the pause after petaso. In Bacch. 799 (Sen.) the
correction is obvious :
Constringe tu Illi(c), Artamo, actutum manus,
and in Amph. 785 (Troch.) necessary :
Tu peperisti Amphitruonem (alium), ego || alium peperi Sosiam.
And so on.
The theory that the first of two like syllables was not elided is upset
by, e. g. Stich. 730 Un(am) amic(am) amamus, Ter. Andr. 324 tu(am)
amat. (Cf. Ter. Heaut. 333 ; Phorm. 1041.) An equally weak theory,
that Alliteration prevented Elision, by Pers. 409 Impur(e), inhonest(e),
iniur(e), inlex ; Mil. 213, etc. That Hiatus is proper between two cases
of the same word (e. g. Mil. 4), by Ter. Phorm. 853 hom(o) hominum,
etc. (On Leo's ' Hiatus after -ae ' see 30 B ; on the imaginary ' Hiatus at
Caesura of Senarius ', 48 ; on ' Hiatus after Abl. Sing.', etc., 10.)
 254 EARLY LATIN PROSODY
(For other examples of well-attested Hiatus in Plautus see Journ. Phil.
27, 208 sqq.)
Before passing from this practice of the Roman Dramatists we would
make yet another appeal to our readers to discard the prejudice that
any departure from Greek practice must be wrong. School-training,
familiarity with Virgil, Horace, and other (non-dramatic) poetry, indeed
1 the revolt of Augustan and Silver Age writers against the earlier
{ Republican poetry (' Accius et quidquid Pacuviusque vomunt ')— all
these have predisposed our readers to find something uncouth, inartistic,
immature in Plautus' use of Hiatus and Syllaba Anceps. But is this
fair ? Let them put themselves in Plautus' place. He found in
Menander such lines as (Her. 40) :
Tl OVV ', (TV TL
IIpaTTets v7T€p (tclvtov ; (one ti with Hiatus, the other with Syll.
Anc. at end of line),
or (Epit. 14) :
to irpay/jb ia-rlv ftpa-X^
Kat paSiov /jLaOeuv (with Syll. Anc. at end of line),
or (Epit. 43) :
(TKvOpOiTTOV OVTa //,€

'l8(ov (with Hiatus and Syll. Anc. at end of line).

Sentences like these efface line-division. The actor could hardly allow
.anything of a pause between «' and garrets, between ftpaxv and Kat
jpaoW, between fU and l&w. Is it not the merest step from such lines
of Menander to Plautus' Hiatus and Syllaba Anceps at the Diaeresis ?
And think of the practical side of the question. Would not an
actor's utterance be shackled by the denial to him of all liberty to
distinguish between a hasty and a slow (reluctant or deliberate) answer,
between flippant and impressive enunciation ? What monotony would
fall on stage-talk if no pauses were allowed ! Surely the only fair judge-
ment is that Plautus' departure from this Greek convention was an
improvement. If the dramas of the classical age had survived, we feel
sure that they would be found to conserve the free, elastic, lifelike
practice of Plautus. We have already (49, end) cited one example of
Hiatus from the fragments of Pomponius (of Sulla's time). Here are
other two (Piscatores, frag. 1 ; Aleones, frag. 2) :
(Troch.) Quid habes in scirpiculis, csdve ? || Omne piscati genus,
(Troch.) At ego rusticatim tSLngam, || urbanatim nescio ;
and one of Syllaba Anceps (Pictores, frag. 5) :
(Iamb.) Quae tuleram mecum mi\\d || decern victoriata.
 HIATUS 255

57. ELISION. Plautus knew nothing of the rule (if it is a rule and
not a mere ' ipse dixit ' of Lachmann) which forbids an Elision like
Alphe(a) ab. The Gen. Sing, suffix -ai (and -ei) is freely elided (30).
Other elisions that deserve mention are per(i) (see 40 A), qu(oi) (see
38 G), m(eb) (see II 33). That an emphatic monosyllable would not
be effaced by Elision, at least in dialogue-verse, seems probable indeed,
but cannot be called certain (53). It is dangerous to infer ancient
usage from our own ideas of what is right and proper. We must let
ourselves be guided by facts and be ready to sacrifice the most cherished
theory the moment that it is opposed by them. We cannot be sure
jt.hat deam and not de(am) was pronounced in Asin. 781 (Sen.):
Deam invocet sibi quam libebit propitiam,
Deum nullum,
since the distinction is certainly obscured by Elision in Cas. 785 (Sen.):
Sed properate istum atque istam actutum emittere.
And before declaring that elision of this or that monosyllable was not
permitted (sim, dem, etc.) we must make sure that the word occurs
often enough to allow us to decide. (On the elision of sim see 53, end.)
The elision of a Greek long vowel is striking, Capt. 880 (Troch.) :

|| pa tov 'A7toAX(w). Et servulum.

(On Elision in Ennuis' Epic see 45, note.)
58- VARIATION OF QUANTITY. Some types of variation in
quantity have been discussed elsewhere and need only the merest
mention here :

(a) Shortening of long vowels or syllables after a ' brevis brevians '
(II 19), e. g. ego, voluptatem.
(p) Shortening of long vowels of monosyllables prefixed to quidem
(II 37), e. g. tuquidem.
(c) Shortening of long vowel (or diphthong) before vowel (25), e. g.
fui, Pellaeus.
(d) Cognate words, like sacro- (O- Stem) and sacri- (I- Stem), riibro-
and rubrica. On these see 42 ; on the Adverb contra, etc., 4 ; on
varied reproduction of Greek words, such as Accheruns of Plautus, etc.,
and apparently Acheront- of Ennius, 28.
(e) Syllaba Anceps (3).
It would be inaccurate to ascribe all elasticity of scansion to Dramatic,
rigidity to Epic Verse. Dramatic scansion is elastic wherever actual
pronunciation was varied. Departures from everyday usage were a
feature of Epic Poetry and the like, not of Dramatic. A short vowel
256 EARLY LATIN PROSODY

is not lengthened (by * position ') before a Mute and Liquid in Dramatic
Verse (Il_39)- Sisenna, we feel, must have erred when he ascribed
1 latronem ' to Plautus (42). Such a variation is possible only in Ennius'
Annals, or Lucilius' Satires, e. g. :
Enn. Ann. 474 It nigrum campis agmen ;
Lucil. 923 (Sen.) At cui? quern febris una atque una d7rei/aa;
379 S nostrum et semigraece quod dicimu' sigma.
The Dramatists confine themselves to nigrum^ febris, etc. (II 37), so
that it is quitejrnlikely that even in an Anapaestic Canticumj^Bacgh.
641?) 'duplicibus' could have been tolerated. In Greek there is
a distinction between Tragedy and Comedy, -n-arpos, l/expos, etc., being
admitted to the former but not to the latter. Not however in Latin.
Lengthening ' in arsi ' is an Epic usage, an imitation of Homer, e. g.
Ennius Ann. 87 ; 147 ; Accius Ann. :
Sic expectabat populus atque ore timebat ;
Et densis aquila pinnis obnixa volabat (see 4) ;
Calones famulique metellique caculaeque.

The old fallacy of Ritschl's time, the scansion of a Dactyl^ (instead of

a Tribrach) in e. g. omnia memini et scio (a Trochaic ending), rises from
its grave every now and then. Nor have all Skutsch's convincing argu-
ments (II 36) succeeded in ' laying ' the phantom by-forms ' Me ' and
1 nempe '. We ourselves believe ' eccruis ' to be another phantom (38).
But there is a puzzling variation of quantity which does seem actual
and not imaginary. The word immo is properly a Spondee, e. g. (to
give a few examples out of a host) :
abiero ;
Bacch. 211 (Sen.) Tanto hercle melior. Immo— Immo hercle

Capt. 933 (Troch.) Proinde ut tu promeritu's de Immo

me et potes
||filio.;

Merc. 297-298 (Sen.) Immo bis tanto valeo quam valui prius.
Bene hercle factum et gaudeo. Immo si scias j
Pseud. 920, 934 (Cret.) Ambula ergo cito. Immo otiose volo.
Iuppiter te mihi servet. Immo mihi ;
Merc. 735 (Sen.) Non tu scis quae sit ilia? Immo iam scio.
But three lines after this last example the word appears as a Pyrrhic :
Merc. 738 (Sen.) Immo sic : sequestro mihi datast. Intellego
(Hardly sequestro-mihi, like molestae-sunt).
And this pyrrhic scansion is by no means unique, e. g. :
Cas. 634 (Choriamb.) Vae tibi ! Immo istuc tibi sit. || Ne cadam,
amabo, tene me (II 32 B) ;
 HIATUS 257

Cist. 565 (Sen.) Immo meretrix fuft ; sed ut sit, de ea re eloquar

(Emended by deletion of mer.) ;
Ter. Heaut. 101c (Iamb.) Immo scis potius quam quidem redeat ||
ad integrum eadem oratio ;
Ph. 936 (Sen.) Immo vero uxorem tu cedo. In ius ambula;
Hec. 437 (Sen.) Immo quod constitui mehodie conventurum eum ;
877 (Iamb.) Immo vero scio, neque hoc impru||dens feci.
Ego* istuc sati ' scio (II 7) ;
Caecilius 128 (Sen.) Immo vero haec ante solitu' sum : res delicat.
(Cf. Amph. 726; Bacch. 672; Cas. 362; Most. 731; 1091 ; Poen.
1 231 ; Ter. Ph. 1047 ; Hec. 726.)
Perhaps the puzzle will not be satisfactorily solved until the etymology
Iof immo has been discovered. For immo (like another common
Particle, igitur) has hitherto baffled philologists.
The variation eheu is not quite a parallel (see 42). For the pronun-
ciation of an Interjection is often arbitrary; and the short e may be
a mere ' correptio vocalis ante vocalem.'
(On the dropping and retention of final s after a short vowel see 16.
On aio see 42. On pro- see 29.)
This seems a suitable place for mentioning that a final short vowel is
[ freely used in Early Latin Verse before sc-, sp-, etc., e. g. Accius
'
I 64 (Anap.) :
Depressum altis || clausere specis (the last word means ' caverns ')
(for instances in Lucilius see Marx' Index, p. 168),
though Ennius apparently admitted to his Epic (Ann. 96) stabilita
scamna solumque.

59. Slurred Pronunciation. We have spoken of ' slurred ' and

1 deliberate ' utterance, instead of the technical terms of German writers
on the subject, ' Schnelltempo ' (or Allegrotempo) and 'Lentotempo '.
For since an English scholar, criticizing these terms, made the extra-
ordinary statement that time could not affect pronunciation, it is clear
that they are not understood in this country. What is it that reduces
' we will ' to the slurred form ' we '11 ' if it is not the hurry of. utterance ?
And we have not deemed it necessary to furnish a phonetic account
of the slurred forms of Latin speech, e. g. voliiptatem, any more than it
is necessary for a treatise on Shakespeare's Verse to explain by what
precise process of articulation ' we will ' passed into ■ we '11 ', ' he has '
into 'he's' (written 'h'as' in the old editions of plays). Probably
a phonetician would tell us that ' we will ' first dropped the vowel i by
Syncope, producing ' we w'll ' and that the second w then passed into
?348 S
 258 EARLY LATIN PROSODY

an indeterminate vowel-sound ; just as ' Eton ' is pronounced not ' Etun '
(with u representing this indeterminate vowel), but has been reduced still
further to ' Etn ' (with n representing the nasal vowel, a vocalic n like
the vocalic r of French ' sabre '). Without committing ourselves to any
precise statement of this sort (it lies outside our province), we may at
least point out that the old explanation of voluptatem by Ecthlipsis,
\ v 'luptatem ' or ' vol'ptatem ', does not tally with the facts. For the
word was actually pronounced as a quadrisyllable with the first two
syllables short. The Roman spelling has fortunately left us a record of
the actual process which transformed calefio, first to calefio, (quadri.
syllable), then to a trisyllable calfio. The second stage (calefio) shows
operation of the Brevis Brevians Law, the third (calfio) of Syncope.
To an Englishman ' slurred ' pronunciation perhaps suggests Syncope
more than any other phonetic process. The Syncope is pretonic in
e. g. ' m'Lady ' for ' my Lady ', ' tis ' for 'it is ' ; posttonic in e. g. ' it's '
for ' it is \ And Syncope is the explanation of many slurrings in
everyday colloquial (not the same as ' vulgar ' or ' plebeian ') Latin,
e. g. cette (for ce-date). Ecthlipsis would perhaps be the best descrip-
tion of e. g. dic(e) mihi, where the short e seems ' squeezed out ' between
the rapidly uttered pair ; nemp(e) tu dixisti ' yes, it was you who said it '
(II 36). How a phonetician would explain cui(u)s (Relative) beside
cuius (Interrogative) we do not know (II 35). An ingenious attempt to
adapt the slurring mebs to the Latin phonetic law * corripitur vocalis ante
vocalem ' has been already mentioned (II 34).
It seems best to include all these echoes of conversational Latin under

the wide term ' slurred ' (as opposed to ' full ' or ' deliberate ') pronun-
ciation. They are appropriate to Dramatic, especially to Comic Verse,
where the actual pronunciation of everyday life was reproduced ; even
though some of them, e. g. viden, forced a way into elevated poetry
(II 25), and a few, e.g. male (II 17), quite ousted the correct form,
j Whether any modern language exhibits the same phonetic process as
( reduced calefio to calefio is uncertain. It may have been a peculiarity
I of Latin. But that it was an actual phonetic process (and no mere
trick of poetry) is absolutely certain (II 21). It was produced by the
sentence-accentuation, and (if phoneticians are right) proves Latin to
have had a stress-accent (not, like Greek, a musical accent or ' pitch-
i accent '). It is the chief difficulty to readers of Plautus nowadays, who,
knowing only the literary prosody of Virgil, are apt to regard such scan-
sions as bon!s (II 23), ab exercitu (II 48) as uncouth attempts of the early
versifier to carve a poetic line out of unyielding material. They should
force themselves to think of these scansions as they think of Shake-
 HIATUS 259

speare's ' i ' faith ', ■ we'll go ', beside the ■ in faith ', 4 we will go ' of
a poet like Milton (I 5). And, to aid the reader, an editor of Plautus
should adopt some such device as editors of Shakespeare, and print
netnp'* tu (like ' i ' faith ') and so on.
A full phonetic treatment of the utterance of the Latin sentence in
Plautus' time (presumably much the same in Cicero's) cannot be
attempted in this book. But if any reader takes the trouble to gather
the hints scattered through our pages, he will see that the pause at the4
end of a line (and, at least often, a hemistich), at the end of a sentence,
\at the end of a clause, is antagonistic to a very large number of these
I slurred forms, indeed to all (properly speaking) which owe their slurred
/ pronunciation to the influence of a following word or a following accent.
\ Thus dice, face, duce, huius (Trochee), uti (disyllabic form of ut\ meos\
(Iambus), etc., are the correct forms at any such pause. They will
usually be found there. And where the traditional text offers the
! slurred forms instead, the suspicion of an editor should be aroused (unless
1 the slurring is the effect of a preceding accent merely, e. g. bonust).

s 2
 IV
EARLY LATIN METRES

i. Unjust Depreciation of Early Latin Metres. In beginning

the last chapter we mentioned Ritschl's poor opinion of the Latin
language in Plautus' time. He believed scansions like apud mensam,
1 feriint non ducunt hominem ', to imply that final consonants and con-
sonant-groups were dropping from the language like branches from a
withered tree. To meet the objection that these endings still appear
hale and hearty in the classical and Augustan periods (and later), he
invoked Ennius as a ' deus ex machina '. The Annales had saved Latin
from premature decay. (Leo's theory of the Elision of -us before a
vowel is an echo of this ; III 17.)
He had an equally poor opinion of Plautus' stock of metres. The
old Umbrian was, in his eyes, a rude versifier who could indeed turn out
dialogue-verse that pleased,1 but was incapable of anything that could be
mentioned in the same breath with a chorus of Euripides or Aristophanes.
While the Greek Dramatists could weave more than one thread of metre
into a line and indulge in variations of the same theme, Ritschl and
those who edited the plays according to his precepts seem to have held
that if Plautus begins a line with, let us say, a Cretic foot or a pair of
Cretic feet, he could not but pursue a plain course to the end of the
line. It must be edited, by hook or by crook, as a plain, unvarnished
Cretic Tetrameter. And they allowed to Plautus Iambics, Trochaics,
Anapaestics, Bacchiacs, Cretics, but hardly anything else, and no varia-
tions of the simple types except the Versus Reizianus. In desperate
cases Ritschl was apt to resort to the Trochaic Octonarius (and some of
his Trochaic Octonarii send a shiver down the reader's back) j though
with the Teubner editors Anapaestic long lines became the usual hospital

1 Ritschl, like most people, made the 'numeri innumeri ' of Plautus' epitaph a
Graecism (fivOfxol appvO/xoi or p\krpa afierpa), savouring less of the early dramatists
than of the ' novi poetae' (e.g. Catull. 64, 83 funera nee funera). Even if the
epitaph be Varro's composition, we prefer the interpretation ' countless metres \
Plautus had adapted a wonderfully large variety of Greek metres. At his death
1 numeri innumeri simul omnes collacrumarunt '.
 EARLY LATIN METRES 261

for incurables, like that choice specimen already cited (with amorem and
nitoribus), Cas. 2 1 7a :
Omnibus rebus ego amorem credo et nitoribus nitidis antevenire
(nee potis).
(See below, 32.) The re-casting of the traditional text of Cantica led
to facetious contrasts between the 'editor's Plautus' of the (large)
Teubner text and the ■ real Plautus ' of the Teubner apparatus criticus.
To Leo we all owe a truer appreciation of Plautine Cantica. And for
any editor inclined to ' slash-cut-and-carve ' editing we cannot prescribe
a more wholesome lesson than a study of Schoell's edition of Cantica
(and Schoell's defence of departure from tradition) in comparison with
Leo's edition of the same Cantica. For ourselves we take as guiding-
principle that if a line cannot be scanned in this or that metre smoothly
and without outrage to the ' pura oratio ' of the old Comedians, that
cannot be the metre of Plautus. Take any Canticum whose metre is
quite certain, say the lover's serenade at the beginning of the Curculio
(For the benefit of any reader who may not be familiar with Cretic
Metre, we indicate the feet by vertical lines), Cure. 147 sqq. :
Pessul(i), heus | pessuli, | vos salujto libens,
Vos amo, | vos volo, | vos pet(o) atq(ue) | obsecro,
Gerit(e) aman|ti mini | mor(em), amoelnissimi,
Fite cau|sa mea | ludii | barbari,
Subsili|t(e), obsecr(o), et | mittit(e) isjtanc foras
Quae mihi | miser(o) aman|t(i) ebibit | sanguinem.
Hoc vid(e) ut | dormiunt | pessuli | pessimi
Nee mea | gratia | commovent | s(e) ocius.
If Plautus * can produce smooth Cretics like these, why should we father
rough Cretics on him ?
2. Dance-metres. A recent attempt at a scheme for Plautus'
Cantica (and Aristophanes' Choruses too) in the light of dance-music
(Sudhaus' Aufbau der Plautinischen Cantica, 1909; cf. Rhein. Mus. 65,
515) has been so annihilated by Leo's criticism (Gotting. Gel. Anzeig.
191 1, pp. 65 sqq.) that it hardly deserves mention here, except as a
specimen of the deplorable results of a ' metrical ' Brevis Brevians Law.
Its author, in order to make Most. 869 suit his theory, takes this variety
(24) of a Bacchiac Tetrameter :
Sincerum atq(ue) | uti || vetem verjberari
1 Modern scholars, under the influence of Horace's preference of Terence's
' Attica elegantia ' to Plautus' ' non astrictus soccus ', take for granted that Terence
avoids song-metres because Plautus' attempts had seemed to him a failure. A more
natural explanation is that Terence had not the gift of writing songs.
 262 EARLY LATIN METRES

and actually scans it (through a brick-wall, as the saying is) as an

Anapaestic Dimeter, with verberari or verb'rari. This is nothing short
of an outrage on Plautus, that well of Latin undefiled (Cicero de Or. 3,
45).
Whether the requirements of dance-music can put us on the right
track for some Cantica remains to be seen. Menander's Chorus-inter-
lude seems to have been a dance (perhaps a song too, though no words
are given) of a band of revellers (kiojjlos, whence Kco/xwSta). There was
the same thing in the Bacchides, for otherwise Plautus would not have
admitted the line which gives intimation (107) :
Simul huic nescioquoi turbae ||quae hue it decedamus hinc ;
and probably in other plays too. Whether the interlude in the Pseudo-
lus at the end of the First Act ($73a) :
Tibicen vos interea hie delectaverit,
combined dance and music we cannot say. (The Curculio interlude,
IV i, was a speech.) At the end of the Stichus the ' tibicen ' sets his
lips to the wine-flagon (meanwhile the metre becomes Iambic Senarii,
Stich. 762-768), and then answers the appeal for a 'cantio nova pro
veteri vino ' by music for the following Iambic lines, in which we hear
the patter of heel and toe :
Sa. Qui Ionicus aut cinaedicust || qui hoc tale facere possiet ?
(Octonarius)
Sti. Si istoc me versu viceris, ||alio me provocato (Septenarius).
Fac tu hoc modo. At tu hoc modo. || Babae ! Tatae !
Papae ! Pax ! (Ditto)
Nunc pariter ambo. omnes voco ||cinaedos contra. (Vers.
Reizianus)

Satis esse nobis non magis ||pote quam fungo imber

In the last Act of the Pseudolus the staggering feet of the carouser(Ditto).
move
)( o Bacchiac (y ) metre at first :
Quid hoc? sijein hoc fit, | pedes ? sta|tin an non?,
and after some Ionic steps (1274 sqq.) :

Ad hunc me mojd(um) intul(i) Mis | sati' facete

Nimis ex dis|cipulina, | quipp(e) ego qui
Prob(e) Ionica | perdidici. | sed palliojlat(im) amictus (A sort of
' skirt-dance ' ?),
shuffle through one line in Palimbacchiac (— ^) :
Plaudunt, ' pa|rum ' clamijtant m(i) ut re|vertar,
 EARLY LATIN METRES 263

and so on. His final ' tour de force ', seven continuous Cretics, must
have left him with little breath to beg applause :
Verum si | vultis ad|plaudere atjque adprobajre hunc gregem et |
fabulam, in | crastinum ||vos vocabo.
The last two words (a Trochaic metrum) we imagine to be spoken 1 ' at
the footlights '.
3. (A) After the unexpected discovery that Menander used a Chorus,
who dares assert that Plautus must have found his Canticum-models in
Mime (or the like) and not in the New Comedy ? And yet the inde-
pendence ofRoman adapters is seen from Gellius' citation (Noct. Att.
2, 23) of an Iambic Trimeter passage of Menander ('Ett' a^oripav plv fj
'irtKk-qpos rj KaA.77, etc.) with Caecilius' adaptation in a Canticum of various
metres (see below, 44). If Caecilius, that closer follower of his Greek
originals, took this independent course, what may we expect from the
abandon of Plautus ? When we read his ' sermones ', his slaves' volleys
of banter, we feel that no icily regular Menander is his guide there.
And why should he not have taken a course of his own in his Cantica
too ? Our own belief is that the more one regards Plautus as an artist
combining much of the skill of a Gilbert and a Sullivan, the better for
one's comprehension (and emendation) of this Musical Comedy of the
Elizabethan Age of Rome, when the Carthaginian foe had been over-
whelmed and his armada shattered and when Spain, Italy, Syria, Greece,
and the rest were rapidly overrun in one triumphant, irresistible progress.
Plautus was the ' flos poetarum ' (Casina pr. 1 8) in these palmy days of
Rome. A far later age, the age of Rome's decay, still clung to him and
(the school-author) Terence and let Naevius and the others go.
(B) Line-division of Cantica in Ancient Texts. The two ancient copies
(one actual, the other hypothetical) from which we derive our knowledge of
Plautus were not cheap copies like the papyrus fragments of Menander. In
a cheap copy space would be saved by writing two (or more) short lines as
one or even by obliterating all the line-divisions of a Canticum and present-
ing it as prose. In the stately pages of A economy of space seems wholly
disregarded ; e. g. at Pseud. 205 we find two separate lines before the Iambic
Dimeter : Sed
Vah
Nimius stultus, nimis fui.
1 We do not forget the Roman practice of having the vocal part of Cantica done
by a professional singer behind the stage, while the actor undertook the movement
(Livy 7, 2, 10). This is the only play of Plautus which does not end with a
Septenarius. If we give rein to our imagination (and why not?), we can find the
usual Trochaic Septenarius here too, the words being repeated by the second and
the third actor :
a. Vos vocabo. b. Vos vocabo. II c. Vos vocabo. grex. Plaudite.
264
EARLY LATIN METRES

And yet one cannot quite shake off a suspicion that towards the end of a
play the scribe may have yielded to temptation now and then (since it was
necessary to begin a new play with a new leaf) and have departed to a slight
extent (e. g. by joining two short lines into one long, two Dimeters into a
Tetrameter) from the line-division of the ' Ambrosian ' edition. For the line-
division of the * Palatine ' edition we are usually (but by no means always) at
a loss in the plays (or portions of plays) for which the Codex Turnebi is not
in evidence. Where both A and T fail us, we have generally no direct means
of knowing how Plautus divided Canticum-lines, unless the metre itself makes
this unmistakable. Where either A or T is available, it seems folly for an
editor to prefer his own judgement to theirs. Our two ancient editions of
Plautus go back to (at least) the second century A.D., when (as we have
remarked, I 6) every care was taken to restore the ancient form of a Republi-
can author.
The scribe of the Codex Bembinus of Terence has not written the Adelphi
Canticum quite so correctly as it appeared in the archetype of the minuscule
MSS., for he has combined the opening ' clausula ' (Discrucior animi, cited
as a 'clausula' by Varro) with the following line. Yet it is possible that the
scribe has not been careless ; for this may be an editor's deliberate choice
(e. g. we may imagine a space left blank between the ' clausula ' portion and
the rest).
In the
sented so Ambrosian
: Palimpsest a bacchiac ' run ' at Cas. 673 sqq. is pre-
Quid cum ea
Negoti tibist ? Peccavi :
Illud dicere vilicum volebam.
Apparently the opening short line was meant for a signal to the reader that
a 'run' ('system') followed. These 'runs' seem to have troubled ancient
editors as they trouble editors nowadays. How can they best be presented
to the reader ? If they are written continuously, without division into lines
(as in the Oxford text : Quid cum ea negoti tibist ? Peccavi : illud dicere
vilicum volebam), the reader may mistake them for prose or, at any rate, may
fail to detect the metre. If they are cut into equal line-lengths, this may in-
volve the breaking up of a word between two lines. And if they do not fall
into lines of the same length, is the shorter line to be put first or to be put
last ? The Adelphi Canticum ends with a Choriambic ' run ' (ending in a
Spondee) which the Bembinus presents as a long line followed by a short :
Sostrata credit mihi me psaltriam hanc emisse ; id anus
Mi[hi] indicium fecit.
but the other edition apparently as one long line. The bacchiac ' run ' at
Pseud. 1329 was in the ' Ambrosian ' edition cut into two halves :
Quid nunc ? numquid iratus es aut mihi aut filio propter has res, Simo ?
Nil profecto.
solent,
I hac : te sequor. Quin vocas spectatores simul ? Hercle me isti hau
 EARLY LATIN METRES 265

while the ' Palatine ' edition made it (or made it look like) a Bacchiac Tri-
meter Catalectic followed by Cretic Tetrameters :
Quid nunc ? numquid iratus es
Aut mihi aut filio propter has res, Simo ?
Nil profecto. I hac : te sequor. Quin vocas
Spectatores simiil ? Hercle me isti hau solent.
The divergence of our two ancient editions of Plautus (and of Terence too ?)
in the arrangement of Cantica seems, so far as the limited material allows a
decision, to have been merely of this slight and trivial extent. Therefore the
arrangement of the Cantica of Plautus (and Terence) seems to have been
settled authoritatively in some standard edition (of the Republican period ?
Cf. Leo, Plaut. Cant., p. 6), and to have been transmitted without any serious
change (even its indentation of lines, h clo-Oeaet and iv iuBwa), Fortunately ;
for if editors of (let us say) the fourth century A. D. had tried to arrange (or
re-arrange) them, our confidence would be shaken. Donatus, a very favour-
able specimen of fourth century learning, does not even know that a sixth-
foot Trochee cannot precede a final iambus-word in the Trochaic Septenarius
(16), for in Terence Adelph. 559 :
Usque occldit. Hem! quid narras ? || Em! vide ut discidit labrum,
he mistakes discidit for discidit. In the beginning of the sixth century
Priscian's remarks on the bacchiac metre of Andr. 481 sqq., and his ascrip-
tion of it to True. 95 sqq. and to a Turpilius fragment (Ionic ?), seem arrant
nonsense (de Metris Terenti, p. 425 K.) : Terentius trochaico mixto vel con-
fuso cum iambico utitur in sermone personarum quibus maxime imperitior
hie convenit, quern, puto, ut imitetur hanc confusionem rhythmorum facit.
Sunt autem et trimetri et hoc plus minusque, et habent paenultimam versus
syllabam in quibusdam longam, in quibusdam brevem, ut in Andria :
Adhuc, Archilis, quae adsolent quaeque oportet, etc.
Hos sequitur dimeter catalecticus finiendi sermonis causa quern ad Archilida
habuit :
Date : mox ego hue revertar.
Similiter Plautus in Truculento eodem metro usus est in sermone ancillae
Astaphii :
Ad fores auscultate atque adservate aedes, etc.
Turpilius in Lindia nautae personam inducit hoc metro loquentem :
Dis advenientem perdant : hicquidem nos perdit.
Festum esse diem hie quartum hodie iterant : ita conventum.
Arrant nonsense too is his description of the metre of Amph. 161, 164 (de
Metr. Ter., p. 422 K.) : utitur tamen in hac ipsa scena et dimetris brachycata-
lecticis, id est a tribus simplicibus pedibus, ut ' Ita peregre adveniens ' et ' Qui
hoc noctis a portu '. But, though we reject it, we accept his line-division,
presumably the traditional division. For his line-division (de Metr. Ter.,
p. 422 K.) of True. 120-121 agrees with AP. Marius Victorinus' account of
1 clausulae ' (nmXapia) is vague, but not nonsensical (Gramm. Lat: vi, 79, 1) :
266 EARLY LATIN METRES

quod vero ad clausulas, id est minuscula cola, pertinet, quot genera versuum
sunt, totidem eorum membra pro clausulis poni possunt et solent in canticis
magis quam diverbiis collocari, et praecipue apud Plautum et Naevium et
Afranium. Nam hi maxime ex omnibus [membris] versuum colis ab his
separatis licenter usi reperiuntur in clausulis.
(C) Ejaculations ' extra metrum '. The ' Ambrosian ' edition (at any rate,
its extant copy, the Ambrosian Palimpsest), as mentioned above, put Sed and
Vah in separate lines at Pseud. 205. They are ejaculations which do not
belong to the line before which they stand. Whether the ' Palatine ' edition
(or any copy) wasted vellum so needlessly we do not know. The minuscule
MSS. merely write the independent word (or words) with the line itself,1 just
as editions of (let us say) Middleton print (Trick to Catch the Old One iv i) :
Tut ! nor him nor them we in this action fear,
and do not honour the interjection with a separate line. (Of course the actual
blank verse begins with Nor him.) Editors of Plautus have not always dis-
lodged these intruders. They should print ' extra metrum ' : Cure. 131 Ah !
(also Trin. 495 ?) ; Mil. 1054 Age ! (also Stich. 723 ; also Cas. 829 ?) ; Most.
9 Em ! (but cf. Ill 52) ; Rud. 177 Hem! ; Mil. 962 Vah ! ; Trin. 318 Quid !
(also 413?); Merc. 884 St!; and perhaps Pseud. 218 (Choriambic ?) Ain ?
(On Stich. 660 see II 32 B.) The Palimpsest itself writes them usually in
this way (Epid. 182, Pseud. 130, 600 St!; Stich. 244 Eu ecastor!; Stich.
259 Au !), but assigns a separate line to Cas. 6i9(Attat!).
(D) Transference from end of line to begimiing of next line. (On Senarii
see II 61-62.)
This is a convenient place for mention of another (ancient) editorial con-
vention. Aristarchus (cf. Schol. ad II. 24, 331) wrote a final elided be at the
beginning of the next line (as in the Laurentian Sophocles at Oed. Col. 17
8' et(r<»). The Cairo papyrus of Menander begins Peric. 161 8' elo-iwv, just as
Priscian (de Metr. Terent. 427, 14) cites a line of Pindar 8' ovbh Trpoo-airecov
tydeyt-dfxav iiri. So the practice was common. Aristarchus also wrote at the
beginning of the next line the second syllable of a final z^va elided, e. g. 11. 24,
331-2 Zrj\v' is irediov. Our ancient texts of Plautus seem to have sacrificed
the metre to the word in such a passage as Epid. 1 73-4 (Cretic) :
Revereor filium. At pol ego te credidi ux-
orem quam tu extulisti, pudore exsequi,
and to have made the first line end with credidi and the second begin with
uxorem. This is really a variety in the presentation of a ' run ' (see above),
and a very plausible variety. But they extended the practice to such a
passage as Cas. 827 (Bacchiac) :
r Facies tun hanc rem mi ex parata iraparatam ? id
Ouaerunt, id volunt, haec ut infecta faciant,

1 Not always at the beginning. Thus St ! St ! Most. 506 are written at the end
of Most. 505, an indication that in the Palatine archetype they occupied a separate
line. •
 EARLY LATIN METRES 267

where they end the first line with imparatam and begin the second with id
qiiaenmt. (Other examples in Lindsay, Ancient Editions of Plautus, p. 79.
Also e. g. hoc Merc. 181 ; id Amph. 792 ? On traces in the Terence tradition
see above, II 60.)
IAMBIC.
4. Horace tells us that when the pure iambic metre became adapted
to the dialogue of Tragedy, it was steadied by Spondees :
spondeos stabiles alterna in iura recepit.
A line like Catullus' :
ait fuisse navium celerrimus
has the restless motion of the yacht riding the Adriatic waves, while
Ennius' (Medea fr. ii) :
antiqua erilis fida custos corporis
echoes the deliberate, dignified address of Tragedy. Greek Comedy
departed from the tragic type by admitting an Anapaest in the Iambus'
place, an admission which threatened to destroy the iambic character of
the dipody. In the opening line of the Clouds :
w Zcv /3acrc\€v, to xprj/xa, etc.,
the first dipody seems unmetrical when compared with e. g. w ZcO 7raTep.
And the danger was acknowledged by the restriction of the Anapaest of
dialogue within limits : it was not allowed the free scope of an Anapaest
in anapaestic verse ; e. g. it could not be divided between words in this
fashion :
w Siofiara Acv/ca, etc.,
nor in this :
a> Sw/xa yepovros, etc.
These restrictions of the dramatic Anapaest are observed even more
strictly by the Roman dramatists. But the Romans, in the adaptation
of iambic metre to conversation, went a step beyond the Greeks ; for
while the Greeks stopped short at the Anapaest, they admitted also the
Spondee. Still this Spondee was restricted too. It was required to
adjust its ictus to the accent of the sentence. Thus e. g. :
antiqua errantis
would be a legitimate opening, since both ictus and accent fall on the
second syllable of errantis, but not e. g. :
antiqua errans,
since the accent falls on the first syllable of errans, while the metrical
ictus falls on the second. It is unfair to assert that this admission of a
(restricted) Spondee was so much more fatal to the iambic character of
268 EARLY LATIN METRES

the line than the admission of a (restricted) Anapaest as to kill the

dipody and change the Trimeter, a line of three dipodies, into a mere
line of six feet (II 4).
5. This regard for Accent is a Roman feature, due to the stress-
character of Roman accentuation ; and no one who is ignorant of Latin
Phonetics can appreciate the rhythm of Plautus' Senarii or even under-
stand their prOSOdy ; thrmgh wp rannpf tfiP ctrnng1V ineUt that- f>|antinp
vprsp j§ qnpnfitgH'y^ w^» a^pnf-||a1 In case any one thinks these
obvious remarks to be uncalled for, we would remind him that it is not
so long since the first foot of the dipody ab exercitu in a Senarius was
declared to be a Bacchius, and Plautus' strange substitution of a
Bacchius for an Iambus in an iambic line was justified by an appeal to
English verse, to accentual (not quantitative) verse ! Any one who has
made an honest attempt to understand Plautine Metre and Prosody by
acquainting himself with Latin Linguistics cannot but feel indignation at
the censure so glibly passed on Roman Dramatic Metre by people who
have indeed adequate knowledge of Roman (and Greek) poetry, but are
profoundly ignorant of the history of the Latin language. Their un-
favourable comparison of the Latin imitation with the Greek model
usually rests on two false assumptions :
(1) that Plautus' Senarii are a mere imitation of Menander's Tri-
meters. Now while Catullus' Iambics proclaim themselves to be a mere
imitation of a Greek pattern, even by their Syntax and their Prosody :
ait fuisse navium celerrimus,
et inde tot per impotentia freta,
Plautus is an independent versifier. He is trying to echo Latin every-
day (educated) talk.
in Latin. He is not trying merely to echo Menander's verse

(2) that the deliberate utterance of each word is the utterance in a

line of Comedy as much as in a line of Epic ; that Virgil's prosody must
be the standard for Plautus ; that ab exercitu is faulty, because Virgil
would scan ab exercitu. By this argument Menander's cyw8a for iyw oTSa
and Shakespeare's ' I'll ' for * I will ' must be faulty too.
6. Cicero's remark may be quoted against us (Orat. 184) : comicorum
senarii propter similitudinem sermonis sic saepe sunt abiecti ut nonnum-
quam vix in eis numerus et versus intellegi possit. But does this do
anything more than point the contrast between elevated and conversa-
tional poetry ? Would not the same remark apply to Middleton's lines ?
We mean lines like (A Trick to Catch the Old One I i) :
Arm your wits then
Speedily ; there shall want nothing in me,
 EARLY LATIN METRES 269
Eith(e)r in behaviour, discourse or fashion,
That shall discredit your intended purpose,
lines which deserve praise for their closeness to everyday speech, since
this is what Middleton was aiming at. Middleton pleads in a preface
that a new type of Comedy required a new type of verse : ' The fashion
of play-making I can properly compare to nothing so naturally as the
alteration in apparel ', etc. In the Romantic Drama, say Love's Labour
Lost, we tolerate or applaud conversation-verse that is far removed from
actual conversation. And when Middleton has a romantic passage, a
lover's protestation or the like, he writes in that style. But his ordinary
blank verse we praise because he can imitate actual, everyday, unim-
passioned speech without having recourse to prose.
Iambic Senarius.
7. In chap. II we have gone through the chief differences between
Plautus' Senarius and Menander's Trimeter. Here we may briefly re-
capitulate them and refer our readers to that chapter for fuller details.
(1) Plautus admits Spondees to the second and fourth feet, but not
y
(as a rule) Spondees with clash of ictus and accent (II 2-7). The same
regard for Accent may be seen in the two (three ?) following rules.
(2) Plautus, while observing scrupulously Menander's rules for the
division of resolved feet between words, adds to them a ban of his own,
against the pyrrhic caesura of a Tribrach (e. g. not ' Ajptjum\ II 44-
47). The trochaic caesura of a Dactyl (e. g. Vendit aim), though not
wholly forbidden by Plautus, is far less used than by Menander (II 53-
54)-
(3) a tribrach word (or word-ending) is not allowed as a foot of the
Senarius, whereas it is quite common in Menander's Trimeter (II 8 ;
58). A dactyl-word,1 though not wholly forbidden by Plautus, is far
less used than by Menander (II 58).
(4) an iambic word at the end of a line must not be preceded by an
Iambus in the Senarius, whereas this is quite a common thing in
Menander's Trimeter (II 58).
(5) Caesura is enforced on the Senarius ; hardly on Menander's Tri-
meter (II 41-42).
(6) the Proceleusmatic (w^w^), a favourite foot with Plautus, is
apparently not used by Menander (II 52).
A fair comparison of the Senarius with the Trimeter must take account
of all six points. Thus the impulse (after reading n°. i ) to pronounce
1 Never a word-ending. Even the scansion of Pseud. 146 as an Iambic Octona-
rius with such a Dactyl in the third foot would be illegitimate : Ut ne peristromata
quidem, etc. (see Class. Quart, xiv, 52).
27o EARLY LATIN METRES

the Senarius to be either irregular or overweighted with Spondees l must

be checked by a consideration of nos. 5 and 6. For the strict observance
of Caesura by the Senarius gives a regularity to the construction of the
line, whereas in a Trimeter like (Sam. 7) :
KaOapa 7TO€U/, 7T€TTCtv, ivdpx^crOat kclvovv,
the words seem to be set down anyhow ; and the interspersing of Pro-
celeusmatics may give a lively tone to the Senarius that the Trimeter
never reaches (II 52). Since the Latin language had a stress-accent
(like our own), the regard for Accent is admirably adapted to bring the
Senarius close to Latin conversation, and (like Caesura, with which it
goes hand in hand), gives an aspect of rule and order, a symmetrical
form to the Senarius which is lacking in such a Trimeter as that just
quoted from Menander.
The Prosody of both authors follows of course the utterance of con-
versation. Sometimes the two languages take the same path : e. g. ttov
\m; and ubist?, monosyllabic Qzov% and deos ; sometimes each takes a
path of its own, e. g. iy&Sa, ab exercitu, etc. And if any one says that
the polished Greek writer would have smiled at a scansion like ab
exercitu, we reply that Plautus would have grinned at a Dactyl like fxrjSe
cv. Imagine neve in a Dactyl in Latin !
No one who reads Plautus' Senarius with due regard to Latin
Phonetics will find it so unworthy of comparison with Menander's Tri-
meter as
2 those find it who know nothing of Linguistics. That both
Plautus and Chaucer should be read by such ' semidocti ' is inevitable,
indeed desirable ; but that criticism should be published by them on
the Metre and Prosody ought to be as inconceivable in the case of
Plautus as it would be in the case of Chaucer.
8. We add some details here for which there was no room in chap. II.
(A) Iambic Word at End of Senarius.
To the rule that it must not be preceded by an Iambus there are three
exceptions :
(1) whena monosyllable precedes the iambic word. Examples will be
1 The six Iambi give a jaunty tone to the Parasite's vapourings in Pers. 352 :
Ferant eantque maximam malam crucem.
A commentator on Horace (Pseudo-Aero, ad Epod. i, 1) says :
Comicus (trimeter) erit qui frequentes tribrachos habet, ut :
Agite (agite) ! quid dubitatis hilares dare choros ?
The source and age of his citation are unknown.
2 The opening Scene of the Heros is no bad specimen of Menander's witty,
lifelike and graceful diction. Yet we would venture to match with it a Latin Scene,
where a woebegone lover is bantered, the opening Scene of the Pseudolus.
Plautus' vein is not Menander's but is as rich in entertainment.
 EARLY LATIN METRES 271

found on every other page, e. g. Trin. 88 Sed istuc negoti cupio scire quid siet ;
Bacch. 27 Nam credo cuivis excantar*? cor potes ; Asin. 768 Vocet convivam
neminem Wla, Hi voces.
(2) when a quadrisyllable with its first three syllables short precedes. The
strong examples in Plautus are :
Bacch. 220 Nam istoc fortasse aurost opus. Philip^ quidem ;
Cure. 86 Quisnam istic fluviust quern non rzexfiiat mare ? ;
Men. 550 Iamne introabiit? abiit, operas/ fores ;
? Most. 57 (incomplete) Stimulis ... si hue rtvtniat senex ;
Poen. pr. 27 Si minu' curassint, quom eri revem'ant domum ;
Rud. 1247 Ne conscii sint ipsi male/fc/ suis;
To which we may add these with quadrisyllable word-groups (or the like ; on
Mil. 1 104, etc., see below) :
Aul. 531 Spes prorogatur militi in sJium diem (cf. Poen. 500) ;
Cist. 148 Meminisse ego hanc rem vos volo. ego abeo domum (Or ego,
with Anapaest as fifth foot? Cf. Pers. 733 ; True. 365) ;
Pseud. 524 Priu' quam istam pugnam pugnabo, ego etzam prius ;
810 Non ego item cenam condio ut a/2'2 coqui (Or uti ; cf. Cist.
Rud. pr. 77 Ad villam illius, exul ubi habitat senex ; 596);
True. 49 Si raras noctes ducit, ab animo perit ;
Cure. 66 Qui quod lenoni nulli est id ab eo petas.
Also the examples mentioned elsewhere (II 46) of the pyrrhic caesura of. a
Tribrach in the fourthJJQQt, e.g.:
Aul. 378 Ita lllis impuris omnibus adzY manum.
A fuller treatment of this second exception (with examples also from
Terence and from the corresponding part of Iambic Octonarii and Trochaic
Septenarii) is provided by Luchs (Studemund's Studien, vol. I, pp. 1 sqq.),
who says happily of this abnormal Iambus : * numerorum varietate rhythmique
celeritate quasi furtim se insinuat.' Since it is abnormal, we prefer to scan
otherwise :
Pers. 733 Redin tu tandem ? Redeo. Ne" ego hodie tibi ;
Aul. pr. 20 Item a me contra factumst : nam item obiit diem ;
Most. 523 Quid faciam ? Cave respexis, fuge, operi caput [fu. atque op.
MSS., perhaps rightly) ;
583 Immo abi domum, verum hercle died, abi modo.
(3) with the phrase malam crucem (apparently a word-group), e. g. :
Cas. 611 Ducas easque in maximam 7nalamcx\icem ;
Pers. 352 Ferant eantque in maximam malam crucem ;
Rud. 176 Sed dextrovorsum avorsa It in malam crucem.
Do not these exceptions point to a regard for accent as the reason 1 for the

1 If the examples of the quadrisyllable were more frequent, they would furnish
a stronger argument for those who deny that a phonetic Brevis Brevians Law can
produce a shortening like balineis and not merely one like balneis (II. 27[5]). For
 272 EARLY LATIN METRES

rule? If so, editors should be guided by this consideration in accepting or

emending the traditional reading in these lines :
Amph. pr. 46 Sed mos numquam (ille) illi fuit patri meo ;
Asin. 64 Omnes parentes, Libane, \\beris suis ;
Men. 480 Ait hanc dedisse me sibi atque earn meae (Uxori surripuisse) ;
750 Negas novisse me ? negas fiatrem meum ? ; Mil. 39) ;
Merc. 331 Huic persuadere quo mo&o fiotis siem ;
Mil. 1 104 Qui tu scis eas adesse? Quia octilis meis (cf. Amph. 914;

Most. 670 Tuus emit aedes films. Bonan fide ? ;

Poen. 447 Ibo atque arcessam testes, quando Amor iubet ;
Pseud. 877 Si credi', nummo ; si non, ne mina quidem ;
Rud. 884 Sicin me spernis ? Sic ago semeiyyfoo (?) ;
Trin. 533 Neque umquam quisquamst quoius ille ager fuit.
(In Pseud. 800 si eras, Anapaest.) And they should be less rigorous with the
Iambics of Cantica, which were not so closely adapted to talk (e. g. in Septena-
rii : Asin. 419 Qui latera conteram tua||; 654. E.g. in Octonarii: Amph.
1058 Animo malest. aquam velim||. E. g. in Versus Reiziani : Aul. 417 Quia
cultrum habes. Coquum decet|| ; 432). (In Amph. 991 me, eum sequor,
with Hiatus at pause.) What other reason can there be for the rule ? Not
the avoidance of monotony ; for Plautus has no hesitation in ending with two
Iambi in other circumstances ; e. g. the prologue of the Amphitruo begins :
Ut vos in vostris volti' merdmonits (Or voltis)
Emundis vendundisque me laetum lucris
Adficere atque adiuvare in rebus omnibus.
When a line ends in an Iambic word the preceding foot is usually a Spondee
» ft. / (the favourite fifth foot in all Latin Senarii), but the Anapaest is not un-
common ; rather less common the Dactyl, rarest of all the Proceleusmatic,
e.g.:
Trin. 576 Di fortunabunt vestra consili{a). Ita volo.
and the Tribrach, e. g. :
Most. 40 Germana illuvies rustica, hirers', hara suis (rusticus MSS.).
A Senarius of Ennius1 seems to defy the law (trag. 197 = Eurip. Iph. Aul.
446):
Plebes in hoc regi antistat loco : licet (Lacrimare plebi).

t while a conclusion like hie balneis opust is absolutely unknown, we cannot quite say
the same of hie balinets opust.
1 Marx' attempt to explain the Roman tolerance of an ending like negotium (he
says nothing of excantare corpotes), but not of one like loco licet, is accompanied by
so suggestive words on Livius Andronicus that we quote the whole passage (Zwei
Auslautsgesetze, p. 196) : Der Ordner der lateinischen dramatischen Poesie, der
Landsmann des beruhmten Aristoxenos, war Tragiker und Komodiendichter zu-
gleich : darum ist die Schranke, die im Griechischen die Metrik der Tragodie und
der Komodie trennt, im Lateinischen gefallen, die Gesetze der Caesur werden
strenger beachtet, als im Griechischen, die correptio Attica ist ausnahmslos zur
Durchfuhrung gekommen. Aus der Technik des Hexameterschlusses war dem
 EARLY LATIN METRES

(B) Cretic Word at End of Senarius. Porson's Law of the Cretir hnlrU,
of course, for the Greek Tragic Trimeter only, not for the Greek Comic
Trimeter. In the Latin Tragic Senarius the Metricians of the Empire >,.(V
declared the fifth foot to be in all circumstances (whether with Cretic line-
ending or not) normally a Spondee (e. g. Diomedes, whojnayjie referring rp
Seneca, etc., in Gram. Lat. iv, 507: Iambus tragicus. hie, ut gravior iuxta
materiae pondus esset, semper quinto loco spondeum1 recipit), though we find
in the fragments an example now and then like Naev. 61 :
Ouam numquam vobis Grail, ntque darb&ri ;
more often like Naev. 2 :
Omnes formidant homines ei(u)s valentiam.
White (Verse of Greek Comedy, pp. 38, 58) finds in the fifth foot of the Tri-
meter of Aristophanes 5,162 Spondees to 3,090 Iambi, and of Menander 376
Spondees to 279 Iambi. The same ratio seems to hold when the Trimeter
ends with a Cretic.
In the fifth foot of the Senarius the Spondee claims a far larger share than
with Menander. Take, for example, the prologue to the Amphitruo. With
quadrisyllable (or longer) ending the Iambus gets a chance, e.g. line I
mercwiomis, line 24 precario. With cretic ending there are only four Iambi
against twenty-five Spondees. With iambic ending (as the last paragraph
showed) only one Iambus (line 46) can be produced. Therefore we should
give the Spondee 2 the benefit of the doubt in lines like :
56 Utrum sit an non vultis ? sed ego stultior (not 'eg5 st.') ;
no Nunc de Alcumena ut rem teneatis rectius (not ' teneati' rect.') ;
137 Quo pacto donis sit donatus plurimis (not 'donatu' pi.').
9. Senarii of Cantica, Tragic Senarii, etc. The symbols used at the head
of Scenes in the ancient texts, DV (i.e. deverbium) at the head of dialogue-
scenes, C (i. e. canticum) before songs or before animated conversation- scenes
(in Iambic or Trochaic Septenarii and Octonarii), have been mostly lost in
our extant MSS. Still we have indication that two Scenes in Senarii were
marked C, Pers. IV vi (Probae hie argenti sunt sexaginta minae, etc.) and
Trin. IV iv (Quid hoc hie clamoris audio ante aedes meas, etc.). These
traces are too slight to reveal to us any metrical or other divergence of such
Griechen die Bedeutung von Wortverbindung und Worttrennung wohlvertraut,
ein Hexameterschluss 'iirnov Kp-qvi) als unertraglich verpont, als fehlerlos aber und
wohllautend bekannt in der Form 'IirirovKprjVTj. Dieselben Gesetze iiber die Milderung
metrischer Harten durch die Weichheit der Wortverbindung finden wir im Lateini-
schen : ein Senarschluss negat virum war, aus Griinden, die uns noch unbekannt
sind, ebenso verpont wie ein Hexameterschluss altas urbes, aber der Schluss
negotium dort ebenso statthaft wie hier der Schluss innuptarum.
1 i. e. rather than an Iambus. Of course an Anapaest in the fifth foot is common
in Seneca, as well as in the Republican Drama.
2 Hardie 'Res Metrica', p. 112: 'The preference of the Roman drama for
spondees (which becomes a rigid rule for the fifth foot in Seneca) only carried
further a tendency which had shown itself in the drama of Athens.'
2348 X
274 EARLY LATIN METRES

Canticum-Senarii from the Deverbium-Senarii. AP)t

Nor do the few isolated
Senarii interspersed in other Cantica avail, e. g. :
Bacch. 669 Ouid vos maestos tarn tristesque esse conspicor (mae. tarn

Cas. 176 Domi et foris aegre quod siet sati' semper est ;
Epid. 24 Percontari. operam da ; opera reddetur tibi ;
177 Quia licitumst earn tibi vivendo vincere ;
Men. 356 Amanti amoenitas malost, nobis lucrost (Lack of Caesura) ;
? 972 Recordetur id, qui nihili sunt, quid eis preti ;
? Most. 899-900 Heus ! ecquis hie est maximam qui his iniuriam (AP.
Emended to his qui in.)
Foribus defendat ? ecquis has aperit fores ? j
?Poen. 1 196 Quid est, fratris mei gnate [gnate], quid vis? expedi ;
Stich. 300 Secundas fortunas decent superbiae (Lack of Caesura ; clash
of accent and ictus in third-foot Spondee).
Indeed the Senarii of Roman Tragedy admit (unlike the Greek Tragic
Trimeters) the Anajmf^t to the second and fourth feet ; and theSprmHpp tog.
They differ from Comic Senarii in their smaller use of resolved feet, in their
elevated diction, in their great (but not complete) avoidance of conversational
prosody (e. g. Breves Breviantes ; see II 26). When a tragic line is cited by
a comedian, it is the different tone of the language which most arrests us :
e. g. in the Poenulus prologue :
Exsurge, praeco, fac populo andientiam,
and
Sileteque et tacete atque animum advortite,
lines from Ennuis' adaptation of the Achilles of the tragedian Aristarchus of
Tegea.
Iambic Septenarius.
10. (A) The admission by the Senarius of Spondees into the second
and fourth feet makes it patently different from the Trimeter. The
Roman Metricians of the Empire are probably right in claiming that
" the divergence was deliberate and not the result of incompetence :
In metra peccant arte, non inscitia (Terentianus Maurus 2237).
Plautus presumably wished to bring the Roman dialogue-metre closer to
actual Roman conversation, Now the Septenarius and the Octonarius
are rather Canticum-metres than Deverbium-metres ; there is no indica-
tion in our extant MSS. that any Scenes written in these long iambic
lines had any symbol but C (i. e. canticum). There would not therefore
be quite the same necessity for departure from the Greek type.
And this a priori inference corresponds to facts. Unluckily we have
not Iambic Tetrameters Catalectic of Menander to compare with the
Iambic Septenarii of Plautus. The self-restrained Menander apparently
eschewed this metre of unrestrained rollicking (est enim iocosis motibus
 EARLY LATIN METRES 275

emollitum gestibusque agentium satis accommodatum. Mar. Victorinus,

Gram. Lat. vi, 135, 25) ; at any rate there is no trace of it in his
fragmentary MSS. We have to fall back upon the Old Greek Comedy
for our comparison.
White's account of Aristophanes' Iambic Tetrameters Catalectic
(Verse of Greek Comedy, p. 65) shows that one in every four lines has
no Diaeresis (at the end of the first hemistich), e. g. Knights 337 :
'Kav Sc fir) ravrr} y' v7rei\\Kr), Xey' otl kolk 7rovr)piov,
and many more, although the hemistich ends at the end of a word, have
no real Diaeresis, e. g. Knights 363 :
'Eyo> 8' i7r€Knrrj8u)V ye tt)v \\/3ov\r)v yStia KVKrjcro),
where the Definite Article cannot be separated from its Noun. In this
respect Plautus' Iambic Septenarii compare favourably. We have |
already mentioned (II 59) one great dissimilarity between the Roman I
and the Greek type, Plautns' free use of Hiatus and Syllaba Anceps at I
the Diaeresis, e.g. Mil. 1226 (with both) :
Namque edepol vix fult copia ||adeundi atque impetrandi,
and the three rival explanations: (1) the influence of Saturnian Metre
(I 9) ; (2) the strong pause at the end of the hemistich; (3) the treat-
ment of the long line as if it were two short lines. In weighing n°. 2
against n°. 3, we have given these details (II 59) : the fourth foot of an
Iambic Septenarius is normally an Iambus ; the sure examples of
absence of Diaeresis are less than one in twenty, and in many of these the
absence is hardly perceptible. The consequence is that the Septenarii
of Plautus have an unretarded, lively movement that makes them very
pleasant reading. We can take no* better specimen than the dance-
accompaniment quoted in the second paragraph of this chapter (Stich.
771):
Fac tu hoc modo. At tu hoc modo. || Babae ! Tatae ! Papae !
Pax!
It is true that there is another great dissimilarity between Plautus and
Aristophanes, that the Roman poet allows the seventh foot to be a
Spondee, e. g. Pers. 286 :
Nam ego me confido liberum ||fore, tu te numguam sfleras ;
a Dactyl, e.g. Pers. 281 :
Dicisne mi ubi sit Toxilus? || Dico ut pevpetuo flereas,
sometimes an Anapaest, e. g. Rud. 409 :
Timidas, egentes, uvidas, ||eiectas, exanimatas,
hardly ever (perhaps never) a Proceleusmatic (see below).
T 2
276 EARLY LATIN METRES

But in practice it will be found to be usually an Iambus (occasionally

the other * pure ' foot, a Tribrach). It must be an Iambus (or Tribrach,
Cure. 520; Pers. 316) if the line ends in a (single) monosyllable, e.g.
Asin. 713:
Atque ut deo mi hie immolas ||bovem ; nam ego tibi Salus sum
(Details in Appendix D).
And of the 1,300 Septenarii of Plautus less than 300 show a Spondee
(we exclude endings like molestu' ne sis ; III 16). Indeed about a
score of these may be challenged ; e. g. Cist. 715 supposita est parva of
the minuscule MSS. may be really supposita parvast (II 38) ; and True.
177 fundi atque aedes has been corrected to fundi et aedes (as in lines
174 and 186); agitabo adversu' divum (Asin. 708) is more natural
(III 41) than adversum of the MSS. n Cure. 519 and 525 see III 29
(s. v. pro).
White states (p. 65) that eight of Aristophanes' lines (out of 362)
have an Anapaest in the fourth foot, Diaeresis being neglected in two
but maintained in six, e.g. Clouds 1427 :
%kQhxi Se rot's aXtKTpvovas \\kcu raAA.a ra (3ora ravri.
Here too Plautus compares favourably (see the statistics in Appendix B).
Until Terence is edited it is rash to estimate his departure from
Plautine usage. The statistics given by Meissner (see Appendix E)
must be used with caution.
A Metrician of the Empire tells us of the great part played by this
metre in the Atellanae (Mar. Victorinus, Gram. Lat. vi, 135, 25 : frequens
hoc comicis metrum, et praecipue antiquae comoediae scriptoribus,
nostri quoque — Plautus et Caecilius et Turpilius — non aspernati sunt,
Atellanae autem scriptores appetiverunt).
(B) Proceleusmatic in Seventh Foot. There is no clear proof of this in
Plautus :
n 52) ;
Asin. 430 || erus in hara, haud aedibus habitat (False Proceleusmatic ;

482 || de nobis detur ? Atque etiam (see II 52 ; 32 B) ; 1307);

Cist. 371 || damno sunt (tui) mihi (simil)es (Emended to mi; cf. Rud.

Pers. 290 || male loquere? Tandem uti liceat (Emended to ut; but see
III 42, s.v. ut).
(Therefore in emending the corrupt lines Most. 174 and Pseud. 153, the
endings in the MSS., hodie aliqui and genera hominum, should not be left as
a Proceleusmatic followed by a syllable.)
The MSS. of Terence offer :
Heaut. 738 || argentum. Ouin ego hie maneo (hie om. A
(In Heaut. 1002 ni[hi]l habeo.)
 EARLY LATIN METRES 277

11. Iambic Octonarius. We have already (II 59) cited a Scene of

the Asinaria (830 sqq. : Numquidnamst tibi molestumst, gna||te mi, si
haec nunc mecum accubat, etc.) for one type of Plautus' Octonarii, in
which the metre is adapted to conversation by disregard of the break in
the middle of the line. For the other type, much more of a jingling
song, the slave's Canticum at the beginning of the Persa was cited
(9 scL9i- Quae ero placere censeat ||praesenti atque absenti suo ; Ego
neque libenter servio ||neque sati' sum ero ex sententia, etc.). Plautus
has only some four hundred Octonarii and often makes sparing use of
them, a line here, a couple of lines there, to break the monotony or
increase the variety of a Canticum. But in the opening Scene of the
Amphitruo they are a good vehicle for the maundering soliloquy of the
bibulous serving-man, with their suggestion of interminable, rambling
remarks (Amph, 153 sqq.) :
Qui me alter est audacior || homo aut qui confidentior, ii.) )
Iuventuti' mores qui sciam, ||qui hoc nocti' solus ambulem ?, etc.
And where whole passages are found in this metre in Plautus, they ar
usually soliloquies. In fact we might call the Iambic Octonariusus a
1 soliloquy ' metre of Plautus. (So print Men. 596 sqq. as Octonarii
This metre (found in Sophocles' Trackers) is not found in (whatrd
remains to us of) Greek Comedy, though we Can find it there if ywea I
imagine two Dimeters Acatalectic to be combined into one long lineJ
Indeed, here and there, Octonarii of the ancient text (or texts) of Plautus
seem to us nowadays better expressed as Dimeters. Thus A at Cist.
45 1 sq. writes as two lines what might also be put thus :
Al. Germana mea sororcula.
Sel. Repudio te fraterculum.
Al. Turn tu igitur, mea matercula.
Me. Repudio te puerculum.
Terence makes great use of Plautus' ' conversational ' type of Iambic
Octonarius, which dispenses with the mid-line Diaeresis and does not
require an Iambus in the fourth foot. He gives it the same Caesura as
the Senarius, viz. penthemimeral or hepthemimeral, e. g. Andr. 209, 211
(Davus' soliloquy) :
Nee quid agam certumst : Pamphilum||ne adiutem an auscultern
seni.
Cui verba dare dimcilest : pri||mum iam de amore hoc comperit.
And (to judge from their fragments) the Tragedians made great use of
it too. Take for example Pacuvius' often-quoted line (trag. 197) :
Mater, te appello, tu quae cuj[ram somno suspensam levas.
278 EARLY LATIN METRES

12. Iambic Dimeter, etc. The Greek trick of putting a Choriambus

among iambic feet is not unknown in Plautus' Cantica, e. g. Trin. 247 sqq.
(some editors make three lines) :
Ibi ilia pendentem ferit : || i(am) amplius oxzX ; non satis
Id est mali, || ni etiam amplius.
And the Greek use of Iambic Dimeters reappears in Plautus. They
play the part of a Clausula to mark the end of a passage, e. g. Men. 1006
(after Iambic Octonarii), Amph. 1073, Men. 967. Also (if the 'bull'
be permitted) to mark the beginning, e.g. Pseud. 187 (before Iambic
Octonarii), Men. 120. Sometimes they convey sallies of banter, e.g.
Epid. 27 sqq. :
Th. At unum a prae||tura tua,
Epidice, abest. ||Ep. Quidnam ? Th. Scies :
Lictores duo, ||duo ulmei
Fasces virga||rum. Ep. Vae tibi !
The combination of Anapaest with Iambus catches the sprightly tone
of the old gallant in the Casina, 708 sqq. (the last Monometer is
catalectic) :
Ly. SI effexis hoc,
Soleas tibi
Dabo et anulum in
Digito aureum et
Bona plurima.
Pa. Operam dabo.
Ly. Face ut impetres.
Pa. E6 nunciam,
Nisi quidpiam
Remorare me.
Ly. Abi et cura.
In both ancient editions the first five lines seem to have been presented
as a \ run ', as a long line which, being too long for the page, was broken
after dabo et.
13. Iambic System. Elsewhere we have clear proof of an Iambic
'run' ('system' is the technical word), e.g. Pers. 40, Stich. 291 ; for
one portion is so dovetailed into the next that it is impossible to make
a break till the end is reached. Similarly e. g. at Amph. 1067, what
editors print ' hypermetrically ' (to prevent the awkward prefixing of the
syllable -bant to the second line) :
Ut iacui, exsurgo. ardere cen||sui aedes, ita turn confulge||bant.
Ibi me inclamat Alcume||na ; iam ea res me horrore afficit.
(See above, 3.)
 EARLY LATIN METRES

14. Versus Reizianus. The 'conversational' type of Iambic

Octonarius, if we may so term the metre in which the dinner-table con-
versation inthe Asinaria (830 sqq.) is carried on, must have been
a Roman invention. ' It is impossible that a Greek should have
tolerated a Spondee in the fourth foot. That is a Latin alteration, like
the admission of Spondees into the even feet of the Trimeter, and
designed to bring the metre farther from song and nearer to talk.
The Versus Reizianus, however closely associated with Plautus, can
hardly be referred to Plautus' creative brain.2 It is true that in the Old
Greek Comedy the Colon Reizianum (see below) goes hand-in-hand with
Glyconic (in stricter parlance Dactylo-trochaic) metre, an affinity less
prominent in Plautus (but cf. e. g. Stich. 3% and perhaps Cas. 959 sqq.,
Epid. 545-6, Pseud. 603). Still we find an isolated Versus Reizianus
even in Aristophanes here and there, e. g. Ach. 840 ; Pax 954 :•
*H <rvKo<f>dvrr)<; a\\os oi||/ao>£wi' Ka^eSetrcu,
^vowti /«u 7rovovfjL€Vio j| 7T/oocr8dj(rcTe hrjirov.

And its ' stichic ' use, i. e. its use in whole passages (as in the Aulularia
Scene already cited, I 9), may have originated on Greek soil. The line
is a blend of an Iambic Dimeter Acatalectic with the Colon Reizianum.

15. The Colon Reizianum. The Glyconic metre is familiar to

every reader of Horace (e. g. Sic te diva potens Cypri, and Cui flavam
religas comam ?) and its catalectic form, the Pherecratean. A Phere-
cratean lacking the first syllable (in technical language an ' acephalous '
Pherecratean) will have these quantities and ictus-beats (we indicate
ictus by the accent-symbol) — — w ^ - -. From the scholar Reiz, who
first revealed the part played by this colon in Plautus and in the Greek
Drama, this acephalous Pherecratean is called the 'Colon Reizianum'.
This colon is blended by Plautus with Iambics usually] A favourite
combination, the Versus Reizianus, in which it follows an Iambic
Dimeter Acatalectic, has just been mentioned. Next to Iambics, we
|;find Anapaestics more often combined with it than any other metre.
And it often appears in independent form, to mark the end of a passage
or to diversify a Canticum. Indeed in the Casina (752 sqq.) the ancient
text (or texts) separated this ' limb ' from the Iambic ' body ' and broke

1 It is earlier than Plautus, e. g. Naevius com. 10, 35 :

Potioremque habui liberta||tem, multo quam pecuniam ;
Ubi vidi, exanimabiliter timi||du' pedibu' protinam me dedi.
(The second line shows the same use of the metre as in Plaut. Capt. IV iv.)
2 Is the solitary fragment of Naevius' Equus Troianus a Versus Reizianus ?
Numquam hodie effugies quin manu || mea moriare.
28o EARLY LATIN METRES

up the Versus Reizianus into two lines, apparently because independent

cola of this type follow :
Gladium Casinam intus habere ait (Iamb. Dim. Acat.)
Qui me ac te interemat. (Col. Reiz.)
Scio. sic sine habere ; (ditto)
Nugas agunt : novi (ditto)
Ego lllas malas merces. (ditto)
Quin tu i modo mecum (ditto)
Domum. [At] Pol malum metuo. (ditto ; II 32 B) ;
I tu modo, perspicito prius (Iamb. Dim. Acat.)
Quid intus agatur. (Col. Reiz.)
Tarn mihi mea vita (ditto)
Tua quam tibi carast. (ditto)
Verum i modo. Si tu iubes, (Iamb. Dim. Acat.)
Em ! ibitur tecum. (Col. Reiz.)
Even this single passage shows how Plautus departed from the normal
form of this colon. In his hands it is as Protean a metre as the Dochmius
of Greek Tragedy. This Casina passage exhibits : (753) - -i - 4 w - ;
(754) Jo*CV-£-; (755) --iww^-; (756) ww^ww-i-; (757)
— - \j w — — • (758) ww— wwow—; (760) w— w w -i — ; (761) — ww
w w -L — j (762) w \j — w w - - ; (764) ^-uu-i-, The endings of the
Versus Reiziani in the Aulularia Scene (415 sqq.) are: (415) quid,
stolide, clamas?; (416) nomen tutim. Quam 6b rem?; (417) quid
comminatu's ? ; (418) quia non latu' fodi ; (419) qui vivat hodie; (420)
male plus libens faxim ; (421) res ipsa testist; (422) magi' quam ullu'
cinaedus ; (423) mendice homo. Quae res? ; (424) quam aequum erat
feci j (425) tuo, si hoc caput sentit ; (426) tuum nunc caput sentit ;
(427) nam erat negoti; (428) volo scire. Tace ergo; (429) quid tu,
malum, curas ? ; (430) nisi tu mi es tutor ; (431) nos coquere hie cenam ;
(432) mean salva futura ; (433) quae ad(te)tuli salva; (434) scio, ne
doce, novi ; (435) nos coquere hie cenam ; (436) tibi secu' quam velles ;
(437) qui angulos in omnes ; (438) mihi pervium facitis ; (439) focum
si adesses; (440) merito id tibi factumst ; (441) iam noscere possis ;
(442) nisi iussero, propius ; (443) mortalis uti (ut ?) sis ; (444) quo abis ?
red! rursum ; (445) (iam) iam, nisi reddi ; (446) hie differamante aedes.
Thus the prevalent forms are--uui- and m w — w w J. — , i. e. the
normal type and the same with thenrst syllable resolved.
One form (^- ^ --), e. g. Trin. 255 (Vers. Reiz.) :
Fit ipse, dum illis comis est, 11 inops amator,
is exactly like a curtailed Bacchiac Dimeter (with Iambus or Spondee
 EARLY LATIN METRES 281

instead of the first Bacchius j 24). When this colon appears (as it
often does) in Bacchiac context, it is difficult (perhaps unnecessary) to
decide how Plautus meant us to take it.

16. TROCHAIC. We have already seen (II 55) that the Roman
regard for Accent makes it impossible to throw Trochaic and Iambic
material into a mass for investigation and to infer from a usage in one
metre the same usage for the other. Thus a tribrach-word can make
a foot of trochaic metre, but not of iambic (II 58) ; the pyrrhic caesura
of a Tribrach (e. g. ' Agit mm ') is avoided in iambic, but not in trochaic
(n 55).
For all that, the same rule usually holds in both, e. g. for the division
between words of an Anapaest (II 55), a Dactyl (II 56), a Proceleus-
matic (II 57). The use of a dactyl-word as a foot (II 58) is as rare in
trochaic as in iambic verse (but perhaps from a different reason), e. g. :
Capt. 558 Hegio : fit quod tibi ego dixi ||.
When an Iambic Senarius ends with an iambus-word, the fifth foot
must not be an Iambus (8). Similarly when a Trochaic Septenarius
ends with an iambus-word, the sixth font must nnt hp p. TrQc{iee. There
are the same two exceptions :

(A) When a monosyllable precedes the iambus-word, e. g. (every

other page will furnish examples) :
Men. 1027 Nee meu' servu' numquam fecit ||tale quale tu mihi ;
Stich. 708 || tibi tute inde, si sapis ;
(B) With the word-group malam crucem :
Cas. 977 ||Casina sum. Tn malam crucem ? ;
Cure. 611 ||quin tu is in malam crucem? (cf. Men. 915);
693 ||(ab)duce istum in malam crucem ;
Poen. 347 Bellula hercle ! i dierecte in || maximam malam crucem ;
511 ||ite hinc in malam crucem.
Mere dislike of monotony cannot be the cause for this usage of
Trochaic Septenarii any more than for the similar usage of Senarii ; for
a precipitate of Trochees is common enough, e. g. :
Capt. 942 || impetrabis atque te.
Other lines where our MSS. offer this irregularity are, e. g. :

Cas. 395 || perdi' me. Lucrum facit ;

Cure. 477 Confidentes garrulique et || malefici supra lacum;
mihist ;
Pseud. 700 Novu' mihist. Nimium est mortali' ||graphicus, evpcTijs
282 EARLY LATIN METRES

Rud. 775 || maximo malo suo ;

True. 586 || egone? Tu. Bona fide? (Or bonan?).
They may be judged by the same criterion as the exceptions to the rule
in Senarii (8).
With the help of P we emend the reading of A in Mil. 204 : Dextera
digitis rationem || computat ferie(n)s femur (ferit A). With the help of
A the reading of P in Pseud. 702 : || quoia vox sonat? Io (resonat A).
In True. 885 some find hiatus at the pause ( ||ubi amicl, ibidem opes),
but it seems better to make the sixth foot a Trochee ( || ubi amici,
ibidem opes). The metre of Pseud. 598 is doubtful. In Poen. 290
scan se amet potest ; III 46. (On Donatus' ignorance of this rule see
above, 3.) A fuller treatment of the subject will be found in Studemund's
Studien, vol. I, pp. 13 sqq, where Luchs provides also a list of the lines
with a Tribrach in the sixth foot, the trochaic equivalents of Senarius-
endings like recipiat mare (8 A [2]).
The Spondee is admitted against Greek usage into the second and
fourth feet of the Senarius, but clash of ictus and accent is avoided.
We may look for the same avoidance in the Spondees admitted against
Greek usage into the first, third and fifth feet of Septenarii. But the
unison of ictus and accent in Plautus' trochaic verse is so prevalent, that
any effort in these particular feet passes unnoticed.1 That there should
be the same regard for Accent in these Trochaic lines is natural, since
they are the vehicle for excited talk, as the Senarii for quiet talk.
Notice the rhythm of Men. 10 15, where each denunciation is accom-
panied bya blow :
Vos scelesti, vos rapaces, || vos praedones. Periimus !
The concluding portion of a play is usually thrown into this lively
metre. The audience would be duly roused when the final appeal came
for applause (2).
It is as hard to find two consecutive spondee-words with clash of ictus
and accent in Trochaics (Cure. 626; Rud. 1423?) as in Senarii (II 7).
In Most. 171 omnes mores is a bad conjecture ; rather res omnes (o. res
MSS.).

17. Trochaic Septenarius. The ancient criticism of Terence's

plays, that they ought to have kept to the Senarius (plus adhuc habitura
gratiaesi intra versus trimetros stetissent, Quintilian 10, i, 99), will not be
applied to Plautus by English readers. If Roman Iambic Septenarii
1 The clash in Bacch. 445 (Ne attingas puerum) is a scribe's error. Nonius cites
the true form Ne attigas pu. (Cf. True. 276.) In Capt. 851 horaeum (III 25). In
Epid. 243 (after Hiatus) || evenlt illi, obsecro. But, e.g. Men. 1159 Fenibunt quiqui
licebunt ||praesenti pecunia.
 EARLY LATIN METRES 283

make pleasant reading (10), his Trochaic Septenarii give us as much

pleasure as English accentual lines, e. g. :
Many a night from out my window-casement ere I went to rest,
a ' Trochaic ' line, by the way, which lacks Diaeresis. For, as we have
seen (II 13), ictus and accent fall into unison in Latin Trochaics with no
perceptible effort, e. g. Men. 859 ; Merc. 833 :
Osse fini dedolabo || assulatim viscera ;
Interemptust, interfectust, || alienatust, occidi ! ;
and English readers find themselves on more familiar ground than with
Virgil. There is no make-believe about it. They feel that they get
much the same sensation as Roman readers would get.
The normal scene-heading for Trochaic Septenarii was C (i. e. Canti-
cum). From the miserable remnant of these symbols in our MSS. we
learn that two (possibly three) Scenes in this metre were marked D V
(i. e. deverbium) : Capt. Ill i (Miser homost qui ipse sibi quod edit ||
quaerit et id aegre invenit, etc.), Cas. IV iii (Age, tibicen, dum illam
educunt || hue novam nuptam foras, etc.) and possibly Epid. I ii (Rem
tibi sum elocutus omnem, || Chaeribule, atque admodum, etc.). But
though we might select from these Scenes some lines which have an
extraordinary number of Spondees, e. g. Capt. 484, 486 ; Cas. 810 :
Nemo ridet ; scivi extemplo || rem de compecto geri,
Saltern, si non adriderent, || dentes ut restringerent,
Illo morbo quo dirumpi || cupio, non est copiae,
the material is too limited to ensure us against self-deception. As a rule
Plautus' lines (though editors spoil them by not dropping -s after a short
vowel ; III 17 C) are not so far removed from the Greek type.1 And the
seventh foot must be (as in Greek) a Trochee (or the other ' pure ' foot
occasionally, a Tribrach).
White (Verse of Greek Comedy, p. 101) finds that Aristophanes
neglects Diaeresis once in every seven lines, while in many others there
is no real Diaeresis. We have seen (II 59) that Plautus' usage in this
respect is stricter, for he (in a selected portion) neglects it once in every
ten or a dozen lines, but that the Roman tolerance of Hiatus at the
Diaeresis (once in about every twenty lines) is a new feature (On Ter-
ence's practice see II 48.) Menander's Trochaic Tetrameters Catalectic
are too few for a satisfactory comparison with Plautus' Septenarii. White
(ibid. p. 104), comparing Menander with Aristophanes, finds in the
1 The Greek tag from the original goes well with the Latin portion in Pseud. 712 ;
Quis istic est ? Charinus. Eugae ! || iam \apw tovto) -now. (Cf. Stich. 707 ; Asin. 866 ;
Merc. 646-7.)
284 EARLY LATIN METRES

younger poet, the Comedian of everyday life, more resolved feet (Ana-
paest, and perhaps Dactyl). We have already remarked that Plautus
went still further in this direction, for, besides the Anapaest, ■ he
welcomes the Dactyl and does not wholly forbid the Proceleusmatic
(II 57).
White finds a Dactyl six times in Aristophanes' dialogue and more
often (three times) in the fourth foot than in any other (e.g. Eccles. 1156
Tots yeA.wo-i S' fjScus Sid ||). Plautus does not show this favour to a
fourth-foot Dactyl. Still examples are not hard to find. E. g. in the
Miles Gloriosus :
292 Lippu' videor ? Medicum istuc tibi || meliust percontarier j
461 Cum Philocomasio oscu/antem, eum *||go obtruncabo extem-
pulo (Or || ego 6b tr.) ;
604 Quippe qui, si xesoXvere *W||mici consilium tuum ;
648 Post Ephesi sum natu', non enim in || Apulis ; Animula non sum ;
745 Serviendae servituti ego || servos instruxi mihi ;
1 139 Quid agi', noster archi/o/i? Egort || architectus ? vah !
Quid est ? ;
1364 Cogitato identidfew tibi || quam fidelis fuerim.
(cf. 248?; 430?; 628?; 660?; 764?; 798?; 1009?). (On a dactyl-
word in this foot see II 58.)
The Trochaic Septenarius was a popular metre at Rome (with soldiers
at triumphs, etc. : Ecce Caesar nunc triumphat qui subegit Gallias).
The Roman Comedians seem to have made far more use of it than
their Greek models and were declared to have (unconsciously) harked
back to the Old Comedy trail (Marius Victorinus, Gramm. Lat. vi, 78,
20 : scio plurimos afiirmare Terentianas vel maxime fabulas metrum
ac disciplinam Graecarum comoediarum non custodisse, id est quas
Menander, Philemon, Diphilus et ceteri ediderunt. Nostri enim in
modulandis metris seu rhythmis veteris comoediae scriptores sequi
maluerunt, id est Eupolin, Cratinum, Aristophanem. Prologos itaque et
primarum scenarum actus trimetris comprehenderunt, deinde longis-
1 The crowd of Anapaests in Cure. 158 is, of course, designed for a special effect.
Presumably they are timed to the foot-fall of the crone :
Placide egredere et sonitum prohibe || forium et crepitum cardinum.
Rude haste is well expressed by the crowd of resolved feet and the neglect of
Diaeresis in Rud. 660 :
Proripite hominem pedibus hue iti||dem quasi occisam suem.
The prattle of the Hetaera is (as Klotz remarks) echoed in Bacch. 83-4 :
Ubi tu lepide voles esse tibi, || • mea rosa ' mihi dicito,
Data qui bene sit : ego ubi bene sit || tibi locum lepidum dabo.
 EARLY LATIN METRES 285

simos, id est tetrametros, subdiderunt, qui appellantur quadrati, etc.).

For instance, Terence Hec. 286 sqq. adapts in Troch. Septenarii a pas-
sage of Apollodorus Carystius in Iambic Trimeters.
18. Trochaic Octonarius. The Trochaic is the metre of excite-
ment :
1/
IIcuc, Trcue rov iravovpyov kcu Tapa^imroa-rpaTov (Knights 247).
When the tone of excitement is highest, when it rises to a shriek of
distraction or shout of exultation, Plautus uses the Octonarius, e.g.
Bacch. 612 sqq. (Mnesilochus tears his hair) :
Petulans, protervo, iracundo || animo, indomito, incogitato
Sine modo et modestia sum, || sine bono iure atque honore,
Incredibilis imposque animi, || inamabilis, inlepidu' vivo.
Bacch. 640 sqq. (Enter Chrysalus in triumph)
auro;:
Hunc hominem decet auro expendl, || huic decet statuam statui ex

Nam duplex facinus feci hodie, || duplicibu' spoliis sum affectus.

Erum maiorem meum ut ego hodie || lusi lepide, ut ludificatust !
Aul. 406 sqq. (The cook runs out shrieking) :
Attatae ! cives, populares, || incolae, accolae, advenae omnes, etc.
(In these three passages of 'raving' Octonarii, eleven lines in all,
there is no neglect of Diaeresis,1 with the doubtful exception of Aul.
410, and three lines show Hiatus at the Diaeresis.) Often however
merely as a thread to vary the texture of a Canticum. Often as an
introduction to a passage in Trochaic Septenarii.
As was said of Iambic Octonarii, the Trochaic Octonarii sometimes
appear in our ancient text (or texts) where we nowadays are inclined to
find Dimeters. Thus we might write Pers. 13-16 thus :
Tox. Quis illic est qui || contra me adstat ?
Sag. Quis hie est qui sic || contra me adstat ?
Tox. Similis est Saga||ristionis.
Sag. Toxilus hi(c)quidem || meus amicust.
Tox. Is est profecto. Sag. Eum || esse opinor (or || Eum esse)
Tox. Congrediar. Sag. Con||tra adgredibor.
Tox. O Sagaristi||o, dl ament te. Sag. O
Toxile, dabunt || di quae exoptes.
19. Trochaic Dimeter, etc. The use of the Trochaic Dimeter (as
of the Iambic ; 12) for Clausulae and the like may be illustrated from
3 Similarly in Caecilius' lines (119 ; 230) :
Nunc enim vero est cum meae morti || remedium reperibit nemo ;
Nunc enim demum mi animus ardet, || nunc meum cor cumulatur'ira.
286 EARLY LATIN METRES

Pseud. '202-202a. And where we find the catalectic form in the rapid
interchange of e. g. Epid. 3 sqq. :
Ep. Respice vero, || Thesprio. Th. Oh !
Epidicumne ego || conspicor ?
Ep. Sati' recte oculis || uteris.
Th. Salve. Ep. Di dent || quae velis,
we can be sure that we actually have Dimeters. (Cf. Most. S85-7.)
But the dividing-line between Dimeter and Octonarius is often obscure.
Compare Pers. 13-16 (just quoted) with these Dimeters Acatalectic of
our ancient texts, Pers. 29ft-3i :
Sag. Quid iam ? Tox. Quia eru' || peregrist. Sag. ain tu ?
Peregrist ? Tox. Si tu || tibi bene esse
Pote pati, veni : || vives mecum.
Basilico accipi||ere victu.
In fact it appears that these passages which modern or ancient editors
arrange differently in lines of different lengths were meant by Plautus
for ' runs ' (' systems '), arid that the only real break in the consecutive
passage is when catalexis appears or when there is a distinct break in the
sense. We are forced to recognize a ' run ' at e. g. Pers. 26 sqq., where
one 'line' is dovetailed into another. But we may recognize it also at
e. g. Bacch. 968 sqq. :
Cepi spolia, is || nunc ducentos
Nummos PhiHppos || militi, quos
Dare se promi||sit, dabit. nunc
Alteris eti||am ducentis
Usus est, qui || dispensentur
Ilio cap||to, ut sit mulsum
Qui triumphent || milites.
(The last ' line ' is a Dimeter Catalectic.) This is what we are accus-
tomed to in Aristophanes, e.g. Knights 621 sqq. The 'run' (printed
as Dimeters) at Amph. 575 sqq. makes good expression of Amphitruo's
excited recriminations and Sosia's excited replies.
20. The Ithyphallic Colon. White (Verse of Greek Comedy, p. 73)
finds this trochaic colon (normally — w - ^ ) in Trochaic, ' Glyconic '
(Dactylo-trochaic) and Enoplic Verse of Aristophanes and describes it
as a syncopated (l protracted') Trochaic Dimeter, e.g. Lysistr. 812 :
OvTOS OVV 6 Tl fXOiV.

He mentions as by-forms of the first half, — w ^ ^ w and w \j w w \^ w

and - w - ; of the second half, - ^ .
 EARLY LATIN METRES 287

As Plautine examples of the normal type we quote three lines of the

Casina :
824 Obsecro, memento;
830 Rem accipe hanc ab nobis ;
888b Savium me sivit (following Cretics).
The admission of a Tribrach is illustrated from Menander's Phasma
by a Roman Metrician (Caesius Bassus, Gramm. Lat. vi. 255, n nam
ithyphallicum metrum saepe recipit hunc tribrachum ut etiam apud
Menandrum in Phasmate et apud Callimachum in Epigrammatibus
ostendi potest). Here is a Plautine specimen (Capt. 207, Cretic
Dimeter with this colon) :
At fugam fingitis. || sentio quam r(em) agitis
(for quam rem agitis, if a First Paeon, would have the ictus abnormally
[II 9] on the second syllable of a tribrach-word ; and if a Choriambus
[25], would hardly be allowed to occupy the fourth foot). Does this
colon ever begin with a Spondee in Plautus (e.g. Cas. 818)? (On
Rud. 945 sq. see 42.)
21. Alternation of Trochaics and Iambics. While Plautus
luxuriates in a great variety of metres (numeri innumeri) in his Cantica,
Terence restricts himself almost wholly to Iambics and Trochaics and
gets his effect by rapid transition from the one to the other metre
(mutatis modis cantica). The alternation of the two is by no means
unknown in the Cantica of Plautus. What a scurry it makes in the
Captivi, when Tyndarus suddenly finds himself in desperate straits.
First come eight Iambic Octonarii (the { soliloquy ' metre ; 11) ; then
this change from an Iambic ' run ' to Trochaics (523 sqq.) :
Operta quae fuere aper||ta sunt, patent praestigiae, om-
ni' res palamst.
Neque de hac re ne||gotiumst
Quin male occidam oppetamque || pestem eri vicem — meamque.
Perdidit me Aristophontes || hie modo qui venit intro.
Is me novlt, is sodali' || Philocrati et cognatus est, etc.,
the Trochaic ' run ' being ushered in by a Catalectic Dimeter and
ending in catalexis. And we recognize a Trochaic Octonarius in the
lifelike description of the thunder-clap in Amph. 1062, especially since
Trochaic Septenarii follow in lines 1064-1065 (the Scene is written in
Iambic Octonarii) :
Strepitu', crepitu', sonitu', tonitrus ; || ut subito, ut prope, tit valide
tonuit ! (with special emphasis on valide ?).
288 EARLY LATIN METRES

The many resolutions, which make anapaestic scansion possible, give a

good imitation of the thunder, and the sudden change of metre strikes
a note of alarm.
English dramatists often change from blank verse to a rhyming
couplet at the end of a dialogue-scene. Have we anything like this in
Plautus ? Some would find it at the end of Bacch. Ill ii (Trochaic
Septenarii) and get rid of the unfavoured patrem (II 27 [4]) by making
line 404 an Iambic Octonarius. Also at Stich. 67 (on Poen. 922 see
II 52 ii). More certain is the two-line change at Epid. 164-5. The
whole Scene is in Trochaic Septenarii, but Plautus ended with two
Iambic Octonarii :
Ibo intro atque adulescenti di||cam nostro erili filio
Ne hinc foras exambulet || neve obviam veniat seni.
When the play was re-staged the ' retractator ' adapted them to the
rest of the Scene, writing apparently :
I, i ! (' extra metrum ' ; 3 C)
Abi mtro atque adulescenti die iam || nostro erili filio
Ne hinc foras ambulet neve usquam ||obviam veniat seni Ad.
(? cf. 854).
Ter.

When a Metrician of the empire ascribes the alternation of Trochaics

and Iambics to Menander (Marius Victorinus, Gramm. Lat. vi, 57, 14:
Nam et Menander in comoediis frequenter a continuatis iambicis versi-
bus ad trochaicos transit et rursum ad iambicos redit}, he seems to be
merely referring to Menander's habit of interjecting here and there a
Scene in Trochaic Tetrameters Catalectic while the other Scenes are in
Iambic Trimeters (17, end).
In a Senarius-scene of the Aulularia the sudden intervention of a
Trochaic Septenarius (Aul. 393) :
Nimirum occidor nisi ego intro hue || propere propero currere,
has roused the suspicion of editors and has been added by the inter-
polation-hunter tohis bag. But undoubtedly it heightens the humour
of a highly humorous situation. The old miser's agony of apprehension
cannot find vent in a mere Senarius ; he has to turn to Trochaic metre
for relief. (Cf. Poen. 1165; Pseud. 1098; Trin. 96?)
Hardly a parallel, but yet a proof that a change of metre within a
dialogue scene was quite a natural thing, is the change to Senarii in
a Trochaic Scene, when a letter is read aloud (Pers. 520; Pseud. 998).
Similarly Iambic Septenarii become Senarii when an oath is adminis-
tered (Rud. 1338). The two Senarii at the beginning of Act V Sc. i of
the Asinaria, and the three at the end of Act III Sc. iv of the Amphi-
 EARLY LATIN METRES 289
truo are a curious feature. (On Stich. 762, where a stoppage of the
music brings a passage in Senarii see above, 2.)
BACCHIAC.
22. Bacchiac Tetrameter. We had to allow the possibility of
Greek priority for the Versus Reizianus (14). But this other ' metrum
Plautinissimum ' we are strongly inclined to claim for the Romans,
though not for Plautus himself. (A Bacchiac line of Naevius is cited
below.)
And if the Bacchius (w - -) was really as rare in Greek (cf.
Hephaestion 43, 1 to Be /?a*x«a/cov <nrdvi6v ccrrt) as its extant specimens
suggest, the claim is justified. In Aristophanes it appears appropriately
in a Bacchic hymn (Frogs, 316) :
Ia/c^' ai IaK^c
"Icucx' ai "louche,

and in a parody of a hymn (Thesm. 1144) :

^avrj6y & Tvpavvovs crrvyovcr Sunrep ei/cds (Or two lines).
But Plautus, putting it to more homely use, strikes a great variety of
notes out of this favourite metre of his, a metre which seems admirably
suited to Roman ' gravitas '. It is used in the Rudens when the digni-
fied priestess (we had almost said ' abbess ') advances to greet her
visitors (259 sqq.) :
Qui sunt qui a patrona || preces m(ea) expetessunt ? , etc.
The old lady in the Aulularia addresses her brother in this metre
(120 sqq.):
Velim te arbitrari || me haec verba, frater, (Or med)
Meai fidei || tuaique rei, etc.
Old Hegio's lament in the Captivi finds expression in it (781 sqq.) :
Quanto in pectore hanc rem || meo magi' voluto,
Tanto mi aegritudo auc||tior est in animo.
Ad ilium modum sub'||litum os esse mi hodie ! , etc.
Less happily perhaps his last song of joy (922 sqq. ; rather conversation
than song). His first song of joy has a quite captivating alternation of
Anapaestics and Bacchiacs. With ' Jean qui pleure ' contrast ' Jean qui
rit' (498 sqq.):
Quid est suaviu' quam || bene rem gerere (Anap.)
Bono publico, sic||ut ego feci heri, quom (Bacch.)
Emi hosce homines ! || ubi quisque videt, (Anap.)
Eunt obviam gra||tulanturque earn rem. (Bacch.)
The capers of Lord Chancellors on the stage of Gilbert and Sullivan
29o EARLY LATIN METRES

would perhaps have offended a Roman audience. But we refuse to

believe that this wealthy burgess of Pleuron (' Sir Marmaduke Ruler ')
maintained a dignified repose during the utterance of these four lines.
There must have been something of a ( pas seul \ As decorous as you
please ; but — it was there.
The use of Bacchiac in 'sermo gravis ' (e.g. Amph. Hi) makes one
almost inclined to regard it as a far-off reflex of the ' syncopated ' Iambic
(' protracted ' Iambic) of the Greek Drama (see below, 42), e. g. :
Birds 229 sq. Itw tis wSe rtov e//,u)i/ bjxoirTtpoiv
"OcrotV r €vcnr6pov<s aypoi^KOiv yvas.
(On its more certain connexion with Ionic, 35.)
23. From the above citations it will be seen that Plautus occasionally
substitutes a Molossus ( , a foot with one more ' mora ') for the
Bacchius. And by the resolution of this or that long syllable, other
varieties1 arise. But the pure Bacchius is always predominant,2
and if a line (or hemistich) ends with a monosyllable, it is invariable
in the fourth (or second) foot. (For details see Appendix D.) That
Diaeresis may be dispensed with is also illustrated by the above
citations. And Diaeresis in this, as in other metres, tolerates Hiatus
and Syllaba Anceps, e. g. :
Men. 968 Ut absente ero rem || eri diligenter ;
True. 463 Vosmet iam videtis || ut ornata incedo.
The Brevis Brevians plays as small a part in the Bacchiac Tetrameter
as in the Tragic Senarius. And for the same reason : because the tone
of flippancy, the careless conversational slurring, is unsuitable.
24. Bacchiac Cola, etc. While the Tetrameter is the usual form,
other sizes of the Bacchiac line are not unknown. And a long * run '
(' system ') of Bacchiacs appears, e.g. Amph. II ii (Alcumena's soliloquy) ;
Men. 571 sqq. (Menaechmus' soliloquy). But the variety which most
calls for notice is the 'curtailed' Tetrameter, with substitution of a
Bacchiac colon for its second half. This colon usually takes the form
^ _ w (more often with iambic than spondaic opening), as in
Cas. 662 :
Insectatur omnes || domi per aedes ;

1 The Choriambus in Cist. 686 (Halisca's agitated song ; 25; is remarkable with
the abnormal ictus on the second syllable of a dactyl-word (II 9) :
Nulla est neque eg6 sum usquam : \\perdtta perdidit me (Notice the symmetry).
2 Nonius MSS. can hardly convince us that esse should stand in Naevius' Danae
fragment :
Earn nunc [esse] inventam 11 probris compotem scis.
(Or monosyllabic Earn ?).
 EARLY LATIN METRES 291

though the ' resolution ' of this or that long syllable may evolve other
forms now and then.
On the other hand it is merely the catalectic Tetrameter which appears,
e.g. at Cist. 692-4:
Sed memet moror quom hoc || ago setius.
Halisca, hoc age, [ad] terram ad||spica et despica (-ce MSS.),
Oculis investiges, || astute augura.
And two catalectic hemistichs at, e. g. Pers. 809 :
Perge ut coeperas. || Hoc, leno, tibi (cf. 811).
Hegio's exultation (in the song quoted in the last paragraph) requires
four of them (Capt. 506) :
Rogo syngraphum : || datur mi ilico : || dedi Tyndaro : || ille abiit
domum.
Did a frisk accompany each ?
So ubiquitous a metre is not restricted to any one companionship ;
but it is often associated with an iambic colon (or sequel), e. g. :
Cist. 37 Suas paelices ess(e) || aiunt, eunt depressum (with Iamb.
Dim. Cat. as colon. Iambic Septenarii follow) ;
Cas. 837 sq. Meum corculum, mel||culum, verculum. Heus tu !
Ma/o, si sapi\ cavebis (Iamb. Dim. Cat.) ;
though some editors find Bacchiacs rather than Iambics here. On the
difficulty of discriminating a Bacchiac colon from a Colon Reizianum
see above, 15.
(On doubtful traces of miuric Bacchiac see 26 B, note.)
CRETIC.
25. Plautus is careful to observe the boundary-line between Comedy
and Tragedy. When the outcries of Trachalio in the Rudens become
too real and moving, they are stopped by a jest (Rud. 632). Why then
does Plautus use Cretics exactly as they are used in Tragedy, as a
vehicle for women's wailings ? Compare the Cretic fragment of Ennius'
Andromacha Aechmalotis :
Quid petam praesidi aut || exsequar ? quove nunc, etc.,
with some of Plautus' Cretics, e. g. Rud. 654 sqq. :
Nunc id est quom omnium || copiarum atque opum,
Auxili, praesidi || viduitas nos tenet, etc.
The answer seems to be that this ' muliebris supellex ' " had of itself a
1 Poen. 1145-6 :
Han. Tace atque parce muliebri supellectili.
Agor. Quae east supellex ? Han. Claru' clamor sine modo.
U 2
292 E ARLY LATI N METRES

comic sound for the unchivalrous Roman. He found it not harrowing,

but rather amusing (cf. Rud. 679-81). These 'wailing' Cretics are
often found (e.g. Rud. 207 sqq. ; 233 sqq. j Cas. 621 sqq.).
The metre however is put to quite other uses too (e. g. for a descrip-
tion of a battle,1 Amph. 219 sqq.). The pretty serenade at the beginning
of the Curculio has already been quoted (1) and the purity of the feet
remarked on. Nearly every one of the eight Tetrameters consists of
four actual Cretics ; twice there is a Fourth Paeon (www -), a very
common form of the foot in Greek. (In fact one usually speaks of
'Paeonic Verse' in the Greek Drama.) The Molossus ( , with
one more ' mora ') is not allowed in the fourth (or second) foot when
the line (or hemistich) ends with a monosyllable. (For details see
Appendix D.) Diaeresis in the Cretic Tetrameter may be dispensed
with on occasion, just as in the Bacchiac (23), and (as in the Bacchiac)
tolerates Hiatus and Syllaba Anceps, e. g. Cas. 190 :
Nee mihi ius meum || optinendi optiost (with both).
In Greek another type of Paeon, the first Paeon (—www) is also
found j similarly, but not to the same extent, in Latin. Plautus even
seems to admit a Choriambus (- w w -) now and then, e. g. Epid. 98 ;
True. 624; and Terence's strange genus hominum, Andr. 629 :
|| genus hominum pessimum in (Denegando),
must be a Choriambus with the first long syllable ' resolved ' (II 27 [5]
note).
As in Bacchiac Verse the Bacchius (23), so in Cretic Verse the Cretic
is always predominant. And yet for special effect, to indicate agitation
of mind, Plautus can turn out a Cretic Tetrameter with most of the
long syllables resolved, Cist. 672 (Halisca in distress over the loss of
the jewel-box which gives the play its name) :
Itaque petujlantia me||a me animi | miseram habet.
(Notice the symmetry of this Paeonic,2 rather than ' Cretic ' line.) Is it
not then futile to make rules for Plautus' Cretics : ' This foot is never
found in these circumstances ; that foot never in those ', and to alter the
text rather than allow for the adaptability of the artist ? The Molossus,
when it shows itself, shows itself usually in the first or third foot of the
Tetrameter; the Tetrameter has usually Diaeresis. A mathematician

1 This strange use of Cretics is in keeping with the ' tragicomoedia ' (Amph.
pr. 51-63).
2 Compare the Bacchants' song in Accius (trag. 239) :
Agite modico gradu ! || iacite thyrsos leves,
a line which flashes on the reader a glimpse of the tripping- footed revellers.
 EARLY LATIN METRES

could calculate the chance of a combination of these two unusual things,

a Molossus in the second foot of a Tetrameter without Diaeresis.
Spengel made a law for Cretics, that when Diaeresis is neglected the
second foot must not exceed the normal by a 'mora'. If we hunt
through all the Cretic passages in the plays, we find this to be the rule.
Usually a Cretic occupies the second foot of a line lacking Diaeresis
occasionally a Fourth Paeon, e.g. Cure. 152 (see above, 1); Most. 728 :
Musice hercle agitis a^||tatem ita ut vos decet.
But are we to sacrifice to Spengel's law the traditional reading of such
lines as these ? : 1
Most. 140 Deturbaz;// *fe/wc||itque a med ilico ;
Pseud. 1300 Quid 'libet'? pergin «w||tare in os mihi? (26) ;
Ter. Andr. 631 Post ubi tempust /w||missa iam perfici (tempus in
some MSS.).
(Cf. Most. 133; 703; Pseud. 1312.)
I The Brevis Brevians plays as small a part in Cretics as in Bacchiacs,
but not quite for the same reason. While Bacchiacs are the vehicle of
1 sermo gravis ' (23), Cretics (like Glyconics ; 32) are a metre of song.
' I would know ' was in ordinary conversation volo-scire (II 8), but in
song:
Cure. 133 Hoc volo scire te : || perditus sum miser.
26. Cretic Cola, etc. (A) Like the curtailed Bacchiac Tetrameter
(24) is the curtailed Cretic Tetrameter, usually with the colon — — - w —
substituted for the second hemistich, e. g. Bacch. 663-4 :
Sed libet scire quan||tum aurum erus sibi
Dempslt et quid suo || reddidit patri, etc. (cf. Most. 108-9).
In Bacch. 621-2 two of these cola are followed by a curtailed
Tetrameter :
Omnibus probris, || quae improbis viris
Digna sunt, dignior || nullus est homo.
A Metrician of the Empire quotes as a specimen of a 'clausula' this
colon of Caecilius (Rufinus 556, 10 K.) :
Di boni ! quid hoc ?
The short second syllable is the rule, the long the exception, for this
colon. But the exceptions make more than a mere negligible quantity
It will not do to claim fecit for Bacch. 665 :
Si frugi est, Herculem || fecit ex patre ;
nor ei for Most. 701 :
Nam et cenandum et cuban||dumst ei. male,
294 EARLY LATIN METRES

and to ignore clear examples of ^ — like Most. 108 ; Rud. 231 :

Atque illud saepe fit || tempestas venit ;
Spes bona, obsecro, || subventa mihi (Or two lines).
The ' responsion ' of the neighbouring lines confirms this form of the
opening colon in Most. 344-5 :

Da illi quod bibat. || Dormiam

ego iam. (With ' miuric ' second
colon ; see below.)
Num mirum aut novum ||quippiam facit ?
At the same time the normal form (— w — kj — ) is always preferable.
We must not read id ' iam ' in Most. 338 :
Iam revertar. Diu || est ' iam ' id mihi ;
we must not prefer the reading of the Palimpsest in Most. 712 :
Nil erit quod deo||rum nullum accusites (ullum A),

especially since the following line too has an ' impure ' third foot :
Te ipse iure optimo || merito incuses licet ;
we must not prefer quaen to quaene in Most. 738 (where the MSS. have
quae nee) :
Quo modo? Pessimo. || Quaene subducta erat?,
nor me to med in Most. 140; and perhaps we must scan :
Asin. 130 At malo cum tuo ; || nam iam ex hoc loco;
135 Nam in mari repperi, || hie elavi bonis;
Most. 149 Cor dolet quom scio ut || nunc sum atque ut fui ;
152 Disco, hastis, pila, || cursu, armis, equo (III 56) ;
315 Nam illi ubi fui, || inde effugi foras.
(B) The ' miuric ' form (with short for the first long syllable of the
final foot) is a curious variety (— w ^ ^ -), e. g. often in Simo's song,
Most. 690 sqq. :
693 Nunc dormitum iubet || me ire minime ;
696 Voluit in cubiculum ab||ducere me anus ;
697 Non bonust somnu' de || prandio. apage ;
703 Siqui' dotatam uxo||rem atque aniim habet (Emended to
(earn) an.) ;
706 Exsequi certa res || est ut abeam (also 708 ?).
Il seems here to have the vulgar tone, the leer, of miuric ' verse. So in
the drunken Scene of the Pseudolus, e. g. :

1 Verrius Flaccus seems to have treated the ' Calabri versus ' (miuric Dactylic
Tetrameters), the vehicle of the rustic banter of South Italian shepherds (e. g. an
 EARLY LATIN METRES 295

Pseud. 1 30 1 Suavi' ructus mihi est. || sic sine, Simo.

The irrational short syllable is not always an accented syllable.1
Like the exceptional form of the other Cretic colon (e. g. Subventa
mihi; e.g. || fecit ex patre), with the second syllable long, is this
exceptional form of the miuric colon :
Rud. 212 || monstret, ita nunc.
(C) The resolution of the first long syllable of the Cretic colon
produces now and then uw ^ — w — (e.g. Pseud. i3ioa Mulier haec
facit ; Most. 116). And there is no denying that other types may have
been struck for a particular occasion by the inventive fancy of Plautus
(or his Greek model). But at the end of a long Cretic passage we
should rather find a Glyconic colon (Horace's Edite regibus; 32) in
Most. 739 :

Tuto in terra. Ei ! 1296).

Quid est ? || Me miserum, occidi ! (cf. Pseud.

And perhaps a mere catalectic Cretic Dimeter, in such an (occasional)

appearance as :
Rud- 953 || non fore infidum ;
955 II quisquis es, audi ;
True. 123 Salva sis. Et tu,

(D) Cretic ' runs ' are as common as Bacchiac ' runs ' (24). A small
1 run ' (of seven Cretics) was cited in the second paragraph of this chapter
from the conclusion of the Pseudolus. A longer specimen will be
found in Pseud. 920-1 (ending in an Ithyphallic Colon).
(E) Cretic Verse has (in Latin as in Greek) affinity with Trochaic.
It takes as colon or sequel a Trochaic Dimeter often (e.g. Cure. 113*-
114; Cas. 237; 628); often an Ithyphallic Colon (20); a Trochaic
Monometer sometimes (e.g. Rud. 237 ; 681). And comic discomfiture
combined with self-reproach is delightfully rendered by a mixture of
Cretic Monometers with Trochaic Septenarii in the first Scene of the
Epidicus (85 sqq.) :
Neque ego nunc
Quomodo

imaginary line, ' Stulte cucule, cucule, quid habes ?1), raised to literary rank in the
Empire by the Neoterici in rural lays, e. g. Quandft flagella ligas, ita liga.
1 Have we a miuric Bacchiac Tetrameter in such lines as Cas. 665, 675 ; a Tri-
meter in Amph. 179?
Metu mussitant. Oc||cidi atque interii (Or interivi ?) ;
Sciens de via in se||mitam degredere ;
Hie qui verna natust queritur (Or queretur, a question ?)..
 296 EARLY LATIN METRES

Me expeditum ex impedito || faciam, consilium placet.

Ego miser
Perpuli
Meis dolis senem ut censeret || suam sese emere filianu:
and so on ; then 94 sqq. :
At enlm tu
Praecave.
At enlm — bat enlm ! nil est istuc. || plane hoc corruptumst caput.
Nequam homo's,
Epidice,
Qui libidost male loqui ? quia || tute te(te) deseris.
ANAPAESTIC.
27. Quintilian's sigh over Terence (above, 17) we may echo for
Plautus : ' si citra anapaesticos stetisset '. Our ears, accustomed to
accentual verse, are offended by the continual clash of accent and ictus
in lines like :

Pers. 753 Hostibu' victis, civibu' salvis ||;

Trin. 239 Blandiloquentulus, harpago mendax ||.
And that the Romans themselves found this unpleasing we see from the
later reform (ascribed by Metricians of the Empire to Pacuvius ; Gramm.
Lat. vi, 77, 7), which avoided this clash; e. g. Pacuv. 350 :
Agite, i(ci)te, volvite, rapite, coma ||tractate per aspera saxa et
humum. (Or ite (e)volv., unless Hector's corpse is referred to.)
As frequently happened in Roman adaptation of Greek Metre, the
pioneers had kept too closely to the Greek model, and it was left to
their successors to prune away the roughness. And even the most
devoted admirers of Plautus feel that there is an excessive use in his
Anapaestics of the slurrings due to a Brevis Brevians, and have to
excuse it by the paucity of short final syllables in his time.
Diaeresis had clearly a great significance to these dramatic pioneers,
whether through the influence of Saturnian verse (I 9) or from some
other reason. In all metres the Roman observance of Diaeresis was
stricter than the Greek. And this Diaeresis the Romans honoured with
the same privilege as any line-ending or any strong pause, for they allow
to it Hiatus and Syllaba Anceps. Now Diaeresis is a marked character-
istic of Greek Anapaests. Not only is the main Diaeresis (at the middle
of the line) so much the rule that some editors of Aristophanes wish to
emend any exception, but there is recurrent Diaeresis throughout the
long line. In Aristophanes (cf. White, p. 125) after the first metrum of
 EARLY LATIN METRES 297

the first hemistich there is nearly always Diaeresis ; after the first metrum
of the second hemistich there is Diaeresis in one out of every three
lines. Pacuvius' reform meant that the Romans had shaken off this
early obsession ; Diaeresis had come to be honoured in the breach as
much as the observance. But to Plautus Diaeresis is the salient feature
}f the Greek Tetrameters, the feature which must — at whatever cost —
De reproduced in the Roman imitation. And so he sacrifices to (re-
:urrent) Diaeresis the regard for accent (a useful reminder, by the way,
that Plautus was a quantitative, not an accentual poet, and did not
assign too high a place to the reconciliation of accent and ictus ; II 14).
If one judges the matter fairly, one has as much right to be offended at
Pacuvius' sacrifice of Diaeresis as at Plautus' sacrifice of the harmony
between accent and ictus. With Greek models before him like :
©wvtiScs oiTTai, <f>via$e<s icp6at,
"Hoi^os, rjcrv)(o<s} rjpefia, kolvOcdv,
*0 Avkc 8eo-7Tora, yetTwv rjpois ||,
*H rpv<pep(i)T€pov rj Seworepov ||,
what could Plautus do but turn out imitations like these ? :
Hostibu' victis, civibu' salvis ||;
Blandiloquentulus, harpago mendax ||.
Indeed Accius, in spite of Pacuvius' reform, found no fault with this
type, e.g. 562 (Philoctetes' lament) :
Heu ! qui salsis fluctibu' mandet ||me ex sublimo vertice saxi ?
lam iam absumor : conficit animam || vis vulneris, ulceris aestus.
28. This censure of Plautus' Anapaestics comes from ignorance of
the history of Latin poetry. The critics forget the veneration for
Diaeresis in Plautus' time. Another censure comes from ignorance of
the phonetics of the Latin language. Critics are offended by a Brevis
Brevians like harpago ; they think it a mispronunciation of harpago — a
mispronunciation admitted only to Anapaestic lines, and due to Plautus'
incompetence to grapple with the difficulties of this metre. They can
reconcile themselves to a scansion like Terence's bonis in a Senarius.
After all, they say, it is much like Virgil's scio, modo (Adverb), viden,
or Horace's puta or Catullus'dabo, volo, or Ovid's cave, or Phaedrus'
vide, or Persius' rogas. But harpago they think a different thing (they
forget Horace's Pollio, mentio, dixero, Catullus' nescio, commoda, etc.),
a thing alien to dialogue-metres. And yet they could find it there if
they look more carefully, e. g. Capt. pr. 8 Alteriim quadrimum j Capt.
558 Hegio; Amph. 391 Dicito; Stich. 223 Hercules; although it is
certainly as rare in dialogue-metres as it is frequent in Anapaestics. But
298 EARLY LATIN METRES

why is it rare? Not because harpago was more of a departure from

actual utterance than bonis. (That is the false assumption one is so apt
to make.) Merely because the rules of Iambic and Trochaic Verse rule
out dactyl-words, — all dactyl-words, not only dactyl-words produced out
of cretic-words by the Brevis Brevians. A dactyl-word like harpagus,
just as much as a dactyl-word like harpago, is rarely allowed as a foot of
Iambic or Trochaic Verse (II 58). The only corner for it is the first
foot of a line (or hemistich) and in this place the harpago-type appears
equally with the harpagus-type. A division of a foot between words like
MacizV vertix. is forbidden in dialogue-metres, and so not merely harpago
but harpagus is not to be found in such a situation in Iambics and
Trochaics, a very common situation for it in Anapaestics. In fact
almost the only place left for a dactyl-word (not elided) in dialogue-
metres, is at the end of the line (or hemistich) ; and of course there is no
opportunity there for a form like harpago.
This paragraph is really a repetition of II 27 (5), etc. But a necessary
repetition. Indeed we doubt whether anything less than a seven-fold
repetition will avail to dispel the mist of prejudice which has been
created on the one hand by Klotz's ' metrical ' theory of the Brevis
Brevians and, on the other, by the Teubner editors' use of the accent-
stroke as an ictus-symbol. When one sees this symbol over the second
syllable of harpago in the Teubner edition :
Blandfloquentulus, harpago, mendax,
he reads (willy-nilly) the word with accent-stress on the second (instead
of the first) syllable. And a long habit of so misreading the Anapaestics
of Plautus leaves a settled impression of a violent, unnatural treatment
of the material by a prentice hand, hacking a way through line after line.
To judge Plautus' Anapaestics more fairly, readers must force themselves
to realize these two things : (1) that harpago reflects conversational
utterance as closely as bonis, (2) that Plautus in this metre deliberately
sacrifices Accent to Diaeresis and, since he is a quantitative (not
accentual) poet, has a perfect right to do so.
Conversational slurrings like bonis, harpago are, as we have seen
(II 26), less favoured (naturally) in the dignified utterance of Tragedy.
As specimen of Tragic Anapaestics we offer a passage (with no Brevis
Brevians) from Ennius' adaptation of Euripides' Iphigenia at Aulis :
Quid nocti' vide||tur in altisono
Caeli clipeo ? || Temo superat
Stellas cogens || etiam atque etiam
Noctis sublim(e) j|iter . . .
 EARLY LATIN METRES 299
Here is the Greek (Iph. Aul. 6) :
Ti? 7tot' ap acrrrjp ||oSc 7ropOjX€V€L ;
Sctpios, iyyvs \\tt}s iTTTcnropov
nAaaSos acrcroiv ||ert paacrrjprjs-
It is after citing an Anapaestic passage of Ennius (0 pater, O patria, O
Priami domus, etc.) that Cicero breaks out with his famous : O poetam
egregium ! quamquam ab his cantoribus Euphorionisxontemnitur (Tusc.
iii, 19, 46). Accius' sea-coast picture, with startled cormorant, etc., is so
modern that it is worth quoting (trag. 571) :
Simul et circum ||merga sonantibus (megna MSS.)
Excita saxis ||saeva sonando
Crepitu clangente cachinnat.
It gives us a picture in words like one of Mr. Peter Graham's paintings.
29. Anapaestic Septenarius. Anapaestics are found in Menander-
citations (by Blass, Hermes 33, 655), but we must take Aristophanes for
comparison with Plautus. White (Verse of Greek Comedy, pp. 1 21 sqq.)
tells us that in Aristophanes' Tetrameters Catalectic of dialogue the
Spondee is the prevalent foot ; the Proceleusmatic (w w w w) occurs only
once (Birds 688), and only once the succession of four short syllables
resulting from collocation of Dactyl and Anapaest (Wasps 397); the
Dactyl is very rare in the fourth foot, seldom in the second (mostly when
a Dactyl precedes in the first) ; the seventh foot is always an Anapaest.
How far these regulations were relaxed or altered in the Middle and
New Comedy we do not know. Plautus does not reserve the seventh
foot of his Septenarius for an Anapaest, even when the line ends in a
(single) monosyllable (for details see Appendix D). Like Aristophanes
he gives free play to the Spondee, but he likes sometimes to accelerate
the metre with Proceleusmatics (e. g. in the seventh foot, Cist. 205, ||cru-
ciabi\ita.ti&us animi), and has no objection to the quartette of short
syllables produced by Dactyl + Anapaest, e. g. Mil. 1024, 1084:
|| nullumst hoc stolidii? saxum ;
lam, iam sat, amabo, est. sinite adesun, ||si possum, viva a vobis.
The Dactyl in the fourth foot is rare (as in Aristophanes), e. g. Bacch.
1 106 (preceded by a Dactyl) ; 1 157 :
[Philoxene] Salve. Et tu. unde agis ? unde homo ||miserfortunatus
atque in-;

Nihili sum. Istuc iam pridem scib. ||sed qui nihili's, id memora.
Beside Aristophanes' abnormal neglect of Diaeresis after the fourth foot,
e.g. Wasps 568 :
K'xv fir) tovtois ava7r€i0(.oiJL€cr\\0a, tu. irai&api evOvs av£\.K€i,
3oo EARLY LATIN METRES

we may place an abnormal line like Pers. 778 :

Solus ego omnibus antideo faci||le miserrimus hominum ut vivam.
But to compile an exact list of lines which violate the law of Diaeresis
I(or other laws) is impossible until the Cantica have been analysed. The
Teubner editors, as we have said, have been reckless in labelling verses
as ' Anapaestic ' ; and since the Teubner edition is the standard critical
text, the avenue through which every one passes to a study of Plautus,
we are all too ready to find this metre where Plautus never put it.
Even the example cited in the last paragraph (Blandiloquentulus harpa-
go mendax) is not above suspicion, for our MSS. seem to offer this line
(Trin. 239) :
Blandiloquentulus, harpago, mendax, cuppes, avarus, elegans, de-
spol[i]ator,
the second half of which, clearly not anapaestic, has not yet been
labelled. (It looks like an Iambic Dimeter followed by a Trochaic
Monometer !) All that we can do at present is to recognize certain
forms of the line as abnormal and to avoid them where they can be
avoided. Thus we must put the hemistich-ending after quaeris (not
after the first syllable of ubi) in Mil. 1012 :
Homo quidamst qui scit quod quaeris || ubi sit. Quern audivi
ego hie?,

since Syllaba Anceps (and Hiatus) are as universal after the Diaeresis
of Anapaestic as of other metres.

30. Anapaestic Octonarius. It seems impossible to detect any

difference of nuance between the Octonarius and the Septenarius in this
metre. The two stand side by side in many anapaestic passages, and
editors are often at a loss for criteria to decide when this or that line
is acatalectic or catalectic. The first Scene of the fifth Act of the
Bacchides illustrates this (1087 sqq.) :
Quicumque ubi sunt, qui fuerunt ||quique futuri sunt posthac
(Catal.)
Stulti, stolidi, fatui, fungi, || bardi, blenni, buccones, (ditto)
Solus ego omnes longe antideo ||stultitia et moribus indoctis.
(Acatal.)
Perii ! pudet : hocin me aetatis ||ludos bis factum esse indigne !
(ditto)
Magi' quam id reputo, tarn magis uror ||quae meu' film' turbavit.
(Catal.),
 EARLY LATIN METRES 301

and so on. Using the criterion mentioned in the last paragraph, we

may make line 1177 acatalectic and save the Diaeresis :
Ego quidem ab hoc certe exorabo. || Immo ego te oro ut me intro
abducas,

with change of speaker at the natural place, the end of the hemistich.

31. Anapaestic Dimeters, etc. Editors have the same difficulty in

deciding -the claim of Dimeter against Tetrameter in this metre as in
Iambic (11) and Trochaic (19). The 'run' at Aul. 721 sqq. is arranged
by some editors in long lines, by others thus (with the last line
catalectic) :
Heu me miserum ! ||misere perii,
Male perditu', pes||sime ornatus eo :
Tantum gemiti et ||malf maestitiaeq(ue)
Hie dies mi obtultt, ||famem et pauperiem.
Per[d]itissimus ego || sum omnium In terra ;
(Superl. of perltus ' perished ')
Nam quid mi opust vi||ta, [qui] tantum auri
Perdidi, quod con||custodivi
Sedulo ? egomet ||me defrudav(i)
Animumque meum ||geniumque meum ;
Nunc eo alii ||laetificantur
Meo malo et damno. ||patl nequeo.

(Similar 'runs' are, e.g. Bacch. 1084 sqq.; Cist. 697 sqq.) The pretty
love-song at the beginning of Act II of the Cistellaria editors set in this
arrangement :
Iactor, crucidr, agitor,
Stimulor, versor
In amori' rota, miser exanimor,
Feror, differor, distrahor, diripior;
Ita nubilam mentem animi habeo,
although, of course, the first line could have no independent existence,
since a ■ tripody ' is as impossible in Anapaestic as in Iambic and
Trochaic verse. (Cf. True. 566 sqq.)
In these Dimeters (as in the hemistichs of Septenarii and Octonarii)
the harmony of accent and ictus is not always sacrificed to Diaeresis.
Lysiteles' song of joy pleases an English reader (Trin. n 15 sqq.) :
Hie homost hominum om||niiim praecipuus, (omn. horn. MSS.)
Voluptatibu' gau||ditsque antepotens ;
Ita commoda quae || cupio eveniunt,
302 EARLY LATIN M E T R E S

Quod ago adsequitur, || subest, subsequitur ;

Ita gaudils gau||dium suppeditat.

Notice the flouting effect of these Catalectic Dimeters, in spite of all

their Spondees (Stich. 315 sqq.):

Ge. Ibo atque hunc com||pellabo.

Salvus sis. Pl. Et ||tu salve.
Ge. lam tu pisca||tor factu's ?
Pl. Quam pridem non ||edisti ?

Ge. Possum scire ex || te verum ?

Pl. Potes : hodle non || cenabis.

(Cf. Cas. 741 sqq.) Do not they recall Livy's description of the earliest
drama of improvisation (vii. 2 imitari deinde eos iuventus, simul incon-
ditis inter se iocularia fundentes versibus, coepere ; nee absoni a voce
motus erant) ?
A quite different rhythm appears in this metre when used (as in
Greek ; Hardie, Res Metrica, p. 57) for sailors' songs (see Servius' note
on Virgil's imitation, Aen. iii, 129 Cretam proavosque petamus). Our
example is of doubtful source (and age), trag. inc. 253 :

Agite, O pelagi || cursores,

Cupidam in patriam ||portate ! (some Mary, Queen of Scots, sailing
from France ?).
GLYCONIC.
32. This term is so familiar to English readers of Horace that we
use it (loosely) in place of the more correct ' Dactylo-trochaic ' x to
include all varieties of this type, not merely the Glyconic proper
(-^-ww — w — Cui flavam religas comam) and the catalectic Glyconic,
the Pherecratean (— ^ — v/ u ), but also lines where a dactyl usurps
the trochee's place or a trochee the dactyl's. Under Glyconic cola we
include Horace's Maecenas atavis (- — - w ^ -) and Edite regibus
(— w w — w — ).
This type is common in Plautus (who favours a strong admixture of

1 Wilamowitz's discovery of the ' Viersilbler ' (I 8) has brought into fashion
a treatment of this verse as Iambic, cui flavam r£|ligas cfimam. Hardie (Res
Metrica, p. 140) vindicates the scansion by Dactyls and Trochees, cui fla|vam
rSli|gas cfimam. He uses the argument that a shortening of a final vowel by
Prosodic Hiatus is alien to all Greek Iambic Verse, but a common occurrence in
this type.
 EARLY LATIN M K TRES 303

Dactyls) ; though the Teubner editors, when they fail to recognize it,
label it 'Anapaestic',1 e.g. Cas. 217 :
Omnibu' rebus amorem ego credo ||et nitoribus nitidis || antevenire
nee potis (eg. am. MSS.).
It is 'par excellence' the metre of song (e.g. the marriage chorus of
Cas. Sco, 808 : Hymen Hymenaee, O Hymen); and Terence, when he
adopts a song-metre for a Canticum, does not forget Dactyls and
Trochees. In the Andria,
625 Hoccinest credibile aut memorabile (four Dactyls),

and perhaps 636 (with Heus ! an ejaculation ' extra metrum ' \ 3 C) :
Proximus sum egomet mihi.
In the Adelphi,
610 Discrucior animi,
and perhaps 614:
Quo modo me ex hac || expediam turba (where some find Dochmii ;

then 615-16 : 41):

Tanta nunc suspicio de || me incidit neque ea immerito.
He associates it in the latter Canticum with Choriambics (3 B), an
association known to Aristophanes and to Plautus too (36), in the
former with Cretics. Cretics are found with it in Lysidamus' song of
despair in the Casina (937 sqq.) :
Maximo ego ardeo nagitio
Nee quid agam mels rebu' scio (Or meis),
Nee meam ut uxorem aspiciam
Contra oculis, ita disperii ;
Omnia(n) palam sunt probra !
Omnibus modis || occidi miser ! (Two Cretic cola ; 26) ;
then after some Cretic lines,
Improbos famulos imiter || ac domo fugiam.
Nam salus nulla est scapulis || si domum redeo,
finally two lines ending in a Colon Reizianum (15) :
Hac dabo protinam et fugiam. Heus 1 || adsta ilico, amator.
Occidi ! revocor : quasi || non audiam, abibo.

1 The Teubner editors, accepting the scribe's transposition, scan this ' run ' (as
far as the third word from the end) as an Anapaestic Octonarius, with the horrid
amorem and nitoribus. What is the sense of the line to which they wed this
unpleasing sound ? ' That Love is the brightest and sweetest of all things bright
and sweet.' Truly an ill-assorted union of sound and sense !
3o4 EARLY LATIN METRES

And the Cretic serenade (i) in the Curculio is followed by three Glyconic
lines :
Re specio nihili meam vos || gratiam facere.
Sed tace [tace] ! Taceo hercle equidem. || Sentio sonitum.
Tandem edepol mihi morigeri || pessuli fiunt.

The first Canticum in the Menaechmi (' Comedy of Errors ') begins
with a Choriambic line, then a Glyconic, then two Cretic, then another
Glyconic :
no sqq. Ni mala, ni | stulta sies, || ni indomita im|posque animi,
Quod viro esse odio videas, || tute tibi odio habeas. (?)
Praeterhac si mihi || tale post hunc diem
Faxi', faxo foris || vidua visas patrem.
Nam quotiens foras ire volo, || me retines, revocas, rogitas.
The duet of Trachalio and Gripus, in the Scene from which the Rudens
takes its name, shows the same connexion of Glyconic with Cretic
(947 sqq.) :
Tr. Eho ! modo est operae pretium || quod tibi ego narrare volo.
Gr. Eloquere(n) quid id est ? Vide num || quispiam consequitur
prope nos ;
then three Cretic Tetrameters ; then two Glyconics with Cretic cola
(26):
Tr. Si fidem modo das mihi te || non fore infidum.
Gr. Do fidem tibi : fidus ero, || quisquis es. Tr. Audi.
Other duets in this metre are, e.g. Bacch. 626 sqq. (associated with
Cretics and Choriambics) :
Pi. Mnesiloche ! quid fit ? Mn. Perii.
Pi. Di melius faciant ! Mn. Perii.
Pi. Non taces, insipiens ? Mn. Taceam ?
Pi. Sanu' satis non es. Mn. Perii.
(628, Troch. Octonar.)
Pi. Heia ! bonum habe animum. Mn. Unde habeam ?
Mortuu' pluri' pretist quam ego sum.
Pi. Militis parasitu' modo
Venerat aurum petere hinc,
Eum ego meis dictis malis, etc.
Bacch. 989 sqq. :
Ni. Ut scias quae Mc scripta sient.
Ch. Nil moror neque scire volo.
 EARLY LATIN METRES 305

Ni. Tamen ades. Ch. Quid opust? Ni. Taceas :

Quod iubeo, id facias. Ch. Adero.
Epid. 533 sqq. (followed by Choriambic and Cretic) :
Pe. Quis lllaec est mulier timido || pectori peregre adveniens (with
Anapaest for Dactyl)
Quae ipsa se miseratur ? Ph. In his || dictu' locis(t) (?) habitare
mirli (dictust MSS.)
Periphanes. Pe. Me nominat haec ; || credo ego Mi hospitio usu'
venit.
Ph. Pervelim mercedem dare qui || (on the rest of the line see
III 18).

33. Versus Priapeus. This combination of Glyconic (proper) and

Pherecratean (— ^ — \j \j - \j — || — ^ — v w ) seems to have been
attempted by Plautus. It is probable that the Glyconic ' run ' just cited
from the Casina (line 217) ends with a catalexis of this kind :
|| antevenire nee potis || quicquam commemorari.
This shows the combination. But a better example, a separate line,
occurs in the same play when the mock bride is lifted over the threshold
(Cas. 815) :
Sensim super attolle li||men pedes, nova nupta (pe. mea no. A).
(Anapaest for Dactyl ; 39.)
34. Versus Diphilius> Versus Eupolideus. These metres may con-
veniently betreated together here. The first (-w^-^^-||^-ww-ww-),
e. g. (Peace 775) Movcra, crv /xkv 7ro\€fiov<s || d7ra>o-a/x,eV?7 fi€T i/xov, gets its
name from Diphilus' frequent use. As an example in Plautus we take
a line from a play adapted from Diphilus' Clerumenoe (Cas. 644) :
lam tibi istuc cerebrum || dispercutiam, excetra tu.
The second, which gets its name from Eupolis, the Old Comedian, was
also affected by Diphilus, Plautus' frequent Greek model. Its normal
scheme is (- ^ - ^ - w ^ - || ^ - - ^ - w -), e. g. (Clouds 518-19) :
TCl 6e(i)fj,€VOi, KarepC!) || 7rpos v/xas eAev^epcos
TaXrjOr} vtj tov Aiovv||ctoj/ rbv iKOptyavra /xe.
An example in Plautus is perhaps :
Bacch. 673 Quid igitur, stulte, quoniam oc||casio ad earn rem fuit?
IONIC.
35. With the minor Ionic foot (y v — -) every reader of Horace is
familiar : Miserarum est | neque amori || dare ludum | neque dulci, etc.
In Horace's Ode each foot is pure (as in Rud. 195 Dati' di ; nam quid
2348 X
306 EARLY LATIN METRES

habebunt || sibi (si)gni impit posthac ?). But a common feature of this
metre is what is called * anaclasis ' : instead of ^ w \w appears
w u - w j- w , e. g. (Wasps 298) Ma At', el Kpefxai(rOe y v/xcts. And
in Aristophanes' Ionics there are all sorts of departures from the normal
forms, by resolution of long syllables, by admission of a long syllable to
the place of a short, and so on (see White, p. 188, for details of these
'free' Ionics). •
The major Ionic foot ( w \S) or by anaclasis (— ^ - ^) was raised
(or degraded) to literary recognition by Sotades, of the Alexandrian
circle. Ennius translated his Greek verses (under the title of ' Sota ',
the pet-form of the name Sotades) in the same metre (ending in a
Spondee), e. g. :
Ibant mala|ci viere | Veneriam co|rollam.
But Plautus' Versus Sotadei in the Amphitruo are probably earlier
(168 sqq. Sosia's drunken soliloquy) :
Noctesque di|esque assidu|o satis su|perque est
Quod facto aut | dicto adest (?) o|pus, quietu' | ne sis.
Ipse dominu' | dives ope|ris, laboris | expers (op. et la. MSS.)
Quodcumque homi|ni accidit ll|bere, posse | retur :
Aequum esse pu|tat, non repu|tat labori' | quid sit.
They are followed by Bacchiacs. Similarly the minor Ionics of True.
448 sqq. :
Puero isti | date mammam. ut j miserae, etc.,
are followed by a passage in Bacchiac Tetrameters. And the substitu-
tion of a Bacchius for an Ionic foot is not unknown in Greek (see
Wilamowitz, Isyllos, p. 151 ; p. 137 sq.).
Plautus' Ionics could hardly be less ' free ' than the Greek. But the
precise rules by which he restricted them cannot be formulated till his
Cantica have all been analysed. The phrase 'Ionica perdidici' in a
passage of the Pseudolus (1275 SQQ*) shows that the metre is Ionic and
not (as recent editors have made it) Anapaestic. From our citation of
the lines (1) it will be seen that there is a great variety of feet.
There is no ' anaclasis ' in Afranius' major Ionic or Sotadean line
(202) :
Multa atque mo|lesta es : potin j ut dicta fa|cessas ?
36. Choriambic. The association (as in Greek, e.g. Thesm. 357) of
this metre with Glyconic has been mentioned and an example quoted
from Menaechmi no. Another example is Epid. 537 (after a Glyconic
duet and before Cretics) :
Noscito ego hanc, | nam videor j nescio ubi | mi vijdisse prius,
 EARLY LATIN METRES 307

with a Spondee as penultimate foot. The Spondee stands at the

beginning in Bacch. 625 (before a Glyconic duet) :
Consolandus hie mist : | ibo ad eum,

as in Horace's Nullam, Vare, sacra vite prius, etc. It stands at the end
of the ■ run ' in Bacch. 651 (after Cretics) :
Nequiu' nil | est quam egens | consiK serjvus, nisi habet | multipotens
There is no Spondee in Cas. 203 (after Cretics) :
1 pectus.
Tu quidem adverjsum tuam amijcam omnia loque|ris. Tace sis,
where the third Choriambus has its final long syllable resolved ; nor in
Cas. 629 (after Cretics) :
Eripite isjti gladium | quae sulst imjpos animi.
DACTYLIC.
37. In comparison with the transference of Greek dramatic metres to
Italian soil, Ennius' capture of Homer's hexameter is so great an
achievement and so momentous for Latin literature that we are tempted
to devote more pages to it than the available material allows. But only
some 600 lines of (or attributed to) the Annales survive, about as many as
the lines of Plautus cited by Nonius Marcellus. If any one constructed
a system of Plautus' versification with the Nonius material alone, how
misleading it would be ! And of these 600 hexameter lines of Ennius
some of the most often quoted may not have belonged to his Epic. He
wrote on Grammar (in his Saturae ?), just as Lucilius wrote on Grammar
in his Saturae in hexameter verse. If he wished to explain what Tmesis
was, how natural would be the remark that it was just as if one were to
say cere- comminuit -brum instead of comminuit cerebrum ; how suitable
to a genial pedagogue ! (Ennius succeeded Livius Andronicus as
headmaster of the great school at Rome.) The shade of Father Ennius
must often chuckle at the over-serious who find in this a line (and a
typical line) of his Epic. And since Lucilius poked fun at Ennius as
well as at his own contemporary, Accius, may not the ' freak-lines ' (pre-
served in the Grammarians' chapters on Solecisms) be caricatures ? 1
1 Cf. Hardie, Res Metrica, p. 4 : 'It has been suggested that Ennius introduced
this effect (Synaphea) by writing ' altisonum cael(um) ' at the end of a line. Ennius
is credited with inventing monosyllabic forms cael and gau for caelum and gaudt'um,
on the analogy of Homer's Kpi and ScD. Elision would not explain gau so readily
as it does cael. But it may be doubted whether Ennius ever really wrote such lines
at all. They may have been fictitious examples of solecisms or eccentricities. We
know (from Pompeius, 289, 10 K.) that Lucilius not merely said that there were 100
kinds of solecism, each with its own appropriate name, but actually described them :
e.g. Tmesis: " as if we* were to write", "as if Ennius wrote Massili- portabant
iuvenes ad litora 'tanas11.''
X 2
3o8 EARLY LATIN METRES

Readers who desire the precarious statistics which a chance collection

of fragments can supply will find them in two excellent articles men-
tioned in Appendix E (s. v. Skutsch $ s. v. Witte).
Of course it is unfair to compare Ennius' hexameter with the finished
form that the line took from Virgil. We must compare it with Homer's.
Ennius found in Homer lines beginning with an Iambus (e.g. II. 22, 379
'E7T€t Srj rovS' avSpa, etc.) ; but we do not know that he imitated this
licence. He found lines beginning with a Tribrach (e.g. II. 21, 352 To.
Trepl Ka\a pU6pa, etc.) and seems to reproduce this type in Ann. 340 :
Veluti ' si quando vinclis venatica velox,
unless Festus has wrongly substituted veluti for sicuti. He found lines
beginning with a Proceleusmatic (e.g. II. 13, 144 'Pea SteAcvVeo-tfai
/cAio-tas, etc.) and may 2 have followed with his :
Ann. 490 Capitibu' nutantis pinos rectosque cupressos.
For his line (un-Virgilian, but with happy union of sound and sense) :
Ann. 478 Labitur uncta carina per aequora cana celocis,
he found a type in e. g. II. 1, 214 :
v/3ptos civc/ca T^crSe. (rv 8' to^eo, 7r€i0co $' fjfuv.
For another (with the true pulse of the oar),
Ann. 230 Poste recumbite vestraque pectora pellite tonsis,
in e.g. II. 1, 488 :
avrap 6 fxiqvu vqvm Trapr}[JL€vo<s wKV7r6poi<rLV.
And so on. He allowed a monosyllabic ending, because Homer
allowed it ; a four-syllabled or five-syllabled for the same reason. It
took the Roman poets a long time to discover that Roman hexameters
sound best when they end in a word of two or of three syllables.
38. Leaving these petty details we insist on the unmistakable and
notable fact that Ennius gave at once to the nascent Roman hexameter
its characteristic Roman form, a form which the metre never lost. Of
the three kinds of Caesura :

/xrjvLv aeiSe, Ocd, \ ILrjXrjidSeo) 'AxiXfjos (penthemimeral),

Sioyeves AaepTtdBr], \ 7ro\vfJLr})(av 'OSvaaev (hepthemimeral),
dvSpa jxoi evv€7T€} Mouo-a, | iroXxtrpoTrov os fidka 7roX\d (trochaic Cae-
sura in third foot),

1 Some scan veluti, but we doubt the use of an Anapaest for a Dactyl in Epic
Verse, although Homeric openings like (II. 17, 461) 'Pt'a pev yelp tpevyea/cev sanction
it. (On Virgil's Fluviorum, etc., see III 23). On its use in the homely hexameters
of Ennius' Hedyphagetica see 1 1.
2 The Old Latin declension seems to have been : Nom. columen, Gen. col(u)minis
(i. e. culminis), Dat. col(u)mini (i. e. culmini, etc.). So that a trisyllable cap(i)tibus
is not quite impossible. On Virgil's abiete see III 23.
 EARL Y LATIN METRES 309

it is the third which is characteristic of the Greek and which gives

Homer's lines their lively movement. But Ennius' unerring judgement
made him prefer the other two, and these two remained to all time the
Roman caesuras of the hexameter.
Another unmistakable feature of Ennius' Epic Hexameter is its
avoidance of Elision, a departure from Homer's practice and from
Ennius' own practice in dramatic verse. Another is the avoidance of
the Bucolic Caesura in Latin. But Ennius, unlike Homer, freely allows
a spondee in the fourth foot whose final is long by ■ position ' only, and
tolerates in the same foot a trochaic division (e. g. corde capessere.
semita nulla pedem stabilibat, Ann. 43). A passage preserved by
Macrobius will, with its Homeric model, illustrate most of these differ-
ences of technique :

(II. 16, 102) Atas 8' ovkct efxifxve' /?ia£eTO yap /JeAeeacriv.
Sdfiva fxiv Zyjvo<s re voos kcu Tpwcs ayavol
(3dX,\ovTc<s' Scivrjv Se 7repi Kpord^oicTL <j>a€Lvr)
Tri]\r)£ /3aWo/X€vr] /cava^^v €%€, fidWzro 8' at€t
K0L7T (fxiXap' eV7TOLrj6 ' 6 8 api(TT€p6v WfJLOV €KO.fAV€V
c/x7TcSoi/ alkv exo)V °"aKO? alokov' ovSe Svvavro
ajxcf> avra> 7reA.c/u£ai epei'SovTcs /SeAeWcrii/.
atct 8' apyaAew cx€T' acr^/xaTt, kclS 8e ol iSpws
rravToOev ii< /xeXecov 7ro\v<s eppetv, ov84 7rrj e?X€V
djXTrveva-at' 7rdvTYj 8e /ca/cov Kaxw eo"T^piKTo.
(Ann. 401) Undique conveniunt velut imber tela tribuno :
Configunt parmam, tinnit hastilibus umbo,
Aerato sonitu galeae (?), sed nee pote quisquam
Undique nitendo corpus discerpere ferro.
Semper abundantes hastas frangitque quatitque.
Totum corpus habet sudor, multumque laborat cor. (su. ha.
MSS.)
Nee respirandi fit copia : praepete ferro
Histri tela manu iacientes sollicitabant.

Ennius' Elegiac Verse is so scanty and usually of so doubtful authen-

ticity that we content ourselves with the remark that the * poeta egregius '
seems entitled to despise the ' cantores Euphorionis ' with their :
(Catull. 73, 6) Quam modo qui me unum atque unicum amicum
habuit.

The homely hexameters of his manual for cooks (and their masters)
have already been treated (I 2).
3ro EARLY LATIN METRES

39. Our judgement of Lucilius' hexameters is coloured by Horace's

verdict "fluit lutulentus'. But Horace's verdict is itself coloured by
party politics. Valerius Cato, the famous professor of poetry who
founded the New School of poets, planned in his later years an edition
of Lucilius (entitled ' Lucilius emendatus ' or the like) and took occasion
for a contrast between the free poetry of the Republic and the servile
poetry of Augustus' court. Horace replied in the opening verses of
Sat. I x (verses discarded in the second edition of the poem), making
rather a poor stroke at the word ' emended ' :
Lucili, quam sis mendosus, teste Catone
Defensore tuo, pervincam, qui male factos
Emendare parat versus, etc.

(See Hendrickson's articles in Class. Phil. 12, 77 and 329.) Lucilius,

whose niece married Pompeius Strabo, the father of Pompey the Great,
and whose poems were edited by grammarians of Pompey's circle (e.g.
Pompeius Lenaeus), would be a ' b£te noire ' to the Caesarians (see
Marx's preface).
Doubtless Lucilius' satiric hexameter fell as far short of Horace's as
Ennius' epic hexameter of Virgil's, but, for a first attempt, it was by no
means contemptible. It is designedly careless, rambling verse, imita-
tive of conversation (sermo) ; and yet* there is much art artfully con-
cealed. How apt is this picture of gladiators (149 sqq.) :
Aeserninu' fuit Flaccorum munere (?) quidam
Samnis, spurcus homo, vita ilia dignu' locoque :
Cum Pacideiano componitur, optimu' multo
Post homines natos gladiator qui fuit unus.

' Occidam ilium equidem et vincam, si id quaeritis ' inquit.

' Verum illud credo fore : in os prius accipiam ipse
Quam gladium in stomacho furia (?) ac pulmonibu' sisto :
Odi hominem, iratus pugno, nee longiu' quicquam
Nobis quam dextrae gladium dum accommodet alter ;
Usque adeo, studio atque odio illi(u)s, efferor ira.'
Clearly the verse is the servant, not the master, of one who can write
like this. Or take the picture of a superstitious man (484 sqq.) :
Terriculas, Lamias, Fauni quas Pompiliique
Instituere Numae, tremit has, hie omnia ponit.
Ut pueri infantes credunt signa omnia aena
Vivere et esse homines, sic isti somnia ficta
Vera putant, credunt signis cor inesse in aenis.
 EARLY LATIN METRES 311

40. We are quite sceptical about Dactylics in Plautus or Terence

and have included the few claimants in our paragraph on Glyconic
Metre (32).
If no apology is needed for the presence of Dactylics in Roman
Tragedy, it may be far-fetched to associate with Aristophanes' use of
the Hexameter for oracles Ennius' use of the Tetrameter (unless it is
rather a dactylic ' run ') for Cassandra's predictions (trag. 50) :
Iamque mari magno classis cita
Texitur : exitium examen rapit :
Adveniet, fera velivolantibu' (Or -bus)
Navibu' complebit manu' litora.
No such association is possible for Accius' Antigona fragment (trag.
140), with its lively movement :
Heus, vigiles, properate, expergite
Pectora tarda sopore, exsurgite !
We can hardly set aside the express testimony of two Metricians of
the Empire and (with Scaliger) assign to Laevius (the predecessor of
the ' novi poetae ') the alternate hexameters and c miuric ' hexameters
of the hymn to Diana cited from Livius Andronicus' Ino, although the
second line in particular hardly suggests to us the Latin of that early
time so well as the fourth (trag. 28) :
Sed iam purpureo suras include cothurno
Balteus et revocet volucres in pectore sinus,
Pressaque iam gravida crepitent tibi terga pharetra.
Derige odorisequos ad certa cubilia canes.
(The Hellenistic miuric hexameters of Oxyr. Papyr. 15 + 1795 are m
four-line stanzas, each line being miuric.)

41. Dochmius. In discussing Plautus' song-metres we have tried to keep

aloof from statements which might be challenged. Some debatable details
we will now touch, and refer the reader to Leo's Plautinische Cantica for a
thorough handling of them (and others).
The Dochmius, the metre that thrills, what place can there be for it in
Comedy ? Yet Aristophanes uses it, e.g. Birds 11 88 sqq. :
HoXenos aiptTai, \\ noXepos ov (paros
Upos efie k<u Oeovs. \\ aXXa (pvXarre nas, etc.

White (p. 206) makes his usual forms w w o — w — and — ^ w — ^ — . The

papyrus published by Grenfeli (An Alexandrian Erotic Fragment, Oxford,
1896) was shown by Wilamowitz to be a love-song, with line after line con>
posed of Dochmii :
Me'XXa) fiaipcadai

ZrjXos yap /x' %X€l) e*C»

3i2 early latin Metres
* Oh ! I shall go mad. I am a prey to jealousy.' And Wilamowitz showed
what light this Hellenistic composition threw on Plautus' Cantica. Its
* stichic' use of Dochmii resembles Plautus' 'stichic ' use of Bacchiacs, of the
Versus Reizianus, etc.
The cola used by Plautus have not yet been all classified or even detected.
But it is doubtful whether the Dochmius has a place among them.
42. Syncopated Feet. The suppression of a short syllable before a long,
the voice dwelling for extra time on the long, is usually expressed by putting
the symbol of short quantity above the line ; e. g. Birds 851 sqq. (two ' synco-
pated Iambic
' Dimeters Acatalectic) :
'Ofxoppodco, ^(jvvOekat,
u aviA7rapaiP€(ra$ e^co.

Some use the term ' syncope ', some prefer ' contraction ' (though this term is
used also of the substitution of a long syllable for two short, e.g. for
ww or for w u, as in Amph. 169 Quod fact(o) aut, a major Ionic;
35). As good a word as any is ' protraction \
We have seen (20) that the Ithyphallic Colon has been referred to a
Trochaic Dimeter Catalectic with 'protracted' third foot, e.g. olros ovv 6
Ti^p,a>v. Some would even refer Bacchiacs to a ' protracted ' type of Iambic
Verse ; and certainly this suits their Plautine use, as vehicle of ' sermo
gravis ' (22).
There are one or two lines in Plautus that strongly suggest 'protrac- vias ;
tion ',e. g. :
Pers. 1 (Iamb.) Qui amans egens ingressus est || ^princeps in Amoris
tam ;
Cure. 103 (Iamb.) Nam ubi tu profusu's, ibi ego me || wpervelim sepul-

Rud. 945-6 (Iamb.) Cave sis malo. quid tu, malum, || wnam me retrahis
Audi. ?

Non audio. At pol qui audies || wpost. Ouin

loquere quid vis ;
Cas. 168 (Iamb.) Nam ubi domi wsola sum, || sopor manus ^calvitur
(Or Cretic Tetrameter, with sopor) ;
? Trin. 250 (Iamb.) Quod ebibit, || wquod comest, etc.
But, if they really are ' protracted ' verse, are they not already-formed types
(like the Ithyphallic Colon) which Plautus transferred, as he found them,
from his Greek originals ? For we have no such wide-spread traces of ' pro-
traction* inPlautus as we have in Aristophanes or in Greek Tragedy.
43. Responsion and Symmetry. After the discovery of the papyrus Men-
ander fragments with their entries xopoy (like the entries xopoy in Aristo-
phanes' MSS. at Eccl. 730, 877 ?, Plut. 322, 627, 802, 959, 1097, and xopoy
KOMMATION at Plut. 771) no one can pooh-pooh the idea that the fishermen's
song in the Rudens was an actual chorus-song, Rud. 290 sqq. (Iamb.) :
Omnibu' modis qui pauperes || sunt homines miseri vivunt, etc.
 EARLY LATIN METRES 313

But it is hard to believe that the fishermen (if all sang, and not merely the
leader) were divided into two companies. There is no trace of strophe and
antistrophe, no ' responsion '. The supposed strophe and antistrophe in the
soldiers' chorus in Ennius' adaptation of Euripides1 Iphigenia at Aulis (it is
a chorus of women in the Greek original) vanishes now that Leo (Trag. Rom.
p. 15) has restored the text.
Plautus is fond of ' responsion ', or rather symmetry, in duets. The examples
quoted in paragraphs 11, 18, 31 show how he likes to make a reply the exact
counterpart of a question, how in the duel of wits stroke is exactly matched
by counterstroke, cut by parry ; e.g. Persa 216 sqq. (Troch. Sept.) :
So. Quoagis? Pa. Quo tu ?
So. Die tu. <Pa. Die tu.> || So. PriSr
rogavi. Pa. At post scies.
So. Eo ego hinc haud longe. Pa. Et quidem ego (eo) haud ||longe, etc.
Pa. Itane est ? So. Itane est. Pa. Mala's. So. Scelestu's. || Pa. Decet
me. So. Me quidem haud decet.
Pa. Quid ais ? certumnest celare || quo iter facias, pessima ?
So. Offirmastin occultare || quo te immittas, pessime ?
Pa. Par pari respondes dicto, etc.
That is a feature of all lively Comedy, English as well as Latin.
But can we make a law of it ? The question comes home to an editor at
the opening of the Persa. Toxilus' six lines exactly correspond to Sagaristio's
six lines, except in one small detail. Sagaristio's first line is an ordinary
Iambic Septenarius :

Qui ero suo servire vult || bene servu' servitutem,

whereas Toxilus' first line (discussed in last paragraph) is an Iambic Octo-
narius with ' protraction ' :
Qui amans egens ingressus est || ^princeps in Amoris vias.
Some editors would 'emend' it into a Septenarius without 'protraction',
although not only A and P but also Donatus (if the Virgil scholium be his)
attest the traditional reading. This seems hypercriticism. The same
reasoning would force qui modus of Asin. 167 (Troch.) into the exact form of
quid modist in line 169 :
Arc Qui modus dandi ? nam numquam || tu quidem expleri potes ;
Cl. Quid modist ductando, amando ? || numquamne expleri potes?

44. Analysis of some Cantica. (A) Stich. 1 sqq. (cf. Classical Review,
August, 1 91 8). On the harp, the violin, flute the same theme is repeated
with variations. The lyric choruses of the Greek Drama echo in words this
mannerism of music. Take any ' Glyconic ' (i.e. Dactylo-trochaic) strophe,
e. g. Eurip. Ale. 962 ('Eyw *a\ 81a fiovaas, etc.) : we find Glyconics proper,
catalectic Glyconics (i. e. Pherecrateans), Glyconics with a syllable added
(989 Km Becov otkotlol <fi&ivovai), Glyconic colons of the Maecenas atav.s type
(990 Halves iv 6avaT(o) or the Edite regibus type, and so on. As a butterfly
flutters round a flower, recurring to the petals vvith new and ever new move-
 3r4 EARLY LATIN METRES

ment, so the poet handles,. leisurely and lovingly, one after another of the
metre's many forms.
The Stichus opens with a duet of the two heroines, the sisters Panegyris
and Pamphila, by way of prologue. The keynote is given by the first line, a
Glyconic colon (— ^ — ^ ^ — ) like Horace's Maecenas atavis, Euripides'
Tiaibis iv Oavara '.
■ i. Pan. Credo ego miseram.
The next four lines have the same metre, but with Anacrusis, the Anacrusis
being a short syllable in three of the lines, a long in one :
2. Fu-isse Penelopam,
3. So-ror, suo ex animo,
4. Quae tarn diu vidua
5. Vi-ro suo caruit.
The sixth line is an acephalous Pherecratean (i.e. a Colon Reizianum, 15) :
6. Nam nos ei(u)s animum (Emphasis on nos).
The next three give the plot of the play in Versus Reiziani (i.e. Iambic
Dimeters with the Colon Reizianum appended ; 14) :
7. De nostris factis noscimus || quarum viri hinc absunt,
8. Quorumque nos negotiis || absentum, ita ut aequumst,
9. Sollicitae noctes et dies, || soror, sumu' semper.
Two Anapaestic Monometers follow (a prelude to the Anapaestic concluding
portion of the Canticum) :
10. Pam. Nostrum oftkium
11. Nos facere aequumst.
Then two lines in the metre of lines 2-5 :
12. Ne-que id magis facimus.
13. Quam nos monet pietas.
Then a conversational line, a Versus Reizianus :
14. Sed hie, soror, adside dum. |[ multa volo tecum.
In a 'Glyconic' chorus of the Hippolytus which begins with a Glyconic
proper with Anacrusis (525 "E-pas, "Epcos, 6 <ar oynmrav), we find a rare
variety (S33"Iwtv i< xeP^u) followed by a Colon Reizianum (534 "Epcos, 6 Aibs
nois). This variety, apparently the metre of lines 2-5, 12-13 witn a Trochee
(or Spondee) substituted for the Dactyl, makes line 15, and is followed (as
in Euripides) by a Colon Reizianum in line 16. The two components are
combined into one verse in the four succeeding lines. By keeping them apart
at their first occurrence a hint is given to the reader that a novel type con-
fronts him :
15. Lo-qui de re viri.
16. Pan. Salvene, amabo ?
17. Pam. Spe-ro quidem et volo ; || sed hoc, soror, crucior
18. Pa-trem tuum meumq(ue) || adeo, unice qui anus
19. Ci-vibus ex omnibus || probus perhibetur,
20. E-um nunc improbi || viri officio uti.
 EARLY LATIN METRES 315
Not merely the Dactyl, but also the Trochee, is effaced in the first half of
the next line (whose second half is a Colon Reizianum), an effacement seen
in Eurip. Ion 501 (followed by a Colon Reizianum) :
2u-pi£eir, & Havi
Tols (Toicriv iv avrpois.

Plautus' Anacrusis is a short syllable :

21. Vi-ris qui tantas || absentibu' nostris.
Finally two variations of the Glyconic proper (cf. Eur. El. 125 "l9t tov avrbv
eycipe yoov, Orest. 83 1-2 Tis voaos t) Tiva daKpva koi Tis ZXeos peifav Kara
yav;) :
22. Facit iniurias immerito
23. Nosque ab eis abducere vult.
Plautus now abandons lyrics and writes the rest of the Canticum in Anapae-
stics. But in two lines, one near the beginning, the other near the end of
this second part, he recalls an echo of the * Glyconic ' portion. Their first
half is ' Glyconic ', their second Anapaestic :
Fac quod tibi tuus pater |j facere minatur,
and
Ne quid magi' sit omnibus || obnixe opibus.
(These two support and defend each other against all the critics and show us
that a passage may not always be so uniform as it seems.)
The time is not ripe for an analysis of all the Plautine Cantica. But this
specimen will suffice to convince our readers that Plautus was a far more
artistic metrician than is often supposed.
(B) Caecilius' Plocium. Aulus Gellius (Noct. Att. 2, 23) has preserved for
us a passage (in Iambic Trimeters) of Menander's Plocium and the Canticum
which Caecilius in his Plocium made out of it :
in dftfpoTepav piv fj ViKXj;pos 17 KaXtj
peXXa KaOevdrjaetv. Kareipyavrai peya

Kai 7repi(36r)TOU epyov' ck Trjs oiKias

e£e/3a\e rqv Xvrrovcrav tjv e&ovXero,
Xv* anofiXeiTGKri ndvres els to KpafivXys
7rp6aa>nop # t evyvaxrTos ova eprj yvvq
beairotva. Ka\ ttjv o\l/iv r\v eKTrjcraro
ovos €V 7TidT)K0ts tovto 8q to Xeyopevov
eo~Tiv. oiairdv fiovXoptu Trjv vvKTa Tqv
7roXXa>v KaKotv dpxrjyov. oipoi Kpa>(3vXr)v
Xafieiv ep €KKai8eKa TaXavra

tt)V piv exovaav nr)Xea)S' eir' eo"r>l TO

<f>pvaypd 7ra>s viroarTarov • pa tov Ata
tov 'OXvpmov Kai tt)V fA&r)vav, ovdapcos.
iraibiaKapiov depanevriKov 8e tov 86pov
BdrcTov dndyoi tis rj Wepav av elardyoi.
 EARLY LATIN METRES
316
Is demum miser est qui aerumnam || suam nesclt occultate (?) (Anap.)
Ferre : ita me uxor forma et factis || facit, si taceam, tamen indicium, (do.)
Quae nisi dotem omnia quae nolis || habet : qui sapiet de me discet, (do.)
Qui quasi ad hostes captu' liber || servio salva urbe atque arce (Troch.)
Dum ei(u)s mortem inhi5, egomet inter || vivos vivo mortuus. (do.)
Quaen mi quidquid placet eo privatum it me, servatum (earn) ? (do.)
Ea me clam se cum mea ancilla || ait consuetum. id me arguit : (do.)
Ita plorando, orando, instando at||que obiurgando me obtudit (do.)
Earn uti venderem. nunc || credo inter suas (Bacch.)
Aequales, cognatas || sermonem serit (do.)
' Quis nostrarum fuit || integra aetatula (Cret.)
Quae hoc idem a viro (do.)
Impetrarit suo || quod ego anus modo (do.)
Effeci, paelice ut meum privarem virum ? ' (Iamb.)
Haec erunt concilia hodie : || differar sermone miser. (' Glyconic ')
 APPENDIX
A. REGARD FOR ACCENT
age age :
Asin. 40 (Sen.) Fiat, geratur mos tibi. Age age, usque exscrea ;
327 (Troch.) || ducere. Age age, mansero ;
Epid. 631 (Troch.) Age age, absolve (me) atque argentum || (numera) ;
Mil. 1024 (Anap.) Age age ut tibi maxime concinnumst ||;
Pers. 606 (Troch.) || age age nunc tu, in proelium ;
766a (Anap.) Mutua fiunt || a me. age age [age] ergo ;
Poen. 717 (Sen.) Age, eamus intro. Te sequor. Age age, ambula.
Therefore do not alter to (age) age :
Stich. 221 (Sen.) Logos ridiculos vendo. age licemini (With Hiatus ' in
Ter. Heaut. 722 (Troch.) Age, age, traducatur Bacchis ||; pausa ') ;
Ph. 559 (Troch.) Age, age, inventas reddam. O lepidum ! ||;
662 (Sen.) Ob decern minas est. Age age, iam ducat : dabo ;
Ad. 8yy (Troch.) Age age, nunciam experiamur ||.
Therefore scan :
Ter. Andr. 310 (Iamb.) Tu si hie sis, aliter sentias. || Age age, ut libet.
Sed Pamphilum,
with an Anapaest in the fifth foot (not a Tribrach).

age si quid agis :

Epid. 196 (Troch.) Age si quid agis. di immortales ! ||;
Mil. 215 (Troch.) Habet, opinor. age si quid agis || ;
Pers. 659 (Troch.) Age si quid agis. ego ad hunc redeo ||;
Trin. 981 (Troch.) Age si quid agis. Quid ego again ? Aurum || redde.
Dormitas, senex.
Therefore emend :
Stich. 715 (Troch.) Bibe, tibicen. age si quid [agis] bi||bendum, hercle hoc
est; ne nega (II 9, note).
Eug(ae) eugae :
Aul. 677 (Sen.) Eugae eugae ! di me salvum et servatum volunt ;
Epid. 493 (Sen.) Eugae eugae ! Epidice, frugi's, pugnasti — , homo es ;
Mil. 241 (Troch.) Dicam hospitio. Eugae eugae ! lepide || ;
Pers. 462 (Sen.) Numquid moror. Eugae eugae ! exornatu's basilice
(Clash)
Rud. 164 (Sen.) Ut afflictantur miserae ! eugae eugae ! perbene ;
Stich. 766 (Sen.) Dare amicum amicae. eugae eugae ! sic furi datur ;
Trin. 705 (Troch.) Non enmi possum quin exclamem eu||gae eugae ! Lysi-
teles, ndXiv !
3i8 APPENDIX A
hdM bon{tini)-animum :
Capt. 152 (Sen.) Nunc habe bonum animum. Eheu ! huic illud dolet
(Hiatus at change of speaker and after Interjection) ;
Epid. 512 (Troch.) Habe bonum animum. Quippe ego quoi li||bertas in
mundo sitast ;
True. 525 (Troch.) Habe bonum animum. Savium, sis ||;
So scan :
Mil. 1325 (Troch.) Quom te video. Habe bonum animum ||(With Hiatus
at change of speaker) ;
Most. 387 (Troch.) Perii ! Habe bonum animum, etc. (Ditto).

keti me-miserum t :
Aul. 721 (Anap.) Heu me miserum ! || misere perii (Clash);
Merc. 624 (Troch.) Mulier. heu me miserum ! Flere o||mitte, istuc quod ;
nunc agis

Ter. Andr. 646 (Troch.) Heu me miserum, qui tuum animum ex || animo
spectavi meo ! ;
Hec. 271 (Iamb.) ||hem, Sostrata. Heu me miseram ! ;
Ennius trag. 164 (Troch.) Heu me miseram ! interii. pergunt || lavere
sanguen sanguine ;
Pacuvius 264 (Anap.) Nudate. heu me || miserum ! excrucior (Clash. But
mis. me var. led.) ;
Accius trag. 346 (Troch.) Heu me miserum ! cum haec recordor ||;
Afranius 409 (Troch.) Heu me miserum ! dum modo doleat ||.
lest ;
hie ille est :
Capt. 518 (Iamb.) Hie ille est dies cum nulla vi||tae meae salus sperabi-

787 (Bacch.) Hie ille est senex doc||tu' quoi verba data sunt ;
Epid. 621 (Troch.) Hie est danista, haec ilia est autem || quam emi de
praeda. Haec inest ? ;
Merc. 268 (Sen.) Nunc hoc profecto sic est : haec illast capra ;
Mil. pr. 155 (Sen.) Ipse exit : hie illest lepidu' quern dixi senem ;
1046 (Anap.) Quin tu huic responde, haec illaec est || ab ilia quam
dudum dixi ;
Most. 162 (Iamb.) Haec ilia est tempestas mea ||;
Pers. 545 (Troch.) Haecin illast furtiva virgo ? ||;
Rud. 771 (Troch.) Cum coniecturam egomet mecum || facio, haec simia illast;

Stich. 196 (Sen.) Hie illest parasitus quern arcessitum missa sum;
Trin. 43 (Sen.) Hie illest senecta aetate qui factust puer;
Ter. Andr. 126 (Sen.) Percussit ilico animum. attat ! hoc illud est,
Hinc illae lacrimae, haec illast misericordia.
The balance of probability therefore inclines to the scansion in Capt. 518 Hie
ille est.
 REGARD FOR ACCENT 319

hoc est (hdc erai) ' that's why ' :

Asin. 863 (Troch.) Hoc ecastor est quod ille it || ad cenam quotidie ;
Cas. 531 (Troch.) Hoc erat ecastor quod me vir || tanto opere orabat meus ;
Men. 1 135 (Troch.) Hoc erat quod haec te meretrix || hum' nomine vocabat ;

Merc. 711 (Sen.) Pol hoc est ire quod rus meu' vir noluit.
mdnc mane ' wait a bit ' :
Asin. 229 (Troch.) Mane man(e), audi, die quid me aequum || censes pro
ilia tibi dare ? ;
Aul. 655 (Troch.) Mane mane, quis illic est ? quis hie intus ? ||;
Men. 179 (Troch.) Mane man(e), obsecro hercle : eapse ec||cam exit, oh !
solem vides ;
Merc. 474 (Troch.) Mane man(e), obsecro, Charine ||.
(Cf. Cist. 704.)
me-miserum or miserum-me :
Amph. 160 (Troch.) Ita quasi incudem me miserum || homines octo validi;
caedant

1056 (Iamb.) || me miseram ! quid agam nescio ;

Aul. 42 (Sen.) Nam cur me miseram verberas ? Ut misera sis ;
69 (Sen.) Queo comminisci : ita me miseram ad hunc modum (With
Hiatus at pause ; cf. Ill 42 ita) ;
409 (Troch.) Ita me miserum et mebs discipulos || fustibus male derunt
contu- ;
Capt. 502 (Troch.) Ita me miserum restitando ||;
Merc. 893 (Troch.) Quid taces ? dice, enicas me ||miserum tua reticentia ;
Most. 739 (Cret.) Tuto in terra. Ei ! Quid est ? || Me miserum, occidi
(Clash) ;
Rud. 682 (Iamb.) Desiste dictis nunciam || miseram me consolari ;
True. 119 (Iamb.) Enicas me mise||ram quisquis es;
Ter. Andr. 240 (Iamb.) Miseram me ? quod || verbum audio I ;
251 (Troch.) Itur ad me. Oratio haec me || miseram exani- ;
mavit metu
761 (Sen.) Di te eradicent : ita me miseram territas ;
788 (Sen.) Me miseram ! nil pol falsi dixi, mi senex ;
882 (Sen.) Me miserum ! Hem modone id demum sensti,
Pamphile ? ;
Eun. 71 (Sen.) Et lllam scelestam esse et me miserum sentio (Emphasis
on seel, and mis. ; 1 1 60) ;
81 (Sen.) Miseram me ! vereor ne illud gravius Phaedria ;
197 (Sen.) Me miseram ! forsan hie mi parvum habeat fidem ; frui ! ;
Heaut. 401 (Iamb.) Hoccin me miserum non lice||re meo modo ingenium

Ph. 749 (Iamb.) Ubi illae sunt ? Miseram me. Hem, quid est ? || (Clash)
Hem ;;
Hec. 205 (Sen) Me miseram! quae nunc quam 6b rem accuser nescio.
32o APPENDIX A

Ad. 291 (Iamb.) Miseram me ! neminem habeo, so||lae surau'hie; Geta autem ;
non adest
305 (Iamb.) Me miseram ! quidnamst quod sic vide||o timidum et pro-
perantem Getam ? ;
cundia ;
310 (Iamb.) Me miserum ? vix sum compos ani||mi, ita ardeo ira-

330 (Iamb.) Me miseram ! quid iam credas ? aut || cui credas ? nost-
rumne Aeschinum ;
486 (Sen.) Scio. Miseram me ! differor doloribus.
Therefore in Plaut. Amph. 897 (Sen.), if a syllable has been omitted, it
must be added after (not before) me miseram :
Sed eccam video qui me miseram arguit.

misir-sum ' I'm out of luck ' (never ' miser sum ') :
Amph. 584 (Troch.) Ut minu' valeas || et miser sis ;
Capt. 993 sq. (Troch.) Et miser sum et fortunatus. ||
E6 miser sum quia male illi ||feci, si gnatus bessit)
meust ;
Pseud. 13 (Sen.) Misere miser sum, Pseudole. Id te Iuppiter (Prohi-
80 (Sen.) Miser sum, argentum nusquam invenio mutuum ;
299 (Troch.) Nimi' miser sum, nummum nusquam || reperire argenti

quid est quod, quid istuc est quod, quid hoc est quod : queo ;
Amph. 502 (Troch.) Quid istuc est, mi vir, negoti || quod tu tamdomo ?;
subito

Bacch. 1 156 (Anap.) Quid est quod pudeat ? Sed amico homini ||;
Capt. 541 (Troch.) Quid istuc est quod mebs te dicam || fugitare oculos,
Tyndare ? ;
Cas. 179 (Anap.) Sed quid est quod tub || nunc animo aegrest ? ;
Cist. 655 (Troch.) Sed eccam eram video, sed quid hoc est || haec quod
cistella hie iacet ? ;
Epid. 560 (Troch.) Mulierem retines. Quid est quod || vultu5
tuus ? ;te(Clash)
turbat

Most. 69 (Sen.) Quid est quod tu me nunc obtuere, furcifer ? ;

1062 (Troch.) Sed quid hoc est quod fori' concrepuit ? ||;
Poen. 867 (Troch.) Quid est quod male sit tibi quoi domi sit ? ||;
884 (Troch.) Quid est quod metuas ? Dum ero insidias ||;
Pseud. 9 (Sen.) Quid est quod tu exanimatu' iam hos multos dies ? (AP) ;
(Clash)
True. 295 (Troch.) Pessimae estis. Quid est quod vobis || pessimae hae
male fecerunt ?
quid hoc clamoris t :
Aul. 403 (Sen.) Sed quid hoc clamoris oritur hinc ex proximo ? ;
Cas. 620 (Sen.) Quid Illuc clamoris, obsecro, in nostrast domo ? ;
Trin. 1093 (Sen.) Quid hoc hie clamoris audio ante aedes meas ?
 REGARD FOR ACCENT

quid istuc vtrbist (est verbi) ? ' what do you mean ? ' :
Cist. 604 (Sen.) Quid istuc est verb! ? Ex priore muliere ;
Cure. 31 (Sen.) Quid istuc est verbi ? Caute ut incedas via ;
Epid. 350 (Iamb.) Quid Istuc est verbi ? Nil moror ||;
321
Pseud. 608 (Troch.) Quid Istuc verbist ? Condu', promu' || sum, procurator
Ter. Ph. 343 (Troch.) Quid Istuc verbi est ? Ubi tu dubites ||.
quid opust verbis ? ' I need hardly say ' : peni.
Amph. 615 (Troch.) Qua ego sum. Quid opust verbis ? geminus ||;
Aul. 468 (Troch.) Circumcirca. quid opust verbis ? ||;
472 (Troch.) Quid opust verbis ? facta est pugna ||;
Bacch. 486 (Troch.) Quid opust verbis ? si opperiri ||;
1 164 (Anap.) Quia flagitium est. Quid opust verbis ? ||; (Clash)
Capt. 937 (Troch.) Quid opust verbis ? lingua nullast ||;
tudine ;
Cist. 94 (Troch.) Ut ego Alum versarem ? Quid opust || verbis ? consue-
Cure. 79 (Sen.) Solet esse. Quid opust verbis ? vinosissima est ;
Mil. 12 13 (Troch.) Quid opust verbis ? libertatem ||;
Most. 993 (Sen.) Perii hercle ! quid opust verbis ? ut verba audio ;
Poen. 436 (Sen.) Neque hercle vero — quid opust verbis ? quippini ;
579 (Troch.) Quid opust verbis ? callum aprugnum ;
Rud. 590 (Troch.) Quid opust verbis ? si invitare ||;
Ter. Andr. 165 (Sen.) Sed quid opust verbis ? sin eveniat quod volo.
sine modo (and sine modo ?), in threats, etc. :
Amph. 806 (Troch.) Sine modo argumenta dicat ||;
Cas. 437 (Sen.) Sine modo rus veniat : ego remittam ad te virum;
Cure. 655 (Sen.) Quae isti committas. Sine modo. Pro Iuppiter ! ;
Most. 11-12 (Sen.) Quia vivis. Patiar. sine modo adveniat senex.
Sine modo venire salvum quern absentem comes ;
Pseud. 222 (Troch.) Tu devincis. || sine modo ;
(Cf. Asin. 897 ; Poen. 11 46.)
! sine
Ter. Eun. 65 (Sen.) Egon illam, quae ilium, quae me, quae non — modo.

tib{i) ego dico ' d'ye hear ? '

nate ;!
dico, heus
Pseud. 243 (Troch.) Hodie nate, heus ! hodie nate, || tibi ego hodie

Cure. 516 (Iamb.) Numquid vis? Bene vale. Vale. || Heus (heus) !^ tu,
tibi ego dico ;
Mil. 217 (Troch.) Tibi ego dico, etc. ;
434 (Troch.) Tibi ego dico, heus ! Philocomasium ||.
tutt-tibi :
Capt. 371 (Sen.) Tute tibi tubpte ingenio prodes plurimum ;
Cist. 563 (Sen.) Tute tibi indigne dotem quaeras corpore ;
Cure. 9 (Sen.) Tute tibi puer es, lautus luces cereum ;
2348 Y
322 APPENDIX A

Men. in (?) || tute tibi odio habeas (AP) ;

Pers. 573 (Troch.) Ferreas tute tibi impingi || iubeas crassas compedes ;
True. 334 (Sen.) Tute tibi mille passum peperisti morae.

vaS capiti-tuo :
Amph. 741 (Troch.) Vae capiti tuo! Tua istuc ||;
Cure. 314 (Troch.) Vae capiti tuo ! Obsecro hercle ||;
Men. 512 (Sen.) Exire vidi pallam. Vae capiti tuo ;
840 (Troch.) Vae capiti tuo ! Ecce Apollo ||;
Mil. 326 (Troch.) Vae capiti tuo ! Tuo istuc ||;
Most. 1002 (Sen.) Modo eum vixisse aiebant. Vae capiti tuo ! ;
Contrast Rud. 375 (Iamb.), where the Noun is emphasized :
Vae capiti atque aetati tuae ! ||.

vat miserd-mi(hi) /, and the like : mihi ! ;

Amph. 1057 (Iamb.) Ita tanta mira in aedibus || sunt facta, vae miserae
mihi ! ;
Aul. 200 (Troch.) De communi re appellare || mea et tua. Ei misero

Capt. 945 (Troch.) || verba. Vae misero mihi ! ;

Cas. 574 (Sen.) Sed uxorem ante aedes eccam. ei misero mihi ! (With
Hiatus at pause) ;
661 (Anap.) Habet. Ei misero || mihi ! cur eum habet? ;
848 (Sen.) Edepol papillam bellulam ! ei misero mihi ! ;
Merc. 181 (Troch.) || vidit ? vae misero mihi ! hoc ;
661 (Troch.) || abiit. heu misero mihi ! ;
681 (Sen.) Disperii, perii misera. vae miserae mihi ! ;
701 (Sen.) Tali viro quae nupserim. heu miserae mihi ! ;
708 (Sen.) Accedam propius. Vae miserae mi[hi] ! Immo mihi ;
759 (Sen.) Immo hercle tu istic ipsus. Vae misero mihi ! ;
770 (Sen.) Cras petito ; dabitur. nunc abi. Heu miserae mihi ! ;
792 (Sen.) Perii hercle ! ecce autem haec abiit. vae misero mihi ! ;
Mil. 1433 (Troch.) Osculari atque amplexari in||ter se. Vae misero mihi ! ;
Most. 549 (Sen.) Dixi hercle vero — omnia. Ei misero mihi ! :
Poen. 1379 (Sen.) eas aliquis, quod nunc factumst. Vae misero mihimihi
! ; !;
Trin. 907 (Troch.) Libet audire. 1111 edepol — 1111 — || ill! — vae misero
True. 342 (Sen.) Ut rem servare suave est ! vae misero mihi ! ;
mihi ! ;
794 (Troch.) Conveniunt adhuc utriusque || verba. Vae misero

Ter. Andr. 302 %(Troch.) Qui scis ? Apud forum modo e Davo aul|divi.
Vae misero mihi ! ;
Heaut. 234 (Iamb.) Cui nil iam praeter pretium dul||cest. Clinia. Ei
misero mihi ! ;
250 (Troch.) Propera: quid stas ? Vae misero mi[hi], || quanta
de spe decidi ! ;
 REGARD FOR ACCENT 323

917 (Sen.) Ni essem lapis, quae vidi ! vae misero mihi ! ;

Hec. 605 (Iamb.) Mi gnate, da veniam hanc mi, red||duc illam. Vae
misero mihi ! ;
Ad. 173 (Iamb.) O facinus indignum ! Gemina||bit, ni caves. Ei mihi
misero! ;

301 (Iamb.) || erili est. vae misero mihi ! ;

327 (Troch.) || aliam. Vae miserae mihi ! ;
383 (Sen.) Fore, si perdiderit gnatum. vae misero mihi !

utfi6te-qni :
Bacch. 511 (Sen.) Amo hercle, opino, ut pote quod pro certo sciam ;
Cist. 317 (Iamb.) (Suspiciost earn esse) Ut pote quam numquam viderim ||;
Mil. 529 (Sen.) Magi'que eandem, ut pote quae non sit eadem, non reor ;
But:
Rud. 462 (Sen.) Satin nequam sum, ut pote qui hodie amare inceperim
(Emphatic hodie).

B. DIAERESIS
Iambic Septenarii (over 1300 in the twenty-one plays of Plautus).
(A) Without (Iambic) Diaeresis :
(a) With Spondee or Anapaest in Fourth Foot :
Asin. 556 Id virtute huiu' collegae ||meaque comitate (Or -gai meaque,
hardly meaque ; II 18 C ; II 32) ;
Cist. 312 Nimi' lepide exconcinnavIthas||ce aedes Alcesimarchus (III 18 ;
Hardly Fut., exconcinnabit) ;
Cure. 125 Omnes, mihi hau saepe eveniunt ||tales hereditates (Emended
to veniunt ; II 54, [2]). (On Cure. 124 see below) ;
508 Vos faenorl, hi male suaden||do et lustris lacerant homines
(Emended to Vos fae. hom., hi) ;
Mil. 361 Respicedum ad laevam : quis lllaec est||mulier? Pro di im-
mortales ! (Emended to illaec quis est) ;
1 23 1 Spero ita futurum, quamquam ilium ||multae sibi expetessunt
(III 33) (Emended to eum) ;
Pers. 319 Enim metuo ut possi(e)m in bubi||le reicere, ne vagentur
(Emended to bovile. But cf. II 52) ;
Poen. 1 23 1 Sed lllud quidem volui dicere — immo herc||le dixi quod vole-
bam (im. dixi he. P; II 49, end; III 58) ;
1245 Et praedicabo quo modS vos ||furta faciatis multa (AP)
(Hardly quomodo; II 37. Perhaps furtificetis) ;
1265 Nam vostra nutrix primum me ||cognovit. Ubi est
ea,? amabo,
(AP) ;

Pseud. 1320 Heu heu heu! Desine. Doleo. Ni ||doleres tu, ego dole-
rem (Heu heu de. A) (II 52);
Y 2
324 APPENDIXB

Iambic Septenarii. (A) Without (Iambic) Diaeresis (con/.)- -

(a) With Spondee or Anapaest in Fourth Foot (cont.) —
Rud. 318 Tortis superciliis, contrac||ta fronte, fraudulentum ;
386 Sed duce me ad lllam ubi est. I sa||ne in Veneri' fanum hue
intro (Emended to I sane — Ubi est ?) ;
700 Ne indignum id habeas neve idcir||co nobis vitio vertas.
(b) With Iambus in Fourth Foot :
hercle ;
Asin. 422 Quin centiens eadem imperem at||que ogganniam, itaque iam

430 Dormitis interea domi at||que erus in hara, haud aedibus

tat (cf. II habi-
52) ;
432 Mea causa ut mittas. Eho, ec||quis pro vectura olivi ;
469 Nemo accipit, te aufer domum abs||cede hinc,(aufer
molestu' ne sis;
te MSS.)
473 Flagitiilm hominis, da, obsecro ar||gentum huic, ne male
loquatur ;
492 Merito meo, neque me(d) Athe||nis alter est hodie quisquam
(Rather al. hodiest) ;
502 Fortasse. Atque etiam tu quoque ip||se, si esses percontatus
(Hardly ips', ||si esses ; II 36) ;
545 Perfidiae laudes gratias||que habemu' merito magnas ;
577 Ut meque teque maxime at||que ingenio nostro decuit ;
585 Manedum. Quid est? Philaenium est||ne haec quae intus
exit atque (Una Argyrippus ?) ;
690 Libertum potiu' quam patro||num onus in via portare ;
733 Argentum ad te. Ut temperi op||portuneque attulistis ! ;
Cure. 124 Nam tibi amantes propitian||tes vinum dant potantes (propi-
nan||tes MSS. ; III 29) ;
491 Causa manu adsereret, mihi omj|ne argentum redditum iri ;
499 Item genus est lenonium in||ter homines meo quidem animo ;
511 Quasi aquam ferventem frigidam es||se, ita vos putati' leges ;
526 Dum meliu' sit mi, des. Dabun||tur : eras peti iubeto ;
Epid. 355 Inveni : nam ita suasi seni at||que hanc habui orationem ;
361 Is adornat, adveniens domi ex||templo ut maritu' fias ;
364 Deveniam ad lenonem domum ego||met solus, eum ego docebo
(Tribrach for Iambus ; II 52 [ii]) ;
Merc. 530 Ego te redeml, ille me||cum oravlt. Animu' rediit ; credo ;
Mil. 359 Credo ego istoc exemplo tibi es||se pereundum extra portam ;
368 Tun me vidisti ? Atque his quidem herc||le oculis — Carebi',

416 Haec mulier, quae hinc exit modo, est||ne erili* concubina?;
417 Philocomasium, an non est ea? Herc||le opinor, ea videtur ;
915 Nam mi patrone, hoc cogita||to, ubi probus est architectus
(Emended to cogita) ;
932 A tua uxore mihi datum es||se eamque ilium deperire ;
 DIAERESIS 325

Iambic Septenarii. (A) Without (Iambic) Diaeresis {cont)—

(b) With Iambus in Fourth Foot {cont.)—
Mil. 1 221 Placide, ipsae dum libitum est mihi, o||tiose, meo arbitratu ;
1238 Istuc curavi, ut opinio||ne illiu' pulchrior sis ;
1246 Nam nulli mortali scio ob||tigisse hoc nisi duobus ;
1254 Tace, ne audiat. Quid adstitis||ti obstupida ! cur non pultas ? ;
Most. 183 Quid aiSj scelesta? quomodo ad||iurasti ? ita ego Tstam amarem ?
(Emphatic ego) ;
188 Tu ecastor erras quae quidem il||Ium exspectes unum atque illi ;
morem (?) ;
200 Nihilo ego quam nunc tu amata sum at||que uni modo gessi

203 Vix comprimor quin involem il||li in oculos stimulatrici ;

205 Illi me soli censeo es||se oportere obsequentem ;
220 eundem animum oportet nunc mihi es||se gratum, ut impetravi ;
247 Si acceptum sat habes, tibi fore il||lum amicum sempiternum ;
Pers. 45 Nempe habeo in mundo. Si iddomiesljsetmihi, iam pollicerer ;
295 Tun me deflgas ? te cruci ipj|sum adfigent propediem alii ;
302 Paratum iam esse dicito un||de argentum sit futurum ;
307 Subnixis alis me inferam at||que amicibor gloriose ;
310 Ecquid, quod mandavi tibi, est||ne in te speculae? Adito (?)
(Sense requires speculae) ;
327 Et mulier ut sit libera at||que ipse ultro det argentum ;
Poen. 1229 Ite in ius, ne moramini. an||testare me atque duce ;
1249 Quid si eloquamur ? Censeo herc||le, patrue. Misera timeo ;
1269 Condamus alter alterum er||go in nervum bracchialem ;
Rud. 309 Sed quos perconter commode ec||cos video adstare. adibomus; ;
349 Capitalique ex periculo or||bas auxilique opumque hue ;
366 (De) nave timidae ambae in scapham in||siluimu', quia vide-
380 Dies noctesque in custodia es||set semper, verum ecastor ;
397 Credo aliquem immersisse atque eum ex||cepisse. maestast
Id misera ;

687 Bonum animum habete. Nam, obsecro, un||de isteinvenitur

animu' mi?;
690 Veneris, quod amplexae modo, un||de abreptae per vim miserae;
701 Si quidpiamst minu' quod bene es||se lautum tu arbitrary ;
1296 Ad Gripum ut veniat. non fere||tis istum ut postulatis ;
1322 Quid dare velis qui istaec tibi in||vestiget indicetque ? ;
Stich. 775 Vos, spectatores, plaudite at||que ite ad vos comissatum ;
True. 151 Hunc nos habemus publicum, il||li alii sunt publicani.
(B) With Hiatus at Diaeresis (An asterisk indicates a combination of
Hiatus and Syllaba Anceps) :
Asin. 381. 390. 411. 424. 425. 440. 478. 554. 582, 583. 591. 6c6. *6i3- 622.
628. 651. 653. 658. 675. 708. 709. 721. 739. Cas. 825. Cist. 730. 734.
744. Cure. 123 (?). 488. 494. Epid. 356.. 358. 365. Merc. 534. 539.
326 APPENDIX B

Iambic Septenarii. (B) With Hiatus at Diaeresis {cont.) —

543. Mil. 338. 370. 394. 397. 399. 415. 910(F). 928. 1216. *I226. 1228.
1236. *I239. 1267. 1269. 1273. 1277. Most. 223. 230. 236. 241 (?).
Pers. 28oa. 314. Poen. 1236. 1243. 1256. Rud. *329. 351. 354. 388.
*692. 1302. *i3i6. 1320. 1337. True. 137. 152. 170. 188. 208 (?). 222.
Some examples with te, etc. (conceivably ted) may have to be added.
Since Hiatus is more likely than lack of Diaeresis in Plautus we scan :
Asin. 478 Male servu' loquere ? Vapula. || Id quidem tibi hercle fiet ; ;
trasti
721 Opta annum hunc perpetuum mihi |jhui(u)s operas. Impe-

Cure. 488 Et aurum et vestem omnem haberet

suam ||esse ai[e]bat quam haec ;

Epid. 356 Ut quom rediisses ne tibi ||ei(u)s copia esset. Eugae ! ;

365 Si quid ad eum adveniam, ut sibi ||esse datum argentum dicat;
Merc. 534 Nunc quando amicum te scio ||esse illi(u)s, indicabo.
(C) With Syllaba Anceps at Diaeresis (not accompanied by Hiatus ; see
above) :
Asin. 382. 400. 404. 417 (?). 419. 420. 435. 438. 442. 443. 454. 463. 488
567. 594. 605. 634. 641. 652. 656. 664. 674. 677. 718. 744. Cist. 38
710. 715. 717. 719. 736. 746. Cure. 497. 504. 510. 512. Epid. 358
371. 378. Merc. 501 (?). 509. 536. 536s. Mil. 355. 414. 878. 912. 927
1229. 1235. 1241. 1275. Most. 158. 162. 166. 170. 174. 194. 224. 245
Pers. 8. 19. Poen. 821. 1257. 1260. 1264. 1267. Rud. 321. 339. 390
408. 694. 699. 1300. 1314. 1319. Stich. 770. True. 132. 147. 156. 158
173. 180. 186. 203. 218. 220.
We ignore the possibility of the false Tribrach (-fulus a- ; II 44) in such a
line as Asin. 400 :
Macilentis malis, rufulus ||aliquantum, ventriosus.
And since Syllaba Anceps is more likely than lack of Diaeresis in Plautus we
scan :

Epid. 378 Nimi' doctus illic ad male|| faciendum, etc.

(D) With Change of Speaker at Diaeresis :
Asin. 385. 392. 393. 394. 395-'4io. 433- 437- 44 1- 456. 474- 475- 477- 478-
480. 488. 493. 589. 596. 597. 615. 617. 625. 626. 638. 639. 652.654. 669.
697. 707. 716. 719- 736- 738. 739- Cist. 366. 716- 724- 727. 732. 734-
745/ Cure. no. 516. 519. Epid. 180. 357. 379. Merc. 510. 515. 529.
538. 540. Mil. 354. 355. 362. 363. 369. 370. 376. 403. 415- 422. 424.
881. 896. 925. 928. 940. 1216. 1222. 1227. 1236. 1250. 1255. 1262. 1267.
1268. 1278. 1281. Most. 176. 208. Pers. 19. 43. 44. 46. 47- 5°- 281.
282. 284. 285. 287. 293. 298. 308. 312. 314. 316. 320. 323. Poen. 1234.
1238. 1243. 1248. 1259. 1262. 1264. 1268. 1270. 1271. Pseud. 915.
Rud. 341. 342. 343. 355. 361. 374. 382. 387. 393. 398. 1289. 1304. 1306.
1311. 1323. 1324. 1327. 1329. 1331. 1333. 1334. Stich. 771. True.
152. 159. 188. 203.
(Of these instances nearly a score have already appeared in the B or C lists.)
 DIAERESIS 327
Iambic Septenarii [cont.) —
(E) With ' Diaeresis ' (Word, not Sense) :
Asin. 413. 466. 571. 576. 649. 678. 679. 681. 688. 689. 695. 701. 705. 710.
711. 720. 726. Cist. 307. 315. 364. 705. 723. 726. 728. 733. Cure. 518.
Epid. 338. Merc. 504. Mil. 365. 875. 880. 899. 901. 911. 1266. 1282.
, 1299. Most. 172. 217. 237. Poen. 1272. Pseud. 156. Rud. 292.
293- 295« 296. 299. 305. 350. 358. 404. 696. 1290. 1292. 1313. True.
141. 165. 187.
Most of these could find parallels at the end of a verse, e. g. :
Asin. 433-4 II Sticho vicario ipsi
Tub. Vah, delenire apparas ||;
585-6 Manedum Quid est ? Philaenium est|(ne haec quae intus
exit atque
Una Argyrippus ? , etc.

Iambic Octonarii (under 400 in the twenty-one plays of Plautus).

(A) With Hiatus at Diaeresis (An asterisk indicates a combination of
Hiatus and Syllaba Anceps) :
Amph. 190 Quod multa Thebano poplo ||acerba obiecit funera;
192 Imperio atque auspicio mei ||eri Amphitruonis maxume ;
195 Me a portu praemisit doraum ||ut haec nuntiem uxori suae ;
199 Nam quom pugnabant maxime,|| ego turn fugiebam maxime ;
208 Redducturum, abituros agro || Argivos, pacem atque otium ;
*2ii Haec ubi Telobois ordine || iterarunt quos praefecerat (Or
ordini; III 30 1) ;
tust ;
250 Perduelles penetrant se in fugam ; ||ibi nostris animus addi-

262 Nunc pergam eri iniperium exsequi ||et me domum capessere ;

1000 Atque illuc sursum escendero : ||inde optime aspellam virum ;
1004 Meo me aequumst morigerum patri, ||ei(u)s studio servire
addecet (Or patri ei||us) ;
Bacch. 930 Sine classe sineque exercitu ||et tanto numero militum ;
933 O Troia, o patria, o Pergamum, ||o Priame periisti senex ;
941 Turn quae hie sunt scriptae litterae, ||hoc insunt in equo
milites (in eq. ins. MSS.)
adhuc ;;
942 Armati atque animati probe. ||ita res successit mi usque

Capt. 780 Speroque me ob hunc nuntium ||aeternum adepturum cibum ;

833 Perlibet hunc hominem colloqui. || Ergasile ! Ergasilum
vocat qui
?;

Epid. 48 Quam amabat, emeretur sibi. ||id ei impetratum reddidi ;

334 Quipp' tu mi aliquid aliquo modo ||alicunde ab aliquibus blatis ;
Men. 1004 O facinus indignum et malum, || Epidamnii cives, erum ;
Merc. 116 Detrude, deturba in viam. || haec disciplina hie pessimast ;
Pers. 256 Danunt, argenti niutui ||ut ei egenti opem afferam ;
328 APPENDIX B

Iambic Octonarii. (A) With Hiatus at Diaeresis (cont.)—

Poen. *8i8 Studeo hunc lenonem perdere, ||ut meum erum macerat
miserum ;

Pseud. 191 Mihi afluam ;

et familiae omni sit meae, ||atque adeo ut frumento

Frag. 145 Quid est ? hoc rugat pallium : ||amictu' non sum commode.
(B) With Syllaba Anceps at Diaeresis (not accompanied by Hiatus ; see
above) :
Amph. 157 Nee causam liceat dicere ||mi, neque in ero quicquam auxili ;
203 Principio ut 1111 advenimus, || ubi primum terram tetigimus ;
207 Si quae asportassent reddere, ||se exercitum extemplodomum ;
215 Propere suis de finibus ||exercitus deducerent ;
? 995 Amat : sapit: recte facit, || animo quando obsequitur suo ;
(III 40 H)

999 Capiam coronam mi in caput, ||adsimulabo me esse ebrium ;

1054 Neque ullast confidentia ||iam in corde, quin amiserim;
1055 Ita mi videntur omnia, ||mare, terra, caelum, consequi ;
1069 Erili' praevertit metus : ||accurro, ut sciscam quid velit ;
Bacch. 934 Qui misere male mulcabere || quadrigentis Philippis aureis
(Not mulcaberis !) ;
947 Mnesilochust Alexander, qui erit ||exitio rei patriae suae (Or
qui erit, Tribrach) ;
Cas. 897 Satm morigera est ? Pudet dicere. || Memora, etc. ;
Cist. 451 Germana mea sororcula. || Repudio te fraterculum ;
452 Turn tu igitur, mea matercula. || Repudio te puerculum ;
Merc. 125 Perii, animarn nequeo vertere : ||nimi' nihili tibicen siem ;
Pers. 255 Quia meo amico amiciter ||hanc commoditatis copiam ;
Pseud. 157 Tu qui urnam habes aquam ingere, ||face plenum ahenum coquosit;

True. 216 Magi' que adeo ei consiliarius ||hie amicust quam auxiliarius ;
Frag. 102 . . . addite ||lopadas, echinos, ostreas. (?)
(C) With Change of Speaker at Diaeresis :
Bacch. 988. Cas. 231. 236. 897. Cist. 451. 452. Epid. 7. 8. 19. 20. 39.
40. Merc. 123. 137. Pers. 23. Pseud. 919. Rud. 944.

Trochaic Septenarii (630 in the Amphitruo).

With Hiatus at Diaeresis :
Amph. 253 Haec illist pugnata pugna ||usque a mani ad vesperum ;
272 Credo ego hac noctu Nocturnum ||obdormivisse ebrium ;
303 Iam pridem videtur factum ||hen quod homines fact um he||r(Or
quattuor i) ;

319 Mirum ni hie me quasi murenam ||exossare cogitat ;

328 Non equidem ullum habeo iumentum. || Onerandust pugnis
probe (Or on. est) ;
 DIAERESIS 329

Trochaic Septenarii (cont.) —

343 Servu'sne an liber ? Utcumque ||animo collibitum est meo ;
350 Quid apud hasce aedes negoti ||est tibi ? Immo quid tibi est ? ;
401 Qui cum Amphitruone hinc una||iveram in exercitum (?);
418 Sed quid ais ? quid Amphitruoni ||a Telobois est datum ? ;
429 Cadus erat vini, inde implevi || hirneam. Ingressust viam
(II 32 B) ;
511 Ego faxim ted Amphitruonem ||esse malis quam Iovem ;
518 Carnifex, non ego te novi ? ||abin e conspectu meo novi
? (Hardly
a||bin) ;

523 Clanculum abil ; a legione ||operam hanc surripui tibi ;

532 Nam qua nocte ad me venisti, ||eadem abis. Cur me tenes ?
(Or venisti e||adem) ;
545 Prlu' tua. opinione|| hie adero : bonum animum habe (?) ;
622 Non soleo ego somniculose ||erf imperia persequi ;
631 Non ego cum vino simitu || ebibi imperium tuum ;
673 Ni ego illi puteo, si occepso, j|animam omnem intertraxero ;
714 Ecastor equidem te certo ||hen advenientem ilico ;
776 Pro cerrita circumferri ? || Edepol qui facto est opus ;
797 Huic dedisti, post hanc rursum || obsignasti clanculum (?) ;
802 Lavisti. Quid postquam lavi ? || Accubuisti. Eugae optime !;
849 Quid si adduco tubm cognatum ||hue a navi Naucratem ? ;
1 01 2 Apud emporium at que in macello, ||in palaestra atque in foro ;
101 5 Nunc domum ibo atque ex uxore ||hanc rem pergam exquirere ;
1032 Quidum ? Quia senecta aetate ||a me mendicas malum ;
1050 Seu patrem sive avum videbo, || obtruncabo in aedibus ;
1086 Amphitruo, piam et pudicam ||esse tuam uxorem ut scias ;
1 128 Ego Teresiam coniectorem || advocabo et consulam.
In the above list the foot in Hiatus is always either a Trochee or a Spondee.
And these are the two feet in which a Trochaic Octonarius commonly ends.
But we find occasionally in Hiatus a Tribrach or an Anapaest, feet which
sometimes stand at the end of an Octonarius (e. g. Aul. 820 Ere, mane, eloquar
iam, ausculta. Age || ergo, loquere. Repperi hodie). In the first eight plays
(which contain over 3,200 Trochaic Septenarii) we find a Tribrach or an
Anapaest in the following lines :
Asin. 243 Interii si non invenio || ego lllas viginti minas (Or ego || illas) \
320 Si istam firmitudinem animi || obtines, salvi sumus ;
894 Die amabo, an foetet anima || uxoris tuae ? Nauteam ;
901 Quid quom adest ? Periisse cupio. || Amat homo hie te, ut prae-
dicat (Or cupio. A||mat) ;
936 Ecastor cenabis hodie, || ut dignu's, magnum malum (Or hodie,
ut || dignus es) ;
Aul. 177 Et tu frater. Ego conveniam || Euclionem, si domi est;
Bacch. 86 Atque ecastor apud hunc fluvium || aliquid perdendumsjt tibi ;
692 Quid vis curem ? Ut ad senem etiam || alteram facias viam ;
33Q APPENDIXB

Trochaic Septenarii (cont.) —

semel ;
757 Numquid aliud ? Hoc atque etiam : || ubi erit accubitum

Capt. 331 eum si reddis mihi, praeterea || unum nummum ne duis ;

843 Bene facis : iube — Quid iubeam ? || Ignem ingentem fieri ;
846 Iuben an non iubes adstitui || ollas, patinas elui ? ;
977 Philocrates, per tubm te genium || obsecro, exi, te volo ;
Cure. 295 Ex unoquoque ebrum crepitum || exciam polentarium ;
619 Ouam ego pecuniam quadruplicem || abs te et lenone auferam ;
684 Pessime metui ne mi hodie || apiid praetorem solveret (Or mihi
hodi||e apiid) ;
Epid. 26 Dices digniorem esse hominem || hodie Athenis alterum ;
640 Lunulam atque anellum aureolum || in digitum ? Memini, homomi ;

716 Vinctus adsto, quoi(u)s haec hodie || opera inventast filia (Or
quoius haec hodi||e op.).
A Trochaic Octonarius never ends in a Dactyl ; and if we could find a
Dactyl in Hiatus at the Diaeresis, we should have a clear proof that this
Hiatus did not imply that Plautus regarded a long line as a pair of short
lines. But the presence of a Dactyl in Hiatus can neither be affirmed nor
denied. Take a line like Epid. 128 :
Salvum te advenisse — Tarn tibi istuc credo quam mihi,
how can we prove that Plautus scanned tarn tibi || istuc (Dactyl with Hiatus),
and not tarn ti||b(i) istuc (Trochee with Elision) ? Or Pseud. 607 :
Tune es Ballio ? Immo vero ego eius sum Subballio,
where ver(o) e||g(o) eius is an alternative to ver(o) eg6 || eius, although the
most likely scansion of all is vero || ego ei(u)s, with emphatic ego (III 31) and
enclitic ei{u)s (II 35). Or Pers. 483 :
Di dent quae velis. Eho an iam manu emisisti mulierem (Only eho an
possible ; II 12),
since man(u) in this phrase, although unusual, is not impossible.
Or Stich. 621 :

Exorabo aliquo modo. veni. Hucine ? Immo in carcerem,

where veni would certainly be natural enough, but the elision ven(i) is an
alternative.
A strong proof is Trin. 330 :
Quid is ? egetne. Eget. Habuifne rem ? || Habuit. Qui earn perdidit,
where total elision of rem is unlikely. Still even this instance might be
challenged.
Since the fragments of Dramatists of the first century B.C. offer examples
of Hiatus at the Diaeresis of the Trochaic Septenarius :
Atta 4 Aquae ita muginantur hodie. || Atqui ego fontem occlusero
(Change of speaker) ;
 DIAERESIS 331

Trochaic Septenarii (cont.)—

Novius 13 lam ego llli subiens sublabrabo || esui illud sinciput ;
Pomponius 171 In terrain, ut cubabat, nudam || ad eum ut conquexi, interim ;
not to mention the tragedian Accius (84, incomplete) :
. . . ut tarn obstinato || animo confisus tuo,
we may look for this Hiatus in Terence. The absence of any clear instance
is suspicious and suggests editorial interference.

C. HIATUS
Monosyllable before Iambic Word (Add the lists in III 53) :
(Plautus) (first eight plays) :
Amph. 403 (Troch.) Quid, malum, non siim ego servus? || (Or ego) ;
450 (Troch.) Quo sigis te ? Domum. Quadrigas || (Or agi') ;
(725 (Troch.) Tu me heri hie vidisti ? Ego, inquam ||) ;
731 (Troch.) Te heri me vidisse, etc. ;
961 (Troch.) Tristi' sit si eri sint tristes ||;
1038 (Troch.) Quid med advocato opiist qui utri || sim nescio ? ||;
advocatus

1 103 (Troch.) Sed puer ille quern ego lavi || (Or ego) ;
Asin. (143 (Troch.) Atque ea s{i) erant, magnas habebas ||) ;
220 (Troch.) || area est, auceps s(um) ego ;
228 (Troch.) I] si eris nactus ; nunc abi (Or eri') ;
354 (Troch.) || si erum vis, Demaenetum ;
355 (Troch.) Quern ego novi adduce || (Or ego) ;
375 (Troch.) (patitor) || tu item quom ego te referiam ;
436 (Iamb.) Sed vina quae heri vendidi ||;
437 (Iamb.) Iam^n? eis sati' fecit Sticho? ||;
658 (Iamb.) Nolo ego te, ^ erus sis, mihi || (Or eru') ;
722 (Iamb.) Ain vero ? Certe inquam. Ad m(e) adi || vicissim
atque experire (Hardly me ad?) ;
(731 (Iamb.) || nunc rem ut est eloquamur) ;
Aul. (89 (Sen.) Abi intro, occlude ianuam. idm ego hie ero)Bacch.(Contrast
1066) ;

(104 (Sen.) Fores ambobus pessulis. idm ego hie ero) ;

134 (Bacch.) Ut tuam rem ego tecum hie || loquerer familiarem
(Rather ego) ;
203 (Troch.) Quo abis ? Iam revertar ad te ||;
(274 (Troch.) Atque aedes occlude, idm ego hie || adero. Quid ego
nunc agam ?) ;

323 (Sen.) Coquum ergo dico. Quid tu ais ? Sic sum ut vides ;
332 APPENDIX C

Monosyllable before Iambic Word (cont.) —

427 (Vers. Reiz.) Sed in aedibus quid tibi meis || nam erat negoti ? ;
444 (Vers. Reiz.) Scis iam meam sententiam. || Quo abisrursum ? redi ;

478 (Sen.) Nam meo quidem animo si idem faciant ceteri ;

574 (Sen.) Tibi quoi decretum est bibere aquam. Scio quam r(em)
agat;

(588 (Troch.) Nam qui ero ex sententia ser||vire servus postulat) ;

(621 (Troch.) Aurum, dtim hie est occupatus) ;
645 (Troch.) Quid abstulisti hinc? Di me perdant || si ego tui
quidquam abstuli (Or ego tui) ;
680 (Sen.) Quamquam hie manere me erus sese iusserat (Or eru') ;
(714 (Anap.) || equidem quo earn aut ubi sim aut qui sim) ;
790 (Troch.) Qui homo culpam admisit in se ||; (Or homo) ;
(797 (Troch.) Quern ego avum feci iam ut esses) ;
Bacch. (5 (Sen.) Cum nassiterna et cum aqua istum impurissimum) ;
72 (Troch.) Ubi mi pro equo lectu' detur ||;
194-5 (Sen.) Animast arnica amanti : si abest, nullus est,
67 adest, res nullast : ipsus est nequam et miser ;
215 (Sen.) Nullam aeque invitus specto, si agit Pellio ;
(329 (Sen.) Quia id signumst cum Theotimo, quieum illi adferet);
364 (Sen.) St ero reprehensus, macto ego ilium infortunio ;
564 (Troch.) || qudm ego mandassem tibi (Or ego) ;
758 (Troch.) || a me erit signum datum (Or med erlt) ;
947 (Iamb.) Mnesilochust Alexander qui erit || exitio rei patriae
suae (Tribrach or Iambus ? Cf. Capt. 65) ;
(988 (Iamb.) || Chrysale, ades diim ego has pellego) ;
1 155 (Anap.) Quid ais tit homo? |[ Quid me vis ? ;
Capt. 224 (Iamb.) Nam .tf erus mihl es tu atque ego || (Or eru') ;
243 (Troch.) Ut qui erum me tibi fuisse at||que ess' nunc
vum conser-
velint ;
489 (Troch.) Omnes de compecto rem agunt ||;
490 (Troch.) Nunc redeo inde, quoniam meibi || video ludincarier
(Or ibi. Or med) ;
627 (Troch.) Quid tu ais ? Me tubm esse servum ||;
807 (Troch.) || qui alunt furfuribus sues ;
(841 (Troch.) I dm ego ex corpore exigam omnes ||) ;
934 (Troch.) || di earn potestatem dabunt ;
(947 (Troch.) ||^Weo argenti ne duis) ;
995 (Troch.) Eheu ! qudm ego plus minusque || (Or ego) ;
Cas. pr. 66 (Sen.) Pontem interrupit qui erat ei in itinere (Or erat) ;
(133 (Sen.) Unde auscultare possis qudm ego illam osculer)
(Rather quom ego) ;
183* (Bacch.) Nee qua in plura sunt
Mi quae ego velim ;
(247 (Cret.) .STego in os meum hodie ||) (Or Si ego) ;
 HIATUS 333
Monosyllable before Iambic Word (cont.) —
269-70 (Troch.) Quid si ego impetro atque exoro a || vilico causa
me a
Ut earn llli permittat ? Quid si e||go autem ab
armigero impetro ?) (Or ego ||; App. B, end) ;
(296 (Troch.) || ciim aqua et sortes. Sati' placet) ;
(476 (Sen.) lam ego uno in saltu lepide apros capiam duos) ;
528 (Troch.) Attatae ! caedendu' tii homo's ||;
(616 (Sen.) Qua ego hunc amorem mi esse avi dicam datum ?) ;
673 (Bacch.) Quid ciim ea negoti ||;
721 (Anap.) || jtfeas ereptum, ilico scindunt;
(880 (Anap.) || ea sunt quag ego intus turbavi) ;
(892 (' Glyconic ') Agedum tii adi hunc. Obsecro ||) ;
933 (Iamb.) Ut senex hoc ebdem poculo || qud eg5 bibi biberet.
Optime est ;
951 (Cret.) Sed e(c)quis est qui homo mu||nus velit fungier
(Rather homo)
ego); ;
(1002 (Troch.) Ne ut earn amasso, si ego umquam adeo ||) (Or

Cist. (1 (Bacch.) Quom ego antehac te amavi ||) ;

(77 (Troch.) Quom ego Ilium unum mi exoptavi || quicumdegerem)
aetatem ;
vivereeo;
85 (Troch.) Ut me quern ego amarem graviter || sineret ciim

(215 (Anap.) Ita m(e) Amor lassum ani||mi ludificat) ;

231 (Troch.) Potine tii homo facinu' facere ? || (Or homo) ;
292 (Sen.) Qui equum me afferre iubes, loricam adducere ;
311 (Iamb.) || volo ego agere, ut tii agas aliquid;
565 (Sen.) Meretrix fuit ; sed ut sit, deea. re eloquar (?) ;
707 (Iamb.) (postremo ille) Plus qui vocat scit quod velit || quam
ego quae vocor. revertor (Or ego q. vocor) ;
(709 (Iamb.) || quam ego hie amisi misera) ;
715 (Iamb.) || quae erae [meae] supposita est parva ;
723 (Iamb.) Quid quaeritabas ? Mi homo et || mea mulier, saluto (?)
vos ;
ego);
773 (Sen.) Dominae. Melaenis. I prae, Mm ego te sequar (Or

Cure. 170 (Troch.) Ipsu' se excruciat qu{i) homo quod amat || (II 32) ;
202 (Troch.) || quoi ego sano serviam (Or ego) ;
(280 (Troch.) || diim ego hie officium meum) ;
320 (Troch.) lam edes aliquid. Nolo hercle ' aliquid ' ||;
(326 (Troch.) Ne me ludas. Ita me amabit || quam ego amo ut
ego haud mentior) ;
362 (Troch.) Rogant me servi quo earn : dico ||;
(386 (Sen.) Edepol ne ego hie med intus explevi probe) ;
412 (Sen.) Ut has tabellas ad eum ferrem. Quis t(u) homo's? ;
334 A P P E N D I X C

Monosyllable before Iambic Word {cont.)—

570 (Troch.) Male loqui, profecto, quoi ego nisi malum nil debeo
(Or ego) ;
595 (Troch.) Peior [quam haec est] quae ubi me hunc habere ||
(Or ubi med) ;
656 (Sen.) Hie est quern ego tibi misi natali die (II 45) ;
Epid. 218 (Troch.) Et cum ea tibicinae ibant ||;
ero);
(304 (Troch.) || priu' quam ego ad te venero) ;
(575 (Troch.) Tii homo insanis. Egone ? Tune ||) ;
580 (Troch.) Ni ego me nego nosse hanc quae sit (Or egfi) ;
585 (Troch.) (equidem hac invita) || tdm[en] ero matris filia (Or

640 (Troch.) || in digitum. Memini, m(i) homo ;

(663 (Troch.) || qu{ae) ego scio Stratippoc (Rather ego scio)
lem (Or quae e. scio) ;
(693 (Troch.) || frugi es til homo, Apoecides) ;
Since Elision is in so marked minority, we may assign Prosodic Hiatus to
all doubtful cases.

Clision (?) of Emphatic Monosyllable :

(Plautus) (first eight plays)
Amph. pr. 17 (Sen.) Nunc quoiu' iussu venio et quam ob rem venerim ;
191 (Iamb.) Id vi et virtute militum ||;
195 (Iamb.) Me 2, portu praemisit domum ||;
206 (Iamb.) Si sine vi et sine bello velint || ;
214 (Iamb.) Respondent bello se et suos || ;
341 (Troch.) Quo ambulas tu ? qui Vulcanum ||;
439 (Troch.) Ubi ego Sosia nolim esse, || tu esto sane Sosia ;
542 (Troch.) || me tuam, te absenti tamen ;
575a (Troch.) Egone ? Tu istic. || ubi bibisti ? ;
592 (Troch.) Quo id, malum, pacto potest nam ? ||;
693 (Troch.) Qui istuc potis est fieri, quaeso ? || ; >.
694 (Troch.) Quid enim censes ? te ut deludam || contra
lusorem meum ;
699 (Troch.) || et te vidi. Quo_ in loco ? ;
766 (Troch.) (nimi' demiror, Sosia,) Qui illaec illi me donatum
es||se aurea patera sciat ;
858 (Troch.) Nimis ecastor facinu' mirum est || qui bitum siet ;
illi volo;
colli-
1 130 (Troch.) || di, obsecro vestram fidem ;
Asin. 42 (Sen.) Etiam amplius. Nam quo usque ? Usque ad mortem

51 (Sen.) Demiror quid sit et quo evadat sum in metu phatic

(Unem-
?) ;
258 (Troch.) Unde sumam ? quern intervertam ? ||quo hanc eeloeem
: - conferam?;
 HIATUS
335
Elision (?) of Emphatic Monosyllable (cont.)—
273 (Troch.) Vae ill i qui tarn indiligenter ||;
302 (Troch.) Quo argumento istuc ? Ego dicam || quo argumento
et quo modo ;
323 (Troch.) Em istaec virtus est, quando usust ||;
336 (Troch.) Em ergo is argentum hue remisit ||;
342 (Troch.) Teneo, atque idem te hinc vexerunt ||;
358 (Troch.) || censes, dice. Em istuc ego ;
431 (Iamb.) Em ergo hoc tibi. Hospes, te obsecro ||;
homini ;
466 (Iamb.) Ego certe me incerto scio hoc || daturum nemini

467 (Iamb.) Hercle istum di omnes perduint ||;

473 (Iamb.) Flagitium hominis da, obsecro, ar||gentum huic, ne
male loquatur ;
660 (Iamb.) Ego baiulabo ; tu, ut decet || dominum, ante inanis
me ito;

716 (Iamb.) Quern te autem divum nominem ? || (Or unemphatic ?);

850 (Iamb.) Em istoc me facto tibi devinx||ti. Quin te ergo
das hilarum
mihi
malo;? ;
909 (Troch.) Ego pol vivam et tu istaec hodie || cum tuo magno

Aul. 71 (Sen.) Nescio pol quae illunc hominem intemperiae tenent;

225 (Troch.) || mihique propter te et tuos ;
229 (Troch.) Te bovem esse et me esse asellum || ;
meu' med) ;
232 (Troch.) Et te utar iniquiore || et meu' me ordo irrideat (Or et ||
263 (Troch.) || me vis ? Istuc. i et vale ;
333 (Sen.) I sane cum illo, Phrygia. tu autem, Eleusium,
(Hue intro abi ad nos) ;
414 (Iamb.) Itaque omnes exegit foras, || me atque hos,fustibus
onustos ;

416 (Vers. Reiz.) Quia ad tresviros iam deferam || nomen tuum

Quam ob rem ? ;
460 (Troch.) Illic hmc abiit. di immortales ! ||;
489 (Sen.) Namque hoc qui dicat * quo illae nubent divites? ' ;
616 (Troch.) Di immortales, quod ego hunc hominem ||;
694 (Sen.) /hac intro mecum, gnate mi, ad fratrem meum ;
743 (Troch.) || apiid me te in nervo enicem (Or unemphatic) ;
744 (Troch.) Ne istuc dixis. Quid tibi ergo || meam tactiost
me invito? ;

756 (Troch.) || habeas me invito meam ? ;

800 (Troch.) I intro, exquaere, etc. ;
Bacch. 127 (Sen.) Etiam me adversus exordire argutias ? ;
132 (Sen.) Iam perdidisti te atque me atque operam meam ;
144 (Sen.) Sperat quidem animus ; quo eveniat dis in manust ;
155 (Sen.) Fiam, ut ego opinor, Hercules, tu autem Linus ;
33^ A P P E N D IX C

Elision (?) of Emphatic Monosyllable (cont.)—

274 (Sen.) Etiamnest quid porro ? Em, accipitrina haec nunc erit ;
412 (Troch.) Nam absque te esset, ego ilium haberem ||;
473 (Troch.) (omnem rem scio) Quern ad modumst, etc. ;
488 (Troch.) || quam me atque illo aequum foret ;
583 (Sen.) E(c)quis exit ? Quid istuc ? quae istaec est pulsatio? ;
624a (Iamb ?) Perdidi me atque ope||ram Chrysali ;
751 (Troch.) Quia mi ita libet. potin ut cures || te atque ut ne
parcas mihi ? ;
830 (Sen.) Die, quo in periclo est meu' Mnesilochus filius?;
843 (Sen.) Per vim ut retineat mulierem ? quae haec factiost ? ;
870 (Sen.) Em illoc pacisce, si potest ; perge, obsecro ;
965 (Iamb.) Item ego dolis me illo extuli || (Or unemphatic) ;
989 (' Glyconic ') Ut scias quae hie scripta sient ;
1065 (Sen.) Vel da aliquem qui servet me. Ohe, odiose facis ;
1 106 (Anap.) [Philoxene,] salve. Et tu. unde agis ? Unde
homo || (II 31);
1 1 13 (Cret.) Perdidit filium || me atque rem omnem meam ;
1 149 (Anap.) || eho, amabo. Quo illaec abeunt ? ;
1 175 (Anap.) /hac mecum intro atque ibi siquid vis ||;
1 181 (Anap.) 1 hac mecum intro ubi tibi sit lepide ||;
1 201 (Anap.) Tua sum opera et propter te improbior. || Nemini'
quam mea mavellem ;
Capt. 150 (Sen.) Tibi tile unicust, mi etiam unico magisunemphatic)
unicus (Or;

175 (Sen.) Propterea a te vocari ad te ad cenam volo (Or un-emphatic) ;

mihi atque) ;
180 (Sen.) Quae mi atque amicis placeat conditio magis (Rather

215 (Troch.) Em istuc mihi certum erat. concede hue ||;

224 (Iamb.) Nam si eru' mihi es tu atque ego ||;
238 (Cret.) Pol ego si te audeam, || meum patrem nom inem (Or;
unemphatic)

249 (Troch.) Scio equidem me te esse nunc et || te esse me. E?n

istuc si potes (Rather me unemphatic) ;
403 (Troch.) || neque me adversatum tibi (Rather unemphatic) ;
441 (Troch.) Serva tibi in perpetuum amicum || me, atque hunc
inventum inveni ;
444 (Troch.) Tu hoc age. tu mihi eru' nunc es, || tu patronus, tu
pater (Or unemphatic) ;
452 (Troch.) Tu intro abi. Bene ambulato || (Or unemphatic) ;
479 (Troch.) ' Salvete ' inquam. ' quo imus una ? ' ||;
533 (Iamb.) Quo ilium nunc hominem proripuis||se foras se dicam
ex aedibus ? ;
liberdomi
543-4 (Troch.) Equidem tarn sum servu' quam tu, et||si ego fui,
 HIATUS 337

Elision (?) of Emphatic Monosyllable {cont.)—

Tu usque a puero servitutem || servivisti in Alide ;
verbio ;
568 (Troch.) Tu enlm repertu's Philocratem qui || superes veri-

569 (Troch.) Pol ego ut rem video tu inventu's ||;

591 (Troch.) ||id quod vi hostili obtigit ;
623 (Troch.) || quam aut ego aut tu. Eho die mihi ;
650 (Troch.) Vae illis virgis miseris quae hodie ||;
669 (Sen.) Sed quid negoti est ? quam ob rem suscenses mihi ?
(Or Relative quam) ;
739 (Sen.) Cur ego te invito me esse salvum postulem (Rather
unemphatic me) ;
799 (Troch.) Quae illaec eminatiost nam ? ||;
817 (Troch.) Ut sciant alieno naso || quam exhibeant molestiam ;
829 (Troch.) Quae illaec est laetitia quam illic ||laetu' largitur mihi ?
863 (Troch.) Quoi deorum ? Mi hercle, nam ego sum mihi|| hercle)
(Rather ;

868 (Troch.) Iuppiter te dique perdant. Te hercle — mi aequum

est gratias ;
885 (Troch.) || esse. Vae aetati— Tuae ;
894 (Troch.) || tu intu' cura quod opus est (Or unemphatic) ;mihi ;
933 (Troch.) Proinde ut tu promeritu's de me et ||filio. Immo potes ;
976 (Troch.) Serva, Iuppiter supreme, et || me et meum gnatum

Cas. pr. 5 1 Paterque filiusque clam alter alterum ;

pr. 54 Sibi fore paratas clam uxorem excubias foris ;
112 (Tun illam ducas ? hercle me suspendio)
Quam tu eiu' potior fias satiust mortuum (Hardly un- emphatic) ;

123 (Sen.) Quae nisi erunt semper plena, ego te implebo flagris ;
246 (Troch.) || rugat. Di me et te infelicent ;
265 (Troch.) Ilium ;/zz aequiust quam me illi ||quae voloconcedere;
279 (Troch.) Qui ilium di omnes deaeque perdant ||;
299 (Troch.) Ibi ego te et suffragatores ||tubs ulciscar. Attamen ;
301 (Troch.) Mi ilia nubet, machinare || (Rather Mihi ilia) ;
359 (Troch.) Uxor, sortes, situla atque egomet. || Te uno adest
plus quam ego volb ;
400 (Troch.) Tu ut liquescas ipse, actutum ||;
587 (Sen.) I tu atque arcesse illam : ego Intu' quod factostopus;
592 (Sen.) Qui me atque uxorem ludificatust larua ;
641 (Cret.) Obtine aures ama||bo. / in malam a me crucem ;
672 (Bacch.) Men occidet ? An quip||piam ad te attinet ? Vah ! ;
677 (Troch.) Plus quam quoiquam. Quam ob rem ? Quia se ||;
766 oportuit (Or;
(Sen.) Properate, cenam iam esse coctam unemphatic)

792 (Sen.) Tu hie cunctas, intus alii festinant. (Or un-

Eoemphatic);
2348 Z
338 APPENDIXC

Elision (?) of Emphatic Monosyllable (cont.)—

814 (Troch.) Di hercle me cupiunt servatum ||;
977 (Troch.) || Casina sum. /in malam crucem (In' P) ;
1016 (Troch.) Qui faxit, clam uxorem ducet ||;
Cist. 3 (Bacch.) Aperuisti, tu atque haec ||;
51 (Iamb.) Di faxint ! Sine opera tua||^z horunc nil facere possunt ;
52 (Iamb.) || sed tu inter istaec verba ;
168 (Sen.) Ill' clam observavit servus, etc. ;
169 (Sen.) Quo aut quas in aedes haec puellam deferat ;
574 (Sen.) Quoi illam dedisset exquisisse oportuit ;
631 (Troch.) Rem elocuta sum tibi omnem ||;
663 (Troch.) || di, obsecro vestram fidem || ;
Cure. 74 (Sen.) Me, te atque hosce omnes. Turn tu Venerem vomere vis ;
20I (Troch.) Auro contra cedo modestum a||matorem : a me aurum
accipe (Or unemphatic) ;
204 (Troch.) || quo usque, quaeso, ad hunc modum?;
303 (Troch.) Te ille quaerit. Quid si adeamus ? ||;
308 (Troch.) Eloquere, obsecro hercle. Eloquere, || te sunt
obsecro,
meae ubi
?;
313 (Troch.) || da, obsecro hercle, obsorbeam ;
370 (Troch.) Dicam quern ad modum conscribas ;
442 (Sen.) Quam ob rem istuc ? Dicam. quia enim Persas,
Paphlagonas (cf. 667) ;
467 (Troch.) Commonstrabo quo in quemque hominem || facile
inveniatis loco ;
Em tibi ;
624 (Troch.) Em ut scias me liberum esse ! er||go ambula in ius.

628 (Troch.) Phaedrome, obsecro, serva me. || Tamquam me et

genium meum ;
711 (Troch.) || quo praesente ? quo in loco ? ;
Epid. 66 (Troch.) Plusque amat quam te umquam amavit ||;
81 (Troch.) || quo in loco haec res sit vides;
143 (Troch.) || quo a tarpezita peto ? ;
(164 On the ' Palatine' reading see IV 21.)
192 (Troch.) Id ego excrucior. Di hercle omnes me || adiuvant,
augent, amant ;
276 (Troch.) || quam ad rem istuc refert ? Rogas ? ;
285 (Troch.) Rein hercle loquere. Etrepperi haec te || qui abscedat
suspicio ;
341 (Iamb.) Pro di immortales mi hunc diem ||;
488 (Sen.) Em istic homo te articulatim concidit senex ;
552 (Troch.) Tuae memoriae interpretari || 7>ie aequumCommode
censes. ;

558 (Troch.) Ego sum. salve. Salva sum quia ||teesse salvum sentio ;
627 (Troch.) Di immortales ! sicin iussi ? ||;
701 (Troch.) In meum nnmmum, in tuiim talentum || pignu5
Enim Tstaec captiostda.;
 HIATUS 339

Elision (?) of Emphatic Monosyllable (cont.)—

710 (Troch.) || quae haec, malum, impudentiast ? ;
714 (Troch.) AM modo intro. /, illuc non temerest||.
Presumably the monosyllable, when emphatic, was not elided, but left in
Prosodic Hiatus, and became a ' brevis brevians \
With regard to less important monosyllables it often seems as if it were
the accentuation of the sentence which determined whether they should be
effaced by elision or not. But the danger of framing laws for such niceties
of ancient utterance is seen from a line like :
Ter. Ph. 816 (Iamb.) Quid fstuc negotist? Iamne operu||it ostium?
lam. O Iuppiter.
How were the last three words pronounced ? Presumably as an Anapaest
followed by an Iambus. And yet Terence's (and Plautus') scansion often
seems to efface iatn (with adero, etc.), e.g. : Eun. 725 • 765 ; Heaut. 238. So
we may scan Plaut. Capt. 585 (Troch.) :
|| i(am) aliquid pugnai edidit (Emended to dedit),
as in True. 510 (Troch.) || i(am) aliquid actum oportuit (ac. al. MSS.).
Finally, for the sake of completeness may be added these examples from
the fragments of the Tragedians :
Enn. 77 (Cret.) || quo accedam ? quo applicem ? ;
Pac. 60 (Cret.) Quae aegritudo insolens || mentem attemptat tuam ? ;
Ace. 230 (Troch.) || quern ore funesto alloquar ?
And a line of Catullus may warn us of the danger of hard and fast rules :
55, 4 T(e) in circo, te in omnibus libellis :
(where it is the Nouns, rather than the Pronouns, which have emphasis).
And since in Virgil elision of a monosyllable before a short syllable is rare
(Aen. 6, 629 sed iam age), Lucilius' elision of the (unemphatic) Relative may
be cited : MSS.)
1029 SicutI te, qui ea quae speciem vitae esse putamus (put. esse

D. FINAL MONOSYLLABLE
This is a much-questioned rule, of much importance to Plautine text,
prosody and metre. To determine it once for all, the widest survey and
the fullest statistics are needed. We include among the monosyllables atque
(ac), neque (nee) ; not est, es where they may be printed st, s.
The rule posited for Plautus is that, with a monosyllabic ending, the final
foot must be 'pure ' : i. e. an Iambus (not a Spondee, etc.) in an Iambic line,
a Bacchius (not a Molossus, etc.) in a Bacchiac line, and so on. The
questions to be settled by the following array of statistics are : (1) Is it
a rule ? (2) If it is, does it apply to all metres, and in particular does it
apply to Anapaestic Metre ? For the other feet, beside the Anapaest, allowed
z 2
340 A P P E N D IX D

in this metre (the Spondee, etc.) have the same number of ' morae ' (four) as
the Anapaest itself. (3) Does the rule hold for acatalectic as well as for
catalectic lines? (4) Does it hold for the end of the hemistich as well as
of the line? (5) Does it hold when the monosyllable is preceded by
elision ?
Every one is agreed that the rule does not hold when the monosyllable is
preceded by another monosyllable, i.e. when the ending is not one mono-
syllable but two. Still, for the sake of completeness, we shall add the
statistics of the double final monosyllable also, but not until the single
final monosyllable has been fully treated . And every one is agreed that it
does not apply to the mere interior of lines (as opposed to the hemistich),
e.g. Merc. 728 (Senar.) Illast — etiam vis nomen dicam? Nil agis ; Aul. 155
(Vers. Reiz.) His legibu' quam dare vis cedo, || nuptias adorna. (Other
examples in Klotz, p. 228.)
In the following lists an asterisk is prefixed to exceptions to the rule.

I. IAMBIC CATALECTIC
Iambic Septenarii :
Asin. 411 Libanum libertum. iam manu || emissu's ? Obsecrote;
*488 Nunc demum ? tamen numquam hinc feres || argenti nummum
nisi me (Emended to ni me) ;
493 Cui credi recte aeque putent. || Fortassis, sed tamen me ;
639 Secede hue, Libane, te volo. || Siquid vis. Obsecro vos ;
713 Atque ut deo mi hie immoles ||bovem; nam ego tibi Salus sum ;
Cist. 705 Quis me revocat ? Bona femina et || malu' masculus volunt te ;
722 Nunc ad te redeo : siquid est || opu', dice et impera tu ;
729 Involvulum, quae in pampini || folia intorta implicat se ;
735 Crepundia una. Est qiiidam homo ||qui illam ait se scire ubi sit ;
Cure. 493 Et nunc idem dico (tibi). Et || commeminisse ego haec volam te ;
512 Tacuisse mavellem. Haud male || meditate maledicax es ;
520 Quid, stulta, ploras? ne time; || bene hercle vendidi ego te
(Tribrach, a ' pure ' foot) ;
Epid. 358 Dedit mi ad hanc rem Apoecidem ; || is apud forum manet me ;
Mil. 1227 Ut tu inclutu's apud mulieres ! || Patiar quando ita Venus vult ;
1234 Ne oculi eius sententiam || mutent, ubi viderit me;
1238 Istuc curavi ut opinio||ne illiu' pulchrior sis;
1253 Ut, quaeso, amore perditast || te misera ! Mutuum fit ;
1261 Ita animu' per oculos meos || (meu') defit. Militem pol ;
Most. 175 Neque patiar te istanc gratiis || laudasse, quae placet mi ;
243 Videas earn medullitus || me amare. oh ! probus homo sum ;
Pers. 43 Alicunde exora mutuum. || Tu fac idem quod rogas me ;
316 Inspicere morbum tubm libet. || Ah, ah ! abi atque cave sis
(Tribrach, a ' pure ' foot) ;
Poen. 1241 Numquam mecastor reperies || tu istuc probrum penes nos ;
1242 Da pignu', ni nunc peieres, || in savium, uter utri det ;
Rud. 326 In navem ascendit, mulieres || avexlt ; hariolus sum ;
 FINAL MONOSYLLABLE 341

Iambic Septenarii {cont.) — te?;

329 eadem, sacerdos Veneria || haec siquid amplius scit ;
342 Sed quam mox coctum est prandium ? || Quod prandium, obsecro

363 Credo hercle anancaeo datum || quod biberet. ut ego amo te ! ;

1283 Abiudicata a me modo est || Palaestra, perditus sum ;
1 301 Ita quanto magis extergeo || rutilum, atque tenuius fit ;
1332 Dabitur talentum. Accede dum hue : Venus haec volo adroget te ;
Trin. ? ^234 Ut utramque rem simul exputem, || iudex sim reu'que ad earn
rem (Rather earn) ;
True. 148 Volo habere aratiunculam || pro copia hie apud vos ;
220 Nos divitem istum meminimus || atque iste pauperes nos.
Result for Iambic Septenarii. That the rule holds is clear. For the two
seeming exceptions are not real. And although it is true that the seventh
foot of an Iambic Septenarius is usually an Iambus (IV 10), such unanimity
can hardly be the result of chance.
Iambic Dimeters Catalectic :
Capt. 233 Dum id impetrant || boni sunt ;
Pers. 854 Sati' sumpsimu' sup||plici iam.
*855 Fateor. manus || vobis do (Emended to do vo.).
IAMBIC ACATALECTIC.
Since the last foot of any acatalectic iambic line must be an Iambus, our
field is limited to :
Iambic Hemistichs.
(A) Septenarii.
The only exceptions (for the fourth foot is normally an Iambus) are the
suspect lines in the list on p. 323, above.
III 18);
Cist. 312 Nimi' lepide exconcinnavlt hasc' || (aedes Alcesimarchus) (see
Mil. 361 Respicedum ad laevam : quis illaec est || (mulier ?, etc.)
(Emended to illaec quis est) ;
Poen. 1245 Et praedicabo quo modo vos || (furta faciatis multa. AP)
(Emended to furtificetis) ;
1265 Nam vostra nutrix primum me || (cognovit, etc. AP)
(Emended to prima) ;
Pseud. 1320 Heu, heu, heu ! Desine. Doleo. Ni || (doleres tu, ego
dolerem) (Heu heu des. A).
(B) Octonarii :
Amph. 260 Qui Pterela potitare rex || (est solitus, etc.) ;
1075 Ibo et cognoscam quisquis est. ||;
Bacch. 936 Non sunt tabellae, sed equu' quern ||(misere Achivi ligneum) ;
950 Iam duo restabant fata tunc ||;
*974 Ouadrigentos filios habet atq' || (equidem omnes lectos sine
probro) (Emended to'quidem) ;
342 APPENDIX D

Iambic Hemistichs (cont,)— Ill 18);

Capt. *I98 Nunc servitus si evenit, ei || (vos morigerari mos bonust) (cf.
yyj Tantum adfero quantum ipsus a || (dis optat, etc.) ;
??*92i Nam hi(c)quidem ut(I) adornat aut || (iam nil est aut iam
nil erit) (Rather Trochaic) ;
Epid. 39 Supersede istis rebu' iam. ||;
Men. 995 Quid stati' ? quid dubitati' ? iam || (sublime raptum oportuit) ;
999 Quo rapid' me ? quo ferti' me ? ||;
Merc. 137 Loquere id negoti quidquid est. ||;
? Pers. 1 Qui amans, egens ingressus est || (princeps, etc.) ;
Pseud. 151 Nempe ita animati (?) esti' vos ||;
Rud. 944 Enicas iam me odio quisquis es. ||.
(C) Dimeters:
Cure. 98 Salve, anime mi || (Liberi lepos) ;
Men. 976 Haec pretia sunt || (ignaviae) ;
1006 Qui liber ad || (vos venerit) ;
Most. 98 [mea] Haud aliter id || (dicetis) ;
Pers. 278 Dicis ubi sit || (venefice ?).
TROCHAIC CATALECTIC AND ACATALECTIC
The seventh foot of the Trochaic Septenarius is in Latin, as in Greek,
always a Trochee (or Tribrach). The final foot of the Trochaic Octonarius
or of a Trochaic Hemistich may of course, in any circumstances be a
Spondee, etc. Klotz (p. 48) seems to think of a monosyllabic ending of a Tro-
chaic line (or hemistich) requiring not the preceding foot but the preceding
metrum to be pure, and finds a proof of iambic ipsa (!) in Ter. Andr. 359
(Redeunti interea ex ipsa re ||). But there is no such rule; as a glance
through any hundred of Plautus' Trochaic Septenarii (or Octonarii) will
show (e.g. Capt. 201 Eiulatione haud opus est ||; Cas. 677 Plus quam cui-
quam. Quam ob rem ? Quia se ||) (see below, at end of section on Elision).
Indeed Klotz elsewhere (p. 228) corrects his own mistake, a mistake into
which he was led by not keeping separate his Iambic and Trochaic material.
An Iambic Septenarius ending like obsecro te implied for him a Trochaic
Octonarius ending (or the ending of any Trochaic hemistich) like obsecro te.
But on p. 228 he makes one of those false quantities which attest the absence
of Latin verse composition from his school-training, ' nltidior ' (!). And on the
preceding page (p. 227) his details of the Iambic Septenarius ending, which
he attempts to connect with a Saturnian ending, are not worthy of attention.
BACCHIAC (ACATALECTIC)
Bacchiac Tetrameters, etc. :
Amph. 551 Age i tu secundum. || Sequor, subsequor te ;
555 Fads ut tuis nul||la apud te fides sit ;
557 Scelestam, scelus, lin||guam abscidam. Tuus sum;
564 Dabo. Istuc tibist in |j manu, nam tuus sum ;
 FINAL MONOSYLLABLE 343
Bacchiac Tetrameters (cont.) —
568 Homo idem duobus || locis ut simul sit ;
569 Profecto ut loquor res || itast. Iuppiter te ;
571 Rogasne, improbe, etiam ||qui ludos facis me ? ;
636 Quin incommodi plus ||malique ilico adsit, ||boni si obtigit quid J
Aul. 128 Tibi proximum me || mihique esse item te ;
Bacch. 1 1 33 Cogantur quidem intro. Hau || scio quid eo opus sit;
Capt. 499 Bono publico, sic || ut ego feci heri quom ;
501 Eunt obviam gra||tulanturque earn rem ;
787 Hie illest senex doc||tu' cui verba data sunt (Fourth 1 purePaeon,
' foot) a;
Cas. 650 Malum pessimumque hie || modo intus (??) apud nos (Rather
modo in|tus apud nos ; Fourth Paeon, a ' pure foot ') ;
655 Tua ancilla, quam tu || tuo vilico vis ;
673 Men occidet ? An quid||piam ad te attinet ? Vah ! ;
692 Habet, sed duos. Quid || duos ? Altero te ;
704 Timor praepedit ver||ba. verum obsecro te ;
Cist. 21 Nemo aHenus hie est. || Merito vestro amo vos ;
*23 Hunc esse ordinem bene||volentes inter se (Accented inter-se) ;
686 Nulla est, neque ego sum usquam. || perdita perdidit me ;
Men. 573 (op)-timi maxime morem habent hunc ;
575 Bonine an mali sint, || id haud quaeritant ; res ;
587 Aut ad populum aut in iujjre aut ad iudicem rest ;
756 Nam pernicitas de||serit ; consitus sum ;
763* Nee quid id sit mihi certius facit, quid ;
764* Verum propemodum iam || scio quid siet rei ;
Merc. 343 Is rescivlt et vi||dlt et perdidit me ;
351 Nunc si dico ut res est || atque illam mihi me ;
352 Emisse indico, quemad||modum existimet me? ;
Most. 86 Ego, atque in meo cor||de, si est quod mihi cor ;
96 Atque hoc vosmet ipsi, || scio, proinde iiti nunc ;
120 Primumdum parentes || fabri liberum sunt ;
784 Heus Theopropides ! Hem ! quls || hie nominat me ? ;
794 Age (i), duce me. Num || moror? Subsequor te ;
871 Malum quom impluit ce||teros, ne impluat me ;
873 Boni sunt, (bonust) ; im||probi sunt, malus fit ;
Poen. 215 Neque is ulla ornandi || satis satietas est ;
227 Poplo cuilibet plus || satis dare potis sunt ;
245* Eius seminis mulieres sunt ;
Pseud. 252 Verte hac, puere, te. Non || licet colloqui te ? ;
254 Licetne, obsecro, bi||tere an non licet ? Vah ! ;
1 1 26 Iamne ilium comesu||rus est ? Dum recens est ;
? 1 1 27 Dum calet, dum datur, de||vorari decet iam ;
1 129 Populo strenui, im||probi mi usui sunt (mi imp. MSS.)' nex;t line) ;
? Rud. 264 Puellae. sed und' vos (Rather unde, with Ire vos beginning the
344 APPENDIX D
Bacchiac Hemistichs :
Amph. 556 Quid est ? quo modo ? iam || ;
558 Proinde ut commodumst et ||;
559 Tamen quin loquar haec ||;
560 Numquam ullo modo me || ;
563 Malum quod tibi di ||;
569 (see above) ;
636 (see above) ;
638 Parumper datast, dum ||;
? 640 || quia ille hinc abest quern || (ego amo praeter omnes) ;
645 Domum recipiat se ||;
Aul. 128 (see above);
130 Et mihi te et tibi (me) ||;
? Bacch. 1 1 Petrae, ferrum ubi fit ;
620 Malos quam bonos par || (magis me iuvare) ;
1 136 Iam illis decidit. non || (vides ut, etc.) ;
Capt. 229 Nam tu nunc vides pro || (tuo, etc.) ;
499 Bono publico sic || (ut ego, etc.) ;
785 Quod quom scibitur, (turn) ||;
Cas. 146 Vir siquid volet me ||;
670 Per omnes deos et ||;
684 Neque est neque fuit me ||;
692 (see above) ;
696 Quid uxor mea ? non || (adit ?, etc.) ;
828 Quaerunt, id volunt, haec ||;
839 Meast haec. Scio, sed ||;
Cist. 4 Qui magi' potueris mi || (honorem ire habitum ?) ;
29 Nostra copia, nil ||;
673 Quae in tergum meum ne || (veniant, etc.) ;
674 SI era mea sciat tarn ||;
Men. 575 (see above) ;
576 Magis quaeritur quam ||;
579 Sin dives malust, is || (cliens frugi habetur) ;
582 Datum denegant quod ||;
753 Ut aetas meast atq' || (ut hoc usu' factost) ;
763 Repente expetit me ||;
764 Velit. quid || (me accersit) ;
764a (see above) ;
766 Ita istaec solent quae ;
? 769 Verum est modu' tamen quoad || (pati uxorem oportet) (Or;
quoad)
968 Ut absente ero rem || (eri diligenter) ;
Merc. 342 Ratus clam patrem (me) || (meum posse habere) ;
344 Neque is quom roget, quid || ;
349 Dum rursum haud placet nee ||;
353 Atque illam abstrahat, trans ||;
 FINAL MONOSYLLABLE 345
Bacchiac Hemistichs (cont.) —
361 Muscast meu' pater, nil ||;
Most. 100 Simul gnarures vos ||;
*ioi Aedes quom extemplo sunt || (paratae, expolitae) ;
787 Quid illi, obsecro, tarn || ;
789 Antiquum obtines hoc ||;
? 868 Ut adhuc fuit mi || (corium esse oportet) ;
Pers. 698 || allatae modo sunt ||;
? 776 Ei qui invidet ml || (et ei qui hoc gaudet) ;
810 Perii ! perculit me ||;
816 Cave sis ne attigas, ne ||;
Poen. 210 Negoti sibi qui || ;
211 Navem et mulierem, haec || (duo comparato) ;
212 Nam nullae magis res ||;
216 At que haec ut loquor nunc || ;
? 222 Binae singulis quae ||;
? 224 (ag)- gerundaque aqua sunt || (viri duo defessi) ;
226 Sed vero duae, sat ||;
227 (see above) ;
? 245 Item nos || (sumus) ;
251 Quiesco. Ergo amo te ||;
? 258 lam num me decet hie || (donari) ;
Pseud. 247 Molestam obtulit. Qui ||;
248 Fuit. Mortuust qui ||;
253 At mi non libet. Sin ||;
256 Dedi dum fuit. Non ||;
1 105 Ex conspectu eri si ||;
1 130 Malum quod tibi di || ;
1252 Profecto edepol ego nunc || (Fourth Paeon, a ' pure ' foot) ;
1266 Dari dapsiles ; non ||;
Rud. 279 Neque hoc amplius [quam] quod || (vides vobis quicquamst) ;
282 Sed haec pauperes res ||;
Trin. 227 Sed hoc non liquet nee ||;
True. 212 Loquar libere quae || ;
554 Nam hoc qui sciam ne || (quis id quaerat ex me) ;
? 569 Quod des devorat (nee) || ((dat)is umquam abundat) ;
712 Ama id quod decet, rem ||.
Residt for Bacchiacs. The evidence, though not so overwhelming as it
seems (for the Bacchius is commoner than the Molossus at the end of any
Bacchiac line or hemistich), shows that the rule holds and makes the isolated
exception (Most. 101) suspect. So we must scan Amph. 572 as a curtailed
hemistich, Merito maledicas || (not ' Merito maledicas mi ||'). And Cas. 818
(Semper sis superstes) is rather an Ithyphallic colon (like 824 Obsecro,
memento) than a Bacchiac Dimeter (but see IV 20).
 APPENDIX D
;46
CRETIC (ACATALECTIC)
Cretic Tetrameters, etc. :
Most. 114 Atque (ea) haud est fabri ||culpa, sed magna pars ;
328 Sine, sine || cadere me (Fourth Paeon, a ' pure' foot) ;
? 336 Num non vis me obviam his ||ire, anime mi (Ditto. But perhaps
a ' miuric ' colon ; cf. 1 1 9) ;
720 Ouom me laudas. Decet. || Certe. quin hercle te ;
? Rud. 273 Unde nos hostias || agere voluisti(s) hue ? ;
276 Ut tuo recipias ||tecto servesque nos ;
Trin. 281 Nolo ego cum improbis||te viris, gnate mi.
Cretic Hemistichs :
Asin. 132 Faxo erunt, capiti' te ||(perdam) (Fourth Paeon, a ' pure ' foot) ;
* ? Bacch. 656 Improbis cum improbu' sit, || (harpaget furibus) (Delete sit ?)
Capt. 210 Unum exorare vos||;
■ 213 ||nos concedamus hue || ;
216 Vobis sumu' propter hanc ||;
Cas. 191 Mira sunt vera si || (praedicas, etc.) ;
232 Obsecro, sanun es ? ||;
621 Nulla sum, nulla sum ||;
Cist. 690 I lie nunc laetus est ||;
Cure. 100 Tu mihi stacta, tu ||;
1 19 Salve. Egon salva sim ? ||;
133 Hoc volo scire te ||;
Epid. 327 Numquam inridere nos ||;
Most. 108 Atque illud saepe fit ||;
697 Non bonust somnu' de || (prandio, etc.) ;
706 Exsequi certa res 11 (est ut abeam) ;
Pseud. 935 Sed vide, ornatus hie ||;
? Rud. *203 Lenior esset hie || (But in A perhaps a longer line. Cf. line
212, and see III 18) ;
207 Hoc quod induta sum || ;
210 Nee loci gnara sum ||;
238 Die ubi es. Pol ego nunc || (Fourth Paeon, a ' pure ' foot) ;
266 Ilico hinc imus haud ||;
True. 584 Ecquid auditis haec? ||;
589 Die ob haec dona quae || (ad me, etc.) ;
599 Illicinest ? Illic est ; || (me intuetur gemens).
Result for Cretics. The result for Bacchiacs is the same for this other
acatalectic Metre. The rule holds for Cretic lines and hemistichs. Although
the occurrences of a final monosyllable are not so numerous as in Bacchiacs,
they are numerous enough, especially when we compare them with the
Molossi found when another monosyllable precedes (see the list at the end of
this section). And we have no difficulty in collecting Molossi, etc. (ending
in one monosyllable) in the first and third feet of the Tetrameter : e. g. Rud.
233 Certo vox muliebris || aures tetigit meas ; Rud. 68oa Me vide. Si modo
id || liceatt vis ne opprimat.
 FINAL MONOSYLLABLE 347

ANAPAESTIC CATALECTIC
Anapaestic Septenarii :
Bacch. 1 160 Sed quid fstuc est ? etsi iam ego tpse (?) || quid sit prope scire
puto me ;
?*n68 Filios et servum ? an ego experior || tecum vim maiorem ?
Abm hinc (Or maiorem, an Octonarius) ;
?*n8o Vidi ego nequam homines, verum te||neminem deteriorem.
Ita sum (Or deteriorem, an Octonarius) ;
Cist. *26 Ut amicitiam colunt atque ut earn || iunctam bene habent inter se
(Accented inter-se) ;
Cure. 131 Quid est? ecquid libet? Libet. Etiam mi||quoque stimulo
fodere libet te ;
? Pers. *7&9 Hoc vide, quae haec fabulast ? pol hie quidem po||tant. ad-
grediar. o bone vir (An Octonarius ?) ;
? Pseud. 937 Nam si exoptem quantum dignu's || tantum dent, minu' nihilo
sit (Not nllo; III 11) ;
True. *6i8 Turn pol ego et donis privatus || sum et perii. Plane istuc est
(Accented istuc) ;
*62i Perii hercle hodie nisi hiinc a te abigo. || Accede hue modo,
adl modo hue (Proceleusmatic).
Anapaestic Dimeters Catalectic :
Cas. 742 Quid nunc ? quam mox || recreas me ? ;
Pseud. *i3i8 Em ! hoc ego num||quam ratu' sum ;
Rud. 2i6a Me nunc miseram es||se ita uti sum.
Anapaestic Octonarii :
Most. *862 Exercent sese ad cursuram, 11 fugiunt ; sed i si reprehensi sunt ;
? Pseud. *I78 Nam nisi mini penus annuus hodie ||con venit, eras poplo
(??) prostituam vos (Rather ann. con. cr. populo, a Septenarius) ;
? Rud *927 Nunc [haec] tibi occasio haec, Gripe, obtigit ut ||liberet ex
populo praetor te (praeter MSS.) ;
Trim *297 Nil ego Istos moror faeceos mores, ||turbidos, quibu' bom dede-
corant se.
Anapaestic Dimeters Acatalectic :
Bacch. 6i6a Nequior nemost ||neque Tndignior cui ;
Cas. 829 Noctuque diu ut ||viro subdola sis ;
Cist. 698 Vestigium m pul||vere : persequar hac ;
699 In hoc iam loco cum al||tero constitit. hie ;
Men. 363 Magi' quam domu' tua ||domu' quom haec tua sit ;
Pers. 772a Paegnium, tarde|| cyathos mihi das ;
??*8oo Litibu'. posterius istaec te (This Brevis Brevians is impossible.
Emended to te istaec) ;
? Pseud. *9i8 Stratioticus homo ||qui cluear iam (Var. lect. in A) ;
??Rud. 212 Aut viam aiit semitam ||monstret ; ita nunc (The first Brev.
Brev. is impossible) ;
348 APPENDIXD

Anapaestic Dimeters Acatalectic {cont.) —

*96i Volo ut dicas. Im||mo hercle etiam plus ;
Stich. 36 Non coliint, quom tu ||tuiim facis. Ita pol.

Anapaestic Hemistichs :
Aul. 150 Domurn ducere. Ei ! ||; Reiz.) ;
? 155 Sed his legibu' si ||;
Bacch. *67o Non placet nee temere est etiam. quin || (followed by Col.

1155* Pudet dicere me ||;

*n8o Vidi ego nequam homines, verum te ||;
1 186 Minime ; nolo, nil moror, sine sic ||; suls) ;
*i2co Quod mihi erat id ||;
1205 Vesper hie est ; sequimini. Ducitenos||;
? 1207 Lepide ipsi hi sunt capti, suis qui ||(filiis fecere insidias) (Rather

Capt. 498 Quid est suaviu' quam ||;

? Cas. *220 ebs eo condimento uno (non)|| (utier omnibu' quod praestat) ;
*722 Ita quoquo adveniunt, ubi ubi sunt ||(Hardly ubi ubi) ;
819 Tuaque ut potior pollentia sit|| ;
889 Emm iam magis adpropero, magi' iam ||;
Cist. 10 Ita m prandio nos ||;
213 Ita mi omnia sunt ||;
647 Sed is hac iit, hac ||;
700 Neque prorsum iit hac ||;
Cure. 97a Ubi ubist, prope mest ||;
Men. 603 Satis. Si sapiam, hinc intro abeam ubi mi ||;
Merc. 339 Ita mihi mala res ||;
?Mil. *I030 Aliquam mihi partem hodie operae des|| (denique iam tandem
ades t ilico t) (om. iam A) ;
? *I046 Quin tu huic responde, haec illaec est||(ab ilia quam dudum
<dixi>) (h. illic est MSS.) ;
*lo56 Eu hercle odiosas res ! quotiens hoc ||;
Most. 902 Ut esse addecet nequam homines, ita sunt ||;
Pers. *i8o Qui nil amat quid ei|| (homini opu' vitast ?) (Proceleusmatic) ;
*492 Ita me dJ anient ut ob istanc rem|| ;
764 ||accede ad me atque amplectere sis ||;
779 Dies datus hodiest ab dis quia te ||;
801 Magi' par ||(agerest) ;
? Pseud. 165 Nam mi hodie natalis dies est ||(Or dies ?) ;
*I75 Quae capiti, quae ventri operam det ||;
182 Cur ego vestem, aurum atque ea quibus est ||;
239 O Pseudole mi ||;
24oa Nunc tu sapis. It ||(dies, etc.) ;
941 Teneo. omnia in pectore condita sunt|| ;
 FINAL MONOSYLLABLE 349
Anapaestic Hemistichs {cont.)—
949 Lepide accipi' me. Iramo si effici', turn ||(followed by Col.
Rud. 927 Nunc [haec] tibi occasio, Gripe, obtigit ut ||;
? *935 Monumentum meae ||(famae et factis) (Or meae ?) ;
Stich. 37 Tace sis, cave sis ||;
*3I3 Defessus sum ||(pultando) ;
*322 Pudor adsit, non ||(me appelles) ;
♦324 Potes ; hodie non ||(cenabis) ;
? 326 Quisnam, obsecro, has ||(frangit fores ? ubtst ?) ;
327 Tun haec faci' ; tun ||;
Trin. *257 Ubi qui eget, quam ||;
*258 Apage te, Amor, non|| ;
1 1 18 Ita commoda quae||(cupio eveniunt) ;
True. ? *553 Satfn, siqui[s] amat, nequit quin nihili sit|| ;
*6o6 Istuc ne mihi responsis. Hoc || ; Libitumst) ;
*6o7 Quid tu ? cur ausa es alium te || (dicere amare hominem ?

? 612 Mebn [ero] tu, improbe, ero male dicere (nunc) ||(audes, fons
viti et periuri ?).
Result for Anapaestics. Clearly the rule does not hold for Anapaestic
Acatalectic lines (nor hemistichs). Nor (presumably) for Catalectic (fuller
evidence will be provided by the list of examples after Elision, below).

II. FINAL MONOSYLLABLE AFTER ELISION

Iambic Septenarii :
Asin. 383 Atque atriensem Sauream, || si est intus, evocato hue ;
*679 Age sis tu in partem nunciam hunc || delude atque amplexare
hanc (Emended to hancque amp.) ;
Men. *98o Propterea eri imperium exsequor, || bene et sedate servo id
(Delete id ? The next line is Atque id mihi prodest) ;
Poen. 1265 Nam vestra nutrix primum me || cognovit. Ubi ea, amabo,
est?
hanc;
Rud. 691 Sedete hie modo ; ego hinc vos tamen || tutabor : aram habete

1304 Verum extergetur. Ut vales? || Quid tu? num medicu', quaeso,

es ?
Iambic Hemistichs : MSS.) ;
Amph. 189 Exstincto duello maximo atq' || (internecatis hostibus) (Due. ex.
202 Prius ipse mecum etiam volo hie ||;
*256 Postridie in castra ex urbe ad || (nos veniunt flentes principes) ;
Bacch. 927 Armis, equis, exercitu atq' || (eximiis bellatoribus) ;
965 Item ego dolis me illo extuli e || (periclo, etc.) ;
Capt. 517 Nunc spes, opes auxiliaque a || (me segregant spernuntque se) ;
35° A P P E N D TX D
Iambic Hemistichs (cant,)—
Men. 133 Meo malo a mala abstuli hoc ||;
Merc. 126 At tu edepol sume laciniam atq' || (absterge, etc.) ;
Most. 127 Sumptu suo et || (labore) ;
?Poen. *H97a Quid est? Est lepida et || (lauta. ut sapit !) (Rather
Anap. Monom. Acat. with Iamb. Monom. Acat.) ;
Pseud. 185 Nunc adeo hoc factust optimum ut ||;
Stich. 291 Atque oratores mittere ad || (me, etc.).
Bacchiacs :
Amph. 561 Scelestissime, audes || mihi praedicare id? ;
Capt. 227 Tanta incepta res est ; || haud somniculose hoc ;
Cas. 650 Malum pessimumque hie ||;
739 Olympisce mi, mi || pater, mi patrone. Em ! ;
827 Facies tun hanc rem mi ex || parata imparatam ? id ;
832 Abite intro. Amabo, in||tegrae atque imperitae huic ;
? Cist.* 1 Quom ego antehac te amavi et || (mi amicam elide esse crevi) (Or;
the Pronoun)
675 Ouamne in manibu' tenui atq' || (accepi, etc.) (Fourth Paeon,

?Curc. *n6 (Adfert potionem et) || sitim sedatum it ;

116 (hemistich) ;
Men. 572 Molesto atque multo atq' || (uti, etc.) ;
Most. 84 Recordatu' multum et ||;
124 Poplo sint sibique haud ||;
314 Volo temperi. audi, em ! || (tibi imperatumst) ;
785 Ero servu' multis || modis fidus. Unde is? ;
Poen. 220 Lavari aut fricari aut ||;
243 Nisi multa aqua usque et ||.
Cretics :
Amph. 244 Equites parent citi : ab || (dextera maximo) ;
Bacch. 1 1 10 Numquidnam ad filiam haec ||;
Capt. 213 Fiat, abscedite hinc ||;
221 Neu permanet palam haec ||;
235 Ex bonis pessimi et || ;
835 Sed quis est ? Respice ad || (me, etc.) ;
Cas. 626 Intu' vidi novam atq1 || (integram audaciam) ;
952 Pro me ? quid nunc agam || nescio, nisi id ;
Cure. 107 Tangere, invergere in || (me, etc.) ;
151 Subsilite, obsecro, et ||;
Epid. 172 Revereor filium. At ||;
Most. 149 Cor dolet quom scio ut || (nunc sum, etc.) ;
700 Res parata est mala in || (vesperum, etc.) ;
708 Atque pol nescio ut || (moribus sient) ;
? 709 Vestrae : haec sat scio || quam me habe[a]t male (et) ;
714 Tempu' nunc est senem hunc ||;
 FINAL MONOSYLLABLE
Cretics (cont.) —
72 iB Heia! mastigia, ad || (me redi, etc.) ;
722 Quid nunc ? quam mox ? Quid est ? || Quod solet fieri hie ;
? 732 Nunc nobis com[mun]ia haec ||;
735 Processerunt. Ita ut ||;
736 Nos profecto probe, ut ||; 35i
Pseud. 259 Quod tibi detuli et ||;
1249 Pergitin pergere ? ah ! ||;
1293 (Vir ma||lus viro) Optim(o) || obviam it ;
1294 Di te ament, Pseudole. Hae! ||;
Rud. 216 Haec parentes meae haud ||;
253 Sed quid hoc, obsecro, est ? ||;
670 Orta in nos est modo hie ||;
68oa Me vide. Si modo id ||;
? True. 118 Si esse vis. Faxo erunt. || respice hue modo. Oh !
Anapaestic Septenarii :
? Bacch. *ii5i Ego ad hunc iratum adgrediar. pos||sumu' nos hos intro
illicere hue (Emended to adg. (si) pos., an Octonarius) ;
Anapaestic Acatalectic :
Bacch. 1082 Ego dare me meo gnato institui, ut ||;
1098 Relicuum id auri factum quod ego ei || (stultissimus homo
promisissem) ;
*i 108 Igitur pari fortuna, aetate, ut || (sumus, etc.) ;
*H99 Hanc veniam illis sine te exorem. Ut || (terebrat ! , etc.) ;
Cas. *228 Tristem adstare aspicio. blande haec ||;
"'825 Noctuque et diu ut ||(viro subdola sis) ;
877 Ita nunc pudeo atque ita nunc paveo atq' || (ita, etc.) ;
879 Operam date, dum mea facta itero : est || (operae pretium, etc.) ;
Cure. *I26 Hoc vide ut ingurgitat impura in || (se, etc.) ;
Men. *36o Nunc eum adibo atque || (adloquar ultro) ;
Mil. *loi3 Socium tubrum consiliorum et ||;
1027 Collaudato formam et faciem et ||;
*I029 Tu cetera cura et contempla et ||;
*I032 Ait illam miseram cruciari et ||;
*io52 Quid nunc vult ? Te compellare et ||;
*io88 Atque adeo (audin ?) dicito docte et ||;
? ? *io9o Hie cum era est ; qui etiam clam nostrum hunc || (sermonem
sublegerunt) (Doubtful reading) ;
*io9i Lepide factumst. iam ex sermone hoc ||;
*io93 lube maturare illam exire hue ||;
Pers. *i69 Nimi' tandem mequidem pro barda et |j;
*I72 Nam equidem te iam sector quintum hunc || (annum, etc.) ;
177 Amas pol, misera ; id ||;
*786 Quern pol ego ut non in cruciatum atq' || (in compedes, etc.) ;
*798 At, bona liberta, haec scivisti et ||;
352 APPENDIX D

Anapaestic Acatalectic (cont.) —

Pseud. *I76 Quam libertam fore mihi credam et ||:
*I77 Facite hodie ut mihi munera multahuc|| ;
179 Natalem scitis mi esse diem hunc ||;
241 || i prae, puere. Heus ! ||;
913 Fuit meum officium ut ||;
13 1 5 Onera hunc hominem ac me consequere hac ||;
Rud. *22i Ita male vivo atque ita mihi multae in ||;
*223 Omnia iam circumcursavi atq' || (omnibu' latebris perreptavi) ;
912 Miroque modo atque incredibili hie ||;
919 Tolerarem ; opera haud || (fui parcu' mea) ;
956* Noveram dominum id || (cui fiebat) ;
Stich. 327 Salve, tuo arcessitu venio hue ||;
True. 124 Fer contra manum et ||;
126 Quid agis ? Valeo et ||;
*7I4 Prome venustatem tuam amanti, ut || (followed by Col. Reiz.).
Result for Elision. Clearly the Elision makes no difference ; so we may
add this new evidence to the old. Even with the addition, the evidence for
Anapaestic Catalectic (the Septenarii) is hardly sufficient for absolute cer-
tainty. We must fall back on a priori reasoning and be content with
probability. Since it seems unreasonable that the Catalectic should have
had a different treatment from the Acatalectic line, we may presume that the
rule does not hold for Anapaestic Septenarii.
And to return to a point not strictly within the limits of our inquiry, Klotz's
supposed rule for Trochaic Acatalectic, that the metrum (not ' foot ') is pure
before a final monosyllable. For hemistich-endings there is abundant
evidence to prove Klotz to be wrong (see above, in the Trochaic section).
But of line-endings there are not a score of Trochaic Octonarii (or Acata-
lectic Dimeters, not always to be discriminated) in all Plautus and Terence
which end in a (single) monosyllable :
Amph. 584 Ut minu' valeas et miser sis,
Salvu' domum si rediero iam ;
Aul. 728 Atque hi(c)quidem Euclio est, ut opinor. || oppido ego interii :
palamst res ;
Bacch. 954 Signum ex arce si periisset ; || alterum etiamst Troili mors ;
969 Nummos Philippos militi, quos
Dare se promisit, dabit. nunc ;
Epid. 25b Nobis praeturam geris ? Quern ;
329 Quid ilium ferre vis qui, tibi cui ||divitiae domi maximae sunt ;
? *Pseud. 164 Versa, sparsa, terta, strata || lautaque coctaque (Or ut!)uti;
sint omnia

Ter. Andr. 245 Adeon hominem esse invenustum aut || infelicem quem-
quam ut ego sum ;
608 Mi obtigisse, quandoquidem tarn in||ers, tam nulli consili
sum (Or tam || in.) ;
 FINAL MONOSYLLABLE 353

*Eun. 21 7 Censeo. Sed heus tu ! Quid vis ? || Censen posse me offirmare

et (Perpeti, etc.) ;
744 Scin tu turbam hanc propter te esse || factam et adeo ad te
attinere hanc ;
Heaut. 581 Recte sane. Syre, pudet me. || Credo: nequeid iniuria : quin
(Mihi molestumst, etc.) ;
Phorm. 156 Quid istuc? Rogitas? qui tarn audacis || facinorisscius
mi con-
sis ;

? *Hec. 613 Quid vis, Pamphile ? Hinc abire || matrem ? minime.

Quid ita istuc vis (Or istuc).
That a different rule should hold for hemistich-ending and line-ending would
be incredible. But the strong testimony of the (very few) line-endings is
disconcerting. Luckily, now that we know that Elision makes no difference,
we can appeal to Ter. Eun. 217.

III. FINAL MONOSYLLABLE AFTER ANOTHER

MONOSYLLABLE
Iambic Septenarii :
Asin. *48i Dabitur pol supplicium mihi || de vestro tergo. Vae te ! ;
*6i4 Oh ! melle dulci dulcior |j tu es. Certe enim tu vita es mime?
; ;
659 Quin tu labore liberas || te atque istam imponis in me ? ;
700 Ten ego veham? Tun hoc feras || (hinc) argentum aliter a

Cist. *7io Nam dudum ut accucurrimus || ad Alchesimarchum, ne se ;sis ;

738 Quae illam cistellam prodidit || cuidam negat esse quod det ;
742 Cistellam habere. At vos Salus || servasslt ! ubi ea nunc est? es;t;
Epid. 345 Erum et Chaeribulum conspicor, || quid hie agitis? accipe hoc

370 Quasi pro ilia argentum acceperit || quae tecum adducta nunc

372 Aliquam dolosam fidicinam, || nummo conducta quae sit ;

*379 Servavit consiliis suis. || Abeamus intro hinc ad me ;
Merc. 503 Amabo ecastor, mi senex, |] eloquere. Exquire quid vis ; est;
520 De lanificio neminem || metuo, una aetate quae sit ;
523 Operam accusari non sinam || meam. Em istaec hercle res

Mil. 358 Quid ais tu, Sceledre ? Hanc rem gero. ||habeo aures ; loquere
quid vis ;
398 Scin te periisse ? Nunc quidem ||domi certost. certa res est ;
415 Palaestrio, o Palaestrio ! || O Sceledre, Sceledre, quid vis ? ;
902 Salve, architecte. Salva sis. ||sed die mihi, ecquid hie te ;
*9I2 Quasique anulum hunc ancillula ||tua abs te detulerlt ad me;es[t] ;
? *92o Si non nos materiarius ||remoratur, quod opus(t) qui det ;
1250 Quin earn intro. Occlusae sunt fores. || Effringam. S,ana non

2S48 A a
354 APPENDIX D

Iambic Septenarii (cont.) —

1270 Verbum edepol facere non potis, ||si accesserit prope ad te ;
Most. 209 Ut te ille amet. Cur, obsecro, ||non curem ? Libera es iam (Or
libera's) ;
*2I4 Numquam ego illi possum gratiam ||referre, ut meritust de me ;
Pers. 46 Hoc meumst, ut faciam sedulo. || Quidquid erit, recipe te ad me ;
♦284 Video ego te : iam incubitatus es. || Ita sum. quid id attinet ad te ? ;
287 Potin ut molestus ne sies ? || Quod dici' facere non quis ;
289 Vadatur hie me. Utinam vades ||desint, in carcere ut sis ;
ut vis ;
328 Sed sequere me : ad earn rem usus est||tua mihi opera. Utere

Poen. 239 Nimia omnia nimium exhibent ||negoti hominibus ex se;

1239 Quia annos multos filias ||meas celavisti' clam me ;
Rud. *303 Atque ut nunc valide fluctuat ||mare, nulla nobis spes est ;
321 Cum istiu' modi virtutibus ||operisque natu' qui sit ;
349 Capitalique ex periculo or||bas auxilique opumque hue ;
*683 Nisi quid re praesidium apparas, ||Trachalio, acta haec res est ;
? *684 Certumst moriri quam hunc pati ||(saevire) lenonem in me ;
1303 Adulescens, salve. Di te ament ||cum inraso capite. Quid fit ? ;
?*I305 Immo edepol una littera|| plus sum quam medicus. Turn tu
(Hardly medicu' ? ; III 17 C) ;
Trin. 234 Ut utramque rem simul exputem, ||iudex sim reu'que ad earn
rem (Hardly earn ; II 33) ;
True. *i87 Per tempu' subvenisti'. sed ||quid ais, Astaphium ? Quid vis ? ;
194 Peperisse audivi. Ah ! obsecro, ||tace, Diniarche. Quid iam ? ;
205 Iboigitur intro? Quippini ? ||tam audacter quam domum ad te ;
207 Quam mox te hue recipis ? Iam hie ero ; ||propest profecta
Iambic Dimeters Catalectic :
quo sum.
? Most. 877 Non eo ; moles||tu' ne sis (Or Trochaic ?) ;
Rud. 284. Veneris fanum, ob||secro, hoc est ?
Iambic Hemistichs :
? Asin. *844 Ea res me male habet at non eb ||(quia tibi non cupiam quae
velis) (Or eo ?) ;
Capt. *I96 Decet id pati animo aequo, si id || (facieti', levior labos erit) ;
Epid. 19 Quid tibi vis dicam nisi quod est ? ||;
Merc. *I2I Curaest negoti quid sit aut || (quid nuntiet, etc.) ;
? Pers. *42 Qui ipsus siti aret. Sicine hoc te ||(mi facere. QuidRogas ?) ;
faciam.

True. *235 Is amatur hie apud nos qui ||(quom dedit, id oblitust datum).
Bacchiacs :
Amph. 552 Scelestissimum te ar||bitror. Nam quam ob rem i
559 Tamen quin loquar haec||uti facta sunt hie ;
562 Domi te esse nunc qui hie ||;
633 Satin parva res est ? ||;

/
 FINAL MONOSYLLABLE
355
Bacchiacs {cont.) —
637 ||domo atque ipsa de me ||;
639 Noctem unam modo, atque is ||;
*639 ||repente abilt a me ||;
Aul. 123 Quamquam haud falsa sum nos||;
Bacch. 1 121 Ego atque hie. Quid hoc est ? ||;
*II32 Merito hoc nobis fit qui j|;
1 138* Ne balant quidem quom a ||;
Capt. 227 Tanta incepta res est ||;
228 Agendumst. Ero ut me ||;
781 Quanto in pectore hanc rem ||;
926 Quae adhuc, te carens dum hic|| (fui, sustentabam) ;
Cas. 689 Quid est ? Est— Quid ? Est quod ||;
690 Moram offers mihi. At tu ||;
739 Olympisce mi, mi ||(pater, etc.) ;
*827 Facies tun hanc rem mi ex ||;
856 Ludos visere hue in ||;
861 Poeta, atque ut haec est ||;
Cist. *3 Aperuisti tu atque haec||;
1 1 Accepisti apud te, ut ||;
21 Nemo alienus hie est ||;
680 Non sum scitior quae hos ||;
682 Nunc vestigia hie si ||;
692 Sed memet moror quom hoc || ;
695 Era ! Hem ! Est— Quid est? Haec ||;
Men. 578 Si est pauper atque haec|| ;
772 Sed id quidquid est iam ||;
971 Potiora esse cui cor ||;
Merc *35i Nunc si dico ut res est ||;
352 Emisse indico quern ad ||(modum existimet me) (Second Paeon,
a ' pure ' foot) ;
354 Scio saeva quam sit ||;
*357 lam hinc olim invitam ||domo extruslt ab se ||;
Most. 92 Quando hie natus est, ei ||(rei argumenta dicam) ;
99 Auscultate, argumen||ta dum dico ad hanc rem ;
*784 Heus Theopropides ! Hem ! quls ||(hie nominat me ?) ;
789 Antiquum obtines hoc||tuum, tardus ut sis ;
794 Age (i), duce me. Num ||;
801 Lucri quidquid est, id ||;
Pers. 497 Tabellas tene has, pellege. Hae quid ad me ? ||;
498 ||-fert, nam ex Persia ad med ||;
814 Atque hoc quod tibi sua||deo facis. Quid est id ? ||;
856 Abi intro in crucem. An me hie ||;
Poen. 251 Quiesco. Ergo amo te|| sed hoc nunc ;
? *256 Dignum Veneri pol cui ||(Or Venere ?) ;
257 Ecquid gratiae, quom hue ||;
A a 2
356 APPENDIX D

Bacchiacs (cent.) —
Pseud. 244 Redi et respice ad nos || ;
246 Quid hoc est ? quis est qui ||;
251 Perdat, quisquis es. Te|| ;
252 Verte hac, puere, te. Non ||;
253 At mi non libet. Sin ||tuamst quidpiam in rem ;
1246 Quid hoc? skin hoc Jit, ||pedes ? statin an non?;
1247 An id vultis ut me hinc ||;
Rud. 911 Salute horiae quae in ||;
918 Paupertatem eri qui et|| ;
Trin. 226 Magister mihi exer||citor animu' nunc est ;
230 Amorin med an rei ob||sequi potiu' par sit ;
232 Utra in parte plus sit ||;
True. 211 Tandem sola sum. nunc|| ;
*249 Transililt ad nos : eum|| (volo convenire) ;
453 Ego prima de me|| ;
457 Mater dicta quod sum, eb|| (magis studeo vitae) ;
? *464 nunc || me)
Puerperio ego nunc med || (esse aegram adsimulo) (Rather ego ;

*554 Nam hoc qui sciam, ne||quis id quaerat ex me ;

720 Molestusne sum ? Nunc ||.

Cretics :
Amph. *23i (Turn) pro se quisque id quod|| ;
Capt. 206* Nostrum officium quod est ||;
210 || sinite nos. Quidnam id est ? ||
Cas. 158 Faciam uti proinde ut est ||;
Cure. 109 Sine, ductim. sed hac ||;
119 ||quae siti sicca sum. At ||;
Epid. 168 Is adeo tu's. Quid est ? ||;
Most. 116 Usque mantant neque id|| ;
722 Quid nunc ? quam mox ? Quid est ? ||;
*723 Intus. Quid id est ? Scis iam ||;
739 Tuto in terra, Ei ! Quid est ? ||;
Pseud. 261 Nosce saltern hunc quis est ||;
262 Qui fuit ; nunc quis est || ;
1 1 12 Convenit neque is ||;
? 1 33 1 Nil profecto. I hac ||;
Rud. *209 Quae mihist spes qua me || (vivere velimj ;
237 Ampelisca ! Hem ! quis est ? ||;
270 Candidatas veni||re hostiatasque ad hoc ;
675 Par moriri est neque est ||; Wth
676 Rebu' miseris. Quid est ? ||;
True. ? 122 Si esse vis. Faxo erunt. ||respice hue modo. Oh ! ;
582 Iusslt orare ut haec ||;
725 Integrum et plenum ador||tast thesaurum. Quis est ?
 FINAL MONOSYLLABLE 357
Anapaestic Septenarii :
Bacch. *i 108 Igitur pari fortuna, aetate ut ||sumus, utimur. Sic est. sed tu ;
1 1 62 Pol vero ista mala et tu nihili. || Quid multa ? ego amo.
Amas ? Nat yap ;
non?;
*u63 Tun, homo putide, amator istac ||fieri aetate audes ? Qui
abs te ;
*H7o Senex optime, quantumst in terra, ||sine (me) hoc exorare

*ny6 Abm a me, sceliis ? Sine, mea Pietas, ||te exorem. tuExores
me t?e>; ;

*II78 Lepidum te ! At scin quo pacto me ad ||te intro

Mecumabducas?
ut sis ;

?Mil. *I049 Nam hunc anuliim ab tui cupientl ||huic detuli, hie porro (ad
ad nos ;
"1069 Quae numquam male de te meritast. || lube eampse exire hue

*io72 Quomq' me oratricem haud sprevisti, ||sistique exorare ex te ;

*io86 Propter (quam) opera est mihi. num quid vis ? || Ne magi'
sim pulcher quam sum ;
Most. *902 Ut esse addecet nequam homines, ita sunt. ||sed eo magi' cauto
est opu' ne hue ;
Pers. *492 Ita me di ament ut ob istanc rem ||tibi multa bona instant a me ;
Pseud. *937 Neque ego hoc homini quemquam vidi || magi' malum et
maleficum. Tun id mi ?
Anapaestic Dimeters Catalectic :
Bacch. 616 Ego esse aliis. cre||dibile hoc est ? ;
*i 155 Quid ais, tu homo? || Quid me vis?
Result of the Whole Inquiry. In tenui labor ! This study of a small
point of Plautus' technique may seem to have been unduly prolonged. But
if we have convinced every one, we have obtained clues to the emendation of
such lines as :
Bacch. 1 1 29 (Bacch. Tetram.) Vetulae sunt tthymiamaef. At || bonas
fuisse credo,
where the word required must be of Bacchiac form (or the like) ;
True. 718 (Bacch. Tetram.) Intromittam tu tpercet ut, etc.,
where tu perge ut must not be made a Molossus at the end of the first
hemistich ;
Men. 360 Nunc eum adibo atq' || adloquar ultro,
where the P-version offers a sound Anap. Dim. Acat., but the A-version
(ult. adl.) an unsound Iamb. Dim. Acat.
Most. 127 Sumptu suo et || labore,
is an Iambic Dimeter Catalectic, not a Bacchiac Dimeter Acatalectic.
Poen. 1197* Quid est ? Est lepida et || lauta. ut sapit !,
begins with an Anapaestic colon, not Iambic.
358 APPENDIX D

We have provided sure examples of some pronunciations, e. g. :

earn rem (not ' earn rem ', as in e. g. Capt. 501)
Trin. 234 (Iamb. Sept.) Ut utramque rem simul exputem, || iudex sim
reu'que ad earn rem ;
f (not ' te ', the emphatic form)
Cist. 1 (Bacch. Tetram.) Quom ego antehac te amavi et || mi
esseamicam
crevi ;

dud (not * dub ')

Epid. 27b (Iamb. Dim. Acat.) Lictores duo, || duo ulmei ;
satis (not the usual sati')
Asin. 446 (Iamb. Sept.) || suo odio. Heus ! iam satis tu (Not with
Hiatus at change of speaker).
But the most satisfactory gain is the correction of the scribes' inveterate
habit of substituting est for 'st, es for 's, etc. We know now that the correct
form is in these endings of Iambic Septenarii :
Asin. 451 non mirum factumst (not ■ factum est ') ;
614 certe enim tu vita es mi (not ' vita's mi ') ;
631 || huic quod dem nusquam quicquamst (not ' quicquam est ') ;
669 || quam vero indignum visumst ? (not ' visum est') ;
710 || proclivi, quamquam nequam's (not 'nequam es') ;
? Cist. 725 id ? quidnamst (not c quidnam est ', unless Idline
was isshort. The ;
defective)
Merc. 543 || ideo quia uxor rurist (not ' ruri est' AP) ;
True. 222 || amavit; aequum ei factumst (not * factum est' AP).
And we get fresh proof of Plautus' nicety in metrical matters. Our rule
may be formulated so : At the end of a line or hemistich an ' irrational ' foot
(i. e. exceeding the normal foot by one ' mora ') was not allowed with a
(single) final monosyllable. Such a Spondee was not regarded as a fair
equivalent for an Iambus; such a Molossus was not regarded as a fair
equivalent for a Bacchius or Cretic. So our rule may stand beside Plautus'
restriction of the ' irrational ' Spondee in the second and fourth foot of the
Senarius, etc. Just as he would allow a Senarius to begin with Antiqua
errantis (with harmony of ictus and accent in the second word), but not with
Antiqua errans (with clash), so he would allow an Iambic Septenarius or
Dimeter Catalectic, an Iambic hemistich, a Bacchiac line (or hemistich), a
Cretic line (or hemistich) to end with e. g. errantes (a single word) or e. g.
hanc nunc (preceded by a long syllable), but not with e. g. errant nunc. The
(single) monosyllable made the ' irrational ' foot too ' irrational ', too far
removed from the normal foot ; just as the clash of accent and ictus did in
the other case. And this is new proof of Plautus' care to weigh exactly the
actual pronunciation, the same care as we found in his regulation of the
Brevis Brevians (e. g. ut accurrit, but not * ut aiidivit ' ; oportebat, but not
' oboedibat ' ; agi, but not ' agri ' ; secas, but not ' aquas ' ; II 27).
 359

E. BIBLIOGRAPHY
If the reader is to use rightly this list (select, not exhaustive) of books and
articles on matters of Prosody and Metre, he must bear in mind certain
dates. The large Teubner edition of Plautus was not completed till 1894 ;
Studemund's apograph of the Ambrosian Palimpsest appeared in 1889 ;
Leo's book on the Plautine Cantica in 1897. Some of the earlier works in
the list are based on a mere trio of plays, the three edited by Ritschl.
Further he must remember that most of the works are dissertations written
by students at the end of their University course, and should not be regarded
as infallible.
To illustrate these two points briefly the two first books on the list may
serve, especially as they are unusually favourable specimens of this type of
literature.
Mr. Audouin's dissertation was written without the help of Leo's epoch-
making work, and includes under the term * Anapaestic ' some lines which
every one now knows to be in other Metres (e. g. Versus Diphilii). Before
accepting Mr. A's statement that this or that irregularity is tolerated by
Plautus in Anapaestic Metre, the reader would do well to satisfy himself that
the line is Anapaestic.
Mr. Ahlberg is often mentioned as having proved that accented syllables
maybe shortened by a Brevis Brevians in trisyllabic words, (1) when the
final is elided, (2) in the first foot of iambic or trochaic metre, (3) in anapae-
stic metre. If the reader will examine Mr. A.'s lists of examples he will
find that they include, e.g. (Corrept. Iamb. p. 18) Bacch. 1017 prius te cavisse
ergo quam pudere aequom fuit. (Mr. A. imagines that the first syllable of
cavisse was short) ; (ibid. p. 30) Cas. 455 ecfodere hercle hie volt, credo,
vesicant vilico (Mr. A. does not know that the e was long). Mistakes like
these are often made even by advanced scholars in countries where Latin
Verse Composition is not practised ; and we mention them merely to
emphasize the need of caution in accepting dissertation-statistics and disser-
tation-verdicts.
Ahlberg, A. W. De Proceleusmaticis Iamborum Trochaeorumque, i. ii.
Lund, 1900, pp. 161 + 30. (chap, iii On shortening of naturally long
vowels in polysyllables by a Brevis Brevians ; chap, iv On shortening
before quidem, etc. ; chap, v On Synizesis.)
De Correptione Iambica Plautina Quaestiones. Lund, 1901, pp. 96. (only
trisyllables and polysyllables are treated.) (with Excursus : De genetivo
pronominali in -ius exeunte.)
Annotationes in Accentum Plautinum. Lund, 1897, pp. 14.
(= Filologiska Fbreningen i Lund.) (on -que, -ne, etc.)
Audouin, E. De Plautinis Anapaestis. Paris, 1898, pp. xii + 290.
De la composition metrique des Cantica de Plaute. Paris.
(= Melanges L. Havet, pp. 3-13.)
Baesey Guilelmus. De Canticis Terentianis. Halle, 1903, pp. 50.
Baier, B. Meletemata Plautina.
(= Abhandl. Martin Hertz zum 70. Geburtstag dargebr., pp. 270-
282.) (on Acchilles, etc.)
 APPENDIX E
360
Below, E. De Hiatu Plautino. Berlin, 1885, pp. 41.
Birt, T. Ueber den Lautwert des Spiritus H.
(= Rhein. Mus. liv [1899], pp. 40-92 ; 201-247.)
Ueber Kiirzung trochaischer Worter.
(= Rhein. Mus. li [1896], pp. 240-272.)
Boemer, A. De Correptione Vocabulorum Natura Iambicorum Terentiana.
Minister, 1891, pp. 69.
Brock, A. Ouaestionum Grammaticarum capita duo. Dorpat, 1897, pp. 184.
JY (Excursus ii : De verbi nisi prosodia.)
Brugman, O. Ouemadmodum in Iambico Senario Romani veteres Verborum
Accentus cum Numeris consociarint. Bonn, 1874, pp. 53.
Brugmann, K. Zur Geschichte der hiatischen (zweisilbigen) Vokalverbind-
ungen in den indogermanischen Sprachen.
(= Berichte sachs. Gesellsch., phil.-hist. Klasse, 65, 139-218.)
Buchhold, L. De Paromoeoseos (Adlitterationis) apud veteres Romanorum
poetas usu. Leipzig, 1883, pp. no.
Buecheler, F. Zu Lucilius und zur altlateinischen Prosodie.
(= Archiv Lat. Lexikograph. iii [1886], pp. 144-146.) (on tiiqui-
dem, etc.)
Prosodisches zu Plautus.
(= Rhein. Mus. xli [1886], pp. 311-313.) (on Chius, Pellaeus, etc.)
Altes Latein xvii.
(= Rhein. Mus. xlvi [1891], pp. 236-238.) (on ter, cor, etc.)
Bursian's Jahresbericht. See all the Reports on Plautine literature, espe-
cially those of Seyffert.
Christ, W. Metrische Bemerkungen zu den Cantica des Plautus. Munich, 1 871.
(= Sitz.-ber. Bayer. Akad. 1871, pp. 41-82.)
Metrik der Griechen und Romer. Leipzig, 1874, pp. xii + 684.
Conradt, C. Die metrische Composition der Comddien des Terenz. Berlin,
1876, pp. 212. Stichische und lyrische Composition bei Terentius.
(= Fleckeisen's Jahrbiicher cxvii [1878], pp. 401-416.)
Ueber einige Eigenthiimlichkeiten des Verssch kisses bei Terenz.
(= Hermes x, pp. 101-110.)
Cretin, M. Ueber die Composition der Plautinischen Cantica. Berlin, 1865,
PP- 53-
Draheim, J. De Iambis et Trochaeis Terentii.
(= Hermes xv [1880], pp. 238-243.)
Esch, J. De Plauti Correptione Secundae Syllabae Vocabulorum Polysylla-
borum, quae mensura iambica incipiunt. Minister, 1897, pp. 116.
Exon, C. The Forms and Scansions of the Gen. and Dat. Cases of is, hie
qui in Plautus.
(= Hermathena xii [1902], pp. 208-233.)
The Relation of Metrical Ictus to Accent and Quantity in Plautine Verse.
(= Hermathena xii [1903], pp. 470-504.)
The Form and Prosody of the Compounds of iacio in the Present-Stem.
(= Hermathena xiii [1904], pp. 129-162.)
Did Plautus use Synizesis ?
(= Hermathena xvi [1910], pp. 121-143.)
Apriorism and some Places in Plautus. (in support of arnica, etc.)
(= Hermathena xvii [191 1], pp. 62-75.)
The Relation of the Resolved Arsis and Resolved Thesis in Plautus to the
Prose Accent.
(= Class. Rev. xx [1906], pp. 31-36.) (Ictus and Accent coincide
in a Resolution.)
 BIBLIOGRAPHY 361
Fabia.
(= Revue Philologie xvii. 32) On monosyllabic endings of Terence's
Senarii.)
Fleckeisen, A. Exercitationes Plautinae. Gottingen, 1842, pp. 54.
(on Perfect-forms of Compounds of eo.)
Francken, C. Woord- en versaccent bij Plautus. Groningen, 1873.
(= Versl. en Meded. afd. Letterk. iv, pp. 40-65.)
Franke, A. De Caesuris Septenariorum Trochaicorum Plautinorum et
Terentianorum. Halle, 1893, pp. 50.
Friedlaender, P. Zum Plautinischen Hiat.
(= Rhein. Mus. lxii [1904], pp. 75-85.)
(connects the Diaereses of Iamb, and Troch. Verse with the
Saturnian Diaer.)
Gottschalk, F. Senarius, qui vocatur, Terentianus comparatur cum Trimetro
Graecorum comoediae novae. Patschkau, 1893, pp. 34.
Gutjahr, E. Terenzische Betonungsfragen. Leipzig, 1888, pp. 17.
Hartenberger. De O finali apud poetas Latinos. Bonn, 191 1.
(shows that the shortening of -o began with iambic words.)
Hauler, E. Edition of Terence Phormio. Fourth edition, Leipzig, 1913.
(with Introduction partly on Terence's Metre and Prosody.)
Havet, L. L/S latin caduc. Paris, 1896.
(= Etudes a G. Paris, pp. 302-329.)
(= Melanges Boisaier, pp. 261-265.)
Hingst, T. De Spondeis et Anapaestis in Antepaenultimo Pede versuum
generis duplicis Latinorum. Leipzig, 1904, pp. 103.
Hoischen, G. De Verborum Accentu in versibus Plautinis observato quae-
stiones novae. Minister, 191 4.
(e. g. Trochaic.)
pro di immortales, an Iambic verse-opening ; di immortales, a

Jachmann, G. Studia Prosodiaca ad veteres poetas scaenicos Latinos

spectantia. Marburg, 191 2, pp. 46.
(on Brevis Brevians, Synizesis, etc.)
Die Prosodie der Baccheen und Cretiker.
(= Glotta 7 [1916], 39-72.)
Bemerkungen zur Plautinischen Prosodie.
(on Brev. Brev.) (= Rhein. Mus. lxxi [1917], pp. 527-547.)
Jacobsohn, H. Quaestiones Plautinae metricae et grammaticae. Gottingen,
1904, pp. 54.
(in support of Diaeresis after fourth foot of Senarius, second of
Troch. Sept., etc.)
Kiessling, A. Analecta Plautina. Greifswald, 1878, pp. 18.
(on some Cantica.)
Klotz,^ Grundziige altromischer Metrik. Leipzig, 1890, pp. x+ 590.
Koehler, H. De verborum accentus cum numerorum rationibus in Trochaicis
Septenariis Plautinis consociatione. Halle, 1877, pp. 86.
Krawczynski, E. De Hiatu Plautino. Breslau, 1906, pp. 56.
Kroll, W. Iambenkiirzung.
(= Glotta vii, pp. 152-160) (a criticism of Jachmann. With statistics
of Resolution in Bacchiacs and Cretics.)
Langen, P. Bemerkungen iiber die Beobachtung des VVortaccentes im
alteren lateinischen Drama.
(= Philologus xlvi [1886], pp. 401-420.)
362 APPENDIXE

Leo, F.
Plautinische Forschungen : zweite Auflage. Berlin, 1912, pp. vi + 375.
(chap, v Auslautendes S und M ; chap, vi Hiatus und Synalbphe
bei auslautendem AE.)
Die Plautinischen Cantica und die hellenistische Lyrik. Berlin, 1897,
pp. 115.
(*= Abhandl. Gesellschaft zu Gottingen. Neue Folge i.)
Der Saturnische Vers. Berlin, 1905, pp. 79.
(= ditto, viii.)
De Tragoedia Romana. Gottingen, 1910, pp. 22.
Lepperniann, H. De Correptione Vocabulorum Iambicorum, quae apud
Plautum in Senariis atque Septenariis Iambicis et Trochaicis invenitur.
Miinster, 1890, pp. 84.
Lieben, W. De Verborum Iambicorum apud Plautum synaloephis. Mar-
burg, 191 5.
(disproves Lachmann's assertion that Plautus leaves Iambic words
before a short vowel in hiatus.)
Lingius, C. Quaestionum Plautinarum i : de Hiatu in versibus Plautinis.
Breslau, 181 7, pp. 80.
Lucks, A. Quaestiones Metricae.
(= Studemund's Studien I [1873], PP- 1_75) (On Iambus in fifth foot
of Senarius, etc.)
Genetivbildung der lateinischen Pronomina.
(ibid. [1890], pp. 317-386.)
Commentationes Prosodiacae Plautinae. Erlangen, 1883, pp. 23.
(on hie, egoquidem, etc.)
Manning, R. C. On a supposed Limitation of the Law of Breves Breviantes
in Plautus and Terence.
(= Harv. Stud, ix [1898], pp. 87-95.)
Marx, F. Zwei Auslautsgesetze der katalektischen iambisch-trochaischen
Verse der altlateinischen Dichter.
(= Bericht. Sachs. Gesellschaft lix [1907], pp. 129-200.)
Mather, M. Quomodo iaciendi verbi composita in praesentibus temporibus
enuntiaverint antiqui et scripserint.
(= Harv. Stud, vi, pp. 84-151.)
Maurenbrecher, B. Hiatus und Verschleifung im alten Latein. Leipzig,
1899, pp. vii + 269.
Parerga zur lateinischen Sprachgeschichte und zum Thesaurus. Leipzig,
1916, pp. 281.
(e. g. on Dat. Sing, of is, hie, qui, res, spes.)
Meissner, C. De Iambico apud Terentium Septenario. Bernburg, 1884, pp. 39.
Die Cantica des Terenz und ihre Eurythmie.
(= Fleckeisen's Jahrbiicher, Suppl. xii [1881], pp. 467-588.)
Meyer, W. Ueber die Beobachtung des Wortaccentes in der altlateinischen
Poesie. Munich, 1884, pp. 120.
(= Abhandl. Bayer. Akad. xvii).
Mohr, P. De Iambico apud Plautum Septenario. Leipzig, 1873, PP- 32«
Mbller, C. Quaestiones Metricae de Synaloephae qua Terentius in versibus
iambicis et trochaicis usus est ratione. Miinster, 1896, pp. 55.
Miiller, C. F. W. Plautinische Prosodie. Berlin, 1869, pp. xi + 800.
Nachtrage zur Plautinischen Prosodie. Berlin, 1871, pp. xvi + 159.
Peters, J. On short Vowels before Mute + Liquid in Plautus. Can they act
as Breves Breviantes ?
(= Harv. Stud, ix [1898], pp. 11 5- 120.)
 BIBLIOGRAPHY 363
Podzaski, O. Die trochaischen Septenare des Terenz, mit besonderer
Beriicksichtigung der Hecyra. Berlin, 1894, pp. 27.
Proskauer, Carola. Das auslautende -S auf den lateinischen Inschriften.
Strasbourg, 1910, pp. 208.
Radford^ R. On the Recession of the Latin Accent in connexion with
monosyllabic words and the traditional word-order, i, ii, iii.
(= Amer. Journ. Phil, xxv [1904], pp. 147-162 ; 256-273; 406-427.)
The Prosody of ille : a study of the anomalies of Roman Quantity.
(= Amer. Journ. Phil, xxvii, pp. 418-437; xxviii, pp. 11-33.)
The Latin Monosyllables in their relation to Accent and Quantity in
Terence.
(= Trans. Amer. Phil. Ass. xxxiv [1903], pp. 60-103.)
Studies in Latin Accent and Metre,
(ibid, xxxv [1904], pp. 33-64.)
Plautine Synizesis: a study of the phenomena of Brevis Coalescens.
(ibid, xxxvi [1905], pp. 158 sqq. ; xxxvii, pp. 15 sqq.)
Notes on Latin Synizesis.
(= Class. Phil, iii [1908], pp. 153-168.)
Ramain, G. Me'trique Plautinienne.
(= Revue Philologie xxix [1905], pp. 205-236.)
La loi du pied ane'tpe'nultieme dans le texte de Terence.
(= Revue Philologie xxx, 31.)
Ritschl, F. Opuscula ii. Leipzig, 1868, pp. xxiii + 813.
Neuen Plautinischen Excurse i. Leipzig, 1869.
Roppenecker, H. De Emendatione Metrica Canticorum Plautinorum.
Freising, 1894, pp. 41.
Zur Plautinischen Metrik und Rhythmik i, ii. Frankenthal, 1900, pp. 36 ;
1901, pp. 47.
Sc/dee, F. De Versuum in Canticis Terentianis Consecutione. Berlin, 1879,
pp. 76.
Schrader, P. De Particularum -ne, anne, nonne apud Plautum Prosodia.
Strasbourg, 1885, pp. 46.
Seyffert, O. Quaestionum Metricarum Particula : de Bacchiacorum Ver-
suum, etc. Berlin, 1864, pp. 46.
Skutsch, F. (Plautinisches und Romanisches : Studien zur Plautinischen
Prosodie.) Forschungen zur lateinischen Grammatik und Metrik i.
Leipzig, 1892, pp. 186.
(On ill' for preconsonantal ille, etc.)
Iambenkiirzung und Synizese. Breslau, 1896, pp. 26.
(= Satura Viadrina, pp. 122 sqq.)
Zur lateinischen Wortgeschichte und Plautinischen Versmessung.
(= Philologus lix, pp. 481-504.)
Ein Plautinisches Canticum.
(= Hermes xlv [1910], pp. 619-623.) (On Epid. 166 sqq.)
Iambenkiirzung und Synizese.
(= repay. Festschrift fur Fick [1903], pp. 108- 151.)
De Lucilii Prosodia.
(= Rhein. Mus. xlviii [1893], pp. 3°3-3°7>)
Ennius (in the Pauly-Wissowa Encyclopaedia).
Em.
(= Arch. Lat. Lexikogr. xi. 429.)
Calefacere.
(= Zeitschr. oesterreich. Gymn. v [1901], pp. 3-5.)
Su alcune forme del verbo latino. Rome, 1903.
(= Atti Congresso Internaz. Stor. ii, pp. 1 91-204.)
364 A P P E N D I X E

Sonnenburg, P. De Versuum Plauti Anapaesticorum Prosodia. Bonn, 1881.

(= Festschrift Bonner Philol. Seminars zum Buecheler-Jubilaum,
pp. 16-29.)
Sonnenschein, E. The Law of Breves Breviantes in the light of Phonetics.
(= Class. Phil, vi [1911], pp. 1-11.)
Accent and Quantity in Latin Verse.
(= Class. Rev. xx [1906], pp. 156-159.)
Spengel, A. Reformvorschlage zur Metrik der lyrischen Versarten bei
Plautus. Berlin, 1882, pp. 430.
T. Maccius Plautus : Kritik, Prosodie, Metrik. Gottingen, 1865, pp. viii
+ 240.
Studemund, IW. Studien auf dem Gebiete des archaischen Lateins, i, ii.
Berlin, 1873-1891.
(dissertations by S.'s pupils.)
De Canticis Plautinis. Halle, 1864, pp. 94.
Duos — duo.
(= Archiv lat. Lexikogr. iii [1886], pp. 550-552.)
Sturtevant, E. The Nom. and Dat. Plural of deus and meus in Plautus.
(= Class. Quart, iii [1909], pp. 8-12.)
The Coincidence of Accent and Ictus in Plautus and Terence.
(= Class. Philology xiv [1919], pp. 234-244.)
Sudhaus, S. Der Aufbau der Plautinischen Cantica. Leipzig, 1909,
pp. viii + 1 54.
(With it read Leo's review in Gotting. gelehrt. Anzeig. 1911,
pp. 65-104).
Ufipgren, A. Ueber sprachliche und metrische Komposition und Kunst
des Terenz. Lund, 1901, pp. 201.
Vahlen, J. Ueber die Versschliisse in den Komodien des Terentius. Berlin,
1900, pp. 60.
(= Abhandl. Berliner Akademie.) (On monosyllabic endings, etc.)
Ueber Fragen der Verstechnik des Terentius. Berlin, 1901, pp. 17.
(= Sitz.-ber. Berlin. Akad.)
Kritische Bemerkungen zur Verstechnik des Plautus. Berlin, 1907, pp. 15.
(ibid.) (on supposed hypermetric lines.)
Vollmer, Fr. Iambenkiirzung in Hexametern.
(= Glotta, viii, pp. 130-137.) (in Ennius.)
Kurzung durch Tonanschluss im alten Latein. Munich, 1917, pp. 32.
(= Sitz.-ber. Bayer. Akad.)
(on siquidem, etc. ; ' slquis ', etc. ; ' Ille ' ; ' quomodo ' ; Immo.)
Wallstedt, E. Enklisis oder Nicht ? Zur Betonung des Possessivums bei
Plautus und Terentius. Lund, 1906, pp. 31.
( = Filol. Foren. i Lund.)
Till fr&gan det dubbeljambiska versslutet hos Plautus.
(= Eranos vii, pp. 27-43.)
Studia Plautina. Lund, 1909. 0
( = Lunds Universitets Arsskrift.) (on Brevis Brevians Law.)
Wedding, G. De vocalibus productis Latinas voces terminantibus.
(Bezzenberger's Beitrage xxvii [1902], pp. 1-62.)
Wengatz, C. De Plauti Senariorum Iambicorum Compositione Artificiosiore.
Marburg, 19 10, pp. 106.
Wiebe, O. De Versus Sententiaeque Concinnitate apud veteres poetas
Romanos. Gottingen, 1909, pp. 76.
Winter, J. Ueber die metrische Reconstruction der Plautinischen Cantica.
Munich, 1 880, pp. 80.
 BIBLIOGRAPHY 365
Witte, K. Der Hexameter des Ennius.
(= Rhein. Mus. 69 [1914], pp. 205-232.)
(e. g. ascribes to the influence of Saturnian Verse some departures
of Ennius from the Homeric type.)
Zander, C. Versus Saturnii, Lund, 1918.
(= Festskrift Lunds Universitet.)
(quantitative theory, but not quite Leo's, though, like Leo's, based on
Wilamowitz's ■ Viersilbler '. Interesting parallel lists of Plautine
cola and Saturnian cola. The first sound attempt to connect
with Saturnian versification the peculiarities of Republican
Dramatic verse.)
[This list was completed for the printer in 1919 ; but two subsequent
publications demand mention (along with the new edition of C. I. L. I, Pars
Posterior, i):
Gagner, A. De hercle, mehercle ceterisque id genus particulis priscae poesis
Latinae scaenicae. Greifswald, 1920.
(a careful and thorough work, with sound arguments for the
quantity edep5l, but pol. [So on pp. 20, 236, above, Most. 376
and Cist. 510 would need no emendation.] Also for the
invariable ictus mehercle, mec&stor.)
Marx, F. Molossische und Bakcheische Wortformen in der Verskunst der
Griechen und Romer. Leipzig, 1922.
(= Abhandl. sachs. Akad., philol.-hist. Klasse, xxxvii, i)]
 INDEX A
(see also pp. 188-221)
-a 116-17 Dactylic Metre 307-11
Accent (see Secondary) 33 ; recon- Dactylo-trochaics 302-5
ciled with Ictus 18-33 J 316-23 ; de-
Dance150 261-3
revealed by Brevis Brevians 25 ; deinde 71
retrogressive in Elision 34; at-
tracted byEnclitic 34 deus 63
Demonstrative Pronouns 163-77
Accheruns 147-8
Achilles 147-8 Diaeresis 107-10; 83; 231-7; 323-
adhuc 119 dice 183
nte
Adverb, Prosody of 185-6 qui
age age 27 ; 317 die ilius 7 1 us
iph s
age siquid agis 27 ; 317 dDiu od6y3 Ver 305
alius 177-8 Dip Law 11
alter 68 ; 146
amicus amici 89-90 ; 130 dives, -vitiae 142
divinus 142
Anapaest, Caesura of 86-92 ; 97-100
Anapaestics 296-302 Division
Caesura)of51Feet between Words (see
ant{e) 149
apud 40 Dochmius 311-12
atque 71 donique 122
duce
Double183 Final Consonant 1 38-9
Bacchiacs 289-91 dudum
duo 63 63
Bibliography 359-65
Brevis Brevians 35-59; 113

Caesura 78-80 ; of foot 80-107 -/dropped 71-3 ; for -is 132-3

Calabri versus 294
ego 158; 23-4; e. eo 18; 22
calefacio &c. 39 ; 47 eho an non 25-6
Cantica 260-5; 3I3"I5 5 *n MSS. 129-33
Elision 255 ; of Diphthong 61 ; of -s
263-5
Chorus 262-3 Enclitic 34-5
clavator 142 Final)
Ending of Line 31-2 ; of Word (see
coagmentum 34
coegi 149
coepulonus 34 Ennius 307-9 ; ' Hedyphagetica ' I ;
Brevis Brevians 42-4
comptionalis 33 -eo- in Synizesis 60-1
Consonants, Final 118-39 eo ego 18; 22
contra 11 6- 17 etsi
esse 187
71 ; (e)st 74-6 ; 127-8
controversia 142
cor 139 eugae eugae 27 ; 317
Cretics 291-6 Eupolideus
extra 117 Versus 305
Cretic Ending of Senarius 273
cuius 65; 174 Extra Metrum 266
cuiusmodi 166
Final (see Vowel; see Consonant;
-d 1 20-1 see Monosyllable) pyrrhic ending
101-3 90-2 ; 98
Dactyl, Caesura of 96-7 ;
-word 106-7 First Foot 56
 INDEX
368 mzser sum 27 ; 320
fortasse 131 miserum me 27 ; 319
Fourth Paeon (Senarius-ending) 31-2
Fragments of Republican Poetry 2 Miuric Verse 294-5 ; 311
modo 36-7 ; 39
Glyconics 302-5 Monosyllable
mulier 125 Final 110-11 ; 339-5*.
Greek Loan-words
46-9 Mute and Liquid 76; 256; resist
Brevis Brevians 45
h 139-40
habe bonum animum 27 ; 318
nempe 71
heu me miserum 27 ; 318
heus ubi est is? 18 nescioquis 173-4
Hiatus 221-57; 331-9; at Diaeresis neve, neu 71-2
io7-!o nihil um, nil 12 1-2
hie 119-20; 163-72; h. illest 27; nimis 126
?zoenu
nolo 185122
o c
h er a t , h. est 27 ; 319
3i8 non tu scis t 18 ; 23
nonne 72
z',/62; 141
Iambics (see Senarius) 267-78 ; Sep- Noun, Prosody of 152-7
tenarii 274-6 ; Octonarii 277 ; numeri innumeri 260
Dimeters 278
iacio (Compounds) 140 obliviscor 142
Iambus-word
ibi 37 105 omnis (after Brevis Brevians) 51-2
optitne (after Brevis Brevians) 35
idem 59-60; 177-8
-ie- in Synizesis 61-2 pane 131
We 55 ; 71 ; 163-72 ; illius 66-8 Parasitic Vowel 144-5
inde 71 Personal Pronouns 157-61
Indefinite and Interrogative Pronoun Philippzis 40
plus 144
172-7
Ionics 305-7 Plautus -text 4 ; Brevis Brevians
ipse 177-8
is 163-72 Possessive Pronouns 161-3
iste 55 ; 163-72 ; istius 66-8 ; istuc possum 184-5
138-9
ita 116; z. fact am 84-5; z. me dz pote
prae- 127-8
151
ament 228-9 Prepositions,
45-8 Prosody of 11 6-7; in
Ithyphallic Colon 286-7 Compounds 149-52
Priapeus Versus 305
-/121-2
lac 139
prius 126
pro- 151
Proceleusmatic 93 ; 276 ; Caesura of
Line, divided between speakers 57-9 ;
and Sentence 1 n-12 ; -division in 94-5 ; 103-5 ; -word 107
MSS. 266-7 profecto 35
Lucilius 310; Brevis Brevians 44 proznde 71
Pronouns, Prosody of 157-79 ; slurred
-m 122-3fore 237 5 l°ng vowel retained be- Gen. Sing. 64-70
proprius 144
machaera 148 prorsus 142
magis 126-8 ; 131 ; 135 Prosodic Hiatus 226-30 ^
malam crucem 271 ; 281 Prosody 11 3-221
malo 185
mane inane 27 ; 319 puncto tempore 1 3 1-2
Pyrrhicsura of)Word (see Tribrach, Cae-
tried 159-60
met 139
Metre 260-316 quadrigenti 77
miy mihi 159
quamvis 186
 INDEX A
369
Quantity 49 ; Variation of 255-7 Slurred Pronunciation 257-9
"uemadmodum 187 smaragdus 40
i amant, q. habet 226-8 Spondee for Iambus 11-18.
aid ego nunc faciatn ? 18 ; 22 ; q. supra 117
' est quod 27; 320; q. hoc clamoris? Syllaba Anceps (at Diaeresis) 107-10
27; 320; q. istuc verbi est? 27; Synaphea 112
321 ; q. it at 95 ; q. ofiust verbis? Syncope of Syllable 145-6 ; of Foot
27 ; 321
is 9-64; 41
quippe 73 Synizes 5 1 ; 143
quisque, quisquis 173
#7^/ ///* rtfor// 25 -/ 135-8
quoins (see cuius) tametsi
312 187
-r 123-5 ted 159-60
tempore (see puncto tempore)
r<?-, ra/- 1 5 1-2
Terence -text 4 ; Brevis Brevians
Reizianus Versus 279 ; -num Colon
279-81 iibi48-9
ego dico 27 ; 321
Responsion 312-13
rest 75 Tribrach, Caesura of 81-6 ; 97-100 ;
retis 131 -word 105-6; -word (or -ending)
with clash of ictus and accent 20-2
-ris, -re 131
Trochaics 281-9; Septenarii 282-5;
-s 126-35 tuteOctonarii
tibi 28 ; 285
321 ; Dimeters 285-6
sagitta 77
satelles 77
satis 126-8 Uy V 14 I "3
uae capiti tuo 28 ; 322 u. fmsero
Saturnian Metre 8-1 1 ; influence on mihi 322
Dramatists 10
ubiestis? 19
Secondary Accent (succumbs to
Brevis Brevians) 52 uelut 186
Senarius 13 ; 269-74 -ui-
unde in72Synizesis 60 ; 62
Shortening of Vowel before Vowel uolo scire 19 ; 23
H3-4
sicut 186 uolup 128
stem, sim 184 uoluptas mea 26
simillimae sunt 52
sine modo 27 ; 321 Verb, Prosody of 178-85
stquidem, &c. 73-4 Vowel, Final 1 15-18

Bb
 INDEX B
LINES WHOSE FORM OR SCANSION IS DISCUSSED

Amph. fr. x (see p. 70) ; 31 (P- 88) ; 55 (p. 178) ; 69 (p. 97) ; 84 (p. 97) ; 91
(p. 112); 94 (p. 83); 141 (p. 70); 145 (p. 253); 161-4 (p. 265); 169
(p. 312); 214 (p. 246); 224 (p. 117); 267 (p. 103); 346 (p. 66); 405
(p. 72) ; 410 (p. 54) ; 429 (p. 58) ; 463 (p. 92) ; 518 (p. 54); 523 (p- 105) ;
572 (p. 205; p. 345); 661 (p. 103); 672 (p. 142); 699 (p. 246); 723
(p. 105); 751 (p. 103); 873 (P- 89); 897 (p. 320); 903 (p. 82); 978
(p. 82) ; 989 (p. 54) ; 991 (P- 272); 1050 (p. 54) ; 1064-5 (P- 287) ; 1 131
(p. 246); 1 135 (p. 153).
Asin. 33 (p. 92) ; 36 (p. 87); 59 (P- 89) I 100 (p. 17); 108 (p. 23); 119(^248);
120 (p. 178) ; 123 (p. 73) ; 126 (p. 37) ; 130 (p. 294); 136 (p. 294) ; 203
(p. 149); 250 (p. 235); 258 (p. 246); 269 (p. 171); 330 (p. 248); 342
(p. 178); 403 (P- 130); 4o5 (p. 89); 446 (p. 358); 45i (P- 358); 478
(p. 326) ; 483 (P- 58) ; 501 (P- 143 5 P- 179) I 54° (P- 209) ; 558 (p. 97) ;
614 (p. 358); 616 (p. 227); 631 (p. 358) ; 666 (p. 145) 5 669 (p. 358);
675 (p. 168) ; 702 (p. 168) ; 710 (p. 74 ; p. 358) ; 7M (p- 97) ; 721 (p. 326) ;
755 (P. 79 ; P. 242) ; 762 (p. 153) 5 781 (p. 255) ; 794 (p- 97) 5 814 (p. 140) ;
831 (p. 84) ; 885 (p. 228) ; 921 (p. 242).
Aul. 19 (p. 88); 135 (p. 84); 140 (p. 123); 198 (p. 105); 305 (p. 97); 321
(p. 54) ; 380 (p. 17) ; 393 (p. 288) ; 395 (p. 87) ; 542 (p. 228) ; 565 (p. 17) ;
619 (p. 227) ; 658 (p. 206) ; 692 (p. 87); 715 (p. 209).
Bacch. 78 (p. 58) ; 83-4 (p. 284) ; 125 (p. 159) ; 142 (p. 130) ; 146 (p. 83) ;
168 (p. 84) ; 507* (p. 87) ; 583 (p.
(p. 89
89 ;; p.
p. 246)
246) ;; 665
665 (p.
(p. 293)
293) ;; 725
725 (p. 51) ;
752 (p. 162); 787 (p. 89); 797 (p. 142); 799 (P- 168); 806 (p. 17); 844
(p. 17); 885 (p. 89); 893 (p. 207); 966 (p. 97); 1006 (p. 53); 1017
(p. 143); 1083 (p. 153); 1097 (p. 143); 1105 (p. 153); 1129 (P- 357);
ii57(P- 53); 1177 (P- 30i)-

(p. 103) ; 942 (p. 71) ; 1033 (p. 166).

Cas. 23 (p. 188) ; 41 (p. 79) ; 68 (p. 87) ; 114 (p- 87) ; 217 (p. 303 ; p. 305) ;
230 (p. 97); 239 (p. 103); 242 (p. 51); 311 (p. 82) ; 432 (p. 95); 488
(p. 55); 496 (p. 192); 578 (p. 75); 634 (P. 58); 673 (p. 264); 734
(P- 163); 75i (P- 9o); 755a (P- 58); 784 (P- 97); 79o (P-23); 795
(p. 227) ; 815 (p. 57 ; p. 305) 5 818 (p. 345) ; 827 (p. 266) ; 829 (p. 266) ;
837 (p. 145); 849-51 (p- 159) ; 917 (p- 145) ; 919 (p- 208).
Cist. 1 (p. 358); 62 (p. 53); 95 (p. 58); 96 (p. 227); 119 (p. 227); 156
(p- 79); 273 (p. 227); 292 (p. 55); 296 (p. 82); 297 (p. 141); 463
(p. 56); 510 (p. 365); 516 (p. 51); 561 (p. 219); 672 (p. 292); 678
(p. 163) ; 686 (p. 290) ; 704 (p. 146) ; 725 (p. 358) ; 766 (p. 87).
Cure. 11 (p. 145); 55 (P- 208); 88 (p. 82; p. 85); 131 (p. 173; p. 266);
142 (p. 227); 158 (p. 284); 170 (p. 57); 179 (P- 161); 220 (p. 203);
244 (p. 203); 271 (p. 89); 344 (p. 150) ; 358 (p. 253) ; 367 (p. 131);
368 (p. 250); 395 (p. 82); 467 (p. 248); 488 (p. 326); 508 (p. 205);
517 (p. 82); 519 (p. 151; p. 250); 520 (p. 90); 525 (p. 151); 531 (p. 175;
p. 246); 557 (p. 174); 7ii (P- 246).
 INDEX B 371

Epid. 27* (p. 358) ; 65 I p. 150); 173 (P- 266) ; 183 (p. 217) ; 202 (p. 158) ;
243 (p. 282); 356(p.69; p. 326); 365 (p. 326); 378(^326); 392 (p. 87) ;
488 (p. 244); 518 (P- 84); 532 (p- 89); 537 (P- 137); 563 (P- 154); 626
(P. 157).
Men. 21 (p. 82) ; 34 (p. 166) ; 236 (p. 81) ; 344 (p. 142) ; 3&> (p. 357) ; 393
(p. 162) : 434 (p. 75); 508 (p. 89) ; 530 (p. 97); 593 (p. 142); 777
(p. 168); 812 (p. 69); 848 (p. 70); 855 (p. 70); 1113 (p. 150); 1135
(p. 69).
Merc. 8 (p. 162); 35 (p. 75); 95 (P- *7) J 108 (p. 170); 227 (p. 82); 259
(p. 240); 270 (p. 168); 330 (p. 244); 352 (p. 187) ; 385 (p. 102); 482
(p. 183); 513 (p. 88); 523 (P- 89); 532 (p. 97); 534 (p. 326); 536a
(p. 89); 540 (p. 97); 543 (P- 358); 580 (p. 89); 619 (p. 159; p. 250);
653 (p. 247) ; 662 (p. 162) ; 683 (p. 51) ; 751 (p. 58) ; 773 (P- 89) ; 827
(p. in); 846 (p. 143); 857 (p. 75); 880 (p. 132); 884 (p. 266); 900
(p. 87) ; 997 (P- 250).
Mil. 4 (p. 252); 12 (p. 87); 18 (p. 145); 24 (p. 241); 136 (p. 61); 192
(p. 103); 251 (p. 180); 262 (p. 61); 311 (p. 74); 373 (P. 82); 391 (p. 61);
448 (p. 166) ; 474 (p. 103) ; 481 (p. 164) ; 483 (p. 92) ; 501 (p. 64) ; 552
(p. 89); 580 (p. 228); 584 (p. 79); 614 (p. 72); 628 (p. 103); 645
(p. 154); 687 (p. 243); 725 (p. 64); 757 (p. 73) \ 758 (p. 178); 761
(p. 103); 883 (p. 154); 962 (p. 266); 1006 (p. 145); 1012 (p. 300);
1038 (p. 143); 1054 (p. 147; p. 266); 1060 (p. 145); 1067 (p. 247);
1 124 (p. 246); 1232 (p. 227); 1259 (p. 92); I3i4(p. 245); 1325 (P- 3i8);
1338 (p. 245).
Most. 9 (p. 266) ; 34 (p. 206); 37 (p. 162); 86 (p. 172); 127 (p. 357) ; 131
(p. 94) ; 140 (p. 294) ; 149 (P- 294) 5 152 (p. 252 ; p. 294) ; 170 (p. 226) ;
171 (p. 282) ; 173 (p. 154) ; 174 (p. 276) ; 280 (p. 71) ; 315 (P- 294) 5 338
(p. 294) ; 376 (p. 365); 387 (p. 318); 475 (p. 79); 484 (P. 240); 513
(P. 93) ; 541 (p. 83) ; 605 (p. 103) ; 626 (p. 73) ; 669 (p. 168); 675 (p. 84) ;
701 (p. 293); 712 (p. 294); 737 (P. 142); 738 (p. 294); 739 (P- 295);
871 (P- 159); 1083 (p. 26).
Pers. 1 (p. 227; p. 313); 57 (p. 82); 60 (p. 15); 150 (p. 72); 162(0. 82);
225 (p. 173); 232 (p. 73); 255 (p. 82); 260 (p. 71); 310 (p. 145); 339
(P- 77); 355 (P. 82) ; 392 (p. 244); 408 (p. 17); 495 (p. 250); 512
(p. 250) ; 769* (p. 51) ; 776a (p. 169) ; 832 (p. 168) ; 834 (p. 143).
Poen. 225 (p. 21); 279 (p. 103) ; 316 (p. 70) ; 388 (p. 206) ; 638 (p. 202) ;
701 (p. 152); 778 (p. 214); 802 (p. 154); 831 (p. 188) ; 835 (p. 252) ;
860 (p. 61); 969 (p. 17); 988 (p. 83); 1014 (p. 145); 1045 (p. 87);
1052 (p. 153) ; 1058 (p. 152); 1065 (p. 92) ; 1 123 (p. 82) ; 1136 (p. 26) ;
U97a(p.357); 1246 (p. 166) ; 1271 (p. 157); 1272(^82); 1312 (p. 17) ;
I3i7(p. 62).
Pseud. 79 (p. 17) ; 133 (p. 206) ; 146 (p. 269); 153 (p. 276); 169 (p. 23);
203 (p. 227) ; 218 (p. 266); 255 (p. 57); 275 (p. 135); 349 (p. 71);
357 (p- 98); 364 (P. 220); 447 (p. 167); 472 (p. 159); 552 (p. in);
558 (p. 162); 577 (p. 2io); 593 (p. 148); 602 (p. 143); 641 (p. 177);
645 (p. 86) ; 668 (p. 178) ; 703 (p. 175) I 7^2 (p. 217) ; 772 (p. 207) ;
774 (p. 227) ; 800 (p. 272) ; 805 (p. 91) ; 872 (p. 26) ; 945 (p. 159) ; 955
(p. 142); 1018 (p. 120); 1071 (p. 202); 1 146 (p. 105); 1242 (p. 169);
1262 (p. 151); 1277 (p. 262); 1329 (p. 264); 1334 (P- 263).
Rud. 127 (p. 166) ; 177 (p. 266) ; 198 (p. 166) ; 200 (p. 166); 212 (p. 295) ;
242 (p. 117) ; 281 (p. 75); 285 (p. 70); 291 (p. 82); 403. (p. 23); 413
(p. 89) ; 462 (p. 94) ; 487 (p. 17) ; 535-6 (p. 242); 578 (p. 26); 596 (p. 82);
372 INDEX B
646 (p. 171); 660 (p. 284); 733 (p. 245); 746 (p. 163); 804 (p. 142);
888 (p. 89) ; nil (p. 180) ; 1170 (p. 145); 1193 (p. 175) ; 1200 (p. 70) ;
1256 (p. 87) ; 1390 (p. 163) ; 1392 (p. 163).
Stich. 1-23 (p. 313) ; 3a (p. 69) ; 21 (p. 56 ; p. 315) ; 25 (p. 192) ; 45 (p. 315) ;
67 (P- 57); 91 (P. 145); 97 (p. 245); 108 (p. 253); 123 (p. 161); 159
(p. 73); 182 (p. 172); 202 (p. 154); 221 (p. 317); 250 (p. 206); 438
(p. 178); 459-61 (p- 242); 473 (P- 75); 483 (p. 17); 660 (p. 58); 686
(P- 57; p. 173); 695 (p. 143); 715 (P-2i); 718 (p. 146); 723 (P- 266).
Trin. 6-7 (p. 173 ; P. 248) ; 56 (p. 79) 5 81 (p. 163) ; 149 (p. 151) ; 163 (p. 68);
176 (p. 166); 197 (p. 203); 218 (p. 173); 234 (p. 358); 239 (p. 300);
244 (p. 227); 259 (p. 73); 3i8 (p. 266); 330 (p. 330); 413 (P- 266);
451 (p. 17) ; 478 (p. 87) ; 495 (p. 266) ; 527 (p. 187) ; 545 (p. 145) i 575
(p. 17); 665-6 (p. 61); 676 (p. 154); 684 (p. 163); 789 (p. 181);
809 (p. 73); 818 (p. 23); 880 (p. 98); 935 (p. 188); 1 108 (p. 238);
1156 (p. 162).
True. 69 (p. 37) ; 114 (p. 250) ; 124 (p. 117) ; 137 (p. 162) ; 193 (p. 172) ;
222 (p. 358) ; 259-60 (p. 162) ; 372 (p. 162) ; 466 (p. 167) ; 467 (p. 166) ;
596 (p. 166) ; 656 (p. 79) ; 696 (p. 239) ; 702 (p. 53) ; 718 (p. 357) 5 741
(p. 162) ; 810 (p. 106) ; 929 (p. 227) ; 942 (p. 145) 5 965 (P- 162).
Vid. 90 (p. 248).

Printed in England at the Oxford University Press

i

\
 PA

LINDSAY, W. M.
Early Latin Verse, 2329,
• L5