THE MECHANISMS OF HUMIDITY REACTIONS

OF TERRESTRIAL ISOPODS
BY NADEJDA WALOFF
Birkbeck College, University of London

(Received 14 February 1941)

(With One Text-figure and Seven Graphs)

INTRODUCTION
RECENT work by Gunn (1937) and Miller (1938) on land isopods clearly shows that
these animals are very sensitive to humidity gradients, aggregating in the areas of
highest humidity. The present series of experiments were carried out mainly on
Porcellio scaber Latr., to analyse the mechanisms of reactions whereby the animals
are retained in the damper zones. Attempts were also made to see whether the
behaviour pattern, shown by Ullyott (1936) to be characteristic of the phobotactic
reactions of the planarian Dendrocoelum lacteum, is also exhibited in the humidity
reactions of the woodlouse. Further comparative experiments on Oniscus asellus Lin.
and Armadillidium vulgare Latr. were undertaken to find whether there are any
differences in the hygrokinetic responses correlated with the natural habitats of the
three species, and how far the differences in behaviour were influenced by the rate
of evaporation of water from the woodlice.

METHODS
The apparatus initially used was a rectangular dish 3 in. deep, covered by a
sheet of glass with a perforation in the centre for the insertion of the animal.
A platform of perforated zinc T.\ in. above the glass floor was supported on glass
rods. The required humidities were obtained by placing dishes with sulphuric acid
of known specific gravity underneath the zinc floor on either side of the dish.
Edney paper hygrometers were used for measurement of humidity. The range of
the gradient obtained was large, namely, from 20 to 90% R.H. However, this
apparatus was found to be unsuitable for work with Porcellio scaber, which is
strongly thigmokinetic and hence tends to keep very close to the sides of the dish
when moving, pressing one of the antennae against the glass wall. Moreover, this
animal showed a tendency to remain motionless at the corners of the dish. It was
interesting to note that when P. scaber was placed in constant humidities in this
type of apparatus, its thigmokinetic response increased with the rise of relative
humidity. There appeared to be a balance, between thigmokinetic and hygrokinetic
behaviour, and further experiments were undertaken to elucidate this point (see
p. 118).
JEB.18,2 8
n6 NADEJDA WALOFF
The second type of apparatus tried was the ' alternative chamber' described by
Gunn & Kennedy (1936). This, in spite of being free from corners, again proved
unsuitable, since the animals instead of moving about freely on the zinc platform
tended to keep to the sides of the glass dish.
A tubular apparatus had therefore to be chosen, since it presents an equal
surface of stimulation throughout its whole length to the strongly thigmokinetic
animals. A modification of Gunn's circular humidity gradient apparatus was tried,
and proved to be suitable. The apparatus used (Fig. 1) consists of a wide glass tube
1 \ in. in diameter and 58^ in. long, bent into an oval. A circular tube would have
been preferable, but this could not be obtained owing to difficulty of construction.

Fig. 1. Tubular humidity gradient apparatus for testing locomotory reactions of thigmokinetic
animals. Alt As, A3, A4, funnels with hygrometers; Blt J52, thermometers; C, opening for insertion
of the animal; Dlt D2, tubes dipping into sulphuric acid; E, perforated zinc platform.

However, the rounded corners of this apparatus did not appear to impede the
progress of the animals or to alter their behaviour in any way. Externally the tube
is divided into inches by thin strips of adhesive paper, and the inches are marked
in white ink. At the two ends of the apparatus, and in the middle of each tube
connecting the two ends are glass funnels Alt Az, A3, Ait designed to hold Edney
paper hygrometers. Holes are. bored in the hygrometers to ensure that the readings
taken in the funnels are identical with those in the tube. The funnels Ax, A2, A3, At
are covered by small glass plates. Tubes Bx and B2 hold rubber stoppers, through
which thermometers are inserted. The animal is inserted through a small opening
at C. Below Ax and A2 are tubes Dx and D2, half an inch in diameter, which can
be inserted into dishes containing water or sulphuric acid of known specific gravity.
The dishes containing sulphuric acid are covered to prevent alteration in specific
gravity due to absorption of moisture.
The apparatus in this condition was used for experiments in which a humidity
gradient was required. When constant humidity was needed, a further modification
was introduced. Aa was connected to a long glass-bead tower, which in turn was
 The mechanisms of humidity reactions of terrestrial isopods 117
connected with two U tubes arranged in series, filled with glass beads. Sulphuric
acid of known concentration was introduced into the tower and the U tubes. Ax was
connected to an air pump. A very slow current of air was then drawn through the
whole apparatus. Tubes Dx and Z)2 were made air-tight by placing their ends into
dishes containing mercury.
All the experiments were carried out in an underground dark cellar at Birkbeck
College, University of London. The annual range of temperature was between
13 and 20° C. Experiments were carried out at between 14 and 18° C, the variation
in temperature at the time of any one experiment not exceeding i° C. Diffuse light
of intensity of 2 Ferranti foot-candles was used throughout the experiments.
Woodlice were isolated and starved for 3 days before the beginning of each
experiment in dishes in which R.H. was maintained between 90 and 95 %, and
probably reached saturation under stones contained in them. To minimize the
Weber-Fechner effect, i.e. the fact that a response depends not on the actual
quantity of stimulus, but on the proportion which the increase bears to the preceding
stimulus, the animals, when placed in the glass tube apparatus, were initially
introduced into about 90% R.H., and left for 10 min. to recover from the effects
of handling. A gentle current of air bubbled through sulphuric acid of known
specific gravity was then drawn through the apparatus for 5 min. Preliminary
experiments had shown that such a slow rate of air flow did not influence the
direction of movement of P. scaber. The apparatus was then made air-tight, and
the speed and the number of turnings performed by the animal were recorded
every 30 sec. for a period of 1 hr. The number of 'rests', or periods of complete
inactivity over a period of 30 sec, was also noted. Although the absolute speed
could be recorded by noting the number of inch divisions traversed per 30 sec,
it was impossible to determine the absolute number of turnings (i.e. klinokinesis,
according to Gunn and the rate of change of direction, or r.c.d., according to
Ullyott), as the shape of the apparatus did not allow for turning in all the possible
directions. The animal moved about freely and could exhibit turnings of three
different types: (a) it could swing itself completely round and then proceed in the
original direction, i.e. undergo an angular deviation of 360°; such movements in
P. scgber were observed at low humidities only, and not once at relative humidities
over 85 %; (b) the animal could swing around and proceed in the opposite direction,
such turning involving an angular deviation of 1800; (c) the animal could turn 900,
and then swing back and proceed in the original direction. For the sake of conr
venience a turning of 900 was taken as a unit of measurement. Although the
absolute number of angular deviations could not be recorded, the relative number
at different humidity conditions could be obtained, and for this reason they are of
value.
EXPERIMENTAL DATA
(1) Reactions o/Porcellio scaber to air currents (100 and 50% R.H., t= 16-170 C.)
It was important to find out whether P. scaber orientated itself to air currents,
since these were used at the beginning of each ' constant humidity' experiment.
8-3
u8 NADEJDA WALOFF
Six woodlice were introduced into a glass tube 12 in. long and | in. in diameter,
and a gentle current of air was drawn through it by means of an air pump. The
number of animals facing the current and of those turned away from it was recorded
every 15 min. for a period of 2 hr. in the case of 100 % R.H., and for a period of 1 hr>
in the case of 50 % R.H. The same animals were not kept for more than 1 hr. at
50% R.H., as they were in danger of being desiccated.
In both humidities forty readings of six animals in each case were recorded.
A new set of six animals was used for each experiment. In the majority of cases
woodlice were seen to aggregate at either end of the tube. The animals 'resting'
or moving at right angles to the current were neglected.
At 100% R.H., £=16-17° C , 108 animals faced the current, n o faced away
from it. At 50% R.H., t =16-17° C., 96 animals faced the current and 101 faced
away. From these results it was concluded that woodlice do not orientate themselves
to slow air currents.

(2) Thigmokinetic behaviour of Porcellio scaber

It was observed that the thigmokinetic response characteristic of woodlice varied

with relative humidity. A series of simple experiments was undertaken in this
connexion. The animals were introduced into the ' alternative chamber' glass dish;
constant humidity was maintained and altered between one set of experiments and
the next by means of different concentrations of sulphuric acid. Temperature
yariation in the experiments ranged between 14 and 17° C , and within a single
degree for any one experiment. Diffuse light of intensity of 2 Ferranti foot-candles
was used.
Since the animals became progressively less active with the rise of relative
humidity, the thigmokinetic response appeared to be closely associated with the
hygrokinetic one. To distinguish between the two, the percentage of the time the
woodlouse spent next to the wall of the glass dish while-it was still moving was taken
as the final indicator of the thigmokinetic behaviour. One woodlouse at a time was
introduced and readings were taken every 15 sec. for a period of half an hour. To
minimize the handling effect, readings were not taken in the first 10 min.
Twenty animals were taken through relative humidities of 20-25, 50-55, 70-75
and 90-95 %. When the animals came to rest they invariably did so touching the
wall of the vessel, with antennae folded ventrally and closely pressed against, the
solid surface. If the animals were placed in a rectangular dish, they usually came
to rest in one of its corners. When the animal walked near the glass side, one of its
antennae always touched the wall, even if the rest of the body was 1 mm. or so
away from it. If one of the antennae be removed, the animal still keeps to the side,
with the remaining antenna pressed against the glass surface. This seems to indicate
that some of the organs of touch are localized in the antennae.
From Table 1 it can be assumed that the thigmokinetic response is dependent
on the degree of humidity, and its intensity rises with the rise in relative humidity,
increasing steeply on the approach to saturation.
 The mechanisms of humidity reactions of terrestrial isopods 119
Table 1. Thigmokinesis of twenty Porcellio scaber in different relative humidities

t= 14-16° C. 2 0 - 2 5 % R-H.» 50-55 % R-H. 70-75 % R.H. 90-95 % R.H.

% of 30 min. spent in
moving next to the 816 79'5 85-9 92-6
wall (thigmokinesis)

(3) Behaviour of Porcellio scaber under constant humidities

Thirty P. scaber were kept at 90-95 % R.H. One at a time was introduced into
the glass tube apparatus and allowed 15 min. before a gentle current of air was
passed through for 5 min. The air current was bubbled through sulphuric acid of
known specific gravity, and thus a known degree of relative humidity was obtained
(see Methods, p. 116). Three sets, of ten animals in each, were experimented on.
Sets I and III were taken through relative humidities of 0-10, 20-25, 40-45, 58-63,
80-85,- 90-95 and 98-100%. Set II was taken through relative humidities of 0-10,
40-45, 58-63, 68-73, 85-90 and 98-100%. Readings were taken every 30 sec. by
a stop-clock, and four factors were noted in the behaviour of each animal:
(a) activity, as shown by the actual number of inches passed through by an
animal in a given period of time;
(b) speed, which differs from activity in that it shows the actual velocity while
the animal is moving, disregarding the periods of rest;
(c) the number of turnings (klinokinesis) in a given time;
(d) the number of rests, i.e. periods of complete inactivity for 30 sec. Clearly
(a) and (d) are two different methods, of expressing the same facts.
The results obtained from these experiments are shown in Graphs 1 and 2.
As can be seen in Graph 1 the number of turns, speed and activity decrease with
the rise of relative humidity, while correspondingly there is a rise in the number
of' rests' Whereas the fall of the number of turns with the rise of relative humidity
is continuous, the maximum speed and activity were obtained at 58-63 % R.H. It is
interesting to note that Gunn (1937) has observed that the intensity of reaction
'was greatest between 35 and 65 % R.H. approximately, and that at 55 % R.H. there
was a definite reaction to a difference of 6 % R.H.' He further notes that the intensity
of the reaction diminishes in the higher and lower humidities.
Speed and activity are almost identical up to 68 % R.H., being greatest at 58-63 %
R.H. (average speed 9-38 in., average activity 9-34 in. per min. for a period of 1 hr.,
when sets I, II and III are taken into account. When the averages of twenty animals
(sets I and III) are taken into account, the greatest speed and activity are obtained
at 20-25 % R.H. (speed 9-98 in., activity 9-88 in. per min.).
. Appreciable difference between speed and activity was first observed at 68-73 %
R.H. (speed = 8-88 in., activity = 6-82 in. per min. from sets I, II and III). This
showed that although speed was maintained at almost the same level as at lower
relative humidities, activity was already decreasing. Speed is, therefore, not an'
indicator of hygrokinesis, but is only one of its factors. Above 68-73 % R-H- both
120 NADEJDA WALOFF

speed and activity fell off with the rise of humidities. However, as stated above,
there was a steady fall in the number of turnings (i turning = 90°) with the rise of
relative humidity. Whereas at 0-10% R.H. the average number of turnings was
61-95 P e r hour, at 98-100% R.H. it was 20-33.
A B
70rl4

12 - o
0 TTurnings-/!

60
\

10 _ Distance between*
turnings—B \

50-9

3
o
u

I
30 - 4

_ % of time at rest—<
20

10 20 30
40 50 60 70 80 90 100%
Relative humidity
Graph i. Behaviour of the woodlouse Porcellio scaber in constant humidities, calculated from the
average of 30 specimens. The number of turnings is read from the ordinate A. The average distance
between turnings, the activity and the speed per minute from the ordinate B. The percentage of
time at ' rest' is read from ordinate C.

Attempts were made to correlate the variables by dividing the total distance
travelled by the animal in 1 hr. by the number of turnings per hour, i.e. to estimate
the average distance between turnings. The figures obtained are those shown in
Table 2 and Graph 1. It is thus seen that not only the total number of turns but
 The mechanisms of humidity reactions of terrestrial isopods 121
the distance, between turns decreases with the rise of humidity. This apparent
discrepancy is due to a large number of ' rests' and the low activity of the animals
at high humidities. The smaller distance between turnings, together with decreased
activity will tend to restrict the woodlice to the regions of higher or 'optimum'
humidity. The significance of this result is discussed later (see p. 134).

10%
10I-R.H. 10

.5

I I
10 20 30 40 50 60
Minutes
Activity and speed coinciding. Q Activity. • • Speed. O---O Turnings.
Graph 2. The behaviour of Porcellio scaber (average of 30 specimens) throughout the 1 hr. experi-
ments at 10% R.H., 68% R.H. and 90% R.H. The average activity and speed per minute, and the
number of turnings are recorded for every 10 min.

Table 2. Distance travelled between turnings obtained from the averages

of thirty Porcellio scaber
O-IO% R.H. 4 ° - 4 5 % R-H. 58-63 % R.H. 68-73 % R-H. 90-95 % R.H. 98-100% R.H.
103 in. 10-4 in. 13'9 in- 135 in- 53 in- 62 in.
122 NADEJDA WALOFF

(4) Comparison of behaviour of Oniscus asellus Lin., Porcellio scaber Latr.

and Armadillidium vulgare Latr. under similar constant humidities
The methods used in observations on Oniscus asellus and Armadillidium vulgare
were identical to those employed in constant humidity experiments on Porcellio
scaber. It was of interest to see whether there were any differences in behaviour

16
Oniscus asellus
14

12

S'O
1 8
6

12 .Porcellio scaber

10
CO

I8
£ 6
4

Armadillidium vulgare

10 20 30 40 50 60 70 80 90 100%
Relative humidity
• • Speed and activity coinciding. • • Activity. • - - - | Speed.
Graph 3. Average activity and speed per minute of Oniscus, Porcellio and Armadillidium
in relation to constant humidjties.

which could be correlated with the differences in their natural habitats. Both
P. scaber and Oniscus asellus are restricted to damp situations under stones, humus,
bark of decaying logs and pieces of wood, but although the habitats of the two
species overlap to a great extent, Porcellio scaber is never found in contact with
actually wet undersurfaces of stones or wood, whereas Oniscus asellus frequently is.
 The mechanisms of humidity reactions of terrestrial isopods 123
Armadillidium vulgare, which may be considered to be one of terrestrial isopods
best adapted to life on land, is restricted to drier calcareous soils.
Ten specimens of Oniscus were observed in relative humidities of 20-25, 4°~~45»
60-65, 7°~75> 80-85, 90-95, 98-100%, and ten specimens of Armadillidium vulgare
600 r-

50C

*•;•

400

.§
'.5

I 300

I 200

100

I I
10 20 30 40 50 60 70 60 90 100%
Relative humidity
© 0 Oniscus asellus. • • Porcellio scaber. H h Armadillidium vulgare.
Graph 4. The number of periods of 'rest' (30 sec. of complete inactivity) in relation to
constant humidities, in Oniscus, Porcellio and Armadillidium.

in relative humidities of 0-10, 25-30, 40-45, 60-65, 80-85, 9°~95 a n ^ 98-100%,

at temperatures between 14 and 180 C. The animals chosen for all experiments were
as nearly as possible of the same size, and varied between 1-25 and 1-65 cm.
In kll essentials the behaviour of the three species was the same, namely, with
increase of relative humidity there was a decrease in activity and speed, i.e. hygro-
kinesis was clear in all cases (Graph'3), e.g.
124 NADEJDA WALOFF
Speed per minute (in.)
% R.H.
Oniscus asellus Porcellio scaber Armadillidium vulgare

20-25 13-24 998 12-04

60-65 15-30 938 13-04
90-95 4"5 4-16 76

The number of 'rests', i.e. 30 sec. of complete inactivity, increased consistently

with the rise of relative humidity in the cases of Oniscus and Porcellio, but remained
at approximately the same level from 60 to 100% R.H. in Armadillidium (Graph 4),
e.g.
'Rests'
% R.H. 10 Oniscus asellus 10 Porcellio scaber 10 Armadillidium vulgare

25 0 4 0
60-65 2 40 155
80-85 38 258 97
343 448 157
90-95 574 457 159
98-100

The smaller and more constant number of 'rests' of A. vulgare may be an

expression of its lesser sensitivity to humidity. This supposition is borne out by
the subsequent experiments which show the greater resistance of this species to
desiccation (see p. 128). In this connexion it is interesting to note that although
A. vulgare can roll itself into a sphere, at no time was 'rolling up' observed at low
humidities; on the contrary, when the isopod was left in the apparatus for several
hours it remained active until it became moribund on desiccation.
The total number of turnings per hour of Porcellio and Oniscus decreased with
the rise of relative humidity, but Armadillidium did not show any appreciable
corresponding decrease (Graph 5). However, the distance travelled between
turnings, obtained by the ratio of total distance/number of turns decreased very
rapidly with the rise in humidity in Oniscus, and to a lesser extent in Armadillidium,
e.g.
Average distance between turns (in.)
% R.H.
Oniscus asellus Porcellio scaber Armadillidium vulgare

25 26-5 10-9 167

45 2I-I 10-4 22-3
60-65 25-9 139 14-5
9O-95 79
98-100 6-6 11 io-8
9'4

See also Graph 6.

It thus seems that the mechanism by which isopods are retained in areas of
higher humidities, namely, hygrokinesis and increase in the frequency of turns,
are operative in all the three species, but are best developed in Oniscus asellus. This
indicates the greatest sensitivity of this species to humidity, and is probably
associated with its most rapid loss of water by evaporation (see experiments on
 The mechanisms of humidity reactions of terrestrial isopods 125
desiccation, p. 128). The lesser sensitivity to humidity of Armadillidium vulgare,
probably correlated with its greater resistance to desiccation, was indicated by the
smaller number of periods of inactivity or 'rests'.

10 20 30 40 50 60 70 60 90 100%
Relative humidity
© 0 Oniscus asellus. | Porcellio scaber. i 1- Armadillidium vulgare.
Graph s. The average number of turnings (angle of 90° taken as unit) per hour in
Oniscus, Porcellio and Armadillidium, in relation to constant humidities.

It was interesting to note that the specimens of Oniscus asellus and Armadillidium
vulgare were consistently more active than those of Porcellio scaber. Moreover,
much less individual variation was seen in the first two species than in the last.
Abbott (1918) in experiments on photic responses also noted that the behaviour
of Oniscus was much more stereotyped than that of Porcellio.

(5) Behaviour of Porcellio scaber in humidity gradients

The me'thod of obtaining humidity gradients has already been described (see
Methods, p. 116).
126 NADEJDA WALOFF
Altogether fifty-five Porcellio scaber were experimented with:
io in 75-95 % R.H. (£=15-18° C), 18 in 26-73 % R.H. (£=14-18° C),
io.in 65-85 % R.H. (£= 14-15° C), 7 in 42-60% R.H. (£=14-15° C ) .
10 in 55-75 % R.H. (£=14-15° C),

10 20 30 40 50 60 70 80 90 100%
Relative humidity
0 0 Oniscus asellus. | Porcellio scaber. H 1- Armadillidium vulgare.
Graph 6. The average distance between turnings in relation to constant humidities,
in Oniscus, Porcellio and Armadillidium.

In each case one animal at a time was introduced into the glass tube apparatus for
a period of i hr. Readings were not taken for the first io min. to allow for recovery
from handling. The results of the experiments are given in Table 3. The number
of turns, speed and the number of 'rests' were recorded every 30 sec, and the
behaviour of the animal in the damp half of the apparatus was compared with that
in the dry half. In each case the following points were calculated:
 The mechanisms of humidity reactions of terrestrial isopods 127
(1) Percentage of total time spent in damp half of the apparatus.
Percentage of total time spent in dry half of the apparatus.
(2) Percentage of total time spent at ' rest' in the damp half.
Percentage of total time spent at ' rest' in the dry half.
(3) Average speed per minute in damp half.
Average speed per minute in dry half.
(4) Total number of turns in damp half.
Total number of turns in dry half.
(5) Average distance travelled between turns in the damp half.
Average distance travelled between turns in the dry half.
(6) Average time interval between turns in the damp half.
Average time interval between turns in the dry half.

Table 3. Behaviour of Porcellio scaber in humidity gradients

Average Average no.

% of total % of total Average Total no. of distance (in.) of minutes
time spent in time spent speed (in.) turns (1=90°) between between
at rest in per minute in in turnings in turnings in
Damp Dry Damp Dry Damp Dry Damp Dry Damp Dry Damp Dry
half half half half half half half half half half half half
75-95 % R.H. 87-17 1283 50-42 0 4-48 7-48 143 41 599 1365 1-32 161
«=iS-i8°C.
10 P. scaber
6 S -8 S % R.H. 73-75 26-25 34-67 0 S-52 800 183 106 8-55 18-89 1 42 2'2I
*=i4-i5°C.
10 P. scaber
55-75 % R-H. 7083 29-17 26-03 6-25 570 656 186 80 6-59 11-89 1-38 2-21
*= 14-15° C.
io P. scaber
20-73 % R-H. 74'49 2551 2903 018 5-38 652 347 255 773 6*05 1-39 099
*=i4-i8°C.
18 P. scaber
42-60% R.H. 6771 3229 2357 0 6-68 760 "3 164 io-54 5 46 i-5i o-73
*= 14-15° C.
7 P. scaber

Examination of Table 3 shows that in all the humidity gradients provided, the
animals spent the greater part of the hour in the damper half of the apparatus.
Speed was consistently higher in the dry half and the number of ' rests' greater in
the damp half.
In the gradients of 75—95, 65-85 and 55-75 % R.H. the average distance between
turnings and the time interval between turnings are shorter in the damper half of
the apparatus. Such behaviour, associated with hygrokinesis (indicated by -the
larger number of' rests' and lower speed in higher humidities), would tend to restrict
the isopods to these conditions. These results are consistent with those obtained
in experiments with constant humidities. No such correlation, however, was
obtained when the animals were subjected to gradients of 20-73 a n ^ 42-60% R.H.,
for although speed was again-higher in the dry half, the greater part of time spent
in the damp half and the periods of 'rest' almost entirely restricted to the damp
128 NADEJDA WALOFF

half,'the average distance, and the average time interval, between turnings, were
greater in the drier part of the apparatus, and such results are inconsistent with any
which were previously obtained. No satisfactory explanation for this difference in
behaviour can be given, but it may be suggested that the normal behaviour pattern
was altered by the continual physiological instability caused by loss of water by
evaporation at low humidities. It is also probable that behaviour alters with the
steepness of gradient, for whereas in four sets pf experiments the gradient was
approximately that of 20%, it was 53 % between 20 and 73 % R.H.

(6) Loss of water by evaporation

It was interesting to see how far the differences in behaviour of woodlice in
different humidities could be associated with the rate of loss of water by evaporation,
and the following series of simple experiments were undertaken in this cqnnexion.
(a) Duration of life of Oniscus asellus, Porcellio scaber and Armadillidium vulgare
under different relative humidities.
Twenty specimens of each species, of approximately the same size (1-25-
1-65 cm.) previously starved in 100% R.H. for 3 days, were placed in desiccators
at o, 25, 50, 75, 85, 93 and 100% R.H. (t= 15-180 C). To prevent conservation of
water due to possible formation of groups, each isopod was placed in a separate
desiccator and the number of hours of its survival was noted (see Table 4). It was

Table 4. Duration of life of Oniscus asellus, Porcellio scaber and Armadillidium

vulgare at different relative humidities (2= 14-180 C). Number of hours of
survival based on the averages of twenty animals in each case

n.H. Oniscus asellus Porcellio scaber Armadillidium vulgare

/o Average Limits Average Limits Average Limits
0 43 3-5 5-27 4-6 7-i5 6-10
25 685 5-7-5 817 6-5-10 9-92 6-16
50 62 5-8 10-17 6-16 3015 16-44
75 16-25 12-20 25-25 18-44 5915 36-87
85 25'5 11-28 290 18-43 6565 35-96
93 33-17 i6-5-75 3917 20-67 114-6 49-240
100 — 32 hr. to over — 47 hr. to over — 76 hr. to over
a month a month a month

found that the resistance to desiccation determined by death-point was consistently

greatest in Armadillidium vulgare and greater in Porcellio scaber than in Oniscus
asellus. The figures in Table 4 are probably purely relative, and do not represent
the exact survival under natural conditions, since in the desiccators the whole ventral
side of the animal, as well as the dorsal, is exposed to uniform humidity. Normally
the ventral surface of the animals at rest is closely pressed against a solid surface,
and undoubtedly a microclimate of higher humidity is thus retained, at least for
some time.
 The mechanisms of humidity reactions of terrestrial isopods 129
(b) Estimation of water loss by evaporation of Oniscus asellus during the 1 hr. experi-
ments.
(i) As before, twenty specimens of Oniscus asellus, starved for 3 days at 100%
R.H., were weighed and placed in desiccators at o, 50 and 75 % R.H. for a period of
1 hr., after which the animals were reweighed. The following figures, taken from
the average of twenty animals, represent the percentage loss of water by evaporation
per hour:
0% R.H., * = 16-18° C. 50% R.H., z = i5-i8°C. 75% R.H., * = i8°C.
9-9% 6-3% 3-6%
After each experiment the animals were returned to 100% R.H., left for 24 hr. and
reweighed. With the exception of four cases out of sixty, the animals gained in
weight. This gain varied widely and could not be correlated either with the original
weight of the animal or its previous percentage loss by evaporation.
(ii) As a check to the above experiments another set was undertaken. Ten or
twenty specimens of Oniscus were placed in 95, 75, 50 and 20% R.H. for 1 hr.,
one woodlouse to a desiccator. Each animal was then weighed and placed in a
weighed specimen tube containing anhydrous copper sulphate, from which it was
separated by a strip of perforated zinc. The tubes were placed in a drying oven for
24 hr.—1 hr. at ioo° C. to stop the action of enzymes, and 23 hr. at 60° C. After
24 hr. the animals and the specimen tubes were reweighed. The water content of
O. asellus after 1 hr. at different relative humidities (2=15-16° C.) was calculated
and is shown in the following figures:
Relative humidity % 95 75 5O 20
Water content % 63 59-6 57-3 57-2
No. of specimens 20 10 20 10

Since the loss of weight is mainly due to evaporation of water on drying, it can be
concluded that at the end of 1 hr. the animals kept at 95 % R.H. have a much higher
water content than those kept at lower humidities; moreover, the evaporation of
water in Oniscus exposed to dry conditions appears to be rapid. It must be remem-
bered that O. asellus may be regarded as an isopod little adapted to a terrestrial
habit, and is found in habitats where humidity must be nearly always at saturation
point.
(iii) The rate of loss of water in O. asellus during one hour was then calculated
at o and 50% R.H. (2=14-17° C). One animal at a time was placed in a small
box of perforated zinc attached to the arm of a balance by a fine wire. The zinc
box was suspended in a small vessel at o and 50 % R.H. This method gave a set
Of continuous readings, but its chief disadvantage was the disturbance of the
initial humidity on the introduction of the zinc box. The rate of loss of weight is
shown in Table 5.
Table 5. Ten Oniscus asellus weighed for 1 hr. at o and 50% R.H.
Minutes 5 10 15 20 25 30 40 5° 60

% loss of weight at o% R.H. 04 I-I 16 2-3 28 3-6 4-4 5-7 64

% loss of weight at 50% R.H. 03 0-7 i-3 16 20 2-4 30 39 56
130 NADEJDA WALOFF

Fipm all the above results it may be concluded that the loss of water by
evaporation in O. asellus is proportional to time and to the percentage of relative
humidity. It may be suggested that the differences in the activity of woodlice are
due to the continuous alteration in their water content, the loss of which does not
appear to be regulated.

(7) The effect of loss of water by evaporation on phototaxis of Oniscus asellus

The present series of experiments are here included, as they indicate that there
is a balance between phototaxis and humidity reactions of Oniscus asellus.
Abbott (1918) in photo tactic experiments on Porcellio scaber and Oniscus asellus
found that the two species are sensitive to a range of light from 100 to o-oi cm.,
the response being the same for all light intensities. He also states that the reaction
of Oniscus was essentially the same whether the animal had been previously kept
in maximum or minimum moisture, but notes that Porcellio was less negative after
living in a dry habitat. The same tendency to reversal of phototactic reaction in
Porcellio, and a definite reversal from negative to positive phototaxis after a period
of desiccation in Armadillidium has been described by Henke (1930).
In the present set of preliminary experiments on phototaxis of twenty specimens
of Oniscus asellus under constant humidity-conditions of ioo, 93, 75, 70, 60, 50,
40, 25 and 10% R.H. at temperature 14-15° C. and at light intensity of 45 Ferranti
foot-candles, it was noted that the animals show definite negative phototaxis above
75 % R.H. Below 75 % R.H. the initial negative phototaxis seems to disappear, and
the animals appear indiscriminately on the illuminated and shaded halves of the
apparatus. Below 40 % R.H. there is an indication that a reversal from negative to
positive phototaxis occurs, as the number of the woodlice at rest in the light half
of the apparatus is equal or greater than in the dark half (see Graph 7). These
experiments, although in their initial stages, suggest that a reversal from negative
to positive phototaxis, probably associated with water loss by evaporation", occurs
in O. asellus.
The effect of humidity gradients on the phototaxis of O. asellus was next
observed. It is known that when land isopods are introduced into an 'alternative
chamber', with low humidity at one end and high humidity at the other, they
aggregate in the damp half of the apparatus (Gunn, 1937). Preliminary and control
experiments have also shown that in saturated air O. asellus invariably comes to
rest in the shaded half of the apparatus. However, when the animals are placed
in the alternative chamber with the range of 50-90% R.H. (t= 18° C), the drier half
being screened with black paper and the damper half exposed to diffuse light of
intensity of 45 Ferranti foot-candles, the isopods at first wander about the apparatus
indiscriminately. The behaviour of forty specimens of Oniscus, previously kept in
darkness at 100 % R.H. for 3 days, was noted under such conditions. One animal
was introduced into the apparatus at a time to prevent the possible conservation of
water due to the formation of aggregations (Allee, 1926). Each animal was weighed
before and after each experiment.
 The mechanisms of humidity reactions of terrestrial isopods 131
Six out of the forty animals came to rest in the dark, dry half of the apparatus
and eventually died of desiccation; five settled in the dark for a short period of
time and then continued to move until they-finally came to rest in the light damp
half;

6
4 -©. .J2>
-©•-©—-0—©

03

.2 12 70% R.H.

10
- a
J 6

2- e flexC.

10- 10% R.H.

45 60 75 90 105 120
Minutes
- 0 At rest in darkness. 0 - - - 0 At rest in light.
Graph J. Tendency to reversal from positive to negative p'hototaxis, at low humidities, in Oniscus
asellus (20 specimens). The numbers of animals at rest in the shaded and illuminated (diffuse light
of 45 Ferranti foot-candles) halves of the apparatus at io, 70 and 100 % R.H. are recorded.

Eight specimens moved continuously in both halves of the apparatus, settling

finally in the light damp half;
Twenty-one others, after much initial activity, came to rest in the dark dry half,
but then moved to the light damp half and came to rest there. 15 min. at rest in
the light half was taken to indicate the final position of each animal.
The initial and final weights of the woodlice, the percentage loss of weight by
evaporation, and the time taken to come to rest in the light damp half were recorded.
The average loss of weight was 6-2% of the original body weight (limits 2-5 and
132 NADEJDA WALOFF

10-2%). No correlation was observed between the weight of the woodlice and the
percentage loss of water. The time taken to come to rest in the light varied between
20 min. and 3 hr. 20 min.:
Twenty-six out of thirty-four settled in the light half in between 50 min. and
2- hr.;
Four in under 50 min.; and
Four in over 2 hr.
It thus seems that the'initial negative phototaxis is stronger than the humidity
reaction of Oniscus. However, the increased activity of woodlice, due to the loss
of water by evaporation and the masking or the reversal of the negative photic
reaction, combine to retain the animals in the regions of greater humidity.

DISCUSSION
The three species of isopods studied, namely, Oniscus- asellus Lin., Porcellio
scaber Latr. and Armadillidium vulgare Latr., are widely distributed and common
in Europe and North America. The animals are restricted to damp situations in
which the microclimatic humidity conditions approach those of saturation. They
are members of a very old stock in which there is but little morphological plasticity.
Their stereotyped behaviour, controlled to a great extent by external stimuli, may
be associated with the primitive scalariform, vermian type of nervous system. The
brain of Porcellio, characterized by only slightly developed sight centres, and
absence of antennal glomeruli and corpora pedunculata, is considered to be one of
the most primitive of Arthropod brains (Hanstrom, 1928).
One of the most important adaptations of isopods to land, which, however, is
not found in all species, is the development of tracheae on the pleopods, and hence
of aerial respiration. The endopodites of woodlice are damp with a distinct film of
water, and retain their ' gill-like' character. These are the only respiratory organs
in Oniscus, which is capable of picking up drops of water by the movements of its
telson and uropods, and conveying them to the respiratory surfaces through minute
channels (Verhoeff, 1920). In Porcellio and Armadillidium while the exopodites are
invaginated to form small branching air tubes, the endopodites are 'gill-like', and
the isopods seek conditions for both methods of respiration.
Miller (1938) has found that the optimum relative humidity for terrestrial
isopods is close to 100 %, and that there is an indication that survival is inversely
proportional to the saturation deficiency. In the present set of experiments it was
seen that the resistance to desiccation was consistently greatest in Armadillidium,
and greater in Porcellio than in Oniscus. However, all the three species survived for
indefinitely long periods only in desiccators at 100% R.H. From experiments on
O. asellus it could be concluded that the peril of desiccation is a very real one to
woodlice, for not only the respiratory organs are affected, but there appears to be
no control of loss of water through the cuticle, the loss of weight due to evaporation
being proportional both to relative humidity and the time. It was noted that
although Armadillidium vulgare is able to roll up into a ball, at no stage in the
desiccation experiments was it seen to do so. It may be that rolling up is a response
 The mechanisms of humidity reactions of terrestrial isopods 133
to rapidly changing conditions, rather than to a continuous stimulus of constant
low relative humidity.
Allee (1926) has suggested that the formation of aggregations by woodlice is' of
survival value, since the loss of water of a group is much lower than that of an
isolated animal. The difficulty of keeping isolated woodlice in laboratory conditions
was continually noted. No doubt the animals not only live in a microclimate, but
produce a microclimate of their own, for while the dorsal surface of the body is
exposed, the damp pleopods of the ventral surface are normally closely pressed
against the substratum, and under such circumstances the air in contact with the
lower surface must be at saturation point. The isopods which survive for any length
of time at 90 % R.H. in laboratory conditions, provided they are in contact with a
solid surface, die of desiccation when the whole body is exposed to this humidity.
The survival of isopods on land has undoubtedly been favoured by the develop-
ment and combination of hygrokinetic, thigmokinetic and negatively phototactic
behaviojjr. It was seen that the degree of thigmokinetic response of Porcellio scaber
rises with the relative humidity of the air. The correlation which exists between
hygrokinesis and thigmokinesis is significant and can also be considered of survival
value, since it prevents the adherence of animals to solid surfaces in dry conditions,
and hence their desiccation.
The effect of water loss on behaviour of woodlice was clearly shown by Oniscus
asellus. According to modern terminology (Fraenkel & Gunn, 1940) this species
shows low hygrokinesis and is negatively phototactic. When these woodlice were
introduced into a humidity gradient, the drier end of which was shaded, and the
damper exposed to diffuse light, the initial negative phototaxis was stronger than
the reaction to humidity, but with the loss of water by evaporation, there was a
masking or reversal of the reaction, which together with increased activity, brought
the animals to the regions of 'optimum' humidity. Such an alteration in the
reactions is of general interest, since it clearly indicates that the behaviour of an
organism is not fixed, but consists of variables depending on the balance between
the external and internal environments.
From the foregoing it is clear that the humidity reactions are of primary
importance to terrestrial isopods. The behaviour of Oniscus asellus, Armadillidium
vulgare and Porcellio scaber in constant humidities, and of the last-named species
in humidity gradients, was observed, and it appears that there are two mechanisms
whereby the animals are retained in the damper zones:
(1) Hygrokinesis, or the decrease of activity with the rise of relative humidity,
was seen in all the three species. Gunn (1937) has demonstrated hygrokinesis in
Porcellio scaber, but further details of this mechanism were obtained, as it was
found that it is expressed not only by the number of minutes during which the
animals are moving, or are at complete rest at any constant humidity, but also by
the variations in the speed of the movements in different humidities. At high
humidities activity and speed are consistently low, and the animals tend to become
completely akinetic. This is true of Oniscus and Porcellio, in the case of Armadillidium,
on the other hand, although the speed again decreases with the rise of humidity,
9-8
134 NADEJDA WALOFF

the number of 'rests', i.e. of periods of akinesis, remains relatively low and constant
above 65 % R.H. In conditions below 85 % R.H. (* = 14-180 C.) all the three species
remained active to the end of the 1 hr. experiments.
Hygrokinesis clearly results in bringing the animals to the regions of 'optimal',
i.e. high humidity. Such a response is considered to be a primitive one (Mast, 1938),
and is widely distributed throughout the animal kingdom.
(2) In addition, there appears to be a more complicated mechanism whereby
the animals are retained in the' optimal' zone, for when the total number of turnings
is divided into the distance passed by the animals in the 1 hr. experiments, it becomes
clear that the distance between turnings decreases with the rise of humidity and
increases with its fall. This change in the behaviour pattern may be associated with
the loss of water by evaporation from the animals, and together with the hygro-
kinetic effect, which in Oniscus and Porcellio results in complete inactivity at
90-100% R.H., is effective in retaining the isopods in moist air. The decrease of
distance between turnings with the rise of humidity has been observed in all the
three species. In humidity gradient experiments on P. scaber this mechanism was
again exhibited in gradients of high, but not in those of low, humidity.
It is interesting to note that Gunn & Pielou (1940) describe a similar mechanism
designated as 'virtual inactivity' in the humidity reactions of Tenebrio molitor, which
in contrast to the isopods aggregate at the drier end of a humidity gradient.
Although the mechanisms of these reactions in Oniscus, Porcellio and Armadilli-
dium are the same, some differences in their intensity have been detected. The
greater activity, at high humidities, of Armadillidium, as compared to that of
Porcellio and Oniscus, may be an expression of its lesser sensitiveness to alterations
in humidity and of its greater resistance to desiccation. The mechanisms are most
clearly expressed in O. asellus, which, of the three species, must be considered the
least adapted to land.
When a comparison is made of the mechanisms of the reactions of isopods with
photophobotaxis described by Ullyott (1936) in the planarian Dendrocoelum lacteum,
two main differences (among others) emerge, viz..
(1) No kinetic effect was shown in D. lacteum, whereas hygrokinesis is clear in
the isopods.
(2) Sensory adaptation is the basic factor determining the photic responses of
D. lacteum. The meaning of sensory adaptation is that the effect of a continuous
stimulus is not uniform, but that the intensity of the reaction produced by it
decreases with the time and gradually dies down (Hecht, 1919). No such adaptation
was satisfactorily demonstrated in the isopods, and it was particularly difficult to
approach this question, since at low humidities woodlice are continually losing
water by evaporation, and hence not maintaining a balanced physiological state.
As to the humidity receptors, none are known in woodlice (Gunn, 1937). It can
be suggested that the rapid loss of water by evaporation and the consequent
concentration of body fluids would in itself have an effect on the reactions of the
animals. Any sense organs connected with this mechanism would be proprioceptive.
If this suggestion is correct and there are no specialized receptors, no sensory
adaptation need take place, and this is borne out by experimental results. This,
 The mechanisms of humidity reactions of terrestrial isopods 135
however, would not account for immediate reactions. Gunn (1937) has suggested
that the receptors probably lie in the thoracic region. It has been continually noted
that the relatively delicate thoracic appendages are the first to show the effects of
desiccation, and it may be that they function as the 'hygrometers' of the body.
In conclusion, it may be said that the terrestrial isopods are the only group of
Crustacea which are able to live on land throughout the whole of their life cycle.
However, they are just able to survive the land conditions, not so much by morpho-
logical adaptations, as by evolving a series of patterns of behaviour, which restrict
them to moist dark habitats. In other words, they are able to survive on land by
avoiding the typical land conditions.

SUMMARY
1. The humidity reactions of Oniscus asellus, Porcellio scaber and Armadillidium
vulgare have been analysed and compared.
2. The mechanism whereby the three species collect in moist air is twofold,
consisting of (a) hygrokinesis, or decrease in activity and speed in moist air, and
(b) of more frequent turnings in space, retaining them in the areas of greater
humidity.
3. These mechanisms are most clearly expressed in Oniscus asellus and least in
Armadillidium vulgare. This sequence may be correlated with the resistance to
desiccation of the three species, which is greatest in Armadillidium vulgare, and
greater in PorceUio scaber than in Oniscus asellus.
4. It is suggested that the humidity reactions of isopods are controlled by water
loss by evaporation from the whole body.
5. A correlation between hygrokinesis and thigmokinesis was observed in
Porcellio scaber.
6. There appears to be a reversal from negative to positive phototaxis in
Oniscus asellus, correlated with the water loss by evaporation.
7. The humidity reactions, low thigmokinesis and negative phototaxis combine
to retain the isopods in damp, dark habitats.
I wish to express my sincere gratitude to Prof. H. G. Jackson for continual help
and encouragement in the carrying out of this work, and my indebtedness to
Mr Alastair Graham and Dr D. L. Gunn for helpful criticism and advice.

REFERENCES
ABBOTT, C. H. (1918). J. exp. Zool. 27, 193-246.
ALLEE, W. C. (1926). J. exp. Zool. 45, 255-77.
FRAENKEL, G. & GUNN, D. L. (1940). The Orientation of Animals. Oxford.
GUNN, D. <L. (1937). J. exp. Biol. 14, 178-86.
GUNN, D.'L. & KENNEDY, J. S. (1936). J. exp. Biol. 13, 450-9.
GUNN, D. L. & PIELOU, D. P. (1940). J. exp. Biol. 17, 307-16.
HANSTROM, B. (1928). Vergleichende Anatomie des Nervensystems der wirbellosen Tiere. Berli
HECHT, S. (1919). jf. gen. Physiol. 1, 545-58.
HENKE, K. (1930). Z. vergl. Physiol. 13, 534-626.
MAST, S. O. (1938). Biol. Rev. 13, 186-224.
MILLER, M. A. (1938). Univ. Calif. Publ. Zool. 43, 113-42.
ULLYOTT, P. (1936). J. exp. Biol. 13, 265-78.
VERHOEFF, K. (1920). Z. tviss. Zool. 118, 366-447.