BIOLOGICAL BULLETIN

OF THE

flDarine Biological laboratory

WOODS HOLE. MASS.

Editorial Staff

GARY N. CALKINS — Columbia University.
E. G. CONKLIN — Princeton University.
M. H. JACOBS — University of Pennsylvania.
FRANK R. LILLIE — University of Chicago.

GEORGE T. MOORE — The Missouri Botanic Garden.
T. H. MORGAN — Columbia University.
W. M. WHEELER — Harvard University.
E. B. WILSON — Columbia University.

lE&itor

C. R. MOORE — The University of Chicago.

VOLUME LIV.

WOODS HOLE, MASS.

JANUARY TO JUNE, 1928

LANCASTER PRESS, INC.
LANCASTER, PA.

Contents of Volume LIV

No. i. JANUARY, 1928.

CHILD, C. M. Experimental Transformations of Bipolar

Forms in Corymorpha palma i

CHILD, C. M. Physiological Polarity and Dominance in the

Holdfast System of Corymorpha 15

HALL, RICHARD P., AND POWELL WILLIAM N. Morphology
and Binary Fission of Peranema trichophorum (Ehrbg.*)
Stein 36

HYMAN, LIBBIE H. Miscellaneous Observations on Hydra,

with Special Reference to Reproduction 65

No. 2. FEBRUARY, 1928.

BECKER, ELERY R. Streaming and Polarity in Mastigina

hylce (Frenzel] 109

TUCKER, MARY ELIZABETH. Studies on the Life Cycles of
Two Species of Fresh-Water Mussels Belonging to the
Genus A nodonta 117

UHLENHUTH, E., AND KARNS, HILDA. The Morphology and

Physiology of the Salamander Thyroid Gland, III 128

QUIGLEY, J. P. Reactions of an Elasmobranch "(Squalus
sucklii) to Variations in the Salinity of the Surrounding
Medium 165

SKOWRON, STANISLAW. The Luminous Material of Micro-

scolex phosphoreus Dug 191

PATTERSON, J. T. Functionless Males in Two Species of

Neuroterus 1 96

No. 3. MARCH, 1928.

PATTERSON, J. T. Sexes in the Cynipida and Male-Producing

and Female-Producing Lines 201

SAYLE, MARY HONORA. Factors Influencing the Rate of

Metabolism of JEshna umbrosa Nymphs 212

iv CONTENTS OF VOLUME LIV.

CLEVELAND, L. R. Further Observations and Experiments on
the Symbiosis between Termites and Their Intestinal
Protozoa 231

BISSONNETTE, THOMAS HUME. Notes on a 32 Millimeter

Freemartin 238

BABKIN, B. P., AND BOWIE, D. J. The Digestive System and

its Function in Fundulus heteroditus 254

No. 4. APRIL, 1928.

ROWLAND, RUTH V. Grafting and Re-incorporation in Actino-

sphcerium eichhornii Ehr 279

WHITING, P. W. Mosaicism and Mutation in Habrobracon. 289
HUMPHREY, R. R. Ovulation in the Four-toed salamander
Hemidactylium scutatum, and the External Features of

Cleavage and Gastrulation 307

HEATH, HAROLD. Fertile Termite Soldiers 324

VISSCHER, J. PAUL. Reactions of the Cyprid Larva of Bar-
nacles at the Time of Attachment 327

VISSCHER, J. PAUL, AND LUCE, ROBERT H. Reactions of the
Cyprid Larvae of Barnacles to Light with Special Reference
to Spectral Colors 336

No. 5. MAY, 1928.

GRAVE, B. H. Vitality of the Gametes of Cumingia tellinoides. 351
FRY, HENRY. Conditions Determining the Origin and Be-
havior of Central Bodies in Cytasters of Echinarachnius

Eggs 363

BODINE, JOSEPH HALL. Action of Salts on Fundulus Egg. I. 396
PEARSE, A. S. On the Ability of Certain Marine Invertebrates

to Live in Diluted Sea Water 405

YOCOM, H. B. The Effect of the Quantity of Culture Medium

on the Division Rate of Oxytricha 410

MANWELL, REGINALD D. Conjugation, Division, and En-

cystment in Pleurotricha lanceolata 417

No. 6. JUNE, 1928.

BELLING, JOHN. Nodes and Chiasmas in the Bivalents of

Lilium with Regard to Segmental Interchange 465

CONTENTS OF VOLUME LIV. V

MACDOUGALL, MARY STUART. The Neuromotor Apparatus

of Chlamydodon sp 471

LOOPER, JAMES BURDINE. Observations on the Food Reactions

of Actinophrys Sol 485

RAU, PHIL. The Honey-Gathering Habits of Polistes Wasps. 503
BURCH, PAUL R. Endodermal Flagella of Hydra oligactis

Pallas 520

KEPNER, WM. A., AND MILLER, LULU. A New Histological

Region in Hydra oligactis Pallas 524

KEPNER, WM. A., AND THOMAS, W. L., JR. Histological

Features Correlated with Gas Secretion in Hydra oligactis

Pallas 529

FALES, DORIS EDNA. The Light-Receptive Organs of Certain

Barnacles 534

*

Vol. LIV January, 1928 No. i

BIOLOGICAL BULLETIN

EXPERIMENTAL TRANSFORMATIONS OF BIPOLAR
FORMS IN CORYMORPHA PALM A.

C. M. CHILD,
HULL ZOOLOGICAL LABORATORY, THE UNIVERSITY OF CHICAC.O.

It has been shown in earlier papers that the localization and de-
velopment of a hasal end in the reconstitution of stem pieces of
Corymorpha may be determined by the dominance of more distal
parts (Child, '26^, '27^:), by contact with or nearness to the bot-
tom (Child, '266, '270, b) and by the inhibiting action of various
agents (Child, '270, b, c). In normal development the base arises
from the low end of the primary gradient (Child, '260), a region
of relatively low metabolism, but contact with, or nearness to the
bottom and various inhibiting agents, or both acting together, may
determine basal ends anywhere, even as mosaic parts on other
body regions (Child, '26^). Evidently these external factors in-
duce physiological conditions similar to those which arise in em-
bryonic development at the low end of the primary gradient.
The present paper gives the results of a few experiments on the
transformation of bipolar forms resulting from reconstitution into
other forms, its purpose being merely to record the occurrence of
such transformations under certain experimental conditions. All
figures are drawn from observed and recorded cases. Regions
inclosed in perisarc are indicated by heavier line than naked re-
gions and perisarcal accumulations not in direct contact with the
body surface are indicated by dotting. Except in case of the
very small individuals, the figures are diagrammatic as regards
numbers of tentacles.

TRANSFORMATION OF BIPOLAR INTO BIPOLAR-UNIPOLAR FORMS.

Bipolar forms are those in which a hydranth develops from
both distal and proximal end of a stem piece. As shown else-
1 . i

C. M. CHILD.

where (Child, '266), bipolar forms may range from extreme apical
ends including merely mouth and hypostome. to complete hy-
dranths with a considerable length of stem between them. Bi-

Vfe >^

2

4

FIGS. 1-6. Bipolar and bipolar-unipolar forms resulting from reconstitu-

tion of stem pieces.

polar forms of the sort used in these experiments are shown in
Figs. 1-3. Fig. i is an earlier, Fig. 2 a later stage of a form with
considerable length of stem between the two hydranths, such as
ordinarily develops from the longer stem pieces. Fig. 3 shows

EXPERIMENTAL TRANSFORMATION IX CORYMORPHA. 3

a bipolar form from a shorter piece, consisting of hydranths only.
Bipolar-unipolar forms (see Child, '26^) are bipolar distally and
unipolar proximally (Figs. 4-6). The two hydranths develop
from the two cut ends of the piece, the unipolar proximal region
from the lateral stem region between the hydranths. The proxi-
mal region represents a new polarity at right angles to the original.

By means of inhibiting conditions it is possible to transform
bipolar forms such as Figs. 1-3 into bipolar-unipolar forms like
Figs. 4-6, although the bipolar forms never develop ,a unipolar
proximal region as long as they are kept under good conditions
Extensive experimentation along this line has not been under-
taken, but several agents have been used, viz., alcohol, two and
three per cent., NH4C1 ;;/ 300 and ethyl urethane 7? 1/300 and
7H/ioo and in case of one lot of pieces sea water which had stood
for some hours in a new galvanized bucket was found very effec-
tive as an inhibiting agent. The same transformations appeared
in all agents used. The bipolar forms were placed in the agent
at earlier (Fig. i) or later stages (Figs. 2, 3) of development, but
only after it was certain that they were developing as symmetrical
bipolar forms.

In most cases they remained in the agent one to two days and
were then returned to sea water. One series, however, remained
in alcohol two per cent, for twelve days, the solution being re-
newed daily. During exposure to the agent the tentacles are usu-
ally reduced to mere stumps or disappear entirely, the distal re-
gions in most cases disintegrating and the stumps undergoing re-
sorption. Hydranths and the stem if it is present, undergo marked
reduction in size, particularly in the longer exposure periods and
in some cases the manubrium of one or both hydranths may show
some disintegration. On return to water the pieces remain undis-
turbed for two days or more and during this time redevelopment
of hydranths and tentacles takes place. If the degree of inhibi-
tion is not too extreme the basal region may develop during ex-
posure to the agent, otherwise it develops after return to water.

Fig. / shows the transformation of a bipolar form like Fig.
2 after alcohol two per cent, for one day, then alcohol three per
cent, for a second day. The figure represents a stage three days
after return to water. A proximal stem region and basal end have

4 C. M. CHILD.

developed from what was originally the lateral stem region between
the hydranths. A similar case is shown in Fig. 8. This form,
originally bipolar like Fig. 2 but somewhat shorter, was in NH4CL
w/3OO one day then in sea water two days. At this time rede-
velopment of tentacles and development of the new base were not
far advanced, as Fig. 8 indicates. Fig. 9 was transformed from a
piece like Fig. i after forty hours in alcohol two per cent, followed
by two days in sea water. In this series the pieces were subjected
to inhibition at a relatively early stage of the bipolar development
and the new basal end was larger in all than when later stages
were used.

In cases in which the degree of inhibition is somewhat greater,
either because of higher individual susceptibility or because of
higher concentration or longer exposure to the agent, modification
of form may be more extreme. In the case of Fig. 10, for ex-
ample the piece was originally bipolar like Fig. 2 but somewhat
shorter. It was subjected to alcohol two per cent, for one day,
then to alcohol three per cent, for a second day and then remained
three days in sea water. As Fig. 10 shows, the proximal regions
of the two hydranths have undergone fusion and instead of the
two complete sets of proximal tentacles originally present, only
one set now appears. In this case there has been not only reor-
ganization of the lateral stem region into a base, but extensive
reorganization of the hydranths has also occurred. Other similar
cases have been observed.

Figs, ii and 12 show two stages of the transformation in the
toxic sea water. The form was originally bipolar like Fig. I,
but under the inhibiting conditions developed the basal region
with holdfast outgrowths within twenty-four hours on the part
in contact with the bottom. The basal region continued to grow
at the expense of other parts and after four days the condition
shown in Fig. 12 was attained. This case was not observed
further, but forms of this sort usually separate sooner or later
into two individuals, the separation occurring through the gradual
atrophy of some portion of the connecting basal region. The re-
gion undergoing atrophy apparently becomes a sort of no-man's-
land between the two individuals and its cells may be at one time
influenced by one, at another time by the other and perhaps some-

EXPERIMENTAL TRANSFORMATION IN CORYMORPHA. 5

times by neither. This absence of definite function probably
determines the atrophy.

The case shown in Fig. 13, originally bipolar like Fig. 2 was in
alcohol two per cent, for twelve days. During this time it de-
creased greatly in size, largely because of disappearance of the
axial parenchyma, the tentacles disappeared completely and the
form of the hydranths approached that of early developmental

10 12 IJ

FIGS. 7-13. Bipolar-unipolar forms transformed from bipolar forms by

inhibiting conditions.

stages. After return to water redevelopment of the two hy-
dranths began and a basal tip developed from the side of the stem
in contact, underwent elongation and became disproportionately
large at the expense of other parts. Fig. 13 shows the condition
seven days after return to water. It will be noted that the hy-
dranth on the left appears to be more inhibited than the other.
As will appear below, the two hydranths of a bipolar form often
show a considerable difference in susceptibility.

TRANSFORMATION OF BIPOLAR INTO UNIPOLAR FORMS.

The first experiments on subjection of bipolar forms to inhibit-
ing conditions were performed in 1910 and it was found that some

6 ( - M. CHILD.

bipolar forms after several days under inhibiting conditions re-
developed as completely unipolar forms. Work had to be dis-
continued before much experiment along this line was possible,
but the observations made on the changes occurring during the
inhibition period showed in some cases complete disappearance of

o

16

20

21

22

FIGS. 14-23. Development of unipolar from bipolar forms determined
by inhibiting conditions. Figs. 14, 15. Apparent obliteration of both
original polarities by inhibiting conditions and determination of new single
polarity by the differential of position. Figs. 16-18. Unipolar from bipolar
forms by death of one and persistence of other axis. Fig. 19. Equal re-
gression under inhibiting conditions of both hydranths of a bipolar form :
the hypostome regions indicated as dotted areas. Fig. 20. Death and
disintegration under inhibiting conditions of one axis of a bipolar form
with persistence of other. Fig. 21. A form in process of becoming uni-
polar by death of most of one axis and development of a base from lateral
stem region. Figs. 22, 23. Unipolar forms from small fragments of
bipolar forms which remained alive while rest was killed by inhibiting con-
ditions.

the more highly specialized structures and the gradual regression
of the form to a rounded mass, usually more or less completely
inclosed in perisarcal secretion (Fig. 14). After return to water
a single apical region developed on the upper surface of this mass
and a complete new axis vertical to the substratum resulted (Fig.
15). Such cases were believed to represent the obliteration of
the double polarity of the bipolar form and the determination of
a new polarity by the differential of position, i.e., the differential
between the upper free surface and the surface in contact with

EXPERIMENTAL TRANSFORMATION IN CORYMORPHA. 7

the bottom which supposedly determines a difference in rate of
oxygen consumption and discharge of CO2 and so initiates a new
gradient (Child, '26b). A case of this sort was described and
figured (Child, '15, pp. 144-6, Figs. 75-78; '24, p. 119 and Figs.
106-109). It was also noted in 1910 that some of the bipolar
forms became unipolar during the inhibition period. This uni-
polarity was observed only after the pieces had undergone con-
siderable reduction and was indicated merely by the somewhat
elongated form tapering toward one end and an accumulation of
perisarc at the slender end (Figs. 16, 17). With long periods of
inhibition, e.g., from twelve days to three weeks, such unipolar
forms, whether derived from bipolar forms or from the reduction
of unipolar forms, might undergo reduction to a length of less
than one millimeter and still give rise to normal individuals after
return to water (Fig. 18).

At the next opportunity in 1922 to continue these experiments
the tank water at the laboratory was apparently somewhat toxic
and there was difficulty in keeping material in good condition long
enough for such experiments, consequently no further informa-
tion on these transformations was obtained at that time. In 1924
and 1926 experiments were made with water taken direct from the
ocean and it was found that some bipolar forms, more commonly
the shorter, underwent regression without actual disintegration
of any parts except the two sets of tentacles. Fig. 19 shows a case
of this sort in which the two hypostome regions are still distin-
guishable by their whitish color at the two ends of the piece. They
are indicated in the figure as dotted areas. In pieces returned to
water at this stage redevelopment of the two hydranths with a new
base between them occurred. When the inhibition was continued
all indications of the original axes disappeared, although the diam-
eter of the mass might be greater in the direction of the original
axes. Such masses may give rise to unipolar forms after return
to water and the polarity in such cases appears to be new and de-
termined by the differential of position. But in 1924 and 1926
it was also found by more frequent observation of the material
that one axis of a bipolar form is often more susceptible than the
other and with a sufficient degree of inhibition it disintegrates,
while the other remains alive (Fig. 20). The axis that remains

8 C. M. CHILD.

alive may undergo regression to a condition like Figs. 16 and 17
and in such cases the polarity is one of the two polarities of the
bipolar form, or when forms like Figs. 16 and 17 result from the
reduction of whole stems or other unipolar forms, it is the original
polarity. In such cases the bipolar form becomes unipolar through
the death of one axis, rather than by the obliteration of the two
original polarities and the determination of a single new polarity.
Even in such cases, however, there is much reorganization. The
hydranth appearing after return to water is not only absolutely
but relatively very much smaller and has a much smaller number
of tentacles than the original and the stem and base may develop
from regions that were originally part of the hydranth.

The occurrence of this method of change from a bipolar to a
unipolar condition is beyond question. It is not properly speak-
ing, a transformation, but merely a persistence of an axis. Since
it does occur, the question must be raised whether the apparent
transformation by obliteration of the old axes and determination
of a single new axis may not actually be merely the persistence
of one of the axes. With respect to this point it may be noted,
first, that in all cases of disintegration of one axis which have been
observed the disintegration occurs early, before the hydranth has
lost its form, and second, that both axes have been identified in
other pieces after the form of the hydranth was indistinguishable
and the only indication of the original axes which was externally
visible was the hypostome regions which differed from the rest
in color (Fig. 19). No disintegration has been seen in such cases
at any later stage and some of them certainly give rise to single
new polarities. To all appearances this single polarity is new and
determined after obliteration of the old axes in these forms, but
in the absence of landmarks after regression of the hydranths it
is difficult to be entirely certain as to what happens in cases of this
sort. It is hoped that there may be opportunity for further work
in the future.

In the light of the data concerning the origin of new polarities
in stem pieces (Child, '27^, b) there is no reason to doubt the
possibility of obliteration of old and determination of new polari-
ties in bipolar forms, particularly if some stem is present between
the hydranths. Moreover the fusion of hydranths, their reorgani-

EXPERIMENTAL TRANSFORMATION IN CORYMORPHA. 9

zation after regression and the development of basal ends from
parts of the hydranth body under experimental conditions indicate
that at least the more proximal levels of the hydranth possess
considerable capacity for giving- rise to other parts, particularly
after some degree of regression.

Figs. 14-18 and 20-23 show observed cases of unipolar forms
which developed from bipolar forms after subjection to inhibiting
conditions. In Figs. 14 and 15 (twelve days in two per cent,
alcohol) the single polarity is believed to be new. In Figs. 16,
17 and 18, also twelve days in alcohol, the polarity is almost cer-
tainly one of the original polarities, but it has developed proximal
and basal regions at the expense of more distal parts. Fig. 19
shows equal regression of both axes, as already noted and Fig. 20
shows a case of disintegration of one axis and persistence of the
other. Fig. 21, a bipolar form two days in water after forty
hours in alcohol two per cent., is almost completely unipolar, but
has a protuberance on one side, which represents the last traces of
the other polarity and which later disappears. In this case the
basal region has developed from the side of the piece, as in the
bipolar-unipolar forms. That is, the unipolar condition has re-
sulted from the death of the more distal regions of one axis and
resorption of the remainder and the development of proximal
region and base from the side of the stem. The unipolar form
in Fig. 22 developed from a bipolar form forty-six hours in al-
cohol, ^'hen returned to water the piece was less than one mil-
limeter in diameter and almost spherical and showed no trace of
polarity. Three days later it had attained the stage shown in Fig.
22. Undoubtedly most of the original bipolar form had died,
but it was impossible to determine what part of the form the small
mass represented, though there is little doubt that it represented
the less susceptible proximal region. The persistence of such
minute fragments after death of the rest of the piece is of common
occurrence under inhibiting condition and in all observed cases,
whether they arose from bipolar forms or from stem pieces these
masses develop, if they develop at all, as unipolar forms, the po-
larity apparently being determined by the differential of position.
Fig. 23 shows another minute form which developed from a bipolar
form after twenty-four hours exposure to NH4C1 111/300 and two

IO C. M. CHILD.

days in water. Here also a large part of the original form disin-
tegrated, but in this case a polarity was evident at the end of the
inhibition period, the form of the minute piece being similar to
that of Fig. 1 6. In this case the final polarity undoubtedly repre-
sents one of the original polarities of the -bipolar form, but com-
pletely reorganized.

To sum up: in the development of unipolar from bipolar forms
all visible traces of the original polarities may disappear, regression
to a rounded mass more or less completely inclosed in perisarc may
occur and the axis of the unipolar form appears to be a new gradi-
ent determined by the differential of position. In other cases one
of the axes of the bipolar form may die and disintegrate under the
inhibiting conditions while the other, or some part of it remains
alive. In such cases polarity is usually visible, at least in the
form of the piece, at all stages and the final single polarity corre-
sponds in direction with this polarity, though it represents exten-
sive reorganization.

DISCUSSION.

The numbers of cases in these transformation experiments are
not large. Among forty bipolar forms resulting from reconstitu-
tion, subjected to inhibiting conditions sixteen (forty per cent.)
transformed into bipolar-unipolar forms, ten (twenty-five per
cent.) became unipolar, six (fifteen per cent.) remained bipolar and
eight (twenty per cent.) died. This does not include preliminary
experiments to determine proper concentrations nor experiments
in which all pieces died. Numerous controls in water under good
conditions have not shown such transformations in any case. The

j

chief purpose thus far has been to establish the fact of these trans-
formations rather than to determine their frequency. As regards
the transformation of bipolar into bipolar-unipolar forms the
data are clear and conclusive, but as regards the question of oblit-
eration of both the original polarities and determination of a new
single polarity as one method of the origin of unipolar from bipolar
forms there is less complete certainty. The course of events in
the cases recorded indicates that the two original axes are obliter-
ated unless one resorbs the other after all visible traces of both
have disappeared, but if they are so much alike physiologically
that they undergo regression equally it is highly improbable that

EXPERIMENTAL TRANSFORMATION IN CORYMORPHA. II

one is sufficiently different from the other to resorh it. "With-
out resorption or death of one axis, there seems to he no escape
from the conclusion that the two original axes are obliterated and
a new one determined. That the original single polarity may be
obliterated has been shown elsewhere (Child, '270, b) and since
this is possible there can be little doubt as to the possibility of
obliteration of the double polarity of bipolar forms.

The question how the inhibiting conditions transform bipolar
into bipolar-unipolar forms is answered very simply from the
gradient standpoint. The middle region of the bipolar form
where the two polarities meet represents the lowest physiological
level present. The inhibiting conditions lower all levels to some
extent and may bring the middle region to the level at which a
basal region arises. It has already been noted that pieces from the
naked stem region commonly secrete more or less perisarc under
inhibiting conditions and may become completely inclosed in it
(Child, '270). Under normal conditions bipolar forms show al-
most continuous motor activity, elongation and contraction of the
stem and movements of hydranth body and tentacles, consequently
no one region is continuously in contact with the bottom. During
the inhibition period, however, and for the first day or two after
return to water there is little or no motor activity and one side of
the piece may be continuously in contact with the bottom. The
effectiveness of contact and nearness to the bottom in determining
basal ends has been shown in many other experiments (Child,
*26b, '270, b). In the cases under consideration the inhibiting
effect of contact and that of the inhibiting agent are additive and
the position of the new basal region one one side of the stem may
be a resultant of the two factors or the effect of either alone. The
important point is that inhibition of the middle region of the bi-
polar form, whether by chemical agents or by contact with the
bottom or by the combined action of both may determine a basal
region there.

Torrey ('10) found that the development of a basal region from
a lateral stem region might be determined by the more distal levels
of the axes already present and without the action of external
factors, in other words, under the dominance of the higher levels
of the gradients. In the experiments with inhibiting agents, how-

12 C. M. CHILD.

ever, the dominance of the more distal levels is undoubtedly de-
creased, consequently it cannot be a significant factor in determin-
ing' the basal region in these cases. This dominance does serve
to determine the middle region as the lowest physiological level
of the bipolar form, but the inhibiting action of the agent or of
contact or of the two combined is required to lower it still further
to the level at which the development of a basal end becomes pos-
sible.

The more proximal levels and the basal end in these bipolar-
unipolar transformations usually attain finally a relatively larger
size than under normal conditions. Under all inhibiting conditions
used a similar change in proportion from the normal has been
observed. This is a result of the differential susceptibility at
different levels of the axis. The most active apical regions are
more inhibited than the less active proximal and basal regions,
consequently the latter are able to obtain a proportionally larger
share of the available nutrition or even to grow at the expense of
more distal levels and so to become relatively larger than in the
normal animals.

The probable transformation of bipolar into unipolar forms
by obliteration of the two original polarities and determination of
a single new polarity is also easily accounted for in terms of the
gradient. Obliteration of the original polarities by the inhibiting
agents is a consequence of differential susceptibility of different
levels of the gradient. The higher levels being more susceptible
are more inhibited than the lower and with certain degrees of
inhibition the normal differences of different levels may be greatly
decreased or completely obliterated, so far as any appreciable ef-
fect is concerned. In the cases of transformation this obliteration
is accompanied by regression, involving complete disappearance
of the more specialized parts. Under these conditions the differ-
ential of position may be effective in determining a new polarity,
the upper, freely exposed portion becoming apical, the region in
contact or near the bottom basal, as in other cases of obliteration
of old and determination of a new polarity by the differential ex-
posure (Child, '270).

The development of unipolar from bipolar forms by the death
of one axis and persistence of the other is not, strictly speaking,

EXPERIMENTAL TRANSFORMATION IX CORY M< 'RIM I A. 13

an axial transformation, but merely persistence of an axis. The
inhibiting conditions kill one axis and not the other because one
is more susceptible than the other. In cases of this sort in which
it has been possible to distinguish distal and proximal axes of a
bipolar form, the distal axis has been found the more susceptible,
as might be expected from its development at a higher level of
the original gradient of the stem than the proximal. This differ-
ence in physiological condition of the two axes is of interest as
indicating that even though one of them, the proximal, represents
a reversal of the original polarity it still retains traces of its origin
from a lower gradient level than the distal axis. Such difference
in susceptibility of the two axes is usually found in the longer
bipolar forms in which the difference in level of stem at which the
two polarities develop is considerable. In the shorter forms with
little or no stem between the two hydranths (Fig. 3) there may
be no appreciable difference in susceptibility between the two hy-
dranths and both axes may undergo regression equally (Fig. 19).

Persistence of one of the two axes evidently means that some
difference in condition persists in the original direction of the
gradient. The reorganization which usually occurs with develop-
ment of a hydranth much smaller than the original and of a basal
region consists in development of stem and basal end from what
was in some cases originally the proximal part of the hydranth.
Apparently the higher levels of the gradient are lowered by the in-
hibiting conditions more than the lower levels so that apical regions
are relatively smaller, proximal and basal regions relatively larger
than under normal conditions. This again is a consequence of the
differential susceptibility of the different levels of the gradient.

The interpretation of the experimental data in terms of gradi-
ents is simple and is in complete agreement with the results of
various other lines of experimentation both on Corymorpha and
on other forms. The existence of the gradients has been demon-
strated by many different methods and the differential suscepti-
bility of different levels has been proved beyond doubt, both by
differences in survival time and by differential modification of
development with many different agents. No other interpretation
has as broad an experimental basis as this.

14 C. M. CHILD.

SUMMARY.

1. Bipolar forms resulting from reconstitution of stem pieces
of Corymorpha palma can be transformed into bipolar-unipolar
forms by subjection to inhibiting conditions for various lengths
of time. The transformation consists of development of a basal
end from the lateral stem region between the hydranths.

2. The development of the base may result from the action of
the inhibiting agent in lowering the physiological level of the re-
gion concerned to the level at which base development is initiated
and the more intimate and continuous contact with the bottom in
consequence of decreased motor activity under inhibiting condi-
tions may act in the same direction and contribute to the result.

3. Bipolar forms may become unipolar by the death of the parts
representing one axis because of higher susceptibility. In such
cases extensive reorganization along the persisting axis occurs.

4. It is believed that inhibiting conditions may bring about
transformation of bipolar into unipolar forms by obliteration of
the two original polarities through differential susceptibility and
the determination of a single new polarity by the differential of
position.

LITERATURE
Child, C. M.

'15 Individuality in Organisms. Chicago.

*26a Studies on the Axial Gradients in Corymorpha pal ma. II. Biol.

Gen., II.
?26b Studies on the Axial Gradients in Corymorpha palma. III. Biol.

Gen., II.
'273 Modification of Polarity and Symmetry in Corymorpha palma by

Means of Inhibiting Conditions and Differential Exposure. I.

Jour. Exp. Zool., XLVII.
'27b Modification of Polarity and Symmetry etc. II. Jour. Exp. Zool.,

XLVII!.
"270 Experimental Localization of New Axes in Corymorpha- without

Obliteration of the Original Polarity. BIOL. BULL., LIII.
Torrey, H. B.

'10 Biolog:cal Studies on Corymorpha. IV. BIOL. BULL., XIX.

PHYSIOLOGICAL POLARITY AND DOMINANCE IN
THE HOLDFAST SYSTEM OF CORY*MORPHA.

C. M. CHILD.
HULL ZOOLOGICAL LABORATORY. THE UNIVERSITY OF CHICAGO.

Cor\morf>ha [>alnnt is found on tidal flats and anchors itself on
the substratum of soft mud or fine sand by means of a large num-
ber of filamentous, stolon-like structures covered with a delicate
perisarc, which extend from the basal region of the hydroid and
serve as holdfasts. Torrey ('04, '07, 'ioa, 'io/>) has described the
structure, the embryonic and reconstitutional development and
some of the reactions of these holdfasts, viz.. their extremely rapid
elongation, the amoeboid activity of their tips, their positive geo-
tropic reaction and the ability of actively developing holdfasts to
inhibit the development of others distal to them.

The holdfasts are much more slender than the usual hydroid
stolon, being onlv about o. i mm. in diameter at the tip which is

O J

the region of greatest diameter except in early stages, but they are
similar to other hydroid stolons in appearance and development.
Apparently, however, they are somewhat more specialized than
the ordinary stolon for, so far as known, they do not transform
into, or give rise to hydranth-stem axes when isolated from the
parent stem, but always remain holdfasts. If they really lack the
capacity to give rise to hydranth-stem axes they should, strictly
speaking, be designated as holdfasts rather than as stolons. Nev-
ertheless it seems evident that they represent merely a specialized
form of stolonic axis. Each holdfast evidently represents a
physiological axis and possesses a definite polarity which appears
in its form, the manner of its development and its final exhaustion.
Moreover, the holdfast system of the individual hydroid shows
certain definite relations to the polarity of the whole, both in its
origin and in the relations of its members. The present paper is
primarily concerned with those aspects of holdfast life which are
related to physiological polarity, viz., the manifestations of polarity
in the individual holdfast and the relations of the holdfast system
to the polarity of the whole animal.

15

16 C. M. CHILD.

THE INDIVIDUAL HOLDFAST AS A PHYSIOLOGICAL Axis.

The Gradient of the Holdfast. — The holdfast originates as a
bud in the basal region of the hydroid and develops as an axis with
a growing tip. Like other buds, it arises as a localized region of
cellular activity which decreases in intensity from a central region
peripherally. As the holdfast develops the most active region
necessarily becomes the tip and the radial gradient in activity
which constitutes the first step in the development becomes an
axial gradient with its high end at the tip. The existence of this
gradient from the earliest localization of the holdfast bud has
been demonstrated by its differential susceptibility to various
agents, by differential vital staining, by differential reduction of
vital dyes and KMnO4 and by the differential in the indophenol
reaction. All the methods used agree in showing that the tip of
the holdfast is the high end of the gradient. The results obtained
on the animal as a whole with these various methods have been
described in an earlier paper (Child, '260.) and since all methods
agree as regards the holdfast, it seems unnecessary to give the
data in detail. Moreover, the gradient is obvious in the general
behavior of the holdfast. It may be noted, however, that the
various agents penetrate the very thin perisarc almost at once and
with those agents which afford a means of following the penetra-
tion directly by staining it is found that the older perisarc farther
from the holdfast tip is usually penetrated before that at the tip.
Evidently the very conspicuous gradient of the holdfast cannot be
ascribed to differences in permeability of the perisarc.

The holdfast bud or primordium may persist for a long time
without appreciable growth beneath the perisjarc of the stem, but
remains capable of very rapid outgrowth when conditions permit.
As a bud it usually tapers slightly toward the tip, but when out-
growth begins its diameter is for a time about the same throughout
its length. With further outgrowth and particularly after it comes
into contact with the substratum, the tip retains its diameter, or
the diameter may appear to increase somewhat because of flatten-
ing of the tip on the substratum, but proximal to the tip the diam-
eter gradually decreases until the proximal portions of the longer
holdfasts are extremely slender threads. This decrease in diam-
eter is the result of gradual atrophy of the cells in the more

POLARITY IN HOLDFASTS OF CORYMORPHA. 17

proximal regions of the elongating holdfast until there is nothing
left except the tube of very delicate perisarc secreted by the hold-
fast as it elongates and within it traces of cellular debris. Figs.
i and 2 show various stages in the development of holdfasts. In
these and all following figures perisarc is indicated by a heavy line.

FIGS, i AND 2. Activation and outgrowth of holdfasts. Fig. i, proximal
cut end of stem piece I 1/2 hours after section, showing activation and
outgrowth of holdfasts from very early bud stages present before section.
Fig. 2, a proximal cut end 24 hours after section, showing a stage of de-
velopment of a second set of holdfasts 3 hours after removal of first set.
A new zone of holdfast buds has begun to develop a short distance distal
to the active holdfasts.

Atrophy of the more proximal levels is characteristic of the true
stolons of other hydroids (Child, '23) as well as of these holdfasts,
another fact which indicates the physiological similarity of true
stolon and holdfast. It was pointed out in the paper just referred
to that this atrophy is clearly a consequence of the gradient in the
stolon. While the outgrowth is short all parts obtain nutrition
from the hydroid body, but it appears that after a certain length
is attained this is no longer possible and from this stage on the
outgrowth is gradually undergoing starvation and reduction. Very
generally, if not always, in the starvation of the simpler animals
the parts which are most continuously and most intensely active
decrease in size less rapidly than the less active because they are
able to live and more nearly maintain themselves at the expense
of the less active. Microscopic examination of the cells of the
longer holdfasts shows that those at or near the tip appear to be
in good condition while evidences of inanition increase with in-
creasing distance from the tip. Toward the proximal end the
2

18 C. M. CHILD.

cells become more transparent and the holdfast has decreased
in diameter more than the perisarc so that only strands of proto-
plasm here and there are in contact with the perisarc. Finally
at a greater or less distance from the tip nothing remains within
the perisarc but a few granules. The transition from the region
of good condition at the tip to that of complete atrophy is usually
very gradual, as indicated in Fig. 3, but occasionally it is rather
abrupt, particularly in the earlier stages of the atrophy (Fig. 4).
The methods used to demonstrate the gradient indicate that such
differences are due to individual differences in length and slope
of the gradient in the holdfast and such differences are undoubtedly
associated with differences in physiological condition and rate of
elongation in the different holdfasts.

The holdfast, like the stolon of other hydroids is obviously a
gradient in which the characteristic process, at least in the earlier
stages, is continued growth of the tip. As a growth gradient it
is also similar to rhizoids, roots and various forms of hairs in
plants and differs from the chief plant axis only in its inability
to give rise to buds.

The Motor Activity of the Holdfast. — Holdfasts growing rapidly
from the cut end of a stem as in Figs, i and 2 often show slight
motor activity before they come into contact with the substratum.
Their movement consists in bending in various directions and in
some cases slight contraction and extension are observed. All
such movements are slow, much slower than tentacle movements.
Motor response of holdfasts to direct external stimulation has not
been tested.

As soon as the holdfast comes into contact with the substratum
it adheres and further motor activity is limited to the tip. In
nature the holdfasts burrow through the mud, but in glass contain-
ers in the laboratory elongation after contact takes place along the
surface of the substratum. (Torrey, 'io&) has noted that the
elongating holdfast is positively geotropic. After attachment,
however, the holdfast tip often shows more or less amoeboid ac-
tivity, as Torrey ('04, p. 416) has also noted. Elongated pseudo-
podia have not been observed, but the distal region may undergo
change in shape, the activity decreasing from the tip proximally.
Fig. 5 shows a series of outline drawings of a holdfast tip at in-

POLARITY IN HOLDFASTS OF CORYMOKPHA.

tervals of one to two minutes and serves to indicate the character
and range of amoeboid activity observed. In general this activity
has seemed to be greater in holdfast tips farther away from the
stem, rather than in the earlier stages of outgrowth, but it cannot
be positively asserted that this is the general rule. Occasionally
the extreme tip has been seen to undergo slight retraction, 0.03-
0.07 mm., followed after a minute or two by renewed advance.
So far as observed, the amoeboid activity occurs chiefly at the ex-
treme tip and decreases rapidly in the proximal direction and at a
distance of 0.3-0.5 mm. from the tip is no longer visible.

r\

FIGS. 3-9. Individual holdfast tips. The shrunken appearance of the
proximal ccenosarcal regions is indicated by dotted lines inside the heavy
lines indicating perisarc. Fig. 3, a holdfast which has covered a distance
of 7-8 mm., showing gradual decrease in diameter and increase in degree of
atrophy in proximal direction. Fig. 4, a holdfast which has progressed 7-8
mm., showing rather abrupt decrease in diameter and increase in degree of
atrophy. Fig. 5, a series of outlines of a holdfast tip sketched at intervals
of 1-2 minutes to show the amoeboid change in form. Fig. 6, a holdfast
which has covered 15 mm. Fig. 7, a holdfast which has covered 20 mm.
and is approaching exhaustion. Fig. 8, an earlier, and Fig. 9 a later stage
of a completely isolated holdfast.

2O C. M. CHILD.

The Later Stages of Holdfast Development. — When the hold-
fast attains a certain length which varies in different holdfasts, the
atrophy of the more proximal regions brings about complete separ-
ation of the distal portion of the ccenosarc from the hydroid stem.
The begining of this atrophy usually becomes evident in the de-
crease in diameter of the proximal region when the holdfast has
attained a length of 2-3 mm. (Fig. 2) and by the time it has
reached a length of 4-8 mm. the living coenosarc of its distal end
is apparently completely separated from the parent stem, the
perisarcal tube being the only connection (see Figs. 12, 13).

Advance of the tip does not cease when this isolation occurs.
The more proximal portions evidently serve as nutrition for the
distal parts and the progress of the tip with the secretion of peri-
sarc continues with gradual decrease in length and diameter of the
ccenosarc until finally exhaustion occurs. Figs. 3, 4, 6, 7 illustrate
the decrease in size of the tip. Figs. 3 and 4 represent the distal
portions of holdfasts which have grown a distance of 7-8 mm.
from the stem, Fig. 6 is a distal end 14-15 mm. and Fig. 7 another
20 mm. from the stem and almost exhausted. Advance does not
continue much beyond the stage of Fig. 7. Apparently the gradi-
ent differences at this stage are no longer great enough in the
small fraction of the original gradient which remains so that the
tip can maintain itself and continue to advance at the expense of
lower levels, i.e., the degree of starvation has become almost as
great at the tip as elsewhere. After its advance ceases the tip
may continue to live and decrease in size still further for at least
a week, perhaps more.

That the continued advance of the tip is entirely independent
of connection with the parent hydroid is evident from the fact that
it is not at all affected by physical isolation of the tip from the
hydroid. Fig. 8 represents a holdfast from which the perisarcal
tube connecting it with the parent stem has been cut away. This
tip continued to advance, secreting the perisarcal tube exactly as if
connected with the stem. Fig. 9 shows a later stage, approaching
exhaustion. When a stem is removed after giving rise to hold-
fasts the latter remain attached to the substratum and continue to
advance until they are exhausted.

In this continued advance of the tip after complete isolation

POLAR I TV IN HOLDFASTS OF CORYMORPHA. 21

from the hydroid the holdfast also resembles the true stolon of
other hydroids (Child, '23). In both the advance continues long
after physiological connection with the hydroid stem is lost and
until the ccenosarc is reduced to a small fraction of its original
size. Even when the advance ceases the cells of the tip are visibly
in better condition than those of lower levels.

In the laboratory holdfasts have been observed to advance over
a distance of 2-3 cm. before exhaustion and the binding of the
mud about the base of the hydroid in nature indicates that under
natural conditions they may cover a somewhat greater distance
than this.

Torrey ('04, p. 416, '07, pp. 277-8) regards advance of the
holdfast tip as clue primarily to amoeboid activity at the tip which
stretches the holdfast. He states, however, that when stems are
suspended free in water the holdfasts extend in all directions.
On cut stems also the holdfasts elongate very rapidly before they
come into contact with the substratum at all (Figs, i and 2).
Such elongation when the tip is not in contact certainly cannot be
due to amoeboid activity for the tip cannot exert tension. More-
over, the volume of the holdfast ccenosarc increases very rapidly
and very greatly during the earlier stages. And finally, the prog-
ress of the tip is not necessarily in a straight line, but the direction
may undergo frequent change, even on smooth glass so that
the perisarcal tube is sinuous or forms a circle or a spiral (see
Figs. 12 and 13). If any actual stretching occurs it must take
place only very near the tip and must very soon reach a limit for
the perisarc is not torn away from the substratum and straightened
and the ccenosarc is not left behind. It seems certain that at least
the earlier stages of holdfast elongation are not due merely to the
amoeboid activity of the tip. To all appearances extensive growth
occurs, although it is conceivable that cells migrate from the
parent stem into the holdfast. Evidence of such migration, how-
ever, is lacking.

To what extent the later advance of the holdfast tip after iso-
lation of its ccenosarc from that of the parent stem may be due to
amoeboid activity and to what extent to actual growth of the tip
at the expense of more proximal levels is less readily determined.
The appearance of the cells at the tip suggests continued growth,

22 C. M. CHILD.

but the very rapid advance suggests amoeboid activity. The peri-
sarc is certainly continuously secreted as the tip advances, rather
than elongated by stretching. The planula of Corymorpha often
progresses for some distance over the substratum leaving a delicate
perisarcal tube behind it as it travels (Child, '260}. In that case
the advance is evidently due to motor activity rather than growth.
It seems improbable, however, that the holdfast changes completely
its activity from growth to amoeboid activity after it comes into
contact with the substratum. Amoeboid activity has been observed
in tips which are approaching exhaustion and have almost ceased
to advance. In the light of all the facts it seems probable that
more or less growth of the tip at the expense of proximal levels
does occur even after isolation from the stem, although amoeboid
activity may play a part in the advance of the tip in later stages.

The continued advance of the tip until exhaustion occurs and
its maintenance in good condition at the expense of more proximal
regions with progressive atrophy of the proximal parts and the
amoeboid activity of the tip are evidently all expressions of the
physiological gradient characteristic of the holdfast. In fact, it
does not seem possible to account for its behavior except in terms
of a gradient. Moreover, the gradient accounts adequately and
entirely for all the phenomena of polarity which the holdfast ex-
hibits.

The Rate of Advance of the Holdfast. — The advance of the
holdfast is so rapid that with a low magnification and an ocular
micrometer the advance from minute to minute can be directly
observed. Different holdfasts show different rates of advance,
but in general the rate decreases very greatly in the later stages
and advance finally ceases. The following table gives a few char-
acteristic measurements of the rate of advance. Nos. I., II. and
III. are the three longest holdfasts of Fig. 2 measured at a stage
a few minutes later than that represented in the figure, i.e., about
three hours after section of the stem and activation of the hold-
fasts. These early stages were elongating at the rate of a mil-
limeter in seventeen to twenty-two and one half minutes, the high-
est rates observed. Nos. IV., V. and VI. are holdfasts twenty-
four hours old, growing from an intact base after removal of the
original holdfasts. In these the rates of advance are much slower

POLARITY IN HOLDFASTS OF CORYMORPHA.

TABLE I.
RATES OF ADVANCE OF HOLDFAST TIPS.

Time in

Distance of
Tip from
Parent Stem
in Mm.

Minutes and
Seconds to
Cover 10
Micrometer
Divisions

(0.33 mm.)

Holdfasts from cut end of stem 3 hrs. after

section

I

3

7:30

II

4

6:40

III

5

5:40

Holdfasts 24 hrs. old from intact basal end

after removal of original holdfasts

IV

8

14

V

IS

20

VI

20

46

and it will be observed that they decrease to a marked extent with
increasing length of the holdfast. These three cases are the three
holdfasts of Figs. 4, 6 and 7 at the stages figured. Fig. 7, No.
VI. in the table, was approaching exhaustion at the time of meas-
urement and its rate of advance is only about one eighth that of
No. I. A number of other measurements made all fall within the
extremes of the table. Torrey ('04) mentions a rate of nine
micra per minute in one case ; this would equal 0.33 mm. in more
than thirty-three minutes, a rate much slower than those of the
earlier stages in the table. Elsewhere (Torrey, '07) he mentions
a rate of 15 mm. per day which is almost midway between the
rates of Nos. V. and VI. of the table. The rates in Table I. are
all from holdfasts in contact with the substratum. The rates of
the earlier stages seem high for a growth process, but rates ap-
proaching these are found in holdfasts which are not yet in contact
with the substratum. Fig. i, for example, shows a basal cut end
i 1/2 hours after the cut was made. Before the section only very
early holdfast primordia were present but within the short time
following section some holdfasts attained a length of i mm. be-
fore their tips came into contact with the substratum. Most of
the length of i mm. was covered during the last half hour of the
period. Except for possible migration of cells from the stem
into the holdfast this elongation appears to be due to growth and

24 C. M- CHILD.

the rate of elongation is nearly as high as in some of the holdfasts
after contact. The holdfast appears to be primarily a growth
gradient with amoeboid activity at its tip.

DOMINANCE AND PHYSIOLOGICAL ISOLATION IN THE ORIGIN AND
DEVELOPMENT OF THE HOLDFAST SYSTEM.

In later stages of the hydroid development the holdfast buds
arise in regular order along the entodermal canals in the basal re-
gion and show in their origin and development certain definite
relations to the polarity of the whole animal and to each other.

Early Normal Development. — The first holdfasts arise as minute
buds at or near the basal end of the perisarcal region of the young
hydroid, but they appear only after the young hydroid has become
attached by its perisarc and they remain as minute buds under the
perisarc for some time after their appearance. Under good con-
ditions the first holdfasts appear only when the hydroid has at-
tained the stage shown in Fig. 10 and often not until still later.
In general holdfast development begins at or near the basal end
and progresses distally from it, but the early holdfasts are more or
less irregularly scattered in position. In vigorous, actively grow-
ing young animals the holdfast buds do not grow out through the
perisarc of the stem at once after their formation, but develop very
slowly for a time, remaining buds and not beginning active out-
growth until the animal attains a considerable length. During this
earlier stage they are merely primordia of holdfasts and play no
part in attaching the animal unless they happen to be terminal.
This very slow development until the animal attains a certain
length suggests that the holdfast buds, like the stolons of other hy-
droids are inhibited in development by apical regions until the in-
crease in length of body brings about a certain degree of physio-
logical isolation.

Premature Development of Holdfasts under Inhibiting Con-
ditions.— Under conditions which inhibit development to a slight
extent holdfasts may not only appear as buds, but may actually
grow out and attach themselves at much earlier stages of develop-
ment than those at which they normally appear. In low concen-
trations of KCN, alcohol, ether, etc. and even when kept in sea
water several centimeters deep without change or aeration the

FIGS. 10-13. Development of holdfasts in young animals. Fig. 10,
young hydroid in stage at which holdfast buds usually first appear under
normal conditions. Fig. n, outgrowth of a holdfast from a blastula under
slightly inhibiting conditions. Fig. 12, a planula with holdfast which ap-
peared in blastula stage under inhibiting conditions. Fig. 13, slightly in-
hibited young hydroid with long holdfast which appeared in the early
planula stage. Hydroid and holdfast are figured in the same plane, though
in nature the plane of holdfast elongation is at right angles to the hydroid
axis.

26 C. M. CHILD.

holdfasts develop in the planula or even in the blastula stage.
Fig. II shows a blastula with holdfast. Fig. 12 a planula with
holdfast which began its development in the blastula stage. Fig.
13 an early hydroid with holdfast which appeared in the early
planula stage twenty-four hours before the stage of the figure.
The spiral growth seen in Fig. 13 is of frequent occurrence in the
holdfasts, but the conditions which determine it are unknown.
These three, as well as numerous other similar cases observed
are from material developing at the bottom of 5 cm. depth of
standing water without aeration or change since early cleavage
stages. As compared with development in frequently changed,
well aerated water, this material was distinctly inhibited in devel-
opment. Many similar cases of premature development of hold-
fasts have been observed with low concentrations of various in-
hibiting agents. In fact, the early formation and outgrowth of
holdfasts is the characteristic effect of a slight degree of inhibi-
tion or retardation of development.

In this connection it may be recalled that development of stolons
instead of the hydranth-stem axis from the planula has been ex-
perimentally determined in a campanularian hydroid, Phialidiwn
gregarium, by various inhibiting conditions (Child, '25). In
that form both ends of the planula may give rise to stolons. In
Corymorplia the premature development of holdfasts has been
observed only from the basal region and the hydranth-stem axis
has developed sooner or later, but always retarded. It has been
pointed out elsewhere that the polarity of the tubularian hydroids
is apparently less readily altered than that of the campanularian
(Child, '19, '25, '260) and the persistence of the planula polarity
in Corymorpha under inhibiting conditions constitutes further
evidence along the same line.

The much earlier appearance and outgrowth of the holdfasts
under slightly inhibiting than under normal conditions shows that
the inhibiting conditions favor in some way their development.
But it is evident that the inhibiting conditions are not in them-
selves necessary for development of holdfasts, since in larger older
animals the holdfasts develop very rapidly in the best possible
laboratory environment (Figs. I, 2). As regards the manner in
which the inhibiting conditions induce premature development

POLARITY IN HOLDFASTS OF CORYMORPHA. 27

of holdfasts, two possible factors suggest themselves. First, the
inhibiting conditions may directly affect the lower end of the
primary gradient which is to become the basal region and lower
its metabolic or physiological level so that conditions favorable to
holdfast development arise earlier than they would under normal
conditions, in which they appear only after the animal has become
considerably older and its metabolism has undergone decrease.
It has been shown elsewhere (Child, '270, b) that basal regions
may be determined and localized by local inhibiting conditions and
it is possible that the direct effect of the inhibiting' conditions in de-
creasing the activity of the lower end of the gradient may be con-
cerned in determining holdfasts at an earlier stage of development
than that at which they normally appear.

A second possible factor in the premature development of hold-
fasts under inhibiting conditions is an increase in the degree of
physiological isolation of the lower end of the gradient from
apical dominance in consequence of the effect of the inhibiting
conditions on the apical region. The fact that the holdfasts are
not only prematurely determined as buds but also develop prema-
turely with great rapidity indicates that the effect of the inhibiting
conditions on their development is, at least in large part, indirect
rather than direct, i.e., an increase in the degree of physiological
isolation of the basal region from apical dominance resulting from
the inhibiting action of the conditions on the apical region. Un-
der normal conditions such physiological isolation does not occur
until a much later developmental stage when the length of the
hydroid begins to exceed the range of that degree of dominance
necessary for complete inhibition of holdfast development, and
even then their development is very slow until a considerably
greater length of stem is attained, which apparently permits a
greater degree of physiological isolation. The dominant apical
region is more susceptible than the less active basal region to the
inhibiting conditions (Child, '260) and its dominance over the
basal region is decreased, consequently holdfasts are able not only
to appear as buds, but to grow out at once. The holdfast evi-
dently represents a new physiological axis which can be formed
in the basal region of the hydroid only when a certain degree of
physiological isolation from the dominance of the apical region

28 C. M. CHILD.

occurs. A similar relation between hydranth region and stolon
exists in Tubularia (Child, '15, pp. 91, 92, Figs. 42, 43) as well as
in other hydroids, but in those forms the stolon tip transforms into
a hydranth-stem axis as soon as the degree of isolation is sufficient.
This interpretation of holdfast development does not conflict
in any way with the fact that new basal ends on which holdfast
buds may develop and grow out are often determined in pieces
undergoing reconstitution by the inhibiting conditions associated
with contact or nearness to the bottom (Child, 26b), especially
with the addition of inhibiting agents (Child, '270). In normal
development and in many cases of reconstitution the basal region
is apparently largely or wholly determined by internal factors as
a secondary gradient at the low end of the primary gradient
(Child, '266). Apparently it represents a region sufficiently iso-
lated physiologically from apical dominance to develop a new
gradient opposite in direction to the old and to grow in length, but
still to some extent under apical dominance since it possesses the
capacity to develop a hydranth-stem axis from its proximal end
(Child, '26b, p. /86), but does not do so except after separation
from the apical and inhibition of its own distal region. The late
appearance and slow development of holdfasts in normal young
animals suggests that a further increase in the degree of physio-
logical isolation of the basal region is necessary for holdfast for-
mation. Since a condition of relatively low metabolism is evi-
dently the essential factor in determining a basal region, local ex-
ternal inhibiting conditions may determine such a region or assist
in determining it, by establishing the proper metabolic conditions.
Whether or when holdfasts shall develop on a basal region thus
determined apparently depends on the degree of physiological iso-
lation from the dominant apical region which is attained in such a
region. In the piece undergoing reconstitution the new apical
region is often not sufficiently developed at the time the basal
region is determined to prevent the appearance of holdfast buds;
but apparently it often does inhibit their further outgrowth for
usually they develop very slowly until the animal attains a certain
length and in the very small individuals from short pieces they
may remain buds indefinitely. In both reconstitution and em-
bryonic development, however, inhibition of hydranth development

POLARITY IN HOLDFASTS OF CORVMORPHA. 2Q

favors their early outgrowth unless the inhibiting conditions are
sufficient to inhibit the holdfast buds directly.

Physiological Dominance and Isolation among the Holdfasts.—
As noted above, the earlier holdfasts are scattered about the basal
end, but after the longitudinal entodermal canals develop, the later
holdfast buds arise in regular order along these canals, visually
two rows to a canal, forming a definite zone surrounding the stem
several millimeters from the basal tip. In this zone the first buds
to appear and the most advanced in development at any given time
are those nearest the basal tip and bud development progresses
distally. Fig. 14 shows the buds on a single canal as they appear
in the older animals and in Fig. 2 a new zone of buds is develop-
ing distal to a cut end.

:o

a

t/.<\

FIG. 14. The system of holdfast buds of a single entodermal canal as it

appears in older animals.

This definite order of development and spatial arrangement also
suggests the existence of a physiological relation between the in-
dividual holdfasts of the system and Torrey has observed that the
removal of the older holdfasts accelerates the development of
those near the wound on the distal side (Torrey, 'ioa, p. 217).
In the normal development of the animal the holdfasts nearest the
basal tip grow out first when the animal attains a certain length.
After protoplasmic connection between their tips and the stem is
severed by atrophy of the connecting region resulting from elon-
gation the buds next in order distally grow out and so on. Mean-
while new buds may begin to develop at the distal end of the bud
zone. This zone therefore gradually changes its position in the

3O C. M. CHILD.

distal direction and in the larger individuals is often entirely above
the surface of the mud in which the basal end is buried. All these
facts indicate the existence of a relation of dominance and subor-
dination in the holdfast system, the holdfasts at the proximal end
dominating those distal to them. In this relation the holdfasts
which develop from the higher levels of the basal gradient dom-
inate those of lower levels, just as in other similar gradients, e.g.,
many plant axes.

After the zone of holdfast buds has developed, the buds of any
level of this zone can be activated at once by section of the stem
just proximal to them. Such section removes the proximal dom-
inant buds or holdfasts and the most proximal buds remaining
grow out and become dominant. This activation and outgrowth
take place very rapidly. Fig. I, for example, shows the proximal
end of a stem 11/2 hours after section through the distal region
of the holdfast zone. Before section the holdfast buds in this
region were in very early stages like those of the upper third of
Fig. 14. Within the short time of i 1/2 hours some of these buds
attained a length of almost I mm.

Fig. 2 shows a stem in which a second set of holdfasts is de-
veloping after removal of the first set which developed after re-
moval of the basal end 24 hours preceding the stage figured. The
first set of holdfasts was allowed to develop for 21 hours at which
time those which developed first had lost ccenosarcal connection
with the stem and others distal to them were developing. Then
all developed holdfasts were removed. Fig. 2 shows the develop-
ment attained by the new set three hours after removal of the
old. Fig. 2 also shows a new zone of holdfast buds developing
at a short distance from the proximal end. This zone has devel-
oped within 24 hours since it was not present at the time of sec-
tion. Such a new bud zone has been observed in other similar
cases and in all, as indicated in Fig. 2, it arises a short distance
distal to the actively growing holdfasts, not immediately adjoining
them. This position is characteristic and again indicates that
these rapidly growing holdfasts dominate a certain length of stem
in such a manner as to prevent development even of other hold-
fast buds within that region. The following experimental data

POLARITY IN HOLDFASTS OF CORYMORPHA. 3!

show that the rapid reaction following section of stem or removal
of holdfasts in the cases above described is not exceptional.

I. Twenty stem pieces comprising the proximal fifth of the
naked region and the distal half of the perisarcal region were cut
from newly collected, relatively young individuals 25-30 mm. in
length. These pieces were approximately one fourth the total
stem length. They possess only very early holdfast buds or no
visible buds. After i 1/2 hours three pieces were attached by
their holdfasts, some of which were I mm. long (Fig. i) and
eight others showed outgrowing holdfasts at the proximal end.
After 14 hours nineteen pieces were attached ; of these fifteen
showed a maze of holdfasts about their bases, some of them
several mm. in length, and four pieces were attached by the peri-
sarc without holdfast outgrowth.

II. Six pieces from the preceding experiment were detached
and all developed holdfasts were removed from the proximal end
after 21 hours growth. One hour later two cases showed out-
growing holdfasts, one a single holdfast 0.23 mm., the other, two,
each 0.17 mm., both attached to the bottom. After i 1/2 hours
four pieces showed outgrowing holdfasts and three were attached
by them. After 2 hours all six showed growing holdfasts and
four were attached. One holdfast was 1.3 mm. in length. After
3 hours four were attached. Fig. 2 represents one of the pieces
at this stage. After 4 hours, five, and after 9 hours, all were at-
tached by holdfasts. Five of the six pieces developed a second
zone of holdfasts buds like Fig. 2 within 33 hours or less.

III. From ten newly collected individuals 50-60 mm. in length
hydranths and basal ends, including all visible holdfast buds, were
removed. The earlier stages of new holdfast development were
not observed, but after 18 hours eight were attached with numer-
ous holdfasts some 3-4 mm. in length. Two pieces were still un-
attached, but show early developing holdfast buds. In this ex-
periment preformed buds were absent but the formation and de-
velopment of new holdfasts to a length of 3-4 mm. has occurred
within 18 hours.

IV. Ten stems similar to III. After 22 hours five were attached
and some holdfasts were 3-4 mm. long. All five showed a new
zone of holdfasts developing 1-2 mm. distal to the outgrowing

32 C. M. CHILD.

holdfasts. Five were unattached, one with early holdfasts, four
without holdfasts. After 46 hours all had developed holdfasts
and become attached.

V. In this experiment four larger animals, 60-70 mm. in length,
which had been 6 days in the laboratory, were used. In these the
holdfast bud zone had already developed above the level of the
mud, see p. 30). These were the largest individuals obtainable
during the summer of 1926, when most of the observations on
holdfasts were made. The basal ends were removed at such level
as to leave 6-8 rows of preformed holdfast buds distal to the cut
end, and the hydranths were also removed. After 2 hours the
most proximal one or two rows of buds were elongating and after
8 hours all four pieces showed holdfasts 0.5-0.8 mm. in length and
two were attached by their holdfasts. The much slower outgrowth
in this, as compared with other experiments is doubtless due to the
six day period in the laboratory without food preceding the ex-
periment.

These few experiments are sufficient to indicate the extremely
rapid activation of preformed holdfast buds and development of
new buds and holdfasts after removal of levels proximal to them.

CONCLUSION.

The Corymorpha holdfast is a tertiary axis which develops from
the secondary axis which constitutes the basal region. It re-
sembles the stolons of other hydroids in appearance and in its
growth and development, but is apparently more specialized than
the true stolon, since it is not known to give rise to a new hydranth-
stem axis as the stolon does when isolated. All experimental
methods applied to the holdfast indicate that it represents a phys-
iological gradient with high end at the tip, like a hydroid stolon, a
plant root or rhizoid and many other physiological axes. More-
over, the developmental behavior of the holdfast, the greater
amoeboid activity of its tip as compared with other parts, the
maintenance in good condition of distal, at the expense of proxi-
mal regions, even during gradual reduction by starvation, all con-
stitute further evidence that it is a gradient. And there is ab-
solutely nothing to indicate, and no ground for assuming that its
axiate character has any other basis than this gradient.

POLARITY IN HOLDFASTS OF CORY MORl'IT A. 33

The secondary gradient which characterizes the basal region
apparently represents a certain degree of physiological isolation
from apical dominance and the slow appearance and development
of the holdfasts under normal conditions, as compared with their
premature development under inhibiting conditions suggest that
a still higher degree of physiological isolation is necessary for their
development than for the determination of the basal end. Since
the apical region is more susceptible to inhibiting conditions than
the basal (Child, '260), these conditions decrease apical dominance
without greatly affecting the holdfasts themselves, i.e., they in-
hibit the internal factors which inhibit or retard the outgrowth of
the holdfasts and so increase the degree of physiological isolation
in the holdfast region.

When active development of the holdfasts does begin, a relation
of dominance and subordination arises within the holdfast system.
The most proximal holdfasts, i.e., those nearest the basal tip, are
the first to appear, as might be expected, since the degree of
physiological isolation must be greater there than in more distal
regions. These first holdfasts dominate buds distal to them and
prevent or retard their development. Removal or outgrowth and
separation of these most proximal holdfasts permits the further
development of those next distally and so on. This relation is
essentially identical with that found in many axiate complexes of
plants.

Concerning the nature of this dominance in Corymorpha noth-
ing definite can be said. The very wide occurrence in both plants
and animals of essentially identical phenomena of dominance, sub-
ordination and physiological isolation and the apparently primary
dependence of dominance on high metabolic activity, rather than
on any particular kind of activity indicate the non-specific character
of dominance and suggest that it is primarily dynamic rather than
a matter of specific chemical substances.

The holdfasts are also of considerable interest because of their
extremely rapid activation and development after the bud stage.
The outgrowth of holdfasts within an hour after removal of more
proximal regions represents an unusually rapid reaction and the
advance of the tip in later stages, even after isolation from the
stem may be as rapid as, or even more rapid than the early elonga-
3

34

C. M. CHILD.

tion. The question whether elongation of the holdfast is due to
actual growth, or, as Torrey ('04, '07) believes, to the amoeboid
activity of the tip, or as to the relative importance of these two
factors, cannot be finally answered at present, but it seems evident
that the earlier elongation of the holdfast, which may take place
free in the water without contact of the tip cannot be due to the
amoeboid activity of the tip. Migration of cells from the stem
into the holdfast is possible, but there is no evidence that it occurs.
To all appearances this elongation is due to real growth. As re-
gards the later advance of the tip after isolation and while it is
undergoing reduction and atrophy of the more proximal regions is
occurring, only further investigation can determine to what ex-
tent growth of the tip at the expense of other parts and to what
extent amoeboid activity is concerned. But whatever the proc-
esses underlying particular aspects of its behavior, the holdfast
originates as a local region of some sort of physiological activity
which decreases peripherally from a center. This activity leads
to outgrowth and elongation and the central, most active region
necessarily becomes the tip of the outgrowth and the radial gradi-
ent becomes an axial gradient. All the behavior of the holdfast
is an expression of that fact.

SUMMARY.

1. The holdfasts of Corymorpha are tertiary axial gradients
developing as buds from the secondary gradient of the basal
region.

2. Their premature appearance and outgrowth in embryonic
development under inhibiting conditions, as compared with their
much later appearance and slower growth in development under
normal conditions, suggest that a certain degree of physiological
isolation from apical dominance is necessary for their development.

3. As the holdfast elongates, the coenosarc of its proximal re-
gion atrophies so that the distal portion becomes completely iso-
lated from the parent stem, but advance of the tip continues until
exhaustion occurs, the distal ccenosarc remaining in good condition
while atrophy continues at the proximal end. The delicate peri-
sarcal tube secreted as the tip advances is the only connection with
the stem in later stages.

POLARITY IN HOLDFASTS OF CORYMORPHA. 35

4. Holdfast elongation is extremely rapid, sometimes I mm.
in 15-20 minutes. After a few hours the rate of elongation de-
creases and exhaustion occurs after 2-3 days. The maximum
lengths observed are 2-3 cm.

5. The holdfast tip in contact shows some amoeboid activity and
while the earlier stages of elongation are apparently due largely
or wholly to growth and may occur with the tip free in the water,
the continued advance over the substratum after isolation from
the stem and while undergoing decrease in size may be due in
part to the amoeboid activity of the tip.

LITERATURE.
Child, C. M.

'15 Individuality in Organisms. Chicago.

'19 The Axial Gradients in Hydrozoa. II. BIOL. BULL., XXXVII.

'23 The Axial Gradients in Hydrozoa. V. BIOL. BULL., XLV.

'24 Physiological Foundations of Behavior. New York.

'25 The Axial Gradients in Hydrozoa. VII. BIOL. BULL., XLVIII.

'263 Studies on the Axial Gradients in Corymorpha pahna. II. Biol.

Gen., II.
'26b Studies on the Axial Gradients in Corymorpha palina. III. Biol.

Gen., II.
'273 Modification of Polarity and Symmetry in Corymorpha palina by

means of Inhibiting Conditions and Differential Exposure. I.

Jour. Exp. Zool., XLVII.
*27b Modification of Polarity and Symmetry etc. III. Jour. Exp.

Zool., XLVIII.

Child, C. M., and Hyman, L. H.
'26 Studies on the Axial Gradients in Corymorpha pahna. I. Biol.

Gen., II.
Torrey, H. B.

'04 Biological Studies on Corymorpha. I. Jour. Exp. Zool., I.
'07 Biological Studies on Corymorpha. II. Univ. of Cal. Publ. Zool..

III.
'ioa Biological Studies on Corymorpha. III. Univ. of Cal. Publ. Zool.,

VI.
'iob Note on Geotropism in Corymorpha. Univ. of Cal. Publ. Zool., VI.

MORPHOLOGY AND BINARY FISSION OF PERANEMA
TRICHOPHORUM (EHRBG.) STEIN.

RICHARD P. HALL AND WILLIAM N. POWELL,

BIOLOGICAL LABORATORIES OF NEW YORK UNIVERSITY (UNIVERSITY COL-
LEGE) AND THE RICE INSTITUTE.

CONTENTS.

Introduction 36

Material and Methods 37

General Morphology 38

Gullet and Pharyngeal-rod Apparatus 39

Nucleus 42

Binary Fission 43

Nuclear Division 43

Kinetic Elements 44

Gullet and Pharyngeal-rod Apparatus 46

Discussion 47

Origin of the Centrosome in Mastigophora 47

The Parabasal Homologue Concept 55

Summary 57

Literature Cited 58

Explanation of Plates 61

INTRODUCTION.

In view of the somewhat unsatisfactory classification of the
colorless euglenoids in use prior to 1925, it has seemed that a
comparative study of Peranema and Hetcroncma (both of the
family Peranemidae Klebs) and of several representative Astasiidse
might lead to results bearing on this problem. Some of the vari-
ous difficulties, however, appear to have been settled satisfactorily
in the classification proposed by Calkins (1926). although Rhodes
(1926) has objected to the changes. In connection with work on
several Astasiidae and on Hetcroncina, this investigation was begun
by Powell in 1925, and has been completed by the senior author at
New York University. A brief account of our findings has been
published elsewhere (Hall, 1926).

The nomenclature of Peranema trichophorum has undergone
several changes since the organism was first described by Ehren-

36

MORPHOLOGY OF PERAXKMA TRICHOPIIORUM . 37

berg in 1838 as Traclielius trichophorus, a supposed ciliate. A
few years later the organism was recognized as a flagellate by
Dujardin (1841), and named Peranema protracta. Overlooking
tbe work of Dujardin, Claparede (cited by Kent) changed Ehren-
berg's terminology to Astasia tricliophora. The flagellate was
replaced in the genus Pcrancma by Stein (1878). and is now
known as Pcrancma trichophorum (Lemmermann, 1913; Conn
and Edmondson, 1918; Calkins, 1926). It is to be noted that
Conn and Edmondson (1918), although listing Peranema trieho-
pJwruin, described an Astasia trichophora with two flagella, one
long and the other relatively short. This organism, even if a valid
species, cannot be regarded as an example of Kent's Astasia tricho-
phora ( = = Peranema trichophorum}, nor may it be called an As-
tasia. since a single flagellum is characteristic of this genus (Kent,
Lemmermann, Calkins). The organism in question might be
placed in either the genus Heteronema or Distii/ma, but reassign-
ment would be rather difficult on the basis of Conn and Edmond-
son's figure.

MATERIAL AND METHODS.

For the past three years Peranema trichophorum has occurred
persistently in various laboratory cultures at The Rice Institute
and later at New York University. A satisfactory culture medium
consists of tap or pond water containing a piece of beef suet. The
medium is allowed to stand for a week or more, and then inocu-
lated with Peranema.

Material has been fixed in the solutions of Altmann and Schau-
dinn. The latter is far more satisfactory for examination of the
nucleus, pharyngeal-rod apparatus and flagellar rhizoplasts, while
the fixative of Altmann gives good results with the flagellum and
gullet, as well as with mitochondria (Causey, 1925). After
Schaudinn's fixative material has been stained in Bordeaux red
followed by iron-alum hematoxylin. and in iron-alum hematoxylin
either alone or followed by a counterstain (eosin, orange G, Congo
red). After fixation by Altmann's method Regaud's hematoxylin
has been used, either with or without a counterstain (eosin or
Bordeaux red). In other Altmann preparations the method of
bleaching described by Causey (1925) has been tried for demon-
stration of mitochondria.

38 RICHARD P. HALL AND WILLIAM N. POWELL.

In observations on the living organism a Zeiss cardioid dark-
field condenser has been used to advantage as a supplement to
the usual microscopic examination. The darkfield condenser has
been especially useful in determining the presence of cuticular
striations and small granules (mitochondria?) which show Brown-
ian movement, and in observing behavior of the flagellum. The
use of a yellow color filter in place of the usual " daylight " glass
of the microscope lamp was found to facilitate the tracing of
chromosomes and rhizoplasts in material stained in Bordeaux
red and hematoxylin. Our optical equipment consisted of a
Spencer binocular microscope equipped with Zeiss compensating
oculars and 2 mm. (Zeiss) and 1.5 mm. (Spencer) apochromatic
objectives.

GENERAL MORPHOLOGY.

Peranetna trichophorum is a colorless, plastic uniflagellate
euglenoid, varying in length from 22 to 70 //,, according to Lem-
mermann (1913). In our stained preparations the length of ex-
tended organisms ranges from 35 to about 60 p.. The living
flagellate shows several peculiarities : a gliding type of forward
locomotion, apparently without spiral rotation; a long and thick
flagellum, only the tip of which seems to be involved in forward
locomotion ; marked plasticity, or metaboly, especially in food-tak-
ing or in changing direction of locomotion ; and a pharyngeal-rod
apparatus lying in the anterior part of the body. With the dark-
field condenser, the periplast seems to be heavily striated and, as
in Jcnningsia (Schaeffer, 1918), the striations begin at the cy-
tostome and extend in spirals posteriorly (Fig. A, i). Occasion-
ally the striations may be detected in both Altmann and Schaudinn
preparations. There are no traces of " movable club-shaped ap-
pendages " such as are found on the periplast of Jenningsia. The
cytoplasm usually contains a number of paramylum bodies and
one or more food masses in vacuoles.

In Altmann-Regaud preparations of Peranema a number of
small cytoplasmic granules are often to be observed, apparently
more numerous in the outer layers than in the deeper cytoplasm.
In some of our material fixed in Schaudinn's fluid and stained in
Bordeaux red followed by hematoxylin, similar granules have been
found ; in such preparations the granules, when present at all, seem

MORPHOLOGY OF PERANEMA TRICHOPHORUM. 39

to be distributed throughout the cytoplasm. Such granules (PI.

•

I., Fig. i ; PI. II., Figs. 10, 13) are possibly to be regarded as
mitochondria, since they are readily demonstrated by Causey's
(1925) methods of mitochondrial fixation and staining. It has
been impossible, however, to demonstrate their mitochondrial na-
ture by the use of Janus green B on living specimens. \Yith the
darkfield condenser granules of similar size are to be seen in the
living organism and, so far as our observations go, they exhibit
continuous Brownian movement. Distinctly rod-shaped mitochon-
dria have not been found, and in this respect our findings agree
with those of Causey (1926), who described only granular mito-
chondria in Euglena gracilis.

GULLET AND PHARYNGEAL-ROD APPARATUS.

In the living flagellate it has not been possible to determine com-
pletely the structure of the gullet and pharyngeal-rod apparatus.
The following description of these structures is, therefore, based
principally upon stained preparations, and secondarily upon con-
firmatory observations on living flagellates. The gullet (Fig. A,
2-5) is a more or less flask-shaped cavity opening to the outside
through the cytostome. The somewhat enlarged posterior portion
of the gullet is often designated as the reservoir (Calkins, 1926;
Baker, 1926; Rhodes, 1926). Near the base of the gullet lies
a large contractile vacuole, which is sometimes apparent in stained
preparations. In watching this structure in several living or-
ganisms, the time between systoles was found to range from 10
to 15 seconds. The earliest stages of diastole could not be fol-
lowed. Soon after systole, a single new vacuole appears, increases
in size, and then discharges its contents into the gullet. A connect-
ing duct could not be traced in living organisms, but indications of
such a canal were found in a few Altmann-Regaud preparations.
The flagellum, on entering the cytostome, passes posteriorly to
end in the visual blepharoplast in the wall of the gullet (PI. I.,
Figs. 1-3; Fig. A, 2, 3). This insertion may also be traced in
the living organisms under oil immersion.

The pharyngeal-rod apparatus (Fig. A, 4, 5) extends along the
wall of the gullet and ends at the rim of the cytostome. In the
majority of organisms examined there are two distinct pharyngcal-

1

FIG. ^4. Pcraneina trichophoruni : (i) diagrammatic sketch of extended
organism showing flagellum, cytostome, gullet, contractile vacuole, pharyn-
geal-rod apparatus, spiral striations of periplast, nucleus, several food
masses (posterior), paramylum bodies, and small granules which exhibit
Brownian movement ; drawing based upon living specimens and stained
preparations; X 1500 approximately. (2) Gullet and cytostome in optical
"frontal" section; pharyngeal-rod apparatus omitted; Altmann-Regaud
preparation ; X 2025, (3) The same organism, camera-lucida drawing of
different focal plane to show "ventral" lip of cytostome; position of
flagellum indicated beyond tip of body; X 2025. (4) Gullet, pharyngeal-
rod apparatus and kinetic elements (blepharoplast, rhizoplast and centro-
some) ; Bordeaux red-hematoxylin preparation; X 1570. (5) Pharyngeal-
rod apparatus and kinetic elements in early prophase ; two flagella, and two
blepharoplasts joined by rhizoplasts to the daughter centrosomes ; Bor-
deaux-red-hematoxylin preparation; X 1570.

MORPHOLOGY OF PER AX KM A TRICHOPHORUM. 4!

rods, perhaps similar to the cesophageal rods of Heteronema acus
(Rhodes, 1926). In other instances (PI. I., Fig. 2) the rods are
so close together that their double nature may be detected only at
the ends ; and in a number of our stained preparations it is possible
to distinguish only a single thick rod (PI. I., Fig. 4). Whether
the occasional appearance of only one rod is due to fusion as a
result of fixation is uncertain, but this does not seem improbable.
Earlier descriptions of the pharyngeal-rod apparatus of Peranema
have been discussed elsewhere (Hall and Powell, 1927). In addi-
tion to the pharyngeal-rods which lie alongside the gullet, another
structure is sometimes to be detected in favorable preparations.
This is a curved cytostomal element which, when it can be demon-
strated, lies along one lip of the cytostome (Fig. A, 4, 5; PI. I.,
Figs, i, 3). From the lip of the cytostome the cytostomal rod
extends in a slightly posterior direction around one side of the
gullet to meet the pharyngeal rods. On the basis of Rhodes'
(1926) description, the "falcate trichite " of Heteronema acus
seems to be similar to this cytostomal element in Peranema.
Furthermore, a curved rod of similar relationships is shown in
Schaeffer's (1918) figures of Jcnningsla dlatomophaga.

The cytostome and gullet of Peranema function in the ingestion
of solid food, which is formed into food vacuoles at the base of
the gullet. The process of feeding has been described by Tann-
reuther (1923), who states that the protoplasmic contents of
captured organisms are sucked in by way of the gullet, after the
pharyngeal-rod has punctured the wall of the prey. Although
confirmation of this description has been attempted repeatedly,
such action of the pharyngeal-rod has so far escaped us. It has
been found, however, that carmine particles, portions of plant cells
and even entire organisms (apparently Mcnoldium incurvum and
Chilonwnas poramechim} are ingested, and that even in stained
preparations there are evidences of the formation of food vacuoles
at the base of the gullet. In watching living specimens in dark-
field, it has been possible to observe the pinching off of food
vacuoles from the base of the gullet. After their formation, the
food vacuoles pass to the posterior half of the body. In its feed-
ing habits, Peranema resembles Schaeffer's (1918) Jenningsia
which, on account of its much greater size, is able to ingest diatoms

RICHARD P. HALL AND WILLIAM N. POWELL.

as much as ioo/x, in length. Contrary to the description of Tann-
reuther for Pcranema, Schaeffer believes that in Jenningsia the
pharyngeal-rods are not protruded in feeding.

THE NUCLEUS.

In Pcrancma trichophorum the nucleus of the interphase con-
tains an endosome which is often irregular in outline or fragmented
(Fig. B), but commonly more or less ovoid in shape (PI. I., Figs,
i, 2). In stained preparations there is usually a clear area im-

9

FIG. B. Camera-lucida drawings of nuclei to show variations in form
of the endosome: 1-8, Pcranema trichophorum; g, 10, Heieroncma acus;
X 1045-

mediately surrounding the endosome. Between the endosome
and the nuclear membrane lie numerous chromatin granules, or
chromomeres, apparently embedded in some sort of an achromatic
network (Schaudinn and hematoxylin preparations) . Belaf ( 1916)
states that the endosome of Peranema is always alveolar in struc-
ture, even in the interphase. Such an appearance seems to be de-
pendent upon methods of fixation and staining, since it is evident
in some of our preparations but not in others. Hartmann and
Chagas (1910) have described a divided " centriole " in the endo-
some of Pcranema during the prophase, but they did not trace the
structure through mitosis and the cytological evidence presented

MORPHOLOGY OF PERANEMA TRICHOPHORUM. 43

does not seem to justify their conclusions. In preparations fixed
in Schaudinn's fluid and stained in iron hematoxylin without a
counterstain, the endosome frequently seems to consist of a deeply
staining central portion and a lighter peripheral layer. The darker
medullary portion might conceivably be interpreted as a centriole.
An interpretation of this sort, however, is confronted with the
difficulty that the endosome (Fig. B) of Perancma is often frag-
mented. The two, three or four fragments may show the same
peculiarity in staining with iron hematoxylin. Three or four such
" centrioles " would seem to be too many for any one nucleus.

NUCLEAR DIVISION.

The nucleus of the interphase is characterized by chromatin in
the form of small granules, or chromomeres, surrounding the en-
dosome and apparently embedded in the nodes of a lightly stained
network. No definite chromosomes can be detected at this stage
(PI. I., Figs, i, 2). In early prophases the chromatin granules
are replaced by thread-like chromosomes (PI. I., Fig. 3), formed
presumably by fusion of these chromomeres. In some early pro-
phases the chromosomes in their arrangement may simulate a
spireme which, except for the greater number of chromosomes,
is similar to that described in Euglcna agilis (Baker, 1926). In
others, the chromosomes appear to be radially oriented (PI. I.,
Fig. 3). In later prophases, in which a second flagellum appears
and division of the gullet takes place, indications of chromosome
splitting are to be seen (PI. I., Figs. 4-6). As in the case of
Eugleua agilis (Baker, 1926) separation of the daughter chromo-
somes begins, in the majority of cases, at one end of the pair and
progresses toward the other end. This process results in the ap-
pearance of V-shaped structures, with a daughter chromosome
forming each limb of the V (PI. I., Figs. 5, 6). The V's gradu-
ally straighten out to form a belt of approximately parallel chromo-
some pairs (PI. II., Figs. 8, 9), the members of each pair still
being joined at one end (the " apex " of the V). This same type
of chromosome splitting has been reported for Mcnoidiuni in-
curvum (Hall, 1923), as well as for Eugleua agilis.

The endosome, during the later prophases (PI. I., Figs. 5, 6;
PI. II., Figs. 8, 9), is gradually drawn out to form a central dumb-

44 RICHARD P. HALL AND WILLIAM N. POWELL.

bell shaped structure surrounded by the belt of chromosomes. Al-
though the endosome of the interphase and earlier prophases fre-
quently consists of two or more fragments, the elongating endo-
some of the later prophase is nearly always a single structure. In
fact, only a single exception was found in our material ; in this
stage, a metaphase, there were two elongated endosomes, one
slightly larger than the other. It seems possible, therefore, that
separate endosomal fragments present in earlier stages may fuse
to form a single mass before the late prophase.

In the metaphase (PI. II., Fig. 10) the daughter chromosomes
are completely separated in the equatorial plane of the elongated
nucleus. During the anaphases (PI. II., Fig. 11) the daughter
chromosome belts draw apart as further elongation and constric-
tion of the nucleus take place. The chromosomes still retain the
parallel arrangement observed in the metaphase, and appear to be
uniformly distributed around each half of the endosome. After
constriction of the nucleus into two daughter nuclei in the early
telophase the chromosomes may retain for some time their parallel
arrangement (PI. II., Fig. 12). Later on the chromosomes be-
come irregularly distributed (PI. II., Fig. 12), and finally chromo-
meres appear as reorganization of the daughter nuclei nears com-
pletion (PI. II., Fig. 13).

The endosome, during metaphase and anaphase stages, becomes
more and more elongated, so that the median portion is drawn out
to form what has been called a centrodesmose (Calkins, 1926;
Baker, 1926). The name centrodesmose is usually applied to a
fibril joining the daughter halves of an intranuclear centriole.
Since there is no evidence for the presence of a centriole in Pcra-
ucina, the use of such a term will be avoided in this case. The
median portion of the endosome finally breaks in the late anaphase
just before the daughter nuclei are completely separated. For a
time the daughter endosomes remain drawn out at their median
ends (PI. II., Figs, n, 12). In later telophases, however, each
endosome approaches the characteristic interphase form as reor-
ganization of the daughter nuclei nears completion.

KINETIC ELEMENTS.

In the typical interphase (PI. I., Figs. 1-2; Fig. A, 2) there
is a single flagellum ending in a blepharoplast in the wall of the

MORPHOLOGY OF PKKA .\ KM A TRTCHOPIIORf M . 45

gullet. In the early prophase two blepharoplasts are apparent (PI.
I., Fig. 4), one of them giving rise to a short flagellar outgrowth,
and the other serving for insertion of the old flagellum. At this
stage the two blepharoplasts are still very close together. In
slightly later prophases (Fig. A, 5) the flagellar outgrowth has
increased in length, but the two blepharoplasts are still contiguous.
The three stages just described are construed as indicating an
early division of the old blepharoplast, followed by outgrowth of
a new flagellum from one of the daughter blepharoplasts. This
interpretation seems better supported by our evidence than is any
hypothesis of nuclear origin of the blepharoplast (Baker, 1926).
If the theory of nuclear origin of blepharoplasts were applied to
Peranema it would necessitate the assumption that, in the earliest
biflagellate stages observed (PI. I., Fig. 4), a granule has already
passed through the nuclear membrane and made its way forward,
to become embedded in the wall of the gullet and give rise to a
new flagellum. Such an assumption is not warranted on the basis
of our observations. In the later prophases (PI. I., Figs. 5, 6. 9)
the two blepharoplasts with their attached flagella gradually draw
apart as widening and fission of the gullet begin.

In interkinetic stages much difficulty has been experienced in
attempting to trace a flagellar rhizoplast to the nuclear membrane,
due to the fact that the interphase nucleus lies typically in the
posterior third of the body. The food vacuoles and fixation
vacuoles often intervening between nucleus and gullet (PI. I., Fig.
i) make the search for a delicate rhizoplast extremely difficult.
Although apparent rhizoplasts have been traced part way in a
number of instances, it has been impossible to demonstrate to our
own satisfaction a rhizoplast joining blepharoplast and nucleus
during interphase stages. Such a rhizoplast, however, has been
demonstrated in vegetative stages of another euglenoid, Phacus
costata (Bretschneider, 1926), and it is not impossible that an
existing interphase rhizoplast of Peranema has merely been over-
looked by us. In the early prophase, however, it has been pos-
sible in a number of instances to trace a flagellar rhizoplast to the
nucleus (PI. I., Fig. 3; Fig. A, 4). In such cases the cytoplasm
happened to be practically free from food vacuoles, so that the
usual difficulties were not encountered. In later prophases (Fig.

46 RICHARD P. HALL AND WILLIAM N. POWELL.

A, 5) two rhizoplasts extend from the daughter hlepharoplasts to
a pair of granules on the nuclear membrane. This stage and the
preceding one indicate that the extranuclear centrosome on the
nuclear membrane divides during the prophase, and that each
daughter centrosome is joined by a rhizoplast to the corresponding
daughter blepharoplast. This arrangement of blepharoplast, rhizo-
plast and centrosome is found also in subsequent stages of binary
fission (PI. I., Fig. 5; PI. II., Figs. 9, n, J2). In flagellates in
which the daughter gullets lie at different focal planes than the
nuclei it is often difficult to trace rhizoplasts to the nuclei, and for
this reason they are not indicated in some of our figures. Baker
(1926) has described a similar rhizoplast joining each daughter
blepharoplast to a so-called " parabasal homologue " at the nuclear
membrane in Euglena agilis.

On the basis of our evidence it is impossible to prove that the
granule, or centrosome, at the nuclear membrane is a permanent
extranuclear organelle of Peraneina trichophoruni, on account of
the difficulty in demonstrating a rhizoplast joining blepharoplast
and centrosome in interkinetic stages. On the other hand, there
is no evidence at all that the extranuclear centrosome ever arises
from the endosome and migrates out of the nucleus. It is neces-
sary, therefore, to leave the question open as to whether this ex-
tranuclear granule is a permanent structure (although so far un-
detected in interphase), or whether it may arise de novo at the
beginning of binary fission.

GULLET AND PHARYNGEAL-ROD APPARATUS.
In the early stages of binary fission the lower portion of the
gullet gradually increases in width (PI. I., Fig. 5). In such
stages a pharyngeal-rod element is to be detected on each side of
the gullet. This might suggest that one of the two original rods
passes to each daughter organism. In the stage mentioned, how-
ever, the rods appear smaller than those of the typical vegetative
stage. Furthermore, Rhodes (1926) states that during "division
the old ' staborgan ' (pharyngeal-rod apparatus) migrates pos-
teriad and is resorbed " in Hetcroncma acns. " In the early telo-
phase new ones differentiate from the cytoplasm." Rhodes' in-
terpretation might be supported by our Fig. 5 of Peranema, in

MORPHOLOGY OF PERANEMA TRICHOPHORUM. 47

which rod-like structures are shown enclosed in two food vacuoles
posterior to the nucleus. Even in this stage, however, pharyn-
geal-rod elements are apparent on either side of the gullet. Hence,
if replacement of the apparatus does occur in Peranema, as might
he expected, the process must take place rapidly and at some time
hefore late prophase. Division of the gullet is completed in late
prophase (PI. I., Fig. 7). In metaphase and following stages
the lower ends of the two gullets lie near the poles of the elongated
nucleus and later at the poles of the daughter nuclei, polarity being
indicated by the orientation of the endosome or its daughter
halves. Separation of the two gullets seems to begin always at
the lower end and to progress anteriorly.

ORIGIN OF THE CENTROSOME IN MASTIGOPHORA.

An extranuclear division center is apparently of common oc-
currence in the Mastigophora, since such a structure has been
described or figured in one or more members of each of the fol-
lowing families (Calkins' system) : Ochromonadidse (Chrysomono-
dida) ; Protodiniferidse, Noctilucidse, Blastodiniidse, Peridiniidse
(Dinoflagellida) ; Euglenidae, Astasiidse (Euglenida) ; Rhizomas-
tigidse (Pantastomatida) ; Choanoflagellidge, Bodonidse (Proto-
mastigida) ; Tribes Monozoa and Diplozoa (Polymastigida) ;
Lophomonadidae, Staurojoenidse, Trichonymphidse and Holomas-
tigotidae (Hypermastigida). On the basis of structural rela-
tionships, avoiding consideration of the highly specialized Hyper-
mastigida. two general types of extranuclear division centers may
be recognized in the flagellates.

The centroblepharoplast (Kofoid),-or centrosome-blepharoplast
(Minchin), combines in a single structure the functions of both
centrosome and blepharoplast. In the interkinetic phases (Fig.
C, i) the flagellum arises from the centroblepharoplast. In nu-
clear division the centroblepharoplast divides, and the daughter
granules take up positions at opposite poles of the nucleus (Fig. C,
2), thus resembling in behavior the centrosomes in Metazoan
mitosis. There is of course need for caution in interpreting such
structures in flagellates, since the granules are often small and are
not always easy to demonstrate. Since morphology and staining
reactions alone are clearly inadequate for confirmatory evidence,

48

RICHARD P. HALL AND WILLIAM N. POWELL.

it seems that the behavior of such a granule during nuclear divi-
sion constitutes the only practicable basis for determining its cen-
trosomal nature in flagellates. In accordance with this assump-
tion, a blepharoplast which divides early in binary fission, and

FIG. C. (i) Centroblepharoplast of interkinetic stages in flagellate; (2)
centroblepharoplast in binary fission; (3) interkinetic stage, separate ble-
pharoplast and extranuclear centrosome ; (4) binary fission, separate ble-
pharoplast and centrosome; (5) Baker's (1926) concept of kinetic elements
in Euglena agilis (after Baker), has. gr., basal granule ("basal body") ;
bl., blepharoplast; bleph., centroblepharoplast; c., centrosome; nuc., nu-
cleus; par. b., parabasal homologue (Baker) ; rh., rhizoplast.

whose daughter blepharoplasts appear and remain at opposite
poles of the nucleus throughout mitosis, may be regarded as a
centrosome. To such a blepharoplast, therefore, the term cen-
troblepharoplast is to be applied in the sense just explained. As
a matter of fact, there would seem to be no logical objection to the
view that a centrosome may exhibit the properties of a blepharo-
plast, in view of the behavior of the centrosome in spermatogenesis
in most Metazoa. The following are some of the flagellates in

MORPHOLOGY OF PERANEMA TRICHOPHORUM. 49

which a division center of this type has been described : Eutricho-
inasti.v scrpcntis, Trichomonas augusta, T. inuris (Kofoid and
Swezy, 1915), T. muris (Wenrich, 1921), T. batrachonnn, Mono-
ccrcoinoiias mclolouthcc ( Grasse, 1926), Hctcromita iinciinita and
Cercomonas longicauda (Wenyon, 1926).

In the second type, centrosome and blepharoplast appear to be
two separate structures. In a number of flagellates, for example,
a rhizoplast extends from the blepharoplast (or blepharoplasts)
to a granule, designated as the extranuclear centrosome, on the
nuclear membrane (Fig. C, 3). At an early stage of binary fission
this centrosome divides, as may also the blepharoplast and rhizo-
plast, and the daughter centrosomes move to opposite poles of the
nucleus (Fig. C, 4). In some instances, each daughter centrosome
is joined by a rhizoplast to a corresponding blepharoplast. In
other forms, such as Giardia, only one of the centrosomes may be
connected with a blepharoplast. A few of the flagellates in which
a division center of the second type has been described are : Giardia
(Kofoid and Christiansen, 1915; Kofoid and Swezy, 1922),
Menoidium incurvuni (Hall, 1923), Polytomella citri (Kater,
1925), Protcromonas longifila (Grasse, 1926), and possibly Eit-
glena agilis (Baker, 1926). In addition to the centrosome-rhizo-
plast-blepharoplast arrangement, some workers have described or
figured granules which appear at the poles of the dividing nucleus
but do not, apparently, show rhizoplast connections with the
blepharoplasts. While such granules are more liable to misin-
terpretation than those which are undoubtedly joined to the ble-
pharoplasts, it seems probable that at least some of them may be
of centrosomal nature.

In the case of the centroblepharoplast of many flagellates it
is possible, with certain obvious reservations, to regard the centro-
some as a permanent organelle which functions as a component
of the locomotor apparatus in vegetative stages and, in addition,
as a division center in binary fission. This interpretation, with
the same reservations, might be applied to flagellates in which cen-
trosome and blepharoplast are separate structures, provided a
rhizoplast can be traced from blepharoplast to centrosome during
interkinetic as well as division stages. Such a concept of con-
tinuity of the extranuclear centrosome is, however, confronted
4

5O RICHARD P. HALL AND WILLIAM N. POWELL.

with the objection that little is known of the fate of this structure
during encystment in flagellates. In Chilo-inasti.r niesnili (Kofoid
and Swezy, 1920), a form characterized by nuclear division within
the cyst, a division center of the second type appears in encysted
stages. This is true also of Giardia ent erica (Kofoid and Swezy,
1922). As for flagellates which do not divide within the cyst,
there is little critical evidence as to whether the centrosome persists
or not. Furthermore, any attempt to solve this problem is handi-
capped by the fact the the criterion of behavior during nuclear di-
vision cannot be employed for identification of a suspected centro-
some. In Polytoniclla citri, however, Kater (1925) has demon-
strated the presence, just prior to excystment, of an extranuclear
granule joined by rhizoplasts to the blepharoplasts (from which
flagella subsequently grow out). Continuity of the centrosome
during encystment might be suggested by the cases cited, but it is
impossible without more evidence, to draw any definite conclusion.

When a connection between blepharoplast and extranuclear
centrosome cannot be traced in interkinetic stages, there arises the
problem as to how and where the centrosome originates at the
onset of binary fission. Several possibilities are apparent, (i)
The centrosome may arise from within the nucleus, becoming
extranuclear at the beginning of nuclear division. (2) The cen-
trosome may be permanently extranuclear, but easily detected only
in stages of mitosis. (3) The extranuclear centrosome may arise
de novo in the cytoplasm at the beginning of nuclear division, and
then disappear after the completion of binary fission.

From various descriptions of nuclear origin of extranuclear
kinetic granules, it would appear that the typical process is as
follows. A small outgrowth first appears on one side of the en-
dosome. This endosomal bud breaks off and moves toward the
nuclear membrane, finally passing through it to become an extra-
nuclear granule. Once in this position, the granule may bud off
blepharoplasts or basal bodies, give rise to flagella, or otherwise
assert its kinetic nature. Three similar descriptions of such a
process may be mentioned. One of the earlier accounts is that
of Wilson (1916), who found that in Diinastigamceba (Ncegleria)
tjniberi the blepharoplast seemed to grow out from the endosome
of the nucleus at the beginning of enflagellation. Wenyon (1926),

MORPHOLOGY OF PERANEMA TRICHOPHORUM. 51

after a critical study of this species, states : " The writer, after
examining many thousands of amoebae at all stages of flagellum
formation, has failed entirely to trace the origin of the blepharo-
plasts from the karyosome. ... It seems more probable that the
blepharoplasts are present in the cytoplasm even during the
amoeboid phases of the organism. They are so minute that, un-
less their connection with the flagella can be traced, it is impossible
to distinguish them from other granules which occur in the cyto-
plasm. . . . There seems to be no real evidence that they are de-
rived from the karyosome of the nucleus."

In a more recent paper Kater (1925) has described such an
origin of the centrosome in Polytomella citri just before excyst-
ment takes place. At a late stage of encystment a small bud ap-
pears on one side of the endosome. In the next step, two chro-
matic masses are shown in the nucleus, one the central karyosome
and the other a " chromatic ball '" just within the nuclear mem-
brane. In the third stage, a granule (" chromatic ball ") is shown
immediately outside the nuclear membrane. Later, a small " basal
granule " separates from the chromatic ball and moves toward
the periphery of the cell, remaining connected by a rhizoplast with
the parent granule. In the later stages, the basal granule and
rhizoplast divide and, at the beginning of excystment, flagella
grow out from the basal granules.

Kater's paper is open to the following criticisms. His critical
stage (Fig. 35), in which an intranuclear granule is shown at the
periphery of the nucleus, offers no proof that such a " chromatic
ball " is on its way to passage through the nuclear membrane. It
is possible that this " chromatic ball " may be nothing more than
a fragment of the karyosome, since Kater's figures show that the
karyosome breaks up in early division stages. In addition, the
mere presence of an extranuclear granule on the nuclear mem-
brane (Kater's Fig. 36) is in itself no evidence at all that such a
position is the result of a migration out of the nucleus. Further-
more, the fact that a fibril apparently joins the karyosome to this
extranuclear granule means, in this instance, precisely nothing,
since the typical interphase nucleus of Polytoinclla shows from
four to six such fibrils extending from the karyosome to the nu-
clear membrane. Most important of all, Kater has not proved

52 RICHARD P. HALL AND WILLIAM N. POWELL.

that the " chromatic ball "' (centrosome) is not permanently ex-
tranuclear during encystment. He actually concludes that, in the
active flagellates, the extranuclear centrosome is present in inter-
kinetic as well as in division stages. In the encysted forms Kater
figures in several stages a large number of " metachromatic gran-
ules " closely surrounding the nucleus. Even if the " chromatic
ball " were extranuclear in such cases, it could not be detected.
In presumably later stages of encystment the metachromatic
granules have migrated to the periphery of the cell, and the
" chromatic ball " is revealed just outside the nuclear membrane.

Baker (1926) has described a similar phenomenon in Euglena
agilis: " One of the first evidences of approaching mitosis is the
budding off from the endosome of a deeply staining granule of
unmistakable character. . . . The bud increases in size . . . and
passes . . . through the chromatin of the outer nucleus to the
nuclear membrane. . . . On the nuclear membrane the mass di-
vides and the daughter bodies separate and pass gradually to op-
posite sides of the nucleus. ... As they separate the bodies again
divide giving rise to the blepharoplasts of the two future cells.
From each blepharoplast a new flagellum arises. When the nu-
cleus has passed anteriorly until it comes in contact with the lower
border of the reservoir, each blepharoplast divides, giving rise to
a basal granule which serves as the basis for the smaller branch
of the bifurcated flagellum. . . . The mass budded from the en-
dosome contains blepharoplasts, basal bodies, and a chromatic
residue . . . (which) remains on the nuclear membrane."

In an early prophase (Baker's Fig. 3) a small darkly stained
granule is shown near the endosome and attached to it by a rhizo-
plast. In succeeding stages (Figs. 4, 5) two granules (supposedly
division products of the first), one of them still joined to the endo-
some, are to be seen near the nuclear membrane. These two
granules look surprisingly like several other granules in the nuclei
of Figs. 4 and 5. In Fig. 6 there are two pairs of extranuclear
granules at opposite sides of the gullet. Each pair of granules is
said to represent a blepharoplast and basal body to which " each
daughter half of the kinetic complex has given rise." The
" daughter halves of the kinetic complex " are no longer to be seen
on the nuclear membrane. Instead, a single granule (" residue of

MORPHOLOGY OF PERANKMA TRICHOPHORU M . 53

the kinetic complex ") invested with a halo is adjacent to and
joined by a rhizoplast to the endosome.

And yet it is stated that each half of the kinetic complex gives
rise to a " hlepharoplast-basal body complex which becomes the
base of the bifurcated flagellum " and to a chromatoid residue
which " remains on the nuclear membrane." After reading this
statement, careful examination of Baker's figures 3 to 8 leaves the
reader somewhat puzzled. There is no apparent difference be-
tween the " kinetic complex " of the early prophase (Fig. 3) and
the " residue of the kinetic complex " shown near the endosome
in Fig. 6 ; in fact, the figures would even suggest identity of the
two structures. How then is it possible to assume that between
these two similar stages the kinetic complex has given rise by
division to tu'o daughter granules on the nuclear membrane, and
that each of these granules has budded off an extranuclear ble-
pharoplast-basal body, leaving a chromatoid residue on the nuclear
membrane.

In Fig. 8 the daughter blepharoplasts are seen at opposite ends
of the nucleus. One of the blepharoplasts is now connected with
the single residue of the kinetic complex, which lies near the en-
dosome. In Figs. 9, 10 and n the single residue of the kinetic
complex has apparently disappeared, and the blepharoplasts are
no longer shown on the nuclear membrane. Instead, two granules
lie at the " poles " of the nucleus, while the blepharoplasts occupy
their normal positions in the wall of the gullet. Each of these new-
granules, joined by a rhizoplast to its respective blepharoplast, is
designated as a " residue from the kinetic complex." Have the
two residues of Fig. 9 arisen by division of the single residue near
the endosome in Fig. 8? Baker fails to make this clear. During
the telophases each " residue " pinches off from the nucleus in
a little nimbus of its own and loses connection with the blepharo-
plast. Each residue persists until some time during the next
mitosis, and while " it does not function as a division center, yet
it sets up a series of concatenated events which result in the final
production of two fully vigorous individuals from a single parent
cell."

Baker's description does not explain: (a) the reason for the
striking similarity between the kinetic complex in Fig. 3 and the

54 RICHARD P. HALL AND WILLIAM N. POWELL.

single residue of the kinetic complex in Fig. 6 and 8; (b) the rela-
tion between the single residue near the endosome in Figs. 6 and
8 and the two residues on the nuclear membrane in Fig. 9 and fol-
lowing stages ; (c ) the shifting of the blepharoplasts from the poles
of the nucleus in Fig. 8 to their usual position (Fig. 9, etc.) ; (d)
the fact that in Fig. 8 a rhizoplast is traced from one blepharoplast
(on the nuclear membrane) to the single residue near the endo-
some, whereas in Fig. 9 each blepharoplast (now in the wall of the
gullet) is joined by a rhizoplast to a residue at one pole of the
nucleus; or (r) the fact that, while the chromatoid residues are
said to remain on the nuclear membrane, a single residue is shown
near the endosome in Figs. 6 and 8. The only possible conclusion
seems to be that the evidence presented fails to prove that the
extranuclear blepharoplast and basal body are derived from the
intranuclear kinetic complex. No critical stages of such a process
are shown in the plates and, on the basis of the evidence presented
in Figs. I to 9, Baker's hypothesis seems to be no more than the
following assumption : given two intranuclear granules in an early
prophase and four extranuclear granules in a later prophase, the
extranuclear granules must have been derived from the intra-
nuclear ones. Furthermore, the plates offer no cytological proof
that the polar chromatic residues (also termed " centroblepharo-
plast homologue " and " parabasal homologue ") are actually de-
rivatives of the kinetic complex. Since Baker's hypothesis of
nuclear origin of kinetic granules is based entirely upon the sup-
posed demonstration of the two phenomena just mentioned, a
demonstration which his plates fail to show, his account cannot be
accepted until verified.

In the case of Pcrancma trichophonnn it is perhaps significant
that the examination of several thousand interkinetic and early
division stages has failed entirely to reveal any positive evidence
for nuclear origin of extranuclear kinetic granules.

In conclusion it may be pointed out that the various workers
cited have presented no critical evidence that the centrosome or
blepharoplast is derived from the intranuclear endosome in flagel-
lates. Wilson's description has been refuted by Wenyon, the
evidence in Rater's paper is entirely inadequate, and Baker's plates
do not contain the proof for his hypothesis. Various other ac-

MORPHOLOGY OF PERANEMA TRICHOPHORUM. 55

counts of nuclear origin of kinetic granules, mentioned by Calkins
(1926) and Wenyon (1926), seem to be no more convincing.
There is no need to point out that most of the evidence against
the theory of nuclear origin is purely negative in character, but
in spite of this fact the whole doctrine of nuclear origin of kinetic
granules in flagellates can only be regarded as " not proven."
And until adequate proof is forthcoming there seems to be little
justification for the statements of various authors that there exists,
hidden away in the endosome of interkinetic flagellates, an intra-
nuclear centriole, or endobasal body, which gives rise to blepharo-
plasts, centrosomes or other kinetic elements during binary fission.

THE PARABASAL HOMOLOGUE CONCEPT.

In the paper just discussed Baker (1926) has introduced an-
other complication in the concept of a homology between his " resi-
due of the kinetic complex " and the parabasal body. " An en-
dobasal body is located in the endosome and gives rise by division
to a chromatoid mass, the kinetic complex. This migrates to the
periphery of the nucleus, where it divides, the daughter halves pass-
ing to the approximate poles of the division figure. . . . The
chromatoid residue left on the nuclear membrane after the ble-
pharoplast-basal body complex is freed ... is homologous to the
parabasal body of many parasitic flagellates as well as to the so-
called kinetonucleus of Trypanosomes." There is no evidence in
Baker's figures for the existence of an endobasal body within the
endosome, or for the formation of the kinetic complex by division
of such a pre-existing granule in the endosome. Obviously, the
endobasal body must be a purely hypothetical structure in this
case.

The parabasal homology of the " residue of the kinetic complex "
is said to be indicated partly by the position of the residue with
relation to the poles of the dividing nucleus. It is impossible
to accept such a statement without reservation. In Belaf's (1926)
figures of nuclear division in Trypanosoma loxicc, for example,
the position of the daughter parabasal bodies is extremely variable
with respect to the poles of the nucleus. In other parasitic flagel-
lates, such as Giardia and TricJwmoiias, the centrosomes (or cen-
tro-blepharoplasts), and not the parabasal bodies, occupy the poles
of the nucleus.

56 RICHARD P. HALL AND WILLIAM N. POWELL.

Another foundation for Baker's concept is his apparent belief
that the kinetic residue and parabasal body show the same sort of
connections with the nucleus and blepharoplast. Thus he stresses
the view that in trypanosomes and crithidial forms the parabasal
body is joined to the nucleus by a rhizoplast. As a matter of fact,
McCulloch (1919) has demonstrated that in Crithidia euryoph-
thalmi and Schizotrypanum crusi, flagellates mentioned by Baker,
this is not the case. Instead, the nuclear rhizoplast extends from
the blepharoplast to the nucleus, while the parabasal body is joined
by its own rhizoplasts to the blepharoplast and not to the nucleus.
In other flagellates as well (e.g., Trichoinonas, Polymastix, Giar-
dia) , the parabasal body seems to be connected with a blepharoplast
rather than with the nucleus.

Another supposed support for this hypothesis is the statement
that "McCulloch (1917) described the origin of a chromatoid
mass from the endosome in Crithidia euryophthalmi. . . . She
identifies the structure as the parabasal body." An examination
of McCulloch's article shows that nowhere in the paper cited does
she mention such an origin of the parabasal body. Nor does she,
in her other papers (1915, 1919). figure or describe the origin of
the parabasal body directly from the nucleus. It would seem that
Baker, instead of reading carefully McCulloch's original papers,
allowed himself to be misled by Calkins (1926). On page 97,
Fig. 46, of his textbook, Calkins has copied several figures of
small zooids of Crithidia and labeled them " origin of parabasal
from endosome of nucleus (after McCulloch)." This interpreta-
tion is somewhat liberal, since McCulloch (1919) herself suggested
a nuclear origin of the blepharoplast, from which in later develop-
ment a flagellum " grows forward anteriorly and a parabasal body
to the side." On the basis of McCulloch's statements, it would
be just as logical to conclude that the flagellum originates from
the endosome within the nucleus as to assume that the parabasal
body is budded off directly from the endosome. It might be men-
tioned also that McCulloch qualified her suggestion with the state-
ment that her search for a complete series of stages in the origin
of blepharoplast and parabasal body had been disappointing; and
that, " owing to the size of these zooids it is extremely difficult
... to form any adequate conception of their structure." Fur-

MORPHOLOGY OF PERANEMA TRICHOPHORUM. 57

thermore, the belief in nuclear origin of the parabasal body is in
opposition to Wenyon's ( 1926) opinion that " there is no conclu-
sive evidence that the parabasal body is of nuclear origin, as some
have supposed."

Baker's parabasal homologue concept thus rests in part upon an
apparent misunderstanding of the relationships of the blepharo-
plast, parabasal body and nucleus in Crithidia and other flagellates,
and in part upon the seemingly erroneous impression that the
parabasal body of Crithidia curyophthalmi is budded off directly
from the endosome. On the basis of Baker's figures of Euglena
agilts it is evident that the so-called " parabasal homologue " is
similar in behavior, for the most part, to the extranuclear centro-
some in a number of the flagellates already mentioned. For this
reason it would seem a more plausible interpretation to consider
this " homologue " a centrosome joined by a rhizoplast to the
blepharoplast, rather than a rudiment of the parabasal body of
parasitic flagellates.

SUMMARY.

Peranema trichophoruni is a colorless, metabolic euglenoid with
a single large flagellum, only the tip of which is employed in for-
ward locomotion. The flagellum arises from a blepharoplast in
the wall of the flask-shaped gullet and leaves the body through the
cytostome. The pharyngeal-rod apparatus (" staborgan ") consists
typically of two pharyngeal-rods and a curved cytostomal element.
The rods extend from the rim of the cytostome posteriorly along
the gullet. The cytostomal element lies along one " lip " of the
cytostome and extends partly around the gullet to join the pharyn-
geal-rods. The apparatus probably operates in expansion of cy-
tostome and gullet in feeding, although this could not be deter-
mined in living material with any degree of accuracy.

In mitosis longitudinal splitting of the chromosomes occurs,
followed by unipolar separation of the daughter chromosomes.
This results in the appearance of V-shaped daughter chromosome
pairs, which unfold to form the chromosome belt of the later pro-
phases. Final separation of the daughter chromosomes (at the
" apex " of each V) occurs in the metaphase. This process is
similar to that described in Menoidium incurvum (Hall, 1923) and
Euglena agilis (Baker, 1926).

58 RICHARD P. HALL AND WILLIAM N. POWELL.

The kinetic elements consist of a blepharoplast, in which the
flagellum ends, a centrosome on the nuclear membrane, and a
rhizoplast joining blepharoplast and centrosome. The blepharo-
plast-rhizoplast-centrosome complex was detected in early pro-
phases, but not in interkinetic stages. In later prophases both
centrosome and blepharoplast apparently divide, and the daughter
centrosomes gradually move apart toward opposite poles of the
nucleus. A second flagellum grows out from one of the daughter
blepharoplasts. There is no evidence for nuclear origin of cen-
trosome or blepharoplast in Perancma trichophorum.

The question of the nuclear origin of centrosome and blepharo-
plast in flagellates is discussed, with special referenence to the
papers of Kater and Baker. It is pointed out that neither paper
contains sufficient evidence to prove a nuclear origin of kinetic
elements.

LITERATURE CITED.
Baker, W. B.

'26 Studies in the Life-history of Euglcna. BIOL. BULL., 51: 321-362,
2 pis., 2 textfigs.

Belaf, K.

'16 Protozoenstudien. I. Amoeba diplogcna, Astasia levis n. sp., Rhyn-
clwmnnas nasuta Klebs. Arch. Protistenk., 36: 13-51, pis. 2-4, 3
textfigs.

'26 Der Formwechsel der Protistenkerne. Ergebn. u. Fortschr. Zool.,

6: 235-654, 263 textfigs.
Bretschneider, L. H.

'26 Uber den feineren Ban von Phacus costata Conrad. Arch. Pro-
tistenk., 53: 131-134, 6 figs.

Calkins, G. N.

'26 The Biology of the Protozoa (Philadelphia, Lea and Febiger),
623 pp., 238 figs.

Causey, D.

'25 Notes on Technique for the Demonstration of Mitochondria in

Protozoa. Tr. Am. Micr. Soc., 44: 156-161.
'26 Mitochondria in Euglcna gracilis. Univ. Calif. Publ. Zool., 28:

217-224, pis. 19-20.

Conn, H. W., and C. H. Edmondson

'18 " Flagellate and Ciliate Protozoa," in Ward and Whipple, Fresh-
water Biology, 238-300, Figs. 362-559.
Grasse, P. P.

'26 Contribution a lY-tude des flagelles parasites. Arch. d. zool. exper.
et gen., 65 : 345-602, pis. 8-19, 76 textfigs.

MORPHOLOGY OF PEKANEMA TRICHOPHORUM. 59

Hall, R. P.

'23 Morphology and Binary Fission of Menoidium incurvum (Fres.)
Klebs. Univ. Calif. Publ. Zool., 20: 447-476, pis. 40-41, 2 text-
figs.

'26 Binary Fission in Pcrancma trichophorum. Anat. Rec., 34: 155-

Hall, R. P., and W. N. Powell.

'27 A Note on the Morphology and Taxonomic Position of Peranema
trichophorum. Tr. Am. Micr. Soc. 46: 155-165, pi. i.

Hartmann, M., and C. Chagas.

'10 Flagellaten-Studien. Mem. do Inst. Osw. Cruz, 2 : 64-125, pis. 4-9.

Kater, J. McA.

'25 Morphology and Life-history of Polytomclla citri sp. nov. BIOL.
BULL., 49: 213-237, 3 pis., i textfig.

Kent, W. Saville.

'8o-'82 A Manual of the Infusoria (London, Bogen), i: 1-472; 2:

473-913; 3: Pis. 1-50.
Kofoid, C. A., and E. B. Christiansen.

'15 On Binary and Multiple Fission in Giardia muris. Univ. Calif. Publ.
Zool., 16: 30-54. pis. 5-8. i textfig.

Kofoid, C. A., and O. Swezy.

'15 Mitosis and Multiple Fission in Trichomonad Flagellates. Proc.

Am. Acad. Arts and Sci., 51: 289-378, pis. 1-8, 7 textfigs.
'22 Mitosis and Fission in the Active and Encysted stages of Giardia

cnt erica (Grassi) of Man, with a Discussion of the Origin of

Bilateral Symmetry in the Polymastigote Flagellates. Univ. Calif.

Publ. Zool., 20: 199-234, pis. 23-26, ii textfigs.

Lemmermann, E.
'13 " Eugleninae " in Die Siisswasserflora Deutschlands, Osterreichs und

der Schweiz. Heft 2 (Jena, G. Fischer), pp. 115-174, Figs. 181-

377-
McCulloch, I.

'15 An Outline of the Morphology and Life-history of Crithidia Icp-

tocoridis sp. nov. Univ. Calif. Publ. Zool., 16: 1-22, pis. 1-4, I

textfig.
'17 Crithidia cnryophthalmi sp. nov. from the Hemipteran bug Euryoph-

tlialmiis coni'irus. Idem., 18 : 75-88, 35 figs.
'19 A Comparison of the Life-cycle of Crithidia with that of Try-

panosoma in the Invertebrate Host. Idem., 19: 135-190, pis. 2-6.

3 textfigs.

Rhodes, R. C.

'26 Mouth and Feeding Habits of Heteronema acus. (The mouth of
the Peranemida Klebs, involving the classification of the Eugle-
noidina relative to the new family Heteronemidae Calkins.) Anat.
Rec., 34: 152-153-
Schaeffer, A. A.

'18 A New and Remarkable Diatom-eating Flagellate, Jenningsia dia-
tnniophaya n.g., n. sp. Tr. Am. Micr. Soc., 37: 177-182, pi. 13.

6O RICHARD P. HALL AND WILLIAM N. POWELL.

Stein, F. R. V.

'78 Der Organismus der Infusionsthiere. Abt. III. Der Organismus
der Flagellaten, Halfte I (Leipzig, Engelmann), x + 154 pp., pis.
1-24.
Tannreuther, G. W.

'23 Nutrition and Reproduction in Euglena. Arch. Entwick. Mech., 52 :

367-383, 52 figs.
Wenrich, D. H.
'21 The Structure and Division of Trichomonas muris (Hartmann).

Jour. Morph., 36: 119-155, 4 pis., i textfig.
Wenyon, C. M.
'26 Protozoology, Vol. I (London, Balliere, Tindall and Cox), xvi

+ 778 pp., 336 figs.
Wilson, C. W.

'16 On the Life-history of a Soil Amoeba. Univ. Calif. Publ. Zool., 16:
241-292, pis. 18-23.

62 RICHARD P. HALL AND WILLIAM N. POWKLL.

DESCRIPTION OF PLATES.

Pemiicina trichophorum: drawings from preparations fixed in Schau-
dinn's fluid and stained in Bordeaux red followed by iron-alum hematoxy-
lin ; magnification of 1570, except Fig. 8.

PLATE I.

FIG. I. Nucleus in interphase ; cytoplasm contains several food masses in
vacuoles near nucleus; cytoplasmic granules (mitochondria?) are nu-
merous.

FIG. 2. Contracted organism ; nucleus in interphase ; pharyngeal rods
appear separate only at anterior ends.

FIG. 3. Early prophase ; rhizoplast extends from blepharoplast to cen-
trosome at nuclear membrane; two endosomal fragments; chromosomes
show radial arrangement.

FIG. 4. Prophase; two blepharoplasts present, and a new flagellar out-
growth has appeared; rhizoplasts not traced to nucleus (see figure A,
5), since the latter lies at'a deeper focal plane than the gullet; only one
thick pharyngeal-rod to be seen ; some indications of chromosome splitting.

FIG. 5. Prophase; optical section of gullet and cytostome; position of
pharyngeal-rods indicated ; separation of daughter chromosomes beginning.

FIG. 6. Later prophase; V-shaped chromosome-pairs evident; portion
of one daughter flagellum and gullet indicated.

FIG. 7. Different focal plane through same organism ; the two daughter
gullets have just separated anteriorly; two pairs of pharyngeal-rods present.

BIOLOGICAL BULLETIN, VOL. LIV

. / vvc , ,

.

'*. -,- ••*

#>:••'.'•'••--

'

<

f/ )

• "- ';

'

*%sr
•

• • ^ *••*'• "^«

; ,.<:. r a

, v*"M-. -,>«--.. i L
^ « •- •JK-... .-••?*•,

/ ^.

• ' -

••«<• p?

hv* ,:

y sr V

'•-• --J -V . ;'

V

H.. . •* , ?
,

v -< >- ,,-V

RICHARD P. HALL AND WILLIAM N. POWELL.

64 RICHARD P. HALL AND WILLIAM N. POWELL.

PLATE. II.

FIG. 8. Late prophase; daughter chromosome pairs unfolding to form
chromosome belt ; the gullets, which lie almost perpendicular to the plane
of the drawing, are omitted ; X 2025.

FIG. 9. Late prophase ; chromosome-V's almost completely unfolded to
form chromosome ring around the e'ongated endosome ; right gullet in
optical section ; left gullet and flagellum, which extend beneath the nucleus,
indicated in their approximate positions ; two pairs of pharyngeal-rods ap-
parent.

FIG. 10. Metaphase ; daughter chromosomes completely separated in
equatorial plane; the gullets, lying at deeper focal plane than nucleus, are
shown in optical section ; pharyngeal-rod apparatus omitted in order to
show distribution of granules (mitochondria?).

FIG. ii. Late anaphase ; daughter chromosome groups and daughter
endosomes well separated; nucleus almost completely constricted; gullets
shown in optical section; rhizoplast shown joining one blepharoplast to
a centrosome.

FIG. 12. Telophase ; in one daughter nucleus the chromosomes still
show a parallel arrangement, in the other they are becoming irregularly
distributed ; from each blepharoplast a rhizoplast extends to a centrosome
at the nuclear membrane; on the left only the lower end of the gullet is
shown, and the pharyngeal rods are omitted.

FIG. 13. Later telophase; daughter nuclei undergoing reorganization;
gullets, which extend beneath the nuclei, are indicated in optical section ;
two pairs of pharyngeal-rods apparent.

BIOLOGICAL BULLETIN VOL. LIV.

PLATE II.

•- - • " ,<

^

' >•• .-=**«:

.

•

r-

•

•a

' ,

'

• ^* ^i^T»^k3fE_r***^Ewrf>-»C?* i

;• 'v^&ggl^^^jl? - *•* .• .

" '• '- :• % , * * - •

f- ••*•• "V *. •• i s - ••'

Sfe<^

^A-**^-

9

.

).
>f

y

v . ' ,.•••-

*• -\ ' '*• S

• -*. \

12

• *. ^'ij¥»

••" *•• JV "--v* * ••

>. ^ '"••»* *' >,

* 'r'f~ >\^'.C ' p^

13

RICHARD P. HALL AND WILLIAM N. POWELL.

MISCELLANEOUS OBSERVATIONS ON HYDRA, WITH
SPECIAL REFERENCE TO REPRODUCTION.

LIBBIE H. HYMAN,

HULL ZOOLOLGICAL LABORATORY, UNIVERSITY OF CHICACO.

The following observations have accumulated during several
years' cultivation of hydra in the laboratory together with numer-
ous collections of the animals from nature.

The taxonomy employed is that of Schulze ('17) who has made
a much needed revision of the hydras and furnished adequate de-
scriptions of the known species. He divides them into three gen-
era: Chlorohydra, with symbiotic algae in the entoderm ; Pclma-
tohydra, column differentiated into body and stalk; and Hydra,
sensu strictu, column not so differentiated. The common green
Hydra then becomes Chlorohydra I'iridissinia ( - = Hydra viridis-
siina Pallas, Hydra 1'iridis Linnaeus) ; the brown Hydra is des-
ignated as PclinatoJiydra oligactis ( = - Hydra oligactis Pallas,
Hydra fused Linnaeus, Hydra dia'da Downing) ; and the genus
Hydra contains a number of species some of which in the past
have been confused under the name of Hydra vnlgarls ( - - Hydra
grlsea Linnaeus). There are at least four, possibly five, species
of hydra around Chicago, of which one does not correspond with
any of Schulze's descriptions and requires to be named and de-
scribed.

The observations in this paper concern Pclinatohydra oligactis
(Pallas) Schulze and Hydra stcllata Schulze.

CULTIVATION.

The cultivation of hydras in the laboratory is not difficult, pro-
vided sufficient food is available. The real problem is therefore
the maintenance of an adequate food supply. Daphnia or other
daphnids furnish the most convenient food and are raised as fol-
lows. Several dishpans or tubs are partly filled with well-aerated
water. To each is added a sufficient amount of boiled aquatic

65

66 LIBBIE H. HYMAN.

vegetation such as El odea or Ceratophyllum to cover the bottom
with a thin layer of debris without inducing visible fermentation.
The water should remain clear. After such cultures have stood
for three or four clays, dapnids are added to them. Later small
pieces of fresh raw liver or the freshly killed bodies of small ver-
tebrates are added but never in sufficient amount to cause cloudi-
ness of the water. Each such culture serves for some time but
new cultures must be prepared at intervals.

Hydras collected from nature are transferred to cultures pre-
pared as follows.1 Jars are filled one-half to three-fourths full of
thoroughly aerated water (avoid water that has been treated with
germicides). To each jar is added a small amount of water plant
such as Elodca that has been boiled for ten minutes. My obser-
vations agree with those of 'Goetsch ('22/7) that hydras flourish
better in cultures containing some organic debris than in clear
water. After jars so prepared have stood for several days, re-
maining clear and not showing any visible signs of fermentation,
hydras may be added to them. Such cultures of hydra must be
fed daily by adding an appropriate number of daphnids to each.
One daphnid per hydra per day is sufficient to keep a culture in
good condition but to induce rapid budding each animal must
receive three or four daphnids daily. At frequent intervals the
ejected remains of daphnids must be removed from the bottom
of the culture and a small amount of water dipped out and replaced
by prepared culture water.

It frequently happens that freshly collected hydras undergo de-
pression (see below) after a short sojourn in the laboratory and
are apt to die, owing probably to the sudden change of environ-
mental conditions. It may therefore be often difficult to start
laboratory cultures from specimens taken from nature. Under
such circumstances it is necessary to resort to " hand-feeding " to
keep the animals alive until they have become adjusted to labora-
tory conditions. By ':< hand-feeding " one means that each hydra
is fed daily by placing a crushed daphnid or Cyclops, preferably
the latter, in contact with its tentacles by means of a fine forceps.
After several days of such care, some specimens will usually have

1 The expression " prepared culture water " or " culture water " used
hereafter refers to water prepared in the manner here described.

OBSKKYATIONS ON HYDRA. 67

recovered and will serve for starting cultures. At intervals new
cultures should be prepared and hydras transferred to these. In
my experience, Pelmatohydra oligactis is much easier to cultivate
than any other species.

The chief difficulty encountered in the continuous cultivation of
hydra in the laboratory is the occurrence of the condition known
to German zoologists as depression.

DEPRESSION.

This phenomenon, well known to European investigators, has
received almost no attention from American zoologists.1 It had
been seen by many of the early students of hydra, beginning with
Trembley, but was first adequately described by R. Hertwig ('06)
who also suggested the name depression, derived from supposedly
similar phenomena in Protozoa. Later Hertwig's student Frisch-
holz ('09) published an extensive paper on the matter and since
that time several workers have considered the phenomenon.

I have observed depression on numerous occasions in various
species of hydra, particularly in P. oligactis. The first symptom
of the onset of depression is a refusal to feed accompanied by a
persistent slight contracted state of column and tentacles. The
tentacles then gradually disappear from the tips proximally.
Meantime, column and tentacles continue to shorten, presenting a
stout, stiff, contracted appearance. Soon the animal is reduced
to a short stout cylindrical body bearing short stumps of tentacles.
At this stage or earlier, the animal frequently ceases to maintain
an erect position, often lying on its side ; the basal disk may lose
its hold on the substratum. As depression continues, the tentacles

1 An example of this is furnished by the recent paper <>f Kepner and
Jester ('27). Finding specimens of the green hydra with shortened ten-
tacles they ascribe this to an ingestion of its own tentacles by the animal.
\Yhile this apparently does occasionally happen it is obvious from their
description and figures that the majority of these specimens with shortened
tentacles as well as those with incomplete tentacles which they found in
nature are simply in a state of depression. Depression and starvation are
independent phenomena and the shortening of the tentacles in depression
has no relation to starvation. The ejection of cells ;md fragments from
the mouth is characteristic of animals in early stages of depression. The
finding of nematocysts in the entoderm ce'ls is no proof that ingesti«n of
tentacles had occurred since in depression the cells pass int<> the interior.

68 LIBBIE H. HYMAN.

completely vanish and the greatly shortened column diminishes
from its distal end towards the basal disk. At length only a small
round ball, constituting the basal end of the hydra, remains and
this soon disintegrates. Hydra may recover spontaneously from
early stages of depression or may be induced to recover by altering
environmental conditions ; but recovery from advanced stages of
depression seems to be impossible.

The first satisfactory account of the histological changes occur-
ring during depression is that recently given by Rehm ('25). Ac-
cording to Rehm, depression begins in the entoderm cells which
gradually become loose and pass into the gastrovascular cavity.
The shortened swollen appearance of tentacles and column in de-
pression is due to the accumulation within the cavity of loose ento-
derm cells which may also be discharged from the mouth in early
stages. Later the mouth closes. After the entoderm has largely
disintegrated, the mesoglcea dissolves and the ectoderm cells pass
through the same changes as did the entoderm. All of the changes
in question begin at the distal end of the animal and progress basi-
petally.

Various external and internal conditions have been assigned as
factors which induce the state of depression. Principal among
them are: long continued rich feeding (Frischholz, '06, Rehm,
'25) ; extremes of temperature (Goetsch, '22b, Grosz, Rehm,
Frischholz) ; lack of oxygen or fouling of the water (Frischholz,
Goetsch, Rehm, Grosz) ; transfer to clear fresh water (Frischholz,
Goetsch) ; infestation with protozoan symbionts (Grosz, '25).
Starvation, moderately low temperatures, and transfer to properly
prepared culture water delay the onset of a depression period and .
lengthen the intervals between depressions (Frischholz).

My own observations on the phenomenon of depression are in
accord with those of my predecessors.1 Temperatures over 20°
C., long continued rich feeding, lack of oxygen, accumulation of
metabolic products and dead daphnids in the culture, and transfer
to fresh clean water are conditions that tend to induce depression,
often very rapidly. Lowered temperature, frequent removal of
organic debris, and small replacements of the water with prepared
culture water at intervals are the chief methods of delaying de-

1 Except that T do not think protozoan symbionts cause depression.

or.sr.KY.YnoNs ON HYDRA. 69

pression. that is, of increasing the period between depressions and
of assisting recovery from depression. In spite of all care, cul-
tures of long standing will undergo depression at intervals. The
best method that I have found for bringing about recovery of
depressed specimens is their transference to a culture in which
hydras are flourishing. The beneficial effect of such transfer is
almost incredible. Elongation of body and tentacles is often evi-
dent within an hour. Depressed specimens that have refused to
feed for several days may accept food within fifteen minutes after
being placed in a culture containing healthy hydras. Other less
effective methods are aeration and partial replacement of the water
with prepared culture water.

In my opinion the primary cause of depression is a general
lowering of the rate of physiological processes, or more briefly,
a condition of senescence. According to Child ('15, p. 58) " sen-
escence is primarily a decrease in rate of dynamic processes con-
ditioned by the accumulation, differentiation and other associated
changes of the material of the colloid substratum." The principle
of indirect susceptibility (Child, '13), according to which animals
in a low metabolic state are less able to adjust to slight adverse con-
ditions than animals in a more active state, probably applies here.
Depression ensues in senescent animals when they are subjected to
slight unfavorable conditions such as lack of oxygen and accumu-
lation of waste materials. In support of this view are the facts
that young animals (personal observations) and regenerated ani-
mals (Goetsch, '21) are less likely to undergo depression that or-
dinary large animals. Evidence that depressed hydras are really
in a low metabolic state will be presented later*

The observations of Trembley (1/44, pp. 182-84), Welch and
Loomis ('24) and Kepner and Jester ('27) leave little doubt that
depression occurs in a state of nature, probably in consequence of
too high temperature.

The relation between depression and the production of abnor-
malities is discussed later.

IRREGULAR HUI>I>IN<; AND COLONY FORMATION.

Ordinarily budding occurs at a definite level of the column and
this fact gives rise to the impression that there exists a fixed bud-

JO LIBBIE H. HYMAN.

ding zone. Such an idea is, however, erroneous, since under
proper physiological conditions, buds may arise anywhere on the
column except within a certain distance of the peristome J and
probably a less distance of the foot. The occurrence of irregu-
larly disposed buds on non-sexual specimens was noted by Trem-
bley (1744), Laurent (1844), Hertwig ('06) and Grosz ('25).
Study of these cases indicates as causes of irregular budding senes-
cence, depression, unusual elongation of the column, and operative
interference with the peristome.

Adventitious budding occurs, however, most commonly on male
specimens which on being fed enter upon a period of asexual re-
production following testis formation. Such irregular budding of
male specimens was noted by Trembley, Laurent, R. Hertwig,
Krapfenbauer ('08) and Grosz ('25) and I shall present numer-
ous instances of it later (see Plates IV. to VI.).

Apart from its occurrence on male specimens I have observed
among thousands of hydras cultivated in the laboratory but one
case of irregular budding, seen in a large P. oliyactis. This animal
(Fig. i, Plate I.), with a column some 20 mm. in length, had six
buds with a possible seventh ; distally there were four buds some-
what normally arranged followed by a stalk-like differentiation
of the column below which was a body region bearing two, possi-
bly three, buds irregularly disposed, followed by the regular stalk
of the specimen. Excessive length of the column probably ac-
counts for this case of irregular budding.

The essential feature of irregular budding is that the budding
region is spread over a greater length of the column than normal,
particularly in the direction of the peristome. Further such bud-
ding is associated with a low metabolic state, such as senescence,
sexual maturity, and depression, or with an excessively elongated
column. These facts suggest that normally the budding zone oc-
curs at a certain distance from the peristome (and probably also
foot) because the peristome usually maintains a certain metabolic
level. Lowering of this level permits buds to form nearer the
peristome and thus to spread over a greater length of column.
The same result may also be brought about by an excessive elon-

: The name peristome is used to designate the tentacle-bearing region of
Hydra.

OBSERVATIONS ON HYDRA. J I

gation of the column without any lowering of the metabolic con-
dition of the peristome. In my observations of abnormal budding
in male specimens numerous cases were observed of such a dis-
tance relationship between the peristome and the location of buds.
Figs. 48 to 50, Plate IV., are typical examples. In Fig. 48 is
shown a portion of a male hydra, separated by transverse fission
and lacking therefore a foot, whose first bud was formed at the
extreme base. Six days later, the specimen was considerably
elongated and had produced a second bud markedly distal to the
first one (Fig. 49). Two days later, the very elongated animal
(Fig. 50) was found with a number of irregularly disposed buds,
the most distal of which is still about the same distance from the
peristome as was the single bud of Fig. 48. Many similar cases
were observed during the study of the budding of males. It was
also noted that buds may occur at the' extreme proximal end when
a foot is lacking but not when a foot is present. These facts sug-
gest that the location of the budding zone on normal hydras rep-
resents a balance between the inhibiting action of peristome and
basal disk on bud formation.

Brief mention may be made here of so-called colony formation
in hydra, due to the retention of buds far beyond the normal pe-
riod of time. Hase ('09), Koch ('u), and Grosz ('25) ascribe
colony formation to such conditions as senescence and depression
and my own observations concur with their opinions. Any low
metabolic condition permits buds to remain indefinitely upon the
parent, because the extent of control of the parent peristome is
then diminished and the peristome of the bud is thus able to gain
control over the adjacent parent column and foot which it uses as
its own. Such retained buds frequently grow out a foot laterally
on the side away from their attachment to the parent and may
fasten to the substratum by such feet. Numerous cases of such
colonies with several retained buds and feet are illustrated in my
Plates III. to V.

ABNORMALITIES.

In addition to irregular budding and colony formation, abnor-
malities in hydra consist chiefly in doubling or multiplication of
tentacles, and apical and basal ends. Such abnormalities occur in
laboratory cultures and are also found in nature. A lowered

72

LIBP.IE H. HYMAN.

metabolic state appears to be the primary cause of these abnormali-
ties, excepting such as may result from regeneration after injury.
The origin of the majority of them is to lie sought in recovery after
depression. At the start of this recovery process, duplication or
multiplication of tentacles, hypostome and peristome frequently
occur, apparently because regeneration begins at two or more
points instead of the usual single point. On several occasions I
have observed split tentacles and doubled and tripled peristomes
with corresponding sets of tentacles arising in specimens recover-
ing from depression. Figs. 23 and 24 illustrate two specimens at
the beginning of regeneration after depression showing split ten-
tacles. This relation of depression to abnormalities was previ-
ously noted by Koch ('12), Boecker ('14), Drzewina and Bohn
('16) and Grosz ('25).

There is also one other common mode of origin of duplicate
apical regions. As previously discussed, buds are likely to be re-
tained on depressed, senescent, or sexual animals. Such retained
buds may grow to be of the same size as the parent or may move
distally, in either case eventually producing an apparently double-
headed hydra. The origin of doubled heads from retained buds
by one or the other method has been recorded by Hase ('09),
Koelitz ('09), Koch ('12), and Grosz ('25) and I have observed
it in my cultures several times. Fig. 21 illustrates a bud, origin-
ally located at the usual budding zone, which is moving distally by
fusion with the parent ; Fig. 22 shows the same specimen later,
simulating a double-headed hydra. Forked tentacles may also
originate from a fusion as well as from a splitting process.

The occasional occurrence in nature of hydras with double or
multiple apical or basal structures or other irregularities has been
recorded by many writers. During the past several years I have
collected annually ten to twenty thousand individuals of an unde-
scribed species of hydra and fixed them within a few hours after
collection. Abnormal specimens have been found in these collec-
tions in the ratio of about one per thousand. Figs. 2 to 20 illus-
trate all of the abnormal individuals occurring among twenty thou-
sand specimens taken in the fall of 1924. Figs. 2 to 1 1 are cases
of doubled apical ends showing various degrees of splitting (or
fusion). As already suggested such duplicities probably arise

OBSERVATIONS ON HYDRA. 73

either during recovery after depression or from the retention of a
bud. The smaller size of one component in Figs. 2, 3, 6, and 7
suggests the latter explanation in these particular cases. Figs. 12
to 1 8 are forms with doubled basal ends in various stages of split-
ting (or possibly fusion). Such basal duplicities may arise in
recovery after depression (Koch, Boecker) or probably more com-
monly through the formation of an accessory foot near the middle
of the column, as in Fig. 12. Such a form as Fig. i 2 by a splitting
process in the apical direction would give rise to cases like Figs.
14, 17, 18, etc. The inequality of the two components in these
cases strongly suggests the correctness of this interpretation. Fig.
19 is undoubtedly a case of a retained bud which has formed a foot
in situ, while Fig. 20 is to be interpreted as a case of fusion of a
bud with a parent.

FISSION.

Longitudinal and transverse fission have been reported for
hydra by a number of observers: Parke ('oo), Leiber ('09),
Koelitz ('o&j, fo8&. '09), Koch ('12), Boecker ('14), Wachs ('19),
Goetsch ('19), and Grosz ('25). Some of them, particularly
Koelitz, regard these processes as normal modes of reproduction
while others consider them as sequels of abnormal conditions.
\Yith this latter viewpoint my own observations concur. Most
cases of so-called longitudinal fission are simple basipetal splittings
of double-headed animals which have arisen through recovery after
depression. I have observed several such " longitudinal fissions "
in my cultures. The splitting process was very slow requiring sev-
eral days to reach the basal disk. After having separated to the
basal disk, the animals remained in this condition for a long time
without further change ; eventually complete separation occurred.
I have also observed numerous cases of transverse fission in male
" colonies," which regularly break up into smaller components by
transverse constrictions.

Thus fission in hydra is not a normal method of asexual repro-
duction but a mode of regulation of previously existing abnormali-
ties. From the literature it appears also that fissions may be
initiated in specimens in a low metabolic state which are not other-
wjse abnormal.

74

LIBBIE H. HYMAN.

METABOLIC STATIC.

The following tests of metabolic state in hydra were carried out
by means of the direct susceptibility method of Child. This
method has been reviewed so many times in papers from this lab-
oratory that an extended consideration of its basis and application
seems unnecessary here. Briefly the method consists in determin-
ing the time between exposure and death in solution of toxic sub-
stances of appropriate concentration. In general this time is
roughly proportional to the metabolic state of the organism, ani-
mals in a high state of physiological activity dying faster than those
in a lower state. The animals to be tested were placed in watch
glasses, usually some time before the test, so that they might at-
tach. Unattached animals exhibit increased activity of the pos-
terior end, which should be avoided if possible. The killing agent
consisted in all cases of potassium cyanide, made up in the same
water as that in which the hydras were cultured, in concentrations
from 1/200 to 1/700 mol., depending on the species. After the
animals had attached, the water was gently removed and replaced
with the cyanide solution. The watch glass was then covered
with a thin plate of glass so as to exclude air, and death observed
under the microscope. Only the time elapsing between exposure
and death of the whole animal will lie considered here as death
differences along the axis were described previously (Child and
Hyman, '19).

The previous observations of Child and Hyman ('19) that meta-
bolic rate as determined by this method is greater the smaller the
animal and is increased by contraction and feeding were confirmed.
In P. oligactis, however, very little difference in susceptibility was
found between animals of different sizes except that mature buds
and the smallest free specimens were more susceptible than all
other sizes. The latter did not differ among themselves. Some
attempt was made to discover how long the metabolic increase
due to food ingestion lasted but without very definite result except
that the increase continues after the prey has been ejected and
seems to persist for about two days. Since size, state of nutrition,
and state of contraction affect susceptibility it is evident that these
factors must be eliminated or equalized before comparisons by the
susceptibility method can be regarded as valid.

OBSERVATIONS OX HYDKA. 75

i. Starvation. — It is probable that many of the observations in
the literature purporting to deal with starved hydras really con-
cern depression. The reason for this is that in starvation experi-
ments hydras are usually isolated from the general cultures into
dishes of clear fresh water. Under such circumstances, hydras
almost invariably undergo depression, a condition frequently mis-
taken as a consequence of starvation, with which in fact it has
nothing at all to do. In order to starve hydras without subjecting
them to depression, it is essential to isolate them in water from the
same culture in which they have been living. They then starve
while remaining normal. They gradually reduce in size and be-
come more slender and paler in color. Small specimens reduce
at a much more rapid rate than do large specimens.

In the following experiments, a number of hydras of a certain
size, that attained when budding begins, were isolated in a dish of
culture water. At intervals during the starvation period, their
susceptibility was compared with that of fed controls. Fed con-
trols were animals of the size just specified, isolated in the same
manner from the same culture two days before the susceptibility
test. Thus all " fed " controls have been starved two days and
are of the same size as were the starved specimens at the beginning
of the starvation period. All of the data concern P. oligactis.
The concentration of cyanide used was 1/600 mol., a concentration
in which .the animals remain well expanded. In these experi-
ments, the animals were not attached.

Four days starvation : 9 starved specimens, dead. 240 to 365
minutes, 9 fed controls. 175 to 515 minutes.

Six days starvation: 6 starved specimens. 170 to 365 minutes,
5 fed controls, 430 to 530 minutes.

Seven days starvation : 10 starved animals, 80 to 370 minutes,

4 fed controls, 280 to 420 minutes.

Eight days starvation : 5 starved specimens, 65 to 290 minutes,

5 fed controls, 335 to 430 minutes.

Ten days starvation: 7 starved specimens. 150 to 450 minutes,

5 fed controls, 430 to 555 minutes.

Eleven days starvation: n starved animals, 30 to 225 minutes.

6 fed controls, 390 to 540 minutes.

76 LIBBIE H. HVMAN.

Fifteen clays starvation : 7 starved specimens, 30 to 300 minutes,
5 fed controls, 315 to 405 minutes.

Eighteen days starvation : 5 starved animals, 50 to 390 minutes,
4 fed controls, 230 to 435 minutes.

These experiments indicate that starvation raises the metabolic
rate of hydra. The effect is not very noticeable until after four
clays of starvation. The difference between starved and fed ani-
mals seems to be greatest in moderate stages of starvation, six to
eleven days, and to diminish in more prolonged starvation, fifteen
to eighteen days. It was noticed that small specimens were very
much more susceptible than large ones starved the same length of
time.

2. Depression. — The earlier comparisons of depressed and nor-
mal specimens gave irregular results, the former being more re-
sistant in some cases, but not in others. It is probable that in
many cases the control specimens were not properly comparable
to the depressed animals. It is in fact somewhat difficult to ob-
tain proper controls for naturally occurring cases of depression,
because generally the entire culture undergoes depression simul-
taneously. This entails the use of control specimens from other
cultures, a questionable proceeding, since different cultures often
differ physiologically. Further, depressed specimens do not feed
and the factor of starvation probably enters into the result. It
was finally decided that conclusive experiments could be secured
only by inducing depression artificially. As already stated this
can usually be accomplished by removing the specimens from the
cultures and placing them in fresh clean water. It is sometimes
necessary to change this water several times to keep the animals
in a state of depression. The procedure then consisted in isolat-
ing a number of specimens of a certain size (size just before bud-
ding) into a finger bowl of fresh clean well water, drawn directly
from a tap, and at the same time isolating others of the same size
from the same culture into a finger bowl of culture water. The
former undergo depression, the latter remain normal. Neither
culture is fed, both being thus in the same stage of starvation. At
intervals specimens were removed from both cultures and their
susceptibility compared. The depressed specimens were mostly
in moderately advanced depression, with short stumps of tentacles

OBSERVATIONS ON HYDRA. 77

or none and with body contracted to a short cylindrical form.
The concentration of cyanide used was 1/600 mol. ; in this con-
centration the control specimens were always well expanded while
the depressed animals tended to some degree of contraction. In
some experiments fed controls (starved two days) were compared
simultaneously with depressed animals and starved controls.
Typical results are :

Two days depression and starvation: 3 depressed, 375 to 510
minutes, 3 controls, 150 to 285 minutes.

Four days depression and starvation: 4 depressed, 415 to 540
minutes, 4 starved controls, 240 to 375 minutes, 4 fed controls
(2 days starvation), 190 to 480 minutes.

Four days depression and starvation: 4 depressed (slight de-
pression), 380 to 665 minutes, 5 starved controls, 255 to 335 min-
utes, 5 fed controls, 175 to 515 minutes.

Five days depression and starvation : 4 depressed, 280 to 350
minutes, 4 starved controls, 165 to 425 minutes.

Six days depression and starvation : 6 depressed, 350 to 470
minutes, 6 starved controls, 170 to 365 minutes, 5 fed controls,
43° to 53° minutes.

Seven days depression and starvation: 12 depressed, 205 to 400
minutes, 10 starved controls, 80 to 370 minutes, 4 fed controls,
280 to 420 minutes.

These experiments show that depressed specimens are generally
in a lower metabolic state than normal control specimens which
have been starved for the same length of time. Occasional ex-
ceptions are met. Depressed specimens are usually not less sus-
ceptible to cyanide than normal fed specimens (starved two days).
It thus appears that starvation is a factor which must be taken into
account. It should further be mentioned that depressed specimens
are always in a state of contraction which commonly increases
when they are placed in cyanide. This factor makes it difficult
to gain a true idea of their metabolic state. Finally it appeared
in the experiments that animals in advanced depression were at
times not very much more resistant to cyanide than the controls,
in fact, in some individuals, less so. This seemed to be due to the
fact that in advanced depression the apical end of the specimens
is already in a state of disintegration and so falls to pieces readily.
6

78 LIBBIE H. HYMAN.

This may account for the fact, generally apparent in the experi-
ments, that the metabolic differences between depressed and normal
animals were small in distal levels and much more noticeable in
proximal levels. One seems justified in concluding that depressed
specimens are in a low metabolic state, which however cannot be
evaluated very accurately owing to the interference of such factors
as starvation, reduction in size, contraction, and preexisting disin-
tegration, all factors which affect susceptibility but which cannot
be eliminated as they are inherent in the depressed condition.

3. Regeneration following Depression. — Depressed specimens
may be caused to regenerate if the water is replaced by culture
water. For proper comparison, starved controls should be used.
One experiment was carried out on P. olic/actis, with KNC 1/600
mol. as follows : 9 specimens, three days depression, seven days
regeneration, death, 150 to 290 minutes; 7 specimens, starved 10
days, 150 to 450 minutes; 5 fed controls, two days starvation, 430
to 555 minutes. This experiment indicates that depressed speci-
mens, which as shown above, are in a low metabolic state, attain
as a result of regeneration, a metabolic rate equal to or greater
than that of specimens in the same state of starvation, and much
greater than that of fed specimens.1

4. Sexual State. — Whenever material was available, suscepti-
bility tests were run on sexual as compared with non-sexual speci-
mens. Here again it is difficult to obtain proper controls or to
decide what constitutes a proper control. Sexual specimens do
not feed as a rule and are therefore generally in a state of starva-
tion, but it is usually impossible to know precisely when they last
fed. If non-sexual controls are taken from the same culture as
the sexual specimens, one cannot be sure that they may not be on
the verge of developing gonads. On the other hand controls from
other cultures are also open to criticism. The data to be pre-
sented were collected some time ago, as in the past few years, there
has not been any available material ; and they are probably not as
carefully controlled as they should have been. In many cases it
was necessary to take specimens from different cultures and the

1 It is worthy of note that of the nine regenerated specimens, eight
budded, despite the fact that they were greatly reduced in size, while only
one of the controls, starved the same length of time, produced a bud. Loss
of control by the peristome thus leads to budding.

OBSERVATIONS ON HYDRA. 79

matter of feeding was not considered at all. Only specimens of
the same size were compared. In some cases the species was not
determined. The data available are as follows :

P. oligactis, KNC 1/400 mol. : 8 females, death 5 to 9 hours;
9 asexual specimens of same size from the same culture, 3 to 7
hours.

P. oligactis, KNC 1/400 mol.: 7 females, 5 to 9 hours; 6 non-
sexual of same size from same culture, 2 to 4 hours.

Comparisons of male and female specimens of the same size
but not always from the same culture, KNC 1/200 mol., species
not determined: male, 115 minutes, female 340 minutes; male 70,
female 340; male 170, female 365 ; two males 55 and 125 minutes,
female 125 plus ; 2 males 80 minutes, female, over 125 ; male 5 1/2
hours, female, over 24 hours ; 2 males 345 and 500 minutes, female,
over 24 hours ; asexual specimen 40 minutes, 2 males 55 and 95
minutes, female 155 minutes.

These figures indicate that sexual specimens are in a lower
metabolic state than non-sexual individuals and that females are
decidedly lower than males. Contrary to what one might expect,
the gonads are the most resistant part of sexual specimens, the
ovaries being much more resistant than the testes. Unfortunately
almost no data were collected on male specimens as compared with
asexual individuals.

The evidence presented in this part of the paper clearly indicates
that the metabolic state of Hydra is subject to wide variation and
that such variation is definitely correlated with such conditions as
age, depression, sexual maturity, and inanition. Metabolic rate
may be a causal factor in the induction of at least some of these
conditions.

SEXUAL MATURITY.

Sexual maturity in hydras seems to depend in part on internal
conditions and in part on external factors, such as temperature.
The formation of gonads has been induced by change of tem-
perature by several investigators : R. Hertwig ('06), Krapfenbauer
('08), Frischholz ('09), Koch ('n), Whitney ('07), Uspenskaja
('21) and Grosz ('25). But since a certain percentage of the
animals remain sterile under the altered condition it must be as-
sumed that some other factor besides temperature is involved in

8O LIBBIE H. HYMAN.

bringing about sexual maturity. Nussbaum's claim that starva-
tion induces males and rich feeding females ('92, '08) has not
been verified and is probably erroneous. All other observers agree
that male and female gonads are induced by the same external
conditions, not by opposite ones.

I have confirmed the result of several investigators that a lower-
ing of temperature causes the development of gonads on P. oligac-
tis. On January 17, 1927, a subculture from a main culture of
P. oligactis that had been maintained in the laboratory for several
months without any signs of sexual development was placed in
the refrigerator (11-13° C.). On February 5, a number of
specimens were observed to be developing testes. Still further
depression of the temperature (to 5° C.) did not increase the
number of male specimens. The culture contained about 400
hydras, of which about 30 developed testes. On March 7 another
subculture was placed in the refrigerator and on March 22, signs
of testis formation were noted on several specimens ; by March
28, of 80 individuals in this culture, 30 possessed testes. Through-
out the period occupied by these experiments and indefinitely
thereafter the main culture and other subcultures kept continu-
ously at room temperature did not show any trace of gonads. The
experimental cultures were fed at frequent intervals while in the
refrigerator. On return to room temperature, the induced gonads
disappeared rapidly.

On the other hand, I can confirm the statements of Grosz ('25)
that hydras may become sexually mature while under laboratory
conditions without any apparent change of temperature. This has
occurred in my cultures on several different occasions. Thus in
April, 1919, a culture of P. oligactis which had been in the labora-
tory for some time developed gonads without any apparent cause ;
all were females. In May, 1921, cultures were started from a
few specimens of P. oligactis ; males soon appeared in these cul-
tures although the species is not sexually mature in nature until
fall. During July, there were many males in all of the subcul-
tures of this line, many cultures containing 50 per cent, or more
male specimens. Males continued to be present abundantly in
these cultures for many months. During my absence from Chi-
cago from April to July, 1922, most of the animals died, but on

OBSERVATIONS ON HYDRA. 8 1

my return some were found in two jars. These were used as
the basis of new cultures, which flourished and in which males
again appeared in September, 1922, continuing throughout Oc-
tober, when the culture was abandoned. Another culture of P.
oligactis, consisting of females, had a similar history. This cul-
ture started from a single specimen, collected in July, 1921. On
July II, one female was noted in the culture. Females continued
to appear in it in small numbers throughout July, August, and
September, 1921, and again in March, 1922. After my return in
July, 1922, the cultures were restarted and females again appeared
in them during September and October.

Contrariwise, cultures may be maintained for long periods of
time in the laboratory and continue completely asexual. Thus for
the last four years, 1923 to 1927, although cultures of hydra have
been on hand almost continuously, not a single sexual specimen
has appeared in them. Since many of these cultures were started
from a few specimens and were continued by subculturing for
many months, it is certain that the animals were of sufficient age.
As already noted, the formation of gonads can be induced in a
certain percentage of specimens in such cultures by subjection to
low temperature (P. oligactis}.

It is apparently widely believed by zoologists and is current in
elementary textbooks that hydra is hermaphroditic. It is now
known, however, that certain species, as P. oligactis, are strictly
dioecious. In the dioecious species, asexually produced offspring
are of the same sex as their parents, as originally determined by
Frischholz ('09) and since verified by others. Nussbaum ('09)
and Goetsch ('220) claim to have observed change of sex in hydra
within a clone but their results are generally considered doubtful.

Most of the observations on sexual reproduction in hydra in the
literature were made on mass cultures. It seemed to me that it
would be of interest to determine the history of individual sexual
specimens. With this in mind several males and females were
isolated, fed, and observed daily. The remainder of the paper
will be devoted to their history.

82 LIBBIE H. HYMAN.

HISTORY OF ISOLATED FEMALES.

In October. 1925, several sexual specimens were brought in
with a collection of hydras made in the lagoon in Jackson Park,
Chicago. On November 18, four females were isolated from
this culture into a finger bowl. They underwent depression and
only one recovered. On November 25, a second female was added
from the general culture and on December 22, a third. The fe-
male culture therefore came from these three females, all of which
belonged to the species P. oligactis.

The three females were fed daily by hand on crushed Daplnua
or Cyclops. Sexual specimens are frequently unable to capture
prey, according to Goetsch ('22a), because they lack sufficient
nematocysts. I cannot agree with this opinion since sexual ani-
mals when in a healthy condition can capture and swallow small
daphnids without difficulty. Depression is probably the cause of
the usual failure of sexual animals to feed.

The females soon began to bud in normal fashion and at the
normal location. By December 22, the culture contained 8 speci-
mens, by December 26, 14. and by the 2Qth, 20 animals. The origi-
nal three females while continuing to bud had retained their gonads
during this budding period. The buds given off were at first
asexual but on January 4, 1926, all of the specimens in the culture
were seen to be developing ovaries. At this time they were able
to catch daphnids by themselves. By January 8 all members of
the culture bore ovaries ?nd had ceased to bud. Many eggs, un-
fertilized of course, were shed and could be seen lying on the bot-
tom. The animals were no longer successful in catching prey
and about January 16, hand feeding was resumed. Budding began
and the young, soon after becoming free, developed ovaries. A
slight depression set in about the middle of February, but recov-
ery occurred almost at once on transferring the animals to a bowl
in which hydras had been flourishing. The ovaries began to dis-
appear and by February 27, were practically gone; budding was
abundant. By March i, rapid budding was in progress and the
gonads had vanished. The animals continued to bud until the ist
of April, when for unknown causes, all of the sexual cultures were
found in a state of advanced depression. The usual procedures

OBSERVATIONS ON HYDRA. 83

were followed and by April 6, some recovery was attained ; but on
April 15 depression again set in and all of the females died.

Throughout their entire history the females were kept at labora-
tory temperature. The general cultures started from asexual
specimens obtained from nature in the same collection as the
original three females had not produced a single sexual specimen
in the period in question.

These observations show : ( I ) that female hydras when fed bud
in the usual manner; (2) that females may bud while still retain-
ing the gonads; (3) that females may lose their gonads and be-
come completely asexual; (4) that specimens which have been
recently female and also the asexual offspring of females tend to
develop ovaries in a relatively short time; (5) that the asexually
produced offspring of female P. olic/actis are of the same sex as
their parents.

HISTORY OF ISOLATED MALES.

The history of the males was very much more interesting than
that of the females. The behavior of budding males completely
surprised me but I found on search through the literature that
certain aspects of it had been seen previously by several observers,
although none of them has furnished a detailed account of the
phenomenon. Trembley and Laurent noted that males bud ir-
regularly; Hertwig ('06) denied the correctness of their obser-
vations but stated that budding males possess two budding zones
instead of the usual one. My observations do not agree with
Hertwig's statement. Krapfenbauer ('08) described and figured
the adventitious budding of male specimens (one of his figures is
reproduced by Schulze, '17) ; and Grosz ('25) again mentions the
phenomenon. None of these observers, however, appears to have
followed out the eventual fate of these curious specimens. All
observers except Laurent agree that the buds arise from the column
of the animal, not from the testes, and with this statement my own
observations accord.

The male specimens were isolated November 18, 1925, from
the same culture as were the females. This culture collected in
October contained a limited number of sexual specimens which
so far as could be ascertained bore gonads when collected. No
additional individuals became sexual in the laboratory. Seven

84 LIBBIE H. HYMAN.

males were isolated which underwent depression; four recovered.
These were fed daily on crushed Cyclops but at times they and
their offspring were able to catch and ingest small daphnids with-
out assistance. The males at all times were noticeably slow and
sluggish in behavior and prone to undergo slight depression, in
which condition they would not accept food. During the entire
period over which they were cultivated (about eighteen months),
the original individuals and their offspring continued to bear testes,
except for one brief interval. The buds of male specimens in
most cases bear gonads before they separate from the parents and
develop additional testes soon after becoming free. In many
cases, testes seemed to migrate from the parent column onto the
bud as has been reported by others ; but most of the testes present
on buds undoubtedly develop in situ.

The species of the four males cannot be stated definitely since
when the experiment was begun it was supposed that all were
P. oligactis. Later examination of offspring showed that some
were P. oligactis and some were Hydra stellata, but I do not know
which of the four belonged to which species. As each behaved
somewhat differently, the history of each will be presented separ-
ately.

History of Male I. — On December 12, this male was found with
two buds, irregularly located as shown in Fig. 25. On December
22, three buds were present, and by December 29, it had given off
two of these and developed two more in the same locations (Fig.
26). The middle region of this male had a stalk-like appearance,
indicated at a in Fig. 26 and there was evidence that a basal disk-
was differentiating at this region. On January 2, 1926, the speci-
men appeared as in Fig. 27. It was attached by a basal disk at the
stalk-like region already mentioned and the old basal disk, at /,
Fig. 27, was being absorbed. The bud marked b in Fig. 26 had
become persistent and had gained control over the basal part of
the animal. On January 4, the specimen was very much larger,
had a number of buds, Fig. 28, and was attached by two basal
disks, the new and the old one, / and f, Fig. 28. The old bud b
was still persistent. Both the original animal and b had normally
located budding zones. On January 6, the animal divided in two
by a transverse constriction, x, Fig. 28. This constriction oc-

OBSERVATIONS ON HYDRA. 85

curred just below the budding zone of the larger component; this
is commonly the place where such constrictions develop. The male
7 was thus divided into two parts which will be designated here-
after as I A and IB, the former being the old animal. The fate
of I A will be followed first. It was left without a foot or stalk
by the constriction. On January 6, it appeared as in Fig. 29, hav-
ing three buds at its base, where buds are usually formed in such
stalkness specimens. Later it gave off two buds and the remain-
ing bud so oriented itself as to give the appearance of a biaxial
animal (Fig. 30). On January n, another bud appeared at the
base of I A; the two buds arranged themselves at right angles to
the parent column as in Fig. 31. The two buds now separated
together by a constriction at the level x, Fig. 31, and were desig-
nated as ID, the old part retaining the name I A. I A subsequently
developed a foot and stalk, became approximately normal, and was
not followed further. ID remained for some time as an ap-
parently biaxial Hydra, heads oriented in opposite directions as in
Fig. 31. Later, January 18, the heads swung around so as to be
approximately parallel, Fig. 32. The gastrovascular cavities of
the two components were perfectly continuous ; the specimen pos-
sessed no foot, lying free on the bottom. Later a foot developed
at the bend and both specimens budded repeatedly. On February
4, Fig. 33, a double-headed bud, which subsequently became free,
was noted on one of the components. On February 9, ID, now
very large, had the appearance shown in Fig. 34. Each of the
components had a definite budding zone at the normal level and
a bud of each had developed a foot in situ as at /. Such produc-
tion of basal disks by persistent buds was very common on these
male specimens ; such feet may or may not be fastened to the sub-
stratum. On February 10, the specimen divided by transverse
constriction at the level x, Fig. 34. It will be noted that the plane
of constriction again passed just below the budding zone of the
larger component. The two parts resulting from this fission were
designated IDi and ID2 and are shown in Fig. 35 and 36, respec-
tively. ID i was left without any stalk or basal disk, while ID2
bore the portion which formerly connected the two specimens.
ID i lost the buds it had at the time of separation, Fig. 35, but de-
veloped a new bud, Fig. 37, on February 16, at its base as is usual

86 LIBBIE H. HYMAN.

in stalkless specimens. It will be noted that this bud again was
double-headed. This bud was given off and other buds were
formed at the same place and became free. On February 23, the
specimen still had no basal disk, appearing as in Fig. 48. The
animal then began to increase greatly in length. The bud shown
in Fig. 48 became persistent, Fig. 49, March i, and the next bud
appeared proximal to it. The specimen continued to increase
greatly in length and on March 3, was found with a number of
irregularly disposed buds, Fig. 50. Attention has already been
called to this specimen as an example of the more distal location
of buds with increasing length of column.

Returning now to ID2, Fig. 36, we find it presenting a compli-
cated appearance on February 16, Fig. 38. On February 18, it
divided by a transverse constriction at the level x, Fig. 38, into
two components, a larger one, still called ID2 and a smaller one,
7Z?3, shown in Fig. 39 as it appeared on February 23. It con-
sisted of two buds which were continuous, similar to the form
shown in Figs. 32 and 33. On February 27, it appeared as in
Fig. 40, each component having budded. It then divided by a
transverse constriction, at x, Fig. 40, the fission again occurring
just below the budding zone of the larger component. The larger
component, Fig. 42, remained in this state for some time, but
eventually again split up by transverse constrictions between the
bases of the components. The smaller part resulting from the
fission shown in Fig. 40, later appeared as in Fig. 41, but regulated
to normal by absorption of the long stalk region. The main part
of ID2 separated by the constriction shown in Fig. 38, appeared
later, February 23, as in Fig. 43, having produced additional ad-
ventitious buds and having almost absorbed the basal process which
it retained from the division shown in Fig. 34. On February 24,
this specimen split up by two simultaneous transverse fissions,
x. x, Fig. 43, into three parts, called ID 4, IDs,, and ID2, from
left to right in Fig. 43. ID^, shown in Fig. 44, consisted of two
buds connected at their bases by a narrow stalk, which subsequently
parted. The history of ID$ is shown in Figs. 45, 46, and 47.
In Fig. 47, it had become nearly normal, but had given rise to a
double-headed bud.

The appearance of double-headed buds several different times

OBSERVATIONS ON HYDRA. 87

in the offspring of male ID was very interesting. The further
history of such buds was followed and is much the same for all
of them. Fig. 51 shows the doubled bud of Fig. 33 as it appeared
on February 16, over two weeks after its separation from the
parent. Each component had budded. The same specimen is
shown in Fig. 52, on February 23. It had additional buds, two
of which had developed feet in situ. This specimen underwent
the usual fate ; it divided by transverse fission at the level x, Fig.

52, into two components, the larger of which is illustrated in Fig.

53, as it appeared on February 25. On February 27, Fig. 54,
it was larger and with additional buds, one of which was very
near the basal disk. On March i, another transverse constriction
occurred, at x in Fig. 54, in such a way as to cut the specimen
into three parts. The subsequent history of the basal region, the
part below .r in Fig. 54, is given in Figs. 55, 56 and 57. It had
given off the bud which it possessed, Fig. 54, and was left with
two feet, /, Fig. 55. It regenerated a hydranth at the free end
and another proximal to a constriction, Fig. 56. The further de-
velopment of the two heads is shown in Fig. 57. The specimen
then divided by a constriction at x, Fig. 57, into a normal small
Hydra and a larger one bearing a process on the side of the
column ; this process was subsequently absorbed and the specimen
became normal. Figs. 65 to 67 show the history of the double-
headed bud of Fig. 37. In Fig. 65, one component had budded ;
in Fig. 66, both components have buds ; and in Fig. 67, additional
buds have appeared, some with feet in situ. It will be noted that
one of these buds is again double-headed.

Returning to male IB separated from male / by the constriction
shown in Fig. 28, we find it on January 6, 1926, presenting the
appearance given in Fig. 58. On January 8, it appeared as in
Fig. 59 and on this day divided by a constriction, at x, Fig. 59,
into two parts, IE and IF, shown in Figs. 60 and 61, respectively.
IE, which retained the basal process, appeared on January 16 as
in Fig. 62; on January 18 it constricted off the basal process, at
x, Fig. 62. The further history of this basal process is shown
in Figs. 63 and 64. After this occurrence IE regulated to a nor-
mal appearance. The further history of IF was not followed.

History of Male II. This male continued to elongate for several

88 LIBBIE H. HYMAN.

weeks without other change. Finally on January 4, 1926, it was
found with two buds, irregularly placed, Fig. 68. On January
6, it had four buds, Fig. 69, but by January n, it had lost three
of these. Buds continued to appear irregularly on this animal
and to be given off, but on January 19, Fig. 70, the largest bud
present at that time became persistent and developed a foot in situ,
/ , Fig. 70. On January 27, another bud had produced a foot in
situ. Fig. 71, and these two buds persisted for some time. Fig.
72 shows male // on February 4 with these two persistent buds,
each with a foot. Finally, the more distal of the two buds was
given off, but additional buds appeared at the same location, Feb-
ruary 9, Fig. 73. February 16, Fig. 74, only the original persistent
bud remained. The animal meantime was elongating greatly. On
February 19, it formed an additional foot, /', Fig. 75, on its
column. It then began to bud again and on March 3, appeared
as in Fig. 76, still retaining the persistent bud, which had itself
budded, and having in addition four buds irregularly arranged.
It was now much elongated. On March 8, it constricted in two
at the level x, Fig. 76, just below the accessory foot. The two
divisions, II A and II B, are shown in Figs. 77 and 81 respectively.
II A produced many buds, Fig. 78, March 25, and became large and
complicated, Fig. 79, March 29, at which time it had several feet,
/, Fig. 79. The other protuberances in Fig. 79 are young buds.
Two of the feet were close together and showed a tendency to
adhere to each other. On March 31, Fig. So, the specimen was
still more complex with several buds and feet. It was lost in the
general depression which occurred April i, 1926. The history of
male II B is given in Figs. 81 to 84. By the division shown in
Fig. 76, this specimen was left with a portion of the original
column, a, Fig. 8r. This developed a foot at its anterior end, the
specimen then appearing with two feet as in Figs. 82 and 83. By
March 29, Fig. 84, the specimen had nearly absorbed one foot
and was approximately normal.

History of Male III. — Male /// behaved somewhat differently
from the two preceding specimens. It continued to grow and
elongate for nearly three months without giving rise to any buds.
On January 6, it formed an accessory foot near the middle of the
column, /, Fig. 85 ; and on January 8, it divided in two just proxi-

OBSERVATIONS ON HYDRA. 89

mal to this foot, at x, Fig. 85. The two divisions, 111 A and II fH
are shown in Figs. 86 and 88, respectively. Ill A absorbed the
basal process and soon became approximately normal, Fig. 87,
January 18. IIIB behaved in an unexpected manner, but one
which was found to be typical for pieces of this character. It
was naturally expected that it would regenerate a head at its apical
end. But this apparently never happens in pieces of this kind.
Instead it formed a basal disk at the apical end and thus appeared
as an elongated piece with a foot at each end, Fig. 88. By Janu-
ary 13, it had formed an additional foot near its middle, thus hav-
ing three feet, /, /, /, Fig. 89, all of which were fastened to the
substratum. On this date, also, a small bud had appeared, b, Fig.
89. This bud was fed with some difficulty on very tiny specimens
of Cyclops. It began to grow rapidly and on January 21, ap-
peared as in Fig. 91, using its three feet in tripod fashion. Feb-
ruary 9, it was similar in appearance, Fig. 92, and on February
24, Fig. 93, the feet appeared to be in process of absorption. By
March i, Fig. 94, it was very much larger and still possessed three
feet by which it was fastened. The specimen now burst into buds,
so to speak, being found on March 3 with seven buds, Fig. 95,
very irregularly arranged, most of them borne on the three basal
regions. On March 8, the specimen divided in two at the level x,
Fig. 95. The distal part later absorbed its basal processes and
became normal. The proximal part could not be recognized
later.

History of Male IV. — This animal behaved in a fashion similar
to male ///. It increased greatly in length for many weeks with-
out budding. Finally on January 18, it developed an acessory foot
on the column and on January 28, another foot distal to this one.
Fig. 96 shows it as it appeared on January 28 with three feet.
On February 4, it divided in two below the middle foot, x, Fig.
96, the two parts IV A and IVB being shown in Figs. 97 and 98,
respectively. The male I]' A, left with two feet, finally produced
a bud, February 27, opposite the distal foot, Fig. 99, but this was
given off and no further buds appeared. On March 12, a third
foot was formed distal to the other two, Fig. 100, and on March
17, the animal divided in two below the distal foot, at x, Fig. 100,
into two parts, IV Ai and IV A2, Figs. 101 and 102, respectively.

9O LIBBIE H. HYMAN.

The further history of IV Ai is given in Figs. 103 to 107. It
budded and also showed a great tendency to produce additional
feet. Thus in Figs. 105 and 106 it has four feet in close prox-
imity and tending to adhere to each other. It finally absorbed the
long basal region and on March 31, Fig. 107, was nearly normal.

We may now follow the history of IVB, Fig. 98, and IV A2,
Fig. 102. Both of these were basal pieces cut off by division and
both behaved alike and in the same manner as did the similar
piece IIIB, Fig. 88. Each regenerated a foot at the distal end.
Fig. in, IVB, and Fig. 108, IV AT. As IV AT already had two
feet at the time of division, Fig. 100, it was now provided with
three feet, Fig. 108, one at each end and one near the middle. A
bud grew out at b, Fig. 109, March 29, and the specimen appeared
March 31 as a small Hydra provided with two feet, Fig. no.
IVB formed a foot at the distal end, Fig. 112 and then gave rise
to two additional feet, Fig. 113. A bud grew out from the central
portion. Fig. 114, February 23. This was provided with three
feet, the fourth having apparently been absorbed. The specimen
was so small that I did not succeed in feeding it.

On April I, as already noted, both the male and female cultures
were found in a state of advanced depression and could not be
restored. Fortunately, however, the fate of the male specimens
had been followed sufficiently to give one an idea of their general
behavior. There remained the offspring of the males which had
been removed as fast as they became free and placed in a separate
dish. This culture, consisting of offspring of the males, escaped
the depression noted and was continued for nearly a year. At
first all of the specimens bore testes and continued to bear them
until March, 1926, when the testes disappeared. Towards the
end of April testes again developed on all of the specimens of this
culture and continued to be present without intermission until
March, 1927, when the culture was discontinued. Some of the
offspring of this male culture had, however, inadvertently been
transferred into a culture of P. oligactis. In July, 1926, these
were wholly asexual but in August testes appeared on all of them
and continued to be present as long as the culture was observed
(until March, 1927).

In the chief subculture of the male line, formed from the buds

OBSERVATIONS ON HYDRA. 9!

of the original four males, the same curious behavior that charac-
terized the parent specimens was repeatedly observed. Irregular
budding, formation of complicated " colonies," persistence of buds,
formation of accessory feet, and occurrence of transverse constric-
tions were prevalent in this subculture for many weeks. Finally,
however, nearly all of the specimens became normal. This seemed
to be due primarily to a reduction in size. The animals after
April, 1926, remained of small size and budded normally, but one
or two specimens exhibiting the behavior described above, could
always be found in the culture.

It seems superfluous to figure and describe any more cases of the
irregular behavior of male specimens, as they were similar to those
already noted. An unusually curious specimen of unknown origin
may be worth mentioning, Fig. 115. It probably represents a
" colony " after recovery from depression, as shown by its doubled
tentacles. Figs. 116 and 117 illustrate further changes in this
specimen, which unfortunately was lost in the general depression
of April i, 1926. Figs. 118 and 119 furnish another illustration
of a distance relation between peristome and buds. This speci-
men, separated by transverse constriction from ID4, Fig. 44,
formed its first bud at the extreme base, as is common in footless
animals, its second bud considerably distal to the first. Six days
later, the specimen, considerably elongated, had lost the more
distal bud of Fig. 118, but had formed a new bud, markedly distal
to the two preceding ones, Fig. 119. Numerous cases of this kind
were observed.

From a consideration of all of these observations on male speci-
mens certain general statements appear justified. Male hydras
when richly fed tend to elongate greatly and sooner or later bud or
give rise to accessory feet. The buds on such elongated males are
always irregularly arranged and may appear at any point over a
considerable length of column. They are always, however, at
some distance from the main peristome. Many of the buds of
such males are set free in the usual manner but some, particularly
those farthest away from the main peristome, tend to become
persistent. Buds, particularly persistent ones, frequently produce
feet in situ on the side opposite to their point of attachment to the
parent. Continued budding of the main individual and of its per-

92 LIBBIE H. HYMAN.

sistent bud or buds results in a complex colony-like form. These
colonies always break up sooner or later by one or more trans-
verse constrictions which commonly pass just proximal to the
budding zone of the main component. The main component is
thus left without any foot or stalk and may remain in this state for
some time, budding repeatedly and again dividing by transverse
fission. It may also grow a foot and stalk and become normal.
When transverse fission occurs the smaller component of the
colony is usually left with a piece of column ; this may be cut off,
whereupon it regenerates to a small hydra, or may be absorbed.
In cases where elongated males do not bud but develop accessory
feet, a transverse division eventually occurs, always just proximal
to one of the feet. The distal portion commonly again develops
accessory feet and divides proximal to a foot or may regulate to
normal by absorption of the feet. The basal region cut off below
a foot does not regenerate a hydranth at its anterior end in any
one of several cases observed but always forms a foot at the distal
end, and may develop accessory feet along its length. Such pieces
then bud somewhere between the two ends ; the bud if fed grows
up at right angles to the long axis of the piece, using all of the feet
as its base, in tripod style. Such specimens may regulate to nor-
mal by absorption of feet or may bud and divide up by transverse
constrictions. Stalkless specimens resulting from transverse fis-
sion always bud first at their extreme basal ends ; but as they in-
crease in length consequent upon feeding, the buds appear at more
and more distal levels.

All of the observations plainly indicate that each peristome con-
trols a certain length of column and that budding does not occur
within the limit of control. They also indicate that the amount
of control is less in these male specimens than in normal animals.
Each persistent bud appears to control a certain length of the ad-
jacent parent column, whose polarity is thereby reversed, as shown
by the fact that when transverse fission occurs, this portion of the
column may become the basal region of the bud in question. Each
foot also appears to prevent the occurrence of a bud within a cer-
tain, rather short, distance, for buds frequently form at the ex-
treme basal end of footless specimens while they never do so when
a foot is present.

OBSERVATIONS ON HYDRA. 93

The remarkable behavior of male specimens is probably depend-
ent on two factors — their low metabolic state and their tendency
to elongate when fed. Evidence was presented earlier in the
paper that sexual specimens have a lower metabolic rate than com-
parable non-sexual animals. This is further evidenced by their
general sluggishness and slow reaction time to stimuli. In a low
metabolic condition the control exercised by each peristome is
diminished. Moreover, as these animals are generally much elon-
gated, a considerable portion of the column will not be under con-
trol of any peristome and thus may bud at any point. Accessory
feet also tend to appear on portions of the column not controlled
by a peristome. This matter of length seems to be the more im-
portant of -the two factors, since male specimens which remain
of small size usually bud normally. Further females bud normally
(although their metabolic rate is lower than that of males) ap-
parently because they do not elongate when fed, so far as ob-
served.

Finally it may be pointed out that the asexually produced offspring
of male P. oligactis and H. stellata are also male and tend to form
testes either before they separate from the parents or shortly after.
Males specimens may become entirely asexual for short periods
but display a marked tendency to develop gonads again in a short
time. During all of the time that the descendants of the males
remained sexual, bearing numerous testes, not a single gonad ap-
peared in the cultures started from asexual individuals obtained
in the same collection with the sexual animals. It appears that
when hydras have once developed sex organs they tend to remain
in the sexual state for long periods of time.

SUMMARY.

1. Directions for the cultivation of hydra are given.

2. The phenomenon of depression, in which there is a shorten-
ing and gradual loss of tentacles and column from the distal end
proximally, is of common occurrence in hydras and probably rep-
resents a lowered metabolic state. It is induced by rich feeding,
high temperature, sensescence, fouling of the culture water, lack
of oxygen, transfer to clean fresh water. Recovery may be spon-
taneous or may by induced.

7

94

LIBBIE H. HYMAN.

3. Irregular adventitious budding tends to occur in senescent,
extremely elongated, depressed, and sexual specimens, that is,
specimens in a low metabolic state. Under such conditions buds
may be retained on the parent far beyond the normal time, giving
rise to so-called colonies.

4. Abnormalities, such as forked tentacles and forked distal
and basal ends arise chiefly in recovery after depression. Forked
distal structures may also arise by retention of a bud.

5. Longitudinal and transverse fission are probably not normal
modes of reproduction of hydra but are methods of regulation of
previously existing abnormalities or occur in specimens in a low
metabolic state.

6. Tests by the susceptibility method have shown that :

(a) Starvation beyond four days increases the metabolic rate.

(b) Depressed animals have a lower metabolic rate than con-
trols starved the same length of time.

(c) Animals recovered from depression have an equal or higher
metabolic rate than controls starved the same length of time.

(d) The metabolic rate of sexually mature animals is lower
than that of asexual controls and that of females is lower than that
of males.

7. The formation of sex organs can be induced in Peltnatohydra
oligactis in about three weeks by lowering the temperature about
ten degrees.

8. In dioecious species of hydra, the sex of asexually produced
offspring is the same as that of the parent.

9. Hydras may become sexually mature in the laboratory with-
out any apparent cause. Such sexual animals and their asexual
offspring tend to remain and to become sexual.

10. Sexual specimens when fed bud while still retaining the
gonads.

n. Sexual specimens when fed may lose their gonads for short
periods but tend to develop them again as do also their asexually
produced offspring.

12. Females when fed remain of normal size and bud normally.

13. Males when fed tend to elongate greatly and thereupon
produce buds irregularly and adventitiously over a considerable
length of column.

OBSERVATIONS ON HYDRA. 95

14. The buds of such male specimens tend to be retained in-
definitely forming complex colony-like specimens.

15. Such retained buds commonly develop a foot in situ with
which they ma}- be fastened to the substratum.

1 6. Colonies formed in this manner split up eventually by one
or more transverse constrictions which pass just proximal to the
budding zone of the main component if any or between the bases
of the components when they are equal.

17. Male specimens when fed may not bud but instead may give
rise to accessory feet at one or more points along the column ; such
tend to divide by transverse fission just proximal to one of the feet.
The proximal piece formed by such a fission regenerates a foot, not
a hydranth, at its distal end ; it then buds near its middle and the
bud if fed grows up at right angles to the axis of the piece using
the two or more feet as base.

19. Reversals of polarity due to the control of parts of the parent
column by buds were frequently observed.

20. All of the facts indicate that each peristome controls a cer-
tain length of column and prevents the formation of a bud within
the limits of that control. The control is diminished in senescent,
depressed, and sexual specimens. Each foot also controls a short
length of column and prevents the formation of a bud within that

distance.

BIBLIOGRAPHY.

All papers in this list have been read except those by Krapfenbauer and
Uspenskaja, which were unavailable, and whose contents were obtained
from other papers.
Boecker, E.

'14 Depression und Missbildung bei Hydra. Zool. Anz., Vol. 44, pp.

75-80.
Child, C. M.

'13 Studies on the Dynamics of Morphogenesis and Inheritance in Ex-
perimental Reproduction. V. Jour. Exp. Zool., Vol. 14, pp. 153-
206.

'15 Senescence and Rejuvenescence. University of Chicago Press.
Child, C. M., and Hyman, L. H.
'19 The Axial Gradients in Hydrozoa. I. Hydra. BIOL. BULL., Vol.

36, pp. 183-223.
Drzewina, A., and Bohn, G.

'16 Phenomenes de reduction et d'activation chez les Hydres, a la suite
de variations de la tencur de I't-au en oxy.nrne. C. R. Soc. Biol.,
Vol. 78, pp. 429-31.

96

LIBBIE H. HYMAN.

Frischholz, E.

'09 Zur Biologic von Hydra. Biol. Zentralbl., Vol. 29, pp. 182-92, 206-

215. 239-55, 267-90.
Goetsch, W.
'19 Neue Beobachtungen und Versuche an Hydra. I. Biol. Zentralbl.,

Vol. 39, pp. 289-302. II, pp. 544-57-
'21 Beitrage zum Unsterblichkeitsproblem der Metazoen. I. Teil.

Unsterbliche Hydren? Biol. Zentralbl., Vol. 41, pp. 374-81.
'223 Same II. Lebensdauer und geschlechtliche Fortpflanzung bei

Hydren. Biol. Centralbl., Vol. 42, pp. 231-40.

'22b Same III. Depressionen und Lebensdauer bei Hydren. Biol. Zen-
tralbl., Vol. 42, pp. 278-86.
Grosz, J.

'25 Versuche und Beobachtungen iiber die Biologic der Hydriden. Biol.

Centralbl., Vol. 45, pp. 321-52, 385-41?-
Hase, A.

'09 tjber die deutschen Siisswasser-Polypen Hydra fusca L., Hydra
grisca L., und Hydra riridis L. Arch. f. Rassen und Gesellsch.
Biol., Vol. 6, pp. 721-53-
Hertwig, R.

'06 Uber Knospung und Geschlechtsentwickelung von Hydra fusca.
Biol. Centralbl., Vol. 26, pp. 489-508.

Kepner, W. A., and Jester, P. N.

'27 The Reaction of Hydra to Inanition. BIOL. BULL., Vol. 52, pp.

173-84.
Koch, W.
'na tiber die Geschlechtsbildung und den Gonochorismus von Hydra

fusca. Biol. Centralbl., Vol. 31, pp. 138-44.
'lib Uber die geschlechtliche Differenzierung und den Gonochorismus

von Hydra fusca. Biol. Centralbl., Vol. 31, pp. 545-75.
'12 Missbildungen bei Hydren. Zool. Anz., Vol. 39, pp. 8-13.
Koelitz, W.

'o8a Fortpflanzung durch Querteilung bei Hydra. Zool. Anz., Vol. 33,

pp. 529-36-
'o8b Zur Kenntniss der Fortpflanzung durch Querteilung bei Hydra.

Zool. Anz., Vol. 33, p. 783.
'09 tiber Langsteilung und Doppelbildungen bei Hydra. Zool. Anz., Vol.

35, PP. 36-46.
Krapfenbauer, A.
'08 Einwirkung der Existenzbedingungen auf die Fortpflanzung von

Hydra. Thesis. Munich.
Laurent, L.

'44 Nouvelles Recherches sur 1'Hydre. Voyage autour le Monde sur

la corvette La Bonite, commandee par M. Vaillant. Zoophytologie

par Laurent. Paris, 1844, A. Bertrand.
Leiber, A.

'09 liber einen Fall spontaner Langsteilung bei Hydra viridis. Zool.

Anz., Vol. 34, pp. 279-84.

OBSERVATIONS ON HYDRA. 97

•
Nussbaum, M.

'92 Die Geschlechtsentwickelung bei Polypen. Sitz. d. Nieclerrhein.
Gesellsch. f. Natur- und Heilkunde zu Bonn. Mcdizin. Scktion.,
Jan. i8th, pp. 13-14.
'09 Uber Geschlechtsbildung bei Polypen. Arch. f. d. gesamm. Physiol.,

Vol. 130, pp. 521-629.
Parke, H. H.

'oo Variation and Regulation of Abnormalities in Hydra. Arch. f.

Entw'mech., Vol. 10, pp. 692-710.
Rehm, W.
'25 tiber Depression und Reduktion bei Hydra. Zeitsch. f. Morphol.

u. Okol. d. Th-ere. Vol. 3, pp. 358-88.
Schulze, P.

'17 Neue Beitrage zu einer Monographic der Gattung Hydra. Arch.

f. Biontologie, Vol. 4, Heft 2, pp. 33-1 19.
Trembley, A.

1744 Memoires pour servir a 1'historie d'un genre de polypes d'eau douce,

a bras en forme de cornes. Leiden, J. and H. Verbeek.
Uspenskaja, W. J.

'21 Der Einfluss von Temperatur und Hunger auf die Bildung von
Geschlechtsprodukter bei Hydra grisea. Nachr. Institut. Exper.
Biol. Moskau, Vol. i.
Wachs, H.

'19 Uber Liingsteilung bei Hydra. Biol. Centralbl., Vol. 39, pp. 1-12.
Welch, P; S., and Loomis, H. A.
'24 A Limnological Study of Hydra oligactis in Douglas Lake, Michigan.

Trans. Amer. Micro. Soc., Vol. 43, pp. 203-35.
Whitney, D. D.

'07 The Influence of External Factors in Causing the Development of
Sexual Organs in Hydra riridis. Arch. f. Entw'mech., Vol. 24,
PP- 5^4-37-

98

LIBBIE H. HYMAN.

PLATE I.

FIG. I. Abnormally elongated specimen of Pelmatohydra oligactis, show-
ing irregular budding; tentacles of the main hydranth omitted.

FIGS. 2 TO II. Abnormal specimens found in nature, showing duplicature
of the apical end and various stages of apparent longitudinal splitting.

FIG. 12. Abnormal specimen found in nature with split basaj end.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE I.

LIBBIE H. HYMAN.

IOO LIBBIE H. HYMAN.

PLATE II.

FIGS. 13 TO 18. Abnormal specimens found in nature showing doubled
basal ends, which appear to be in process of splitting in the apical direction.

FIG. 19. Specimen found in nature with a retained bud, having also a
foot of its own.

FIG. 20. Specimen found in nature, probably resulting from the partial
fusion of a bud with the parent.

FIG. 21. Laboratory specimen of P. oligactis in which due to a slight
depression, the bud b, originally at the normal budding zone, is moving
apically by fusion with the parent.

FIG. 22. The same specimen as Fig. 21 at a later time, the bud having
become equal in size to the parent, simulating a double-headed Hydra.

FIGS. 23 AND 24. Specimens at the start of recovery from depression,
showing split tentacles.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE II.

LIBBIE H. HYMAN.

IO2 LIBBIE H. HYMAN.

PLATE III.

FIG. 25. Male /, on December 12, 1925.

FIG. 26. Male /, on December 29, showing three buds; b, bud which
later became persistent ; a, region where a foot was developing.

FIG. 27. Male /, January 2, 1926, showing persistence of the bud b.
f, the old foot, /', new foot formed at the point marked a in Fig. 26.

FIG. 28. Male /, January 4, b, the persistent bud, /, the old foot, /', the
new foot, .r, the level at which transverse fission occurred, on January 6.

FIG. 29. Male I A, distal product of the fission in Fig. 28, on January 6.

FIG. 30. Male IA, January 8, with one bud, oriented opposite to the
parent.

FIG. 31. Male I A, January 11, with two buds at its base, x, the plane
of transverse fission, January 11.

FIG. 32. Male ID, January 18, formed from the two buds cut off at x
in Fig. 31.

FIG. 33. Male ID, February 4, one component with a double bud.

FIG. 34. Male ID, February 9, with four buds, two of which have a
foot in situ at /. x, plane of transverse fission on February 10.

FIG. 35. Male IDi, right hand product of the fission shown in Fig. 34.

FIG. 36. Male ID2, left hand product of the fission shown in Fig. 34.

FIG. 37. Male IDi, February 16, with a doubled bud at its base.

FIG. 38. Male ID2, February 16, with three buds, each with a foot in
situ, x, plane of transverse fission, February 18.

FIG. 39. Male /£>3, right hand component of Fig. 38, on February 23.

FIG. 40. Male ID^, February 27, with several buds, x, plane of trans-
verse fission.

FIG. 41. Right hand part of Fig. 40 after the fission.

FIG. 42. Left hand part of Fig. 40 after the fission.

FIG. 43. Main part of male ID2, formed by the fission shown in Fig.
38, as it appeared February 23. x, x, planes of the two transverse constric-
tions which occurred February 24.

FIG. 44. Male 7£>4, the left hand part cut off by the fission shown in
Fig. 43-

FIGS. 45 AND 46. Further history of male IDs, middle part cut off by
the fission shown in Fig. 43.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE III.

40

•46

LIBBIE H. HYMAN.

8

104 LIBBIE H. HYMAN.

PLATE IV.

FIG. 47. Male IDs later, nearly normal but with a double-headed bud.

FIG. 48. Male ID i, continued from Fig. 37, as it appeared on February
23, still without a stalk.

FIG. 49. Male IDi, March I, with the bud of Fig. 48 persistent, and a
new bud formed considerably distal to it.

FIG. 50. Male IDi, March 3, very elongated and with many buds, still
more distally located.

FIGS. 51 AND 52, further history of the double-headed bud given off from
the specimen shown in Fig. 33, as it appeared February 16 and February
23, respectively, .r. Fig. 52, plane of fission.

FIG. 53. Right part of the fission in Fig. 52, February 25.

FIG. 54. Same as Fig. 53, on February 27. .r, plane of fission, dividing
specimen into three parts.

FIGS. 55, 56, AND 57. History of the basal piece, having two feet, /, /,
separated off by the fission .r, Fig. 54.

FIG. 58. Male IB, left part of the fission shown in Fig. 28, as it appeared
January 6, 1926.

FIG. 59. Male IB, January 8. .v, level of fission.

FIG. 60. Male IE, right part from the fission of Fig. 59.

FIG. 61. Male IF, left part from the fission of Fig. 59.

FIG. 62. IE. January 16. .r, plane dividing the basal process off on
January 18.

FIGS. 63 AND 64. Fate of the basal process separated off in Fig. 61.

FIGS. 65 TO 67. History of the double-headed bud given off from the
specimen shown in Fig. 37.

FIG. 68. Male II, January 4, 1926, with two buds, irregularly placed.

FIG. 69. Male //, January II, with additional buds.

BIOLOGICAL BULLETIN, VOL. LIV

65

69

LIBBIE H. HYMAN.

IO6 LIBBIE H. HYMAN.

PLATE V.

•

FIG. 70. Male II, January 19, the lowermost bud persistent and with
a foot, /.

FIG. 71. Male //, January 27.

FIG. 72. Male //, February 4, with two buds, each with a foot.

FIG. 73. Male 77, February 9.

FIG. 74. Male II, February 16, with the same persistent bud.

FIG. 75. Male II , February 19, much elongated, with the same persistent
bud and with a new foot at /'.

FIG. 76. Male //, March 3, same bud still persistent, and with additional
buds, .r , plane of the fission of March 8.

FIG. 77. Male II A. distal part resulting from the fission of Fig. 76.

FIG. 78. Male II A, March 25, with several buds.

FIG. 79. Male 1 1 A, March 29, with several buds and feet. /, /, /.

FIG. 80. Male II A, March 31, with many buds and feet; lost by depres-
sion.

FIG. 81. Male IIB, proximal part from the fission of Fig. 76. a, the
part of the old column left on IIB by the fission.

FIGS. 82 TO 84. Further history of IIB. The process a in Fig. 81 has
formed a foot at its tip, so that the specimen has two feet, both of which
are functional. FIG. 83. By absorption of the basal region, the specimen
became nearly normal, Fig. 84.

FIG. 85. Male ///, January 6, with an accessory foot, /, near the middle
of the column, .r, plane of transverse fission, January 8.

FIG. 86. Male III A, distal part from the fission of Fig. 85.

FIG. 87. Male III A, January 18, nearly normal.

FIG. 88. Male IIIB, proximal part from the fission of Fig. 85.

FIG. 89. IIIB, January 13, with a foot at each end and one near the
middle, /, /, /, and a bud forming at b.

FIG. 90. IIIB, January 18. It is shorter and the bud is larger.

FIG. 91. 77/5, January 21, the bud larger and using the three feet as
its basal region.

FIG. 92. 7775, February 9.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE V.

LIBBIE H. HYMAN

IO8 LIBBIE H. HYMAN.

PLATE VI.

FIG. 93. IIIB. February 24, still with three feet.

FIG. 94. IIIB, March i, much elongated.

FIG. 95. IIIB, March 3, with many buds, x, plane of fission of March
8.

FIG. 96. Male IV , January 28, with two accessory feet, /. /. .v, plane
of fission, February 4.

FIG. 97. Male IV A, distal part from the fission of Fig. 96.

FIG. 98. Male IV B, proximal part from the fission of Fig. 96.

FIG. 99. Male IV A, February 27, still with two feet and having formed
a bud.

FIG. 100. Male IV A, March 12. with a third foot, .r, fission plane of
March 17.

FIG. 101. Male IVAi, distal part from the fission of Fig. 100.

FIG. 102. Male II' A 2, proximal part from the fission of Fig. 100.

FIGS. 103 TO 107. Further history of IV Ai. It formed additional feet
and budded, but finally by absorption, became nearly normal, Fig. 107.
March 31.

FIGS. 108 TO 1 10. Further history of male IV A2 (Fig. 102). It had three
feet; the bud formed at b. Fig. 109, used these feet as base. Fig. no.

FIGS, in TO 114. Further history of IV B (Fig. 98). It developed ac-
cessory feet, making four in all, Fig. 113. and a bud grew out from the
middle of the specimen, which used these feet as base, Fig. 114.

FIGS. 115 TO 117. A peculiar specimen found in the male culture, show-
ing regulatory changes in it.

FIG. 118. Specimen without a stalk resulting from transverse fission
of ID4, Fig. 44. The specimen has formed two buds, one at the extreme
base, the next more distal.

FIG. 119. The same specimen as Fig. 118, six days later, showing another
bud formed still more distal to the preceding two. A foot has also formed
at the proximal end.

BIOLOGICAL BULLETIN. VOL. LIV.

PLATE VI.

118

LIBBIE H. HYMAN.

Vol. LIV

February, IQ2S

No. 2

BIOLOGICAL BULLETIN

STREAMING AND POLARITY IN MASTIGINA HYLJE

(FRENZEL).

ELERY R. BECKER,
IOWA STATE COLLEGE.1

Mastigina hylcs, a member of the order Rhizomastigina, and an
inhabitant of the rectum of the bullfrog tadpole, is from a strictly
morphological standpoint a flagellate. The nucleus lies in the
extreme anterior end, and near it are to be found the basal body
which gives rise to a single, inconspicuous flagellum, and certain
intracytoplasmic structures which the writer (1925) has previously
described. From a physiological standpoint, however, the or-
ganism is an amoeba, for the cytoplasm is continuously engaged
in a process of " fountain streaming," and the single flagellum is
inactive so far as can be observed. Before this flagellum was
noted, the writer (1923) published a note in which he referred to
this organism as " an entozoic amoeba." This miscomprehension
of the structure of the cell led to certain misinterpretations of the
experiments which were reported in that note.

It is convenient to divide the substance in a streaming Masti-
gina hylcs into three zones: (i) the anterior zone, consisting of
the nucleus, basal granule and associated structures, and the gelled
cytoplastic matrix in which these structures are embedded (Fig.
ia) ; (2) a semi-fluid posterior or tail region, whose substance
takes no part in the general cytoplasmic streaming (Fig. ic) ; (3)
the more extensive intermediate streaming region (Fig. ib). The
cell substance and ingested materials of this intermediate zone
are reflected posteriad from the posterior surface of the nucleus
and remaining portions of the anterior zone and anteriad from the
concave anterior surface of the tail region.

1 The experiments reported in this paper were begun while the writer was
located at Princeton, and completed at Iowa State College during the
summer of 1926.

8 IOQ

HO ELERY R. BECKER.

It is then the intermediate zone which is almost entirely involved
in the process of streaming. We may, like Prof. John A. Ryder
(1893), conceive of this streaming system as having the configura-
tion of an elongated smoke ring, a system in which " some of
itself may continuously flow through itself in order to make the
progressive forward movement of itself possible." There is an
axial stream which flows swiftly and continuously toward the an-
terior end (Fig. i). Upon encountering the posterior surface of
the nucleus and nearby region the materials in this axial stream
are deflected radially to the periphery and then reflected back-
wards about the axial stream, forming a hollow cylinder of " ec-
toplasm " about it whose contents flow counter to it. The cylinder
substance enters the axial stream again in the posterior region,
although some part of it may have done this before the posterior
region is reached. Actual timing shows that it takes from 8 1/2
to ii 5/6 seconds for a particle to make the complete circuit in
healthy streaming cells. It is important to keep in mind that the
nucleus is constantly at the anterior end of the long axis of the
cell, where it may appear to oscillate slightly from side to side.

At any instant the living Mastigina has somewhat the shape of
a helix of from one and one fourth to two complete coils. The
coils of the helix are being constantly formed at the anterior end,
while they fade out at the posterior end. The shape of the body is
due to the fact that the nucleus and surrounding medium is jostled
about somewhat after the fashion of the ball in the vertical jet
of water which is sometimes seen in a shooting gallery. This
accounts for the apparent oscillation from side to side as observed
under the low power of the microscope. As a consequence the
materials in the axial stream are deflected radially somewhat un-
uniformly, and the body is curved toward the arc of the cross-
section receiving the lesser amount at any instant. Since the
streaming is uninterrupted and steady, the natural resultant of
the various factors is the construction of a helix. Sometimes,
however, the organism is almost spatulate in shape.

The progressive movement of the organism as a whole is not
usually the maximum to be expected for it often remains attached
by its tail. It was found by actual timing that the maximum
forward locomotion actually obtainable was about 480 microns per

STREAMING AND POLARITY IN MASTIGINA HYL/E. I I I

minute. The rate may be reduced to approximately zero, de-
pending on how much the attachment of the uroid impedes
progress.

The application of the colloid theory of protoplasm to amoeboid
movement is now so well recognized that it is hardly necessary
to point out that the axial stream is the plasmasol (See Mast,
'23, '26), the protoplasm about is the plasmagel. A thin tough
elastic plasmalemma encloses the entire protoplasmic substance.
If this is suddently broken, as by an abrupt jab with a glass
needle, the entire protoplasmic mass except the nucleus and gelled
anterior end goes into solution in the water medium. Flakes of
the plasmalemma can also be seen floating about.

Here is a cell with a definite polarity. Should this polarity be
ascribed to some factor inherent within the general cytoplasm?
Or are the direction of streaming and consequently the cell axis
maintained by virtue of some property of the nucleus, whether
physical, chemical, or physico-chemical, which determines that the
direction of the axial stream shall always be toward the nucleus?
Or does the differential factor which determines the direction of
the streaming operate in the anterior zone independently of the
nucleus ?

A number of individuals were cut squarely either across the
middle or slightly anterior to the middle by means of a finely
drawn hard-glass needle operated in a Chamber's micromanipula-
tor. In every case the anterior fragment is able to resume the
normal shape and streaming almost instantly (Fig. 3). The pos-
terior fragment always rounds up somewhat, undergoes some
fitful and uncoordinated cyclosis, extends and retracts a number
of pseudopods from its cortical region, all without accomplishing
any appreciable change of position (Figs. 2 and 4). The ability
to carry on the gelation-solation process (the ectoplasmic-endo-
plasmic proces) does not exist within it. In about eight minutes
it rounds up and streaming ceases completely. The anterior region
continues to behave quite normally, even though it be as small as
one fifth the area of the entire organism. It is not even necessary
to sever the two areas completely, for the application of pressure
by a needle held transversely to the large axis at any point in the
intermediate zone will bring about the same result (Fig. 6).

112 ELERY R. BECKER.

When the pressure is released the two areas flow together and
streaming is resumed.

If only the nucleus and a negligible amount of cytoplasm be
amputated, the result as regards the enucleated amoeba is the same.
These observations decide definitely against the view that the direc-
tion of movement is determined by some factor inherent in the ul-
tramicroscopic structure of the cytoplasm. Thus the differential
factor which determines the direction of streaming is either in the
nucleus or in the gelled cytoplasm at the extreme anterior tip. It
cannot be held that streaming was arrested simply by removal of
nuclear influence, for the writer (1926) and others have shown
that cessation of streaming is not an immediate effect of the re-
moval of the nucleus in amceb?e.

Oftentimes in unsuccessful attempts to cut off the entire anterior
end (Fig. 7), the pressure applied served to dislocate the nucleus
(and all the other anterior structures with it as was subsequently
proved), and it commenced to travel toward the posterior end in
the outer gel. A new anterior tip was temporarily established
(Fig. 8). But always after a moment the streaming was ar-
rested, a bulge appeared at the point where the nucleus had stopped,
endoplasm flowed toward the nucleus, and the organism traveled
away at right angles to its former course (Fig. 9). The uroid
or tail piece seems also to be a permanent structure, for it again
resumes its position at the posterior tip, and is not formed anew.
Centrifuged specimens likewise show similar temporary distur-
bances of polarity, but these are more difficult to interpret. Ex-
periments of this nature lend further support to the view that the
factor that determines the direction of streaming is located some-
where in the anterior tip, for it persists in resuming its natural
position at the anterior pole of the cell even though the polarity
be temporarily disturbed.

But does this differential factor lie in the nucleus as the writer
(1923) formerly supposed? In the summer of 1926 while at-
tempting to cut off the anterior tip of one of these organisms by
the free-hand method the nucleus was accidentally squeezed from
the vacuole which it normally occupied, while the remaining an-
terior structures were undisturbed. The vacuole which the nucleus
formerly occupied was plainly visible as a less refractile space in

STREAMING AND POLARITY IN MASTIGINA HYL^. 113

the anterior end. The nucleus traveled with the food particles,
making time after time the circuit from axial stream to anterior
end, backwards in the outer layer of gelled protoplasm to the
posterior end, and then again into the axial stream. The polarity
of the cell was not in the least disturbed. This makes necessary
a reversal of the former conclusion of the writer that " the
nucleus seems in some way to play a role in directing endoplasmic
streaming, presumably by its ability to slightly liquefy the ecto-
plasmic gel, lowering the resistance at certain points to the force
of the endoplasmic stream." This conclusion, as stated above,
was based upon a misconception of the structure of the anterior
end of the organism.

The more probable status of affairs is that it is the entire an-
terior tip of permanently gelled protoplasm which prevents
gelation of the " endoplasm " behind it, and which by imperfect
continuity with the outer gelled layer of the protoplasm of the
intermediate zone creates a circular zone of weakened elasticity
or lowered resistance to the internal pressure of the plasmasol
When the anterior tip is amputated the remainder of the organism
tends to round up. In such cases the plasmasol is imprisoned
within a wall of plasmagel for the penetration of which it cannot
muster enough internal pressure, thus making impossible the sola-
tion-gelatin process and the resultant streaming. If a normally
streaming individual be agitated mechanically, it too rounds up,
just as an enucleated fragment. This is known as the stimulated
condition. But if left quietly for a moment one or several pseu-
dopods are thrust out, but it is always the one with the nucleus in
its distal tip which becomes the permanent anterior end of the
streaming organism. The possession of the anterior zone is thus
the sine qua iwn for normal streaming in this protozoon.

The writer has in rare instances observed in fragments of
Atnceba dubia which he had enucleated streaming strikingly similar
to that in Mastigina hylce. In such cases the contractile vacuole
must be in the enucleated fragment and lie at the edge. It some-
times comes to occupy a position similar to that pf the nucleus in
Mastigina hylcs (Fig. 5). This type of streaming commences an
instant after enucleation and may continue for from one to two
minutes before the contractile vacuole bursts. The streaming is

U4 ELERY R. BECKER.

usually unbelievably rapid while it lasts. Such instances would
certainly justify the assumption that polarity is determined by the
physical presence in the gel of a foreign body which lowers the
resistance of the peripheral zone to the internal pressure in its
immediate vicinity.

CONCLUSION.

Mastigina hylcc shows typical fountain streaming, and during
locomotion assumes the shape of a helix with the nucleus at the
anterior end. If the organism be cut into an anterior and a pos-
terior fragment, the anterior fragment continues to stream nor-
mally, while the posterior fragment rounds up, shows uncoor-
dinated streaming and pseudopod formation, and soon dies. If
just the extreme anterior extremity be severed from the cell, the
entire protoplasmic mass behaves just as the above mentioned
posterior fragment. This indicates that the factor which de-
termines the direction of streaming, and hence polarity, is located
in the anterior end. That the nucleus is not immediately responsi-
ble for polarity is indicated by the fact that polarity is not dis-
turbed if it is dislodged from the vacuole it occupies without dis-
turbing the remaining anterior structures. If the entire anterior
end is made to travel posteriad with the outer gel, and a new an-
terior end formed, the original polarity is resumed by a cessation
of streaming, and renewed streaming toward the nucleus.

LITERATURE CITED.
Becker, E. R.
'23 The Role of the Nucleus in the Locomotion of an Entozoic Amceba.

Proc. Soc. Exp. Biol. and Med., 21: 155-156.
'25 The Morphology of Mastigitta hyla (Frenzel) from the Intestine

of the Tadpole. Jour. Parasit, n : 213-216, I pi.
'26 The Role of the Nucleus in the Cell Functions of Amoebae. BIOL.

BULL., 50: 382-392, i pi.
Mast, S. O.
'23 Mechanics of Locomotion in Amoebae. Proc. Nat. Acad. Sci., 9:

258-261.

'26 Structure, Movement, Locomotion, and Stimulation in Amoeba.
Jour. Morph. and Phys., 41 : 347-425.

Il6 ELERY R. BECKER.

DESCRIPTION OF PLATE.

FIG. i. A normally streaming Mastigina hylce showing the three zones
into which the body may be divided. Arrows show direction of streaming.

FIGS. 2 AND 3 show diagrammatically the difference in the behavior of
fragments from the posterior and anterior portions of the body respec-
tively. The latter streams normally.

FIG. 4. An enucleated fragment sending out broad pseudopods from
the cortical gel (ectoplasm). The sol (or endoplasm) is confined closely
within the gel.

FIG. 5. An enucleated fragment of Amoeba dubia streaming after the
fashion of Mastigina hylcc with a contractile vacuole in the anterior tip.

FIG. 6. The effect of pressure at right angles to the longitudinal axis
upon the streaming of Mastigina hylcc.

FIG. 7. Illustrating the method by which a break was made in the outer
gel. This permitted the substance of the axial stream to rush out.

FIG. 8. The nucleus, etc., has passed to the posterior end, and a new
anterior end temporarily formed.

FIG. 9. The reversal of streaming, in which the nucleus resumes its
former position and the former polarity is restored.

BIOLOGICAL BULLETIN, VOL. LIV.

n

ELERY R. BECKER.

STUDIES ON THE LIFE CYCLES OF TWO SPECIES

OF FRESH-WATER MUSSELS BELONGING

TO THE GENUS ANODONTA.1

MARY ELIZABETH TUCKER,
UNIVERSITY OF ILLINOIS.

Following the pioneer work of Lefevre and Curtis (1911) ir*
this country, much interest has centered around the investi-
gation of the developmental cycles of the fresh-water mussels.
Numerous studies have demonstrated that all members of the
family Unionidse undergo sexual development through a larval
stage known as the glochidium. While Surber (1912) and others
have called attention to the differences in the glochidia of many
species, there are as yet many untouched possibilities of differ-
entiating easily confused species and varieties on the basis of
glochidial characters. Until quite recently it has been held that
in all but two of the numerous species within this family, the
glochidium must pass through a period of parasitic existence on
an aquatic vertebrate before transformation to the juvenile
mussel is possible. With few exceptions, fishes serve as the
hosts essential for the parasitic glochidia.

It was in 1866 that Leydig made the discovery that the
glochidium, after leaving the parent, completed its development
as a parasite on a fish. Following that disclosure, it was believed
that the parasitic life was a necessary phase in the life cycle of
any fresh-water mussel. Lefevre and Curtis (1911) observed a
species, Strophitus edentuhis, which undergoes metamorphosis
without parasitism, the first case cited of such a condition among
the Unionidae. This report marked an important advance in
the history of the knowledge of these forms.

Three years after the discovery by Lefevre and Curtis that
Strophitus edentulus passes through its metamorphosis in the
entire absence of parasitism, Howard (1914) reported a second

1 Contributions from the Zoological Laboratory of the University of Illinois,
No. 305.

II?

Il8 MARY ELIZABETH TUCKER.

case of such development. He declared (1914: 353): "One of
the mussels for which I found no natural infection and for
which none have been reported was Anodonta imbecillis (Say)."
Howard stated further (p. 354): "I have tested the reaction of
the glochidia in the presence of fish and obtained evidence that
they do not respond as other known parasitic forms."

As a result of further investigations, Howard (1911: 355)
presented the account of obtaining both infection and encyst-
ment of Anodonta imbecillis glochidia. He also reported having
obtained infection and encystment of Strophitus edentulus, and
for this species in addition, declared actual metamorphosis to
occur on fish. He concluded, therefore, that Strophitus edentulus
glochidia may be facultatively parasitic, while for Anodonta
imbecillis, "there is at least a persistence of the parasitic reaction
when the glochidia are artificially brought in contact with fish."

As a result of investigations carried on at the University of
Illinois, some observations have been made on the life histories
of two species of fresh-water mussels belonging to the genus
Anodonta, Anodonta imbecillis x and Anodonta grandis. At-
tention has been directed especially to a study of the glochidia
and to their metamorphoses. In this connection, the occurrence
of metamorphosis without parasitism has been restudied in
Anodonta imbecillis and observations have also been made of
both glochidia and juvenile mussels of Anodonta imbecillis and
Anodonta grandis.

This work has been carried on under the supervision of Pro-
fessor H. J. Van Cleave at the University of Illinois. The
writer wishes to express to Professor Van Cleave appreciation
for suggestions and guidance in this study. Further indebtedness
is hereby acknowledged to R. E. Richardson for the identification
of fishes used, and to F. C. Baker for aid in ascertaining the
species of mussels concerned.

1 Baker (19276) created a new genus, Utterbackia, with A. imbecillis as type.
Evidences presented in this paper eliminate one of the principle characters for
differentiating this genus from Anodonta. Under these circumstances it seems
desirable to use the older name in this publication until the exact status of Utter-
backia is determined.

LIFE-CYCLES OF FRESH-WATER MUSSELS. 1 19

METAMORPHOSIS WITHOUT PARASITISM.

Development of Strophitus edentulus. — When Lefevre and Curtis
(1911) recorded their observations on the direct development of
Strophitus edentulus, they stated that the young mussels had
developed inside cords within the marsupia, reaching the stage
"attained by any other unionid at the time of liberation from the
host fish." The young at this stage were liberated from the
marsupia by the disintegration of the cords and showed "charac-
teristic features resembling those of Anodonta at the close of the
parasitic period." The authors reported that they were unable
to bring about the attachment of these young to fish. Since the
juveniles crept about on the bottom of the dish after the manner
of the young of other species in post-parasitic life, Lefevre and
Curtis concluded that no subsequent parasitism was possible.

Case of Anodonta imbecillis. — Howard's discovery of Anodonta
imbecillis passing through its metamorphosis in the absence of
parasitism, brought to light a second case of this newly dis-
closed means of development. He secured several specimens of
Anodonta imbecillis the latter part of November (1913). Since
the species is hermaphroditic (Sterki 1898: 87), his expectations
were confirmed in that all the specimens were gravid. The
marsupial contents upon examination showed what at first
appeared to be glochidia to be in reality "juvenile mussels with
organs developed to the stage usually seen at the end of parasitism
when the young mussel escapes from its host" (1914: 353~4)-

In confirmation of his conclusion as to the development of
Anodonta imbecillis, Howard stated (1914: 354): "I have tested
the reaction of the glochidia in the presence of fish and obtained
strong evidence that they do not respond as other known parasitic
forms." He does not record, however, the species of fish used.
Mature glochidia taken in March were employed which "in an
exposure to fish for an hour failed to give the usual infection.
A few glochidia lodged in the mouths of the fish, bat no encyst-
ment could be detected. The fish showed no response." He
placed especial significance upon the fact that the glochidia of
Symphonota complanata rapidly became attached to the fish
which showed considerable uneasiness in marked contrast to the
behavior of fishes in the presence of Anodonta imbecillis glochidia.

I2O MARY ELIZABETH TUCKER.

Upon this premise, Howard thought that he had conclusive
evidence that Anodonta imbecillis passes through its meta-
morphosis in the absence of parasitism.

As a result of further investigation, Howard was forced to
modify his conclusions, for he reported (1914: 355) both infection
by and encystment of Anodonta imbecillis glochidia. Howard
does not record the host used in his experiments, but, since the
susceptibility of fishes is significant in this connection, it is
possible that in his earlier experiments he used a species entirely
immune to Anodonta imbecillis.

OCCURRENCE OF PARASITISM IN Anodonta imbecillis.

In the present investigations, it has been found that Anodonta
imbecillis not only discharges glochidia but that they are
capable of attachment upon a fish (Apomotis cyanellus). The
period of parasitic life is of practically the same duration as is
necessary for the metamorphosis of Anodonta grandis.

In the course of this investigation, constant reference has
been made to a comparison between conditions in Anodonta
imbecillis and Anodonta grandis for the developmental cycles of
these two species has been studied from a comparative point
of view.

Three individuals of Anodonta imbecillis obtained from Lake
Decatur at Decatur, Illinois, in November, 1926, were placed in
a tank of running water and kept under observation. Of these,
one died December 3 and was found to contain only a few living
glochidia in the outer gills. The remaining two specimene wers
placed in separate containers January 27. On February i, one
of these was found to have given off living glochidia which were
recovered by a pipette from the bottom of the jar. During a
period of twenty-four hours following, approximately 200 glo-
chidia were discharged, which were followed by similar numbers
until February 8, when the mussel was found with gaping valves.
Upon examination the marsupia showed a very few of the larvae'
In the meantime, the second specimen of Anodonta imbecillis
(2/4/27) also gave off living glochidia. After two weeks, these
ceased to be discharged, the mussel living, however, until March
17, at which time an examination of the marsupia showed a

LIFE-CYCLES OF FRESH-WATER MUSSELS.

121

few glochidial shells present but no living glochidia. It is
possible that these glochidia were given off unnaturally as a
result of changed environmental conditions. At any rate, the
discharge was unquestionably of mature glochidia and not of
young juvenile mussels.

A fairly large proportion of the glochidia discharged were
active, opening and closing their valves in characteristic fashion,
but in so far as these observations went were incapable of
locomotion. A small green sunfish (Apomotis cyanellus] was
subjected to a suspension of some 450 of the glochidia which
had been discharged by the second specimen of Anodonta imbe-
cillis. A count made after thirty minutes of exposure showed
more than fifty of the larvae attached in fairly equal distribution
to the several fins of the fish. After twenty-one hours, the
attached glochidia were surrounded by host tissue, encystment
having been complete for at least part of the larvae.

The infected fish was observed almost daily for the numbers of
glochidia remaining. During the first few days a considerable
dropping off of originally attached larvae was noticeable. It is
probable that individuals having obtained poor attachment were
thus early lost.

On the eighteenth day after infection, two juveniles were
recovered from the bottom of the battery jar containing the host
fish. On the following days additional juveniles were secured.
The following table records the number of attached larvae and
the duration of parasitism.

TABLE I.

RECORD OF THE NUMBERS OF INFECTION'S AND THE DURATION OF PARASITISM

IN Anodonta imbecillis.

Feb.

March.

Apomotis

Duration of

cyanellus.

Infection.

19

21

23

25

28

2

4

5

7

8

9

10

ii

12

15

Glochidia attached.

5C

50

40

35

30

25

23

22

22

19

12

6

2

I

0

18-22 days

Juveniles recovered

2

4

3

I

I

While Howard has reported Anodonta imbecillis to undergo
metamorphosis without parasitism, the results of the present

122 MARY ELIZABETH TUCKER.

study do not confirm his results. This divergence in results
indicates the possibility of recognizing two distinct physiological
varieties of Anodonta imbecillis. Howard does not give a de-
scription of the species, but the characters of the glochidia
presented by Surber for Anodonta imbecillis agree in detail with
those of the species here studied. Since preserved specimens
which were collected at various times of the year yielded only
glochidia, showing no evidences of transformation to the juvenile
stage, it is necessary to conclude that the young of Anodonta
imbecillis forming the basis of this study are discharged as
glochidia. According to the present observations, the glochidia
were given off under the conditions already stated, and some
actually underwent metamorphosis upon the green sunfish
(Apomotis cyanellus).

In this connection, it may be worth mentioning that Howard
presumably worked with specimens from a large stream (the
Mississippi River), while the individuals forming the basis of
this study were all from relatively small streams and an artificial
lake formed by damming a small stream. Utterback (1916: 2)
and Baker (i92ya: 112) have shown that stream conditions
profoundly influence the features of the adult shell in Unionidae.
Later work supplementing that given in this paper may lead
to the recognition of correlation between the differences in
breeding habits and environmental factors.

THE LIFE CYCLE OF Anodonta grandis.

As mentioned earlier in this paper the observations and experi-
ments on the development of Anodonta imbecillis have been
paralleled by a series of similar studies on another species,
Anodonta grandis.

A number of living specimens of Anodonta grandis were
obtained from Lake Decatur, Decatur, Illinois, in November,
1926. A few were examined at once and found to be gravid.
The marsupia were greatly distended bearing enormous numbers
of living glochidia. Measurements of the shells were made
from living glochidia supplemented later by measurements from
permanent microscopic mounts. Table II. records the measure-
ments taken.

LIFE-CYCLES OF FRESH-WATER MUSSELS.

123

TABLE II.

RECORD OF MEASUREMENTS OF GLOCHIDIA OF Anotionta

Place and
Date of
Collection.

Extremes of
Measurement.

Most Frequent
Size.

Length.

Depth.

Length.

Depth.

Specimen No. I. . .
Specimen No. II .

Decatur, 111.,
11/6/26
Decatur, 111.,
11/6/26

0.350 mm.-
0.300 mm.
0.350 mm.-
0.398 mm.

0.343 mm.-
0.382 mm.
0.358 mm.-
0.390 mm.

0.358 mm.
0.366 mm.

0.350 mm.
0.358 mm.

In the literature, there is a marked discrepancy between the
measurements recorded for the glochidia of Anodonta grandis.
Thus Surber (1912: 8) states that specimens which he studied
were 0.41 mm. by 0.42 mm., while Ortmann (1919: 139) found
the glochidial shells to be 0.36 by 0.37 mm. The latter author
called attention to the difference between his measurements and
those of Surber but offered no explanation for the disagreement.

Conchologists have long recognized a subspecies of Anodonta
grandis characteristic of the large rivers under the name of
Anodonta grandis gigantea. The differentiation of this variety
from the typical Anodonta grandis has been largely upon shell
characters. It is known that the material examined by Ortmann
(1919: 140) was from a small stream as was also the material
upon which the present investigation was based. On the other
hand, it is probable that Surber 's material, though recorded as
Anodonta grandis, was from the Mississippi River and conse-
quently belonged to the variety Anodonta grandis gigantea.
Thus it seems reasonably certain that Anodonta grandis and its
variety Anodonta grandis gigantea may be differentiated with
greater certainty on the basis of glochidial measurements than
on shell characters of the adult.

Under conditions of the experimentation, specimens of Ano-
donta grandis kept in tanks of running water continued to retain
living glochidia within the marsupia from November 22 until
January 6. Following this date, some of the females bore
glochidia in the anterior region of the marsupia only, while
after February 10 all the females examined lacked living glo-
chidia. According to observations of Coker, Shira, Clark, and
9

124

MARY ELIZABETH TUCKER.

Howard (1921: 142), females of Anodonta grandis bearing glo-
chidia have been observed in December, January, February, and
March only. Thus while there is a slight difference between
the observations recorded by these authors and those of the
present study, the length of the gravid period is almost identical
in the two instances.

Four days after the living specimens of Anodonta grandis had
been placed in the tank, material examined from the bottom
contained living glochidia. In two weeks the number of living
ones was greater in a similar quantity of material, which con-
tained also many glochidial shells. The dates of gravidity for
those specimens kept under observation in running water have
already been indicated. Three living specimens brought in from
the field on May 7, 1927, contained no glochidia.

In order to observe the results of artificial infection with
Anodonta grandis for a comparison with the earlier study on
Anodonta imbecillis, several specimens of small fishes were used.
One species of minnow (Pimephales notatus), two of darters
(Etheostoma cseruleus, Beleostoma nigrum}, one sucker (Moxostoma
breviceps?), and the green sunfish (Apomotis cyanellus) were
placed in the tank with the living Anodonta grandis December 6,
1926. At this time, a fairly large number of living glochidia
were present in the material at the bottom of the tank. A
record was kept of the numbers of glochidial attachments for
three weeks. The results are shown in the following table.

TABLE III.

RECORD OF THE ARTIFICIAL INFESTATIONS BY GLOCHIDIA OF Anodonta grandis.

Fish Introduced into the
Tank Dec. 6, 1926.

Numbers of Infestations.

Dec. 8.

Dec. 13.

Dec. 16.

Dec. 18.

Dec. 23.

Jan. 3.

Pimphales notatus:
Specimen No. I

i
o

0
0
0

o
o
I

i

i
3

2
I

O
O

I

3
3

i

14
4

21
2 •

I

2
2
I
17

5
37

2

4

i
i
i
ii

2

5
o

5

5

2
0
0
O

2

I

5

Specimen No. II

Specimen No. Ill

Etheostoma cceruleus

Beleostoma nigrum

Moxostoma breviceps ?
Specimen No. I

Specimen No. II

Apomotis cyanellus

LIFE-CYCLES OF FRESH-WATER MUSSELS. 125

The results in the case of the minnows, darters, and sucker
showed considerable variation in numbers of infestations. In-
creased attachments were evident in all cases after the first
infection, which condition was especially marked in the case of
one darter (Etheostoma cxruletis) and one sucker (Moxostoma
breviceps?}. Particularly in the case of the darter with its habit
of remaining near the bottom of the tank darting about actively
at times, it is not surprising to find that fairly heavy infection
resulted. In the cases of the fish first mentioned, the increased
infections were shortly followed by a dropping off of the glochidia,
as indicated by the reductions of the numbers carried. Exami-
nation of the fish showed parts of the fins sloughed off presenting
noticeably ragged appearances especially marked in the case of
the darters and suckers. The sunfish, on the contrary, did not
show such a reduction in the numbers of glochidia. The fins
remained intact, and examination soon showed the glochidia
encysted. The sunfish was the only host encountered in these
experiments which retained the parasites for any appreciable
length of time. Lefevre and Curtis (1910: 103) recorded several
species of Centrarchidae as natural hosts of Anodonta grandis.

In order to observe the results of attachment of the glochidia
to the sunfish more accurately than was possible by leaving the
fish free to swim about in the tank, two small uninfected speci-
mens were used for experimentation. Of these, one was a small
green sunfish (Apomotis cyanellus), the same species which had
appeared earlier to be subject to infection, and the other a rock
bass (Ambloplites rupestris'). A fairly large quantity of living
glochidia from one of the gravid females of Anodonta grandis
was placed with each of these fish in separate small containers.
Infection was evident at once, and after fifteen minutes of
exposure, each fish was placed in fresh water, again in the small
containers which were now set in the tank in order to keep the
water at a temperature favorable for the existence on the fish,
at the same time preventing additional infection.

The two fishes kept under observation were kept in the same
temperature conditions, the glochidia which they carried under-
going metamorphoses in eighteen days and in seventeen to
twenty days respectively. A record of the duration of parasitism
on the different fishes is given in Table IV.

126

MARY ELIZABETH TUCKER.

TABLE IV.

RECORD OF THE NUMBERS OF INFECTIONS AND DURATION OF PARASITISM

OF Anodonta grandis.

Feb.

March.

Duration of

Infection.

16

18

19

21

23

25

28

2

4

5

7

8

Apomotis

cyanellus:

Glochidia

attached

2

2

2

2

2

2

2

2

2

0

1 8 days

Juveniles

recovered

2

Ambloplites

*

rupestris:

Glochidia

attached

40

40

35

30

25

20

17

6

4

I

i

0

17-20 days

Juveniles

recovered

2

3

SUMMARY.

1. Anodonta grandis and Anodonta imbecillis have been studied
with reference to the glochidia and later stages in the life cycles.

2. Through experimental infestation, glochidia of Anodonta
imbecillis have been reared under laboratory conditions through
the period of transformation to the juvenile stage.

3. An examination of mature individuals collected at various
times of the year yielded only glochidia, showing no evidences of
transformation to the juvenile stage.

4. The claim that metamorphosis in Anodonta imbecillis occurs
in the marsupia of the parent has not been confirmed by this
investigation.

5. Apomotis cyanellus served as host to the glochidia of
Anodonta imbecillis.

6. The transformation of glochidia of Anodonta grandis has
been followed in experimental infestations using Apomotis
cyanellus and Amploplites rupestris as hosts.

7. Anodonta grandis may be differentiated from Anodonta
grandis gigantea on the basis of glochidial measurements.

REFERENCES CITED.
Baker, F. C.

The Naiad Fauna of the Rock River System: A Study of the Law of
Stream Distribution. Trans. 111. St. Acad. Sci., 19: 103-112.

LIFE-CYCLES OF FRESH-WATER MUSSELS. 127

'27b On the Division of the Sphaeriidae into Two Subfamilies: and the Descrip-
tion of a New Genus of Unionidae, with Descriptions of New Varieties.
American Midland Naturalist, 10: 220-223.
Coker, R. E., Shira, A. F., Clark, H. W., and Howard, A. D.

'21 Natural History and Propagation ot Fresh-Water Mussels. Doc. No. 893.

Bull. Bur. of Fisheries, 37.
Howard, A. D.

'14 A Second Case of Metamorphosis Without Paiasitism in the Unionidae.

Science, 63 (n. s. 40): 353-355-
Lefevre, G. and Curtis, W. C.

'10 Reproduction and Parasitism in the Unionidae. Jour. Exper. Zool., 9:

79-H5-
'n Metamorphosis Without Parasitism in the Unionidae. Science, 33: 863-

865.
Ortmann, A. E.

'19 A Monograph of the Naiades of Pennsylvania. Part III. A Systematic

Account of the Genera and Species. Mem. Carnegie Mus., 8: 1-384-
Sterki, V.

'98 Anodonta imbecillis, Hermaphroditic. Nautilus, 12: 87-88.
Surber, T.

'12 Identification of the Glochidia of Fresh-Water Mussels. Bull. Bur. of

Fisheries. Doc. No. 771.
Tucker, Mary E.

'27 Morphology of the Glochidium and Juvenile of the Mussel Anodonta im-
becillis. Trans. Amer. Micros. Soc., 46: 286-293.
Utterback, W. I.

'16 The Naiades of Missouri. American Midland Naturalist, 4: 1-200.

THE MORPHOLOGY AND PHYSIOLOGY OF THE
SALAMANDER THYROID GLAND.

III. THE RELATION OF THE NUMBER OF FOLLICLES TO DE-
VELOPMENT AND GROWTH OF THE THYROID

IN Ambystoma maculatum.

E. UHLENHUTH AND HILDA KARNS.

From the University of Maryland Medical School and the Marine Biological
Laboratory, Woods Hole, Mass.1

Owing to the profound changes which the follicular pattern
of the human thyroid gland undergoes in various pathological
states, the follicular composition of the thyroid has been given
careful attention by many students of the thyroid gland.

Martin Heidenhain was one of the first to examine the thyroids
of various mammalian species with reference to the follicular
pattern. He made the noteworthy discovery that profound
differences exist between the different species of this group of
animals, regarding the development and morphology of the
follicular design of the thyroid. According to Heidenhain two
types of follicular design must be distinguished, an association
type to which belong the thyroids of the dog and cat, and a
dissociation type to which belong the thyroids of the cattle
and of man. In the association type the follicles which develop
from the primary cell columns remain permanently in epithelial
continuity, while in the dissociation type they become separated.
In both types secondary "fusion" may take place. Thus in
the thyroids of dog and cat secondary communication (canali-
zation) between the adjacent follicles of an epithelial column
may take place, if the quantity of secretion increases, and lead
to the formation of tubuli possessing a continuous lumen. The
same process may, at times, occur in man, at the stage of the
primary cell columns, before splitting up into separate follicles
has taken place; if it does occur, it acts inhibiting upon the

1 Part of the work reported in this article was carried out with the aid of the
"Julius Friedenwald Fund for Medical Research" of the University of Maryland.

128

MORPHOLOGY OF SALAMANDER THYROID GLAND.

process of splitting up. Heidenhain made some observations
also on the growth and multiplication of follicles. He states
that real budding of follicles is very rare; normally each cell of
the primary cell columns is destined to develop sooner or later
into one individual follicle. No new follicles, beyond the original
number of cells, develop. In cattle, however, budding may be
observed during the period of most rapid growth, just after
birth. What appear to be buds in man, are only young follicles
developing out of cells which failed to separate at the time when
the other cells separated, the result of an incomplete dissociation.
As to a multiplication of follicles Heidenhain observed that even
in case of real buds these never become separated; likewise a
cleavage of larger into smaller follicles does not occur.

In 1923 Williamson and Pearse (14, 15) published a study of
the follicular pattern of the human thyroid gland and reported
conditions essentially different from previous observations. Ac-
cording to these investigators the human thyroid is composed of
gland-units a certain number of which make up one thyroid
lobule. These units consist of an endothelial capsule the lumen
of which is part of the general lymphatic system and communi-
cates with other units and with the general lymphatic system, by
means of a stalk-like lymph capillary. Very oddly in this
lymphatic capsule are contained blood vessels, the capillaries of
the gland-unit, which enter the capsule through the stalk-like
hilus. In addition to the blood vessels each capsule contains
probably one, possibly several coiled and freely floating epithelial
columns. These epithelial columns of each gland unit, being
the real glandular elements of the thyroid, are completely in-
dependent of the columns of other gland units, except for the
lymphatic and blood-vascular communications. The solid epi-
thelial columns represent the primary resting condition of the
gland-unit. When functioning either of two completely different
states — and only one of them at a time — may develop from these
solid cell-columns. In both cases a lumen appears in the center
of the cell column; in one case it becomes filled with colloid,
in the other case with a more fluid substance. The colloid is
not a secretion; the secretion is represented by the more liquid
substance which, it is asserted, leaves the lumen of the epithelial

I3O E. UHLENHUTH AND HILDA KARNS.

tube by way of "intracellular" tubules. Normally units in
each of these three states are found together in one thyroid.
But a unit which elaborates colloid cannot, at the same time,
elaborate secretion. Before this is possible, it must revert to
the original condition of the primary solid cell column.

So far these assertions, together with many others to be found
in the publications referred to and adhered to by Williamson
and Pearse in their more recent articles (16, 17) are mere state-
ments ; to support them the authors have furnished a surprisingly
small number of facts. For reasons outlined specifically in a
previous article we cannot subscribe to these theories at large;
some of them are rather incoherent and others are based upon
grossly wrong observations and interpretations. But in con-
nection with the present article we are mostly interested to know
whether in the human thyroid what impressed most observers as
separate vesicles are in reality only parts of continuous tubules.
It is well known and has been pointed out by Heidenhain and
others that the human thyroid, in certain states, does seem to be
composed of numbers of coiled, continuous tubules rather than of
isolated vesicles.

Upon inspection of the thyroid of an exophthalmic goiter
gland l we have convinced ourselves that within the larger lobules
smaller areas can be distinguished, which are separated from
the vicinity by finer connective-tissue strands; the follicles
located within these areas suggest frequently that they might be
merely parts of one larger, continuous tubule and sometimes
resemble in a striking manner the condition found in Ambystoma
opacum, illustrated in Fig. 39, Plate I. of a previous paper (n);
as will be mentioned presently, in this species there really exists
a condition somewhat similar to the one described by Williamson
and Pearse.

Since 1922 the senior author of this article has been engaged
in a study of the thyroid gland of the salamander Ambystoma
opacum; the results of this work have been published in a number
of previous articles (2 to 13). In this species of salamander the
thyroid, during development, may follow either of the two types
distinguished by Heidenhain (i) in mammals. But whichever

1 This gland was received from Dr. A. S. Blumgarten, of the German Hospital
in New York; it gives me pleasure to thank Doctor Blumgarten for his courtesy_

MORPHOLOGY OF SALAMANDER THYROID GLAND.

type it follows, the final result is, in the main, the formation of a
large and coiled tube, distended with secretion to a varying
degree in different places of its course, and resembling the con-
dition described by Williamson and Pearse for the thyroid of man.
The whole gland of the marble salamander may be compared, at
that stage, with one single unit of the human thyroid gland.
This continuous tube forms through a process of fusion of the
smaller follicles, very much as described by Heidenhain, except
that in case of previous dissociation really separated follicles fuse
with each other. It was found that the development of this tube-
like follicle, together with the elaboration and storage of the
secretion product (a mixture of unstainable and, mainly, stain-
able colloid) coincides with the larval period. It represents,
what was called, the developmental or storage phase of a func-
tional cycle the second phase of which is the "functional phase"
or phase of colloid release, which coincides with metamorphosis
and is characterized, as regards the follicular pattern, by a
complete collapse of the follicular tube, caused by the sudden
release of the secretion product from the lumen of the follicles.
The functional phase is followed again by the storage phase
which now is initiated by an enormously increased elaboration of
predominantly unstainable colloid. Like in man and other
mammals- — and unlike the assertions made by Williams and
Pearse — the two secretion products, the chromophobe and the
chromophile colloid, are always elaborated in and excreted from
the same cells at the same time and into the same secretion-
space, although in a varying proportion.

The important feature in the thyroid of the marble salamander
and the one of especial interest in connection with the present
article is the existence of a phase during which the small primary
follicles fuse with each other to form a large tube, sometimes
resembling a coiled tube, sometimes exhibiting a more bag-
like appearance, which is distended with the secretion product
of the cells. At the time when these observations were made the
thyroid gland of Ambystoma tigrinum and of the axolotl, in
addition to the thyroid of the marble salamander, were studied
on sections and it was noticed that the follicular patterns of
these glands differ markedly from that of the thyroid of the

132 E. UHLENHUTH AND HILDA KARNS.

marble salamander; the presence of a particularly large follicle
was not noticed in these glands.

It was thought desirable to make a comparative study of
the follicular pattern of several different species of salamanders,
to see what was essential in the follicular pattern of the thyroid
in general and whether differences in the follicular pattern
corresponded to certain specific physiological differences of the
thyroid apparatus. As at the same time large numbers of
thyroids of experimental material had to be examined, the need
was felt of developing a method which would enable us to make
accurate examinations of the follicular pattern without the
laborious work of sectioning and as soon as possible after the
removal of the organs from the animals. It was found that
upon maceration with 40 per cent, nitric acid the thyroid can
be taken apart completely into its individual follicles, the size
and shape of the follicles can be studied and the exact number of
follicles can be ascertained for each individual thyroid gland.
A brief account of the results thus obtained was published in the
Anatomical Record (Uhlenhuth, 12).

In the present article the results obtained with this method
for the thyroid of the salamander Ambystoma maculatum only
will be reported; the thyroids of other species of salamanders
have been successfully investigated with this method and
publication of these investigations will be made soon. It should
be mentioned here that the same method can be applied very
successfully also to the thyroids of warm-blooded animals (dog)
and of man.

METHOD.

The animals were anesthetized with chloretone (10 cc. of an
0.5 per cent, aqueous solution of chloretone plus 40 cc. of a 30 per
cent. Ringer solution) and pinned up in a tray under 30 per cent.
Ringer solution. The thyroids were removed in the usual way
and placed at first into a dish containing a 30 per cent. Ringer
solution. From this solution they were placed directly into a
40 per cent, nitric acid (Merk's Blue Label). It is best to leave
them in the acid for from 12 to 24 hours and then to transfer
them into a large dish filled with tap water. The water should
be changed at least once. The glands treated in this fashion

' % MORPHOLOGY OF SALAMANDER THYROID GLAND. 133

should be dissected as soon as possible to avoid softening of the
epithelial walls of the follicles. If this procedure is employed
the connective tissue will be found disintegrated sufficiently to
give way to the slightest pulling, while the follicles show great
resistance and can be separated from one another without the
least injury to the epithelial walls. The colloid assumes a
yellowish tint, the cells are whitish. In case a part of the
epithelial wall is scraped off or components belonging to one
single follicle are torn asunder, the defect of the epithelium and
the smooth, yellowish surface of the denuded colloid can be
noticed at once.

Each thyroid was completely dissected into its individual
follicles and the absolute number of follicles was ascertained
(Table I., column 6). Representative follicles as to shape and
size were then selected and their outlines drawn at a magnification
°f X 33 (Zeiss Binocular Dissection Microscope, Oc. 2, Obj. az}.

In order to determine the relative number of the follicles,
i.e. the number of follicles per unit of thyroid volume, it would
have been necessary to determine the volume of each thyroid
gland. As only in a few cases the thyroid volume was ascertained
another, less accurate method was resorted to. Of each thyroid
the largest optical section was drawn with the aid of a camera
lucida, at a magnification of X 33, after the thyroids had been
removed from the acid into the water. With the aid of a
planimeter the area in cm2, of each outline drawing was de-
termined (Table I., column 8). By dividing the number of
follicles over the area the number of follicles per cm2, area was
found for each thyroid (Table I., column 7). This method
would give correct results only if the cross section of each thyroid
were a circle. As some thyroids are less cylindrical than others,
the values for the relative numbers of follicles as recorded here
are only approximately correct. No conclusions, however, were
drawn from these values beyond those suggested by the con-
tinuity and persistency of the change of this value with the
progressing increase in the size of the thyroids and age of the

animals.

MATERIAL.

The investigations reported in the present article were carried
out on the species Ambystoma maculatum, in spring and summer

134 E- UHLENHUTH AND HILDA KARNS. «

1925 and 1926. With the exception of the larvae of the series
CCLXII. and the sex-mature animals of the series CCLXX. the
animals were reared from eggs in the laboratory. All animals
were kept under the same conditions except where mention is
made of the contrary. All animals of the same series are the
offspring of the same female, except the animals of series CCLXII.
and CCLXX. The entire material was collected by Mr. George
Gray in the vicinity of Woods Hole, Mass.

THE NUMBER OF FOLLICLES.

I . Average number of follicles : The average number of follicles
of the thyroid gland of the spotted salamander, as calculated
from the thyroids of 88 normal animals examined since 1925, is
39.7 follicles. There is, howrever, a considerable variability from
the average; the smallest number of the follicles in one thyroid
was 5 follicles, the largest number 72 follicles (Table I., column 6).

EXPLANATION TO TABLE I.

The early larval stages (from No. i to No. 29) are arranged according to body
length in mm. ; the older larval stages and the metamorphosed animals are arranged
according to stage, as during these periods of life the stage and not the size of the
animal determines the condition of the thyroid gland.

The areas of the thyroids of the first seven animals (column 8) were drawn
from the fresh gland, in all the other glands the areas were drawn after removal of
the thyroids from the nitric acid into water. Comparison of the area before and
after the acid treatment showed a shrinkage of the thyroid, which decreased the
area by approximately 18 per cent, of the area of the untreated thyroid; therefore
1 8 per cent, should be subtracted from the values of the area of the first seven
animals, to obtain the correct measurement.

Abbreviations.

E.pr.sl Eyes protrude slightly.

E.pr.d . Eyes protrude distinctly.

E.pr.c Eyes protrude considerably.

Yell.N . . .The even coloration of the earlier larval stages has been

replaced by a yellow network on dark background.

Op.sh Opercular pouch, on ventral side, has started to grow

to the body wall and has become shallow as compared
to the previous larval stage.

G Gill slits.

Pieces i day Cast off, for the first time, some small pieces of epi-
thelium, I day before examination of thyroid.

Sh.Sk Shed his skin completely.

L Placed on land (moist filter paper).

Eats Started taking food again.

MORPHOLOGY OF SALAMANDER THYROID GLAND.

135

c^^-'c-awg^

,>>

"1 ° £ _o § '§ 5 '£ ^

*O ^2 M fO ^1 ^ ^" *^ *^ *'"• *^ ^^ *^ ^*

o

Cy c* J™ ,— . .M s Cy "y*

flj c3 i3 5 t-j **^ o ^^

OOOOOOOOOO >nO»-*MCSC^wwCNrO^WW

"o

c/; rt 'H °

. "5

i* x 4-

"c <~ S o S

I | OrN^|«Ni^rONO^roroOtC-ir-

0)

,3 "o S <l

I I GO^rol^rO'1-<NM(N^M^MM^^

si

2 fe a

o rt

1*1

i i svjaasu****!^***

£ k

!^ "o

X 03
^> 73

c/5

^s cu

J5 .2 £

2 2 M n 5 * N^S 5^" ^"Sx01Ov§>00>xOOO>1?-?-t2

t^ £

< Q

o 2

I «•>

. 2

cq

r 2
To .2

x-3 g

100 ^-1000 1-1 ^t 11 Tfmo O\O ro^">Noooo ^}-oo O r~

o -

1— 1

•o se 5

O C «

r°"» oo oo ON Ov O\ i (^ 10 ^3 ^O t^* t^« o ^ n i w Oi CN ro ro ^^ ^^

8

W

pg a) _,

c

*7

H- J - —

•r -3

s

z -

C/3

a s*

X _

D

a

1^

z

"— > w

S

J3 rt

X

3 "3

. w

1— I

1-8

8

M ^

_O Z>

IN

5 1

CQ •§

< V

^C

4) O

H 1

a

1

'S

. -*->
•LJ 4_* --r
333,
. O O as

v. r2

'~. p

0 r^

• ij <ii c/3 w w

_o

^

i! o u "C T; -a . ....

O~"^333 >. X>>XX

O c/3

^

o
o

^88^5^ 5 5^55

£ S

ctf <->

1

^

j

w

ccrjScsc ^s ~^:^n^:
rt -2 — s = s =« " • •-« K

15^^-a^-a « -£-S^

•° 0 0 .S .2 .2 3 S 3 3 3

1 a

S 5

•^ ."

............ . . O . O O C C

"rt 5

0

"c ^

£

* 1

U CJ

cs cs rorofO^O O\oo C\W O\ J *J »— Ji-4>— 3i— Jt-JnJWWUW

'"5 "3

c c

-*-> 3 "O

Co O CJ

a U »->

1- 03 -2

CO ^ t*™ ON O ^ W M "^ O O OO ^T ^ CN ^C CT\ O r** OO >H LO CS H

O O •"-"

005

• • S 3

- '_ z

03

' '** '** ' "^

- 5 i=

_OJ

xxxxxxx>x>^>>pxpxp5xxxxxx

*^ **^

"o ° si

01 U 1-

S ° <u

UUUUUUUUUUUUUUUUUUUUUU'o^>

uouuuuooouuuuooouuuoouoc

Si -= •:

3 i

^ 3 QJ

"S fl !S

~ CJ -. -«

O ~ "o

s i

M cs t*O ^" *^3 O t^~ OO O\ O •-* C^ f*3 ^" ^O ^Ot^QOO^O^^NfO^I

i? IT "E

z

I

136

E. UHLENHUTH AND HILDA KARNS.

c •_

— O

cfl «•

<u a

•— -

U; rt C

- ° a

o £<

SB

•§

t^ooosOMt— i-iOM

r^-ot^cq

W
nJ

ffl

1
£

M

n)

C/3

•

UJ
to

6

1

s s

"i •i

> >•

""

sea

El?

3 "^

W W W U W U U U U CU OH

W

d C

° S

.. . ^

^sj|§-is§s

OQ CQ

(f)

J _] J J 1-J H-! H-) J iJ J hJ hJ P-J J J

u uuu uuuu uuu uou u
u uu uu uuuuuuuuuu

J iJ J J J H-! J ,4 iJ hj

uuuuuuuouu
uouuuuuuuo

MORPHOLOGY OF SALAMANDER THYROID GLAND.

137

'•*- i r-1

c _ « - o 2 $ § ^

. o u t? o ^ -H ^

rt «• £«•- .0 j- 25 rt x

Q> M *""* "•*• -*-* ^* . M ' N

U — M-

E'-fc-i ^ r !ooo>^O OO r** O w wi w o oO ~O I ^ ^

O "-3 S In H- p-H hHH-l MM MM |CSM

2 (T^ ^

^^~

O ? ir>iOO1OoOMfOOOcCO\OCN^^^OLOO';1'Or^-O*OOO

<-S^

Its

PQ ^ C

J~ I

r-> • r^ *o *O

cfl TU C* CJ O

^T3 T'^'g^lt'l|"|^ll^^
co tfltccftw .OOOCf^OOfK^i

fTi. ,-H —^. . i ^-- ^^ ^J O O CJ O CJ O

^H •— ' ^"~^ U M CJ O CJ O *rt

fH O *C O 'O 'O *O *C " *^

"rt

I

^ *o c^ o ^ o c o N

O r< -r3 /£**' jj.^.-.-.^ •• . - - . ^ __*• .».».» . w

O -7; J5£""'2"""'""K'tt!tBMa503c's ?'?';"'**

M 0.0.-^^ '^^2^43 .^'HJS— .J2^ . .j3J3^5-3"^

jfi ^^O«|-|O|-1'-1'-11-' MNNCN roro

t", ^ * J J X J J J J X J J J J >, J J >, X J J J J J

• • ° • • •••>•.•.•,•.•.• j .; .; .; j

^MiJOTC/3W3U3C/3C/3C/5C/)C/5MC/3C/3C/3C/3C/3t/3U3C/3C/3U3 X X

« ^ i^ >

XJJJJJH-IJJJ-:^-;^-:--:-:-;-!^
lUUUUUUUUUUUUUUUUUUUUU
uuuuuuoouoouooouooooouuuo

138

E. UHLENHUTH AND HILDA KARNS.

^ ' £

•s. "s 1 g e 1 ^ •! S

*3" ^3" O ^J" ^ O ON M i/l *o M O\Osrow CN OO O IJ^ 00 O M

P) loro^t^fON 'S'rO'O '3'^}''<trOM M MOO romo O\O\

<^ Ut_]^ <u r^ .+j •••* c

c/5 oj c

l-glld

u-OxiNOOTl-u-jOOt^OMr^cero LO 01 >o^

S

= 5 S<

Z fc a

IN

^^^^^^^^.^^^^^

t»

o\

00

cu w

ss**s8jg«3Sfaggi

>>« E

MOvMO<NfOK>^Otn0^iONOu-iOOrONir>roro

"II

rO f^5 ^O *^5 OO rO f^5 ^1" ^3" ^~) ^t" ^/1 ^f ^"s ^^ O O\ 00 O CO 00 ON OO

M M

i I . (continued) .

73

\

**? ^? 2 .n 'C M be _o o _o 'C

a^add'Jlj'iD,

H^- S fc<«fcfcfc< . . . »oo^£r

"OrOt!t3'arO'1O'nO,^vO^^ (NCN(NO\

QJflJflJdJQJflJflJO 1^ iJ ,J L.I O\ ON O\ *~*

«

UW W 33333333^^^^-s^^y

^

H

0
cu

5?

w to « co giW co ^X^^^^^^i^r^r^^i 3 3Mn<i

03 03 03 03 03 03 03 "U 'C *O *^3 *O "C *U 'O rti !i i ni ni (""1 rt i

cu

2

0)

W

<

3 cu aj cu cu cu cu

T3T3T3-OT3T3*d'OT3 -a -a §555^^25
ccccccccc c G ceEco3o3e£
rtoSrtoJtartnJrtoS nS 03 ^^.iJiJESJiJi

C/3C/)C/)(/}(/3(/3(/)C/}C/3(/) C/) CD 33333333

wcowcccflwwwwco co co <•<«•«;«:

CO

hHC^MMC)CSOlC<CSM M M M M

•n

C/)

uuuuuuuuuuuuuuuuuuuuouu
uuuuuuuuuuuuuouuuuuuouu

i

S £

too t^-oo ^O M w n ro^^oO t^-r^oo O\O w w ro^io

r*r-t^r*r^oooocooooooooooooooooooo o\O\O\O\O\O\

MORPHOLOGY OF SALAMANDER THYROID GLAND.

139

The difference of the number of follicles is considerable not
only between the thyroids of different individuals, but also
between the two thyroids of the same individual (Table 1 1. a).

TABLE II. a.

DIFFERENCES IN THE NUMBER OF FOLLICLES BETWEEN THE Two THYROIDS OF

THE SAME INDIVIDUAL.

Series Number of Animal.

Left Thyroid.

Right Thyroid.

s-s--.

- '= -:

~ .;

— ~~*

sfi £

MO

:" ^ C

1+4

^ X

u ^

sc ~ r

? •_

=• -^

— ~^ '~

Z. ^

- /. ~

B ?

- -/ ^

y t

nj ^ *•*
^ ^ -**•

^ fc

-I G L*

w *^ ^'

i-1 C-i «

•< C H

< 5?

i.

2.

CCI XXI. a 23

7-1
39

2-9

32

3--'

CCI.XXI.a 24

3-

CCXXXIY. 13

1 1

5.0

19

6.5

4-

( < XXXYI. 10.

34

4-9

37

4.2

5-

CCXXXYI. 12

26

3-3

23

2.8

2. Relation of the number of follicles to the size of the gland:
The differences in the number of follicles are independent of
the differences in the size of the glands. The number of follicles
is not determined by the size of the gland. This fact may be
ascertained by direct observation. It is also evident from com-
parison of the number of follicles with the values of the area of
the thyroids. One thyroid, for instance (No. I in Table 1 1. a)
measured 2.9 cm2., but contained 72 follicles, while the thyroid
of the other side of the same animal contained only 50 follicles,
although it measured 3.2 cm2.

In Table II. b are recorded the number of follicles, the thyroid-
volume 1 and the number of follicles per cm3, of thyroid-volume
for both thyroids of each of two animals. It will be noticed

1 The thyroid-volume was calculated in the following manner. The largest and
the smallest longitudinal optical sections of each gland were outlined and the areas
measured as before. The area of the largest longitudinal optical section was
multiplied by the largest diameter of the smallest longitudinal optical section and
the area of the smallest longitudinal optical section multiplied by the largest
diameter of the largest longitudinal optical section. The two products were
added and the sum was divided over 2.

10

140

E. UHLENHUTH AND HILDA KARNS.

that the left thyroid of animal No. i contained fewer follicles
than the right one, yet the volume of the left gland was larger
than that of the right gland. In animal No. 2 of Table II. & the

TABLE II. b.

DIFFERENCES BETWEEN THE Two THYROIDS OF THE SAME INDIVIDUAL IN RESPECT

TO THE ABSOLUTE NUMBER OF FOLLICLES AND TO THE NUMBER

OF FOLLICLES PER CM'. OF THYROID VOLUME.

Left Thyroid.

Right Thyroid.

•V

•73

. 6

^ *n 'o

.. e

<— h '3

Series Number of Animals.

Number o
Follicles.

Volume o:
'hyroid in c

Number o
Follicles pe
m3. of Thyr
Volume.

Number o
Follicles.

Volume o
'hyicid in c

Number o
Follicles pe
m3. of Thyr
Volume.

t-i

o

o

i. CCXXXIV. 13

ii

7.16

1.5

IQ

n. 'i

1.8

2 CCXXXVI 12

26

4.2

6.2

23

3-2

7-i

larger thyroid contained more follicles, absolutely, than the
right one, but relatively the number of follicles was smaller in
the left one than in the right one.

The variability of the number of follicles of the thyroids of
different animals will become evident from an inspection of
Table I. There are a number of thyroids among the smallest
ones (less than 3 cm2, area), which are composed of over 50
follicles, while among the largest thyroids there is, for instance,
one which measures 13.6 cm2., but contains only 30 follicles,
and another one measuring 8.6 cm2, and containing 24 follicles.

In Table II. c the absolute number of follicles, the thyroid-
volume and the number of follicles per cm3, of thyroid-volume are
recorded. As will be noticed, the variability of the number of
follicles, independent of the size of the gland, is not a merely
apparent one caused by an inexactness of the method of ex-
pressing the size of the thyroid by the area of the largest optical
section; a similar variability is noticeable, if the size of the
thyroid is expressed in volume.

3. Relation of the number of follicles to the growth of the
thyroid: From an inspection of Table I. it will be obvious

MORPHOLOGY OF SALAMANDER THYROID GLAND.

141

that, although the variability of the number of follicles is con-
siderable, it does not show any relation to the growth of the
gland. In a general way the number of follicles remains un-
changed during the entire life of the animal. This circumstance
is born out most clearly by the values for the relative number of
follicles (Table I., column 7). The relative number of follicles
is at first high (80, 42, 62), decreases rapidly with the size of the
gland and in adult animals is found to be between 2 and 5.

TABLE II. c.

VARIABILITY OF THE THYROID GLAND IN RESPECT TO THE ABSOLUTE NUMBER

OF FOLLICLES AND TO THE NUMBER OF FOLLICLES PER

CM3. OF THYROID VOLUME.

L

eft Thyroic

1.

Series Number of Animal.

Number of
Follicles.

Thyroid
Volume in cm3.

Number of
Follicles per
cm3, of Thyroid
Volume.

Approxi-
mate Age
of Animal.

i CCLXXI a 24

39

5-3

7-4

ist year

2 CCLXXI a 25

36

4-7

7-7

3 CCXXXIV. 13. .

ii

7-2

i-5

4 CCXXXVI. 10

34

7-9

4-3

2d year

5 CCXXXVI. 12.

26

4.2

6.2

6 CCXXXIX 10

40

21.7

1.8

3d year

7 CCLXX 7

52

23-7

2.2

Adult

8 CCLXX 8

59

21.3

2.8

The thyroid gland of the spotted salamander does not grow by
multiplication of its follicles, but by an increase in the size of
its individual follicles.

There is, in this species, no definite relation also between the
number of follicles and the development of the gland, as will be
discussed presently.

SIZE AND SHAPE OF THE FOLLICLES.

I. Types of follicles: The size and shape of the follicles
shows an extreme degree of variation ; nevertheless a number of
general types of follicles can be distinguished.

142

E. UHLENHUTH AND HILDA KARNS.

In a general way it may be stated that the shape of the follicles
is the more irregular the larger the follicles are.

(a) The smallest follicles (primary follicles) are usually round
and in most thyroids strictly spherical (Fig. I, a). But in
nearly every thyroid there is one or several large follicles which
are fairly round, possessing no diverticula at all (Fig. I, b and c).

c

•m.

FIG. i. Various types of follicles found in the thyroid gland of Ambyslorna
maculatum. Outlines drawn with the aid of a camera lucida, at a magnification of
X 33 (Zeiss Binocular Dissecting Microscope, Oc. 2, Obj. az).

The cross-lined areas indicate solid follicles.

(&) Another kind of follicles with smooth surface are the

elongate, tube-like follicles of cylindrical shape (Fig. i, d and e),

(c) The simplest type of diverticulated follicle is represented

MORPHOLOGY OF SALAMANDER THYROID GLAND. 143

> K.

FIG. 2. Various types of follicles, found in the thyroid gland of Ambysloma
maculatum. Drawn with the aid of a camera lucida, at a magnification of X 90
(Zeiss Binocular Dissecting Microscope, Oc. 4, Obj. 03). a, side view; b, view of
the base of the same follicle; c, side view; d, view of the base of the same follicle.
f, a pollicle taken from the thyroid of Nolophthalmus torosus (CCXLVI. 2».

144 E- UHLENHUTH AND HILDA KARNS.

by fairly large, round follicles bearing on their surface one or
several bud-like, spherical diverticula of the same size as the
spherical primary follicles (Fig. I, h and i). This type is rare
in the spotted salamander, but is found frequently in the tiger
salamander as will be shown in a subsequent paper.

(d) A very conspicuous type is represented by follicles of
conical or pyramidal shape (Fig. i, j, k, I and Fig. 2, a and &).
The apex tapers into a point; from the base project two or more
finger-like diverticula. In the simplest case these follicles consist
of two component follicles communicating at the apex by a
small hole and diverging towards the base. But usually the
distance along which the component follicles are fused is much
more considerable; a conical follicle results, which, at the base,
is split into two finger-like processes. In many thyroids,
especially noticeably so in the thyroids of sex-mature adults, a
dozen or mpre finger-like diverticula may project from the base,
while the apex is still single and pointed (Fig. i, k and Fig. 2,
a and b).

(e) Another type of follicles, especially frequent in animals
metamorphosed for some time and in adults, is represented by
follicles bearing so many diverticula that their shape has become
completely irregular (Fig. i, m and n, Fig. 2, c, d, e and/).1

(/) While in each of the previous types one major follicle is
recognizable, bearing on its surface a varying number of smaller
follicles, there is another type, the "composite follicle," which
consists of two or more equally large major component follicles
(Fig. i, o to s, Fig. 2, g and h). The component follicles may
communicate with one another along a considerable area of
their circumference (Fig. i, o and s, Fig. 2, g and h), or communi-
cation may be established merely by a tiny hole (Fig. i, p, q
and r). Particularly interesting are those follicles whose com-
ponent follicles are in close apposition with one another along a
considerable portion of their entire circumference, yet communi-
cate only by a very small hole (Fig. i, p) or by a short stalk.
Each of the component follicles may possess a relatively smooth
surface or may bear, in its turn, numerous smaller diverticula

1 /has been taken from the thyroid of Notophthalmus torosus, but is representative
for many follicles of the spotted salamander.

MORPHOLOGY OF SALAMANDER THYROID GLAND. 145

on its surface. In the latter case extreme irregularity and
complexity of the follicle results.

(g) The smaller follicles of each type may be without a lumen,
consisting of a solid mass of cells. Solid follicles are found
more often among the spherical primary follicles than among
the other types. In the large follicles one or several diverticula
may be found solid (Fig. i, h).

(h) Not infrequent are flat follicles; they still contain some
colloid, but the lumen is very small, slit-like.

2. General arrangement of the follicles: Sometimes no definite
design can be found in the arrangement of the follicles. But
in most thyroids the follicles exhibit very definite relations.

The follicles are frequently arranged in perfect rows which
follow the course of the main vessel (external jugular vein);
the original columnar arrangement which the thyroid shows
before the differentiation into follicles may be preserved thus in
the thyroid of the adult (Fig. 3).

FIG. 3. Diagram of an entire thyroid gland (Zeiss Binocular Dissecting
Microscope, Oc. 2, Obj. 02; magnified X 33), showing the arrangement of the
follicles in rows parallel to the external jugular vein.

FIG. 4. Diagram of a cross section through a thyroid gland (Zeiss Binocular
Dissecting Microscope, Oc. 4, Obj. 03; magnified X 90), showing the arrangement
of the follicles. White, follicles; cross lined, connective tissue; black, large veins.

The follicles are frequently so arranged, within the cross-
section of the organ, that each follicle not only borders, with its
central extremity, on one of the large vessels, but also comes to
lie, with its peripheral surface, within the periphery of the
whole organ (Fig. 4).

146 E. UHLENHUTH AND HILDA KARNS.

At each (caudate and craniate) end of the thyroid is usually
one particularly large, fairly round follicle. The small spherical
follicles — which in A. maculatum are always scarce — are usually
located in the periphery of the organ (in A. tigrinum some
thyroids contain as many as 25 per cent, extremely small follicles;
here they are often crowded together into nests surrounded on all
sides by larger follicles).

Characteristic is the location of the large, finger-like follicles.
The apex of these follicles is always located in the center of the
cross-section of the gland (Fig. 4). The finger-like diverticula
point towards the periphery and their free ends come to lie,
with their entire surface, within the periphery of the gland.

3. Relation of size and shape of follicles to growth and develop-
ment of the thyroid: While the number of follicles bears no
relation to either growth or development of the thyroid, the
size and shape of the follicles is distinctly correlated to the
size of the gland and, therefore, indirectly to the developmental
stage of the gland.

The development of the thyroid of the spotted salamander,
during the very early stages, up to the completion of the limb
development, is very similar to the development of the thyroid
in the marble salamander, as described in a previous paper (n).
If the entire organ is removed and examined in fresh condition
under the microscope, it will be seen that it consists, at first,
of a solid mass of cells, devoid of any orderly arrangement of
the cells. Soon this mass of cells assumes a columnar shape.
The column is at first a continuous mass of epithelial cells.
Gradually it becomes differentiated into small spherical groups
of epithelial cells and finally each of these groups develops a
lumen in its center, and differentiates into a primary follicle.
At that stage maceration in nitric acid was attempted, but was
unsuccessful, in as much as separation of the primary follicles
was impossible. Microscopical examination shows that the
connective tissue has not yet invaded the epithelial anlage;
the follicles are not yet separated by connective tissue, but are
in perfect epithelial continuity with each other. The youngest
animal in which isolation of the follicles was partly successful
was animal No. 9 in Table J., a larva possessing nearly the

MORPHOLOGY OF SALAMANDER THYROID C.LAMD. 147

Tx

complete number of toes. The right thyroid of this animal was
isolated into 38 portions. Of these 29 were recognized to be
small individual spherical globules; although, as microscopical
examination of the left thyroid showed, these globules contained
a distinct lumen, the contents must have been different from the
colloid of later stages, as it did not turn yellow in the nitric acid.
Eight other portions of the same thyroid were of irregular shape
and consisted probably partly of primary follicles and partly of
solid cell-groups, all of which were connected directly by epi-
thelial tissue. The last portion was quite large constituting a
considerable part of the entire organ and being partly columnar
in shape, representing in its entirety a solid mass of epithelial
cells (Fig. 5, A).

The youngest larvae in which complete isolation of the follicles
is possible measure about 20 mm. body-length (No. 14 in Table
I.) and are in the possession of completely developed toes. In
the thyroids of these animals only a small number of spherical
primary follicles is present (about 10 to 15 per cent.), the other
follicles vary in size up to five times the size of a primary follicle
and are slightly irregular in outline (Fig. 5, B).

By the time metamorphosis approaches (Fig. 5, C) the smallest
follicles are still of the size of primary follicles, they are almost
invariably spherical or at least roundish and their relative
number is unchanged (10 to 15 per cent.). The largest follicles
have not only increased considerably in size, but have become
irregular in shape; among them may be represented all the
types enumerated previously. Between these two extreme types
of follicles transitional forms are found.

About at the time of the first skin shedding flat follicles
containing almost no colloid and corresponding to the collapsed
follicles found regularly in A. opacum after the first skin shedding
are conspicuous; an exact study of that condition has not been
possible with the maceration method.

The follicular pattern as found in the late larval stages remains
almost unchanged during the entire life of the individual. In
the adult animal the largest follicles are still larger and more
irregular. Among the smallest follicles there may be found, in
the first and second year, still one or several small, spherical

148

E. UHLENHUTH AND HILDA KARNS.

o

C D

FIG. 5. Four thyroid glands of different developmental stages of A. maculatum,
showing the changes of the follicles in size and shape, a, outlines of the largest
optical section of each thyroid gland; below the outlines of representative follicles
of each gland. Cross-lined areas indicate a solid cell mass. Outlines drawn with
the aid of a camera lucida, at a magnification of X 33 (Zeiss Binocular Microscope,
Oc. 2, Obj. 02).

A. Thyroid of a larval animal (No. 9 in Table I.) of 44 days of age and 15.1 mm.
body length.

B. Thyroid of a larval animal (No. 14 in Table I.) of 65 days of age and 20.0 mm.
body length.

C. Thyroid of a larval animal (No. 29 in Table I.) of 92 days of age and 32.3 mm .
body length, approaching metamorphosis.

D. Thyroid of an adult, sex-mature male (No. 92 in Table I.) of 86.2 mm. body
length (age unknown).

MORPHOLOGY OF SALAMANDER THYROID GLAND.

149

or at least roundish follicles of the size of a primary follicle.
In the sex-mature adults, however, the smallest follicles are
larger than the primary follicles and may be five or more times
as large as a primary follicle (Fig. 5, D}.

FIG. 6. Thyroid and follicles composing it, of specimen No. 47 in Table I.,
an animal which was examined one half hour after the first skin shedding. The
whole thyroid gland was composed of only five follicles, one especially large and
making the largest part of the entire organ.

Outlines drawn with the aid of a camera lucida, at a magnification of X 33
(Zeiss Binocular Dissecting Microscope, Oc. 2, Obj. 02).

a, outlines of largest optical section of the whole thyroid gland.

b, c, d, e and/, outlines of the follicles composing the gland.

THE MODE OF THYROID GROWTH.

As may be calculated from the values recorded in column 8
of Table I. (area in cm2.) the volume of the adult thyroid gland
of A. maculatum is about 680 times as large as the volume of the
thyroid of a larva at the time at which the first toes are de-
veloped.

This considerable growth might take place in two ways,
either by an increase in the size of the follicles or by an increase
in the number of the follicles.

The diverticulation of the follicles, increasing with the size
of the gland, the presence of follicles which consist of several

150 E. UHLENHUTH AND HILDA KARNS.

component follicles communicating often only by a narrow stalk
and finally the presence of relatively very small follicles in the
thyroids of practically every stage, if taken together, would
suggest that the follicles develop buds which, after attaining a
certain size, separate from the mother follicles. If this assump-
tion were correct, the number of follicles should increase with
the size of the gland and more follicles should be found in the
adult than in the larval gland. In reality the number of follicles
stays unchanged. The growth of the thyroid is effected entirely by
an increase in the size of the follicles .

Nevertheless it could be imagined that the follicles composing
a particular thyroid at a certain stage are not the same follicles
composing the same thyroid at another stage. The follicles
could divide continuously without increasing in number, if just
as many follicles disintegrate as new ones are forming. Although
under certain experimental conditions, as will be described in a
subsequent article, disintegration of many follicles may be
produced, positively no atrophic or degenerating follicles were
ever found in the normal thyroid of A. maculatum. This
observation is fully in accord with the condition found through
histological study in the thyroid of A. opacum.

Although the shape of the follicles suggests strongly a process
of budding and cleavage of larger into smaller follicles, it is
obvious that in reality separation of the bud-like diverticula from
the mother follicles and cleavage of larger follicles into smaller
follicles does not take place.

The presence of bud-like diverticula and of composite follicles
may be explained in two different ways. The diverticula may
form as an outgrowth from the wall of the original follicles or
they may be the result of fusion of smaller follicles with larger
follicles. Histological study of the thyroid of A. opacum showed
convincingly that follicular fusion plays a very important role
in the growth of the thyroid of this animal. Conditions in
A. maculatum are somewhat less favorable to analyzing this
phase of the problem. Several facts, however, suggest, that
in the spotted salamander processes of follicular fusion are
going on during a large part of the animal's life. As pointed
out, small 'spherical follicles are present in the thyroids of almost

MORPHOLOGY OF SALAMANDER THYROID GLAND. 15!

every stage. A complete analysis of the problem hinges around
the question whether the primary follicles found in any particular
stage of an individual thyroid gland are the same as those present
at a later stage or whether new primary follicles are developing
continuously during the growth of the thyroid. The primary
follicles retain practically the same size from an early larval
stage up to the winter of the second year of the animal's life.
If the primary follicles found during the second winter are the
same as those present at the early larval stages, it would mean
that they have not been taking part in the general follicular
growth. It would be difficult to explain this peculiarity.
Neither pressure nor lack of nourishment can be invoked as
the cause of it ; the primary follicles are located almost invariably
at the periphery of the organ and are frequently in close vicinity
of large vessels. Moreover the primary follicles do not by any
means lack the potential ability of growing; on the contrary
they may finally attain a considerable size as may be seen in
the thyroids of completely adult, sex-mature animals.

If the primary follicles seen in the thyroid during the later
stages are not the same individuals as those seen at the earlier
stages, then it would mean that new primary follicles are forming
continuously during the growth of the thyroid, but do not
increase in number, because a corresponding number fuse with
the follicles already present, either while they are still small and
spherical (bud-like diverticula) or after they have reached a
larger size (composite follicles). Two facts were observed, which
would fit in well with this interpretation. As has been mentioned
above, follicles are found frequently, which consist of several
large component follicles communicating with each other some-
times merely by a small hole. Since cleavage of follicles does
positively not occur, this condition may indicate a fusion in its
earliest beginning. Particularly suggestive are such follicles
which are in closest apposition along a considerable surface, but
communicate nevertheless only by a small opening in the adjacent
walls (Fig. I, p). The small spherical follicles present very
frequently a relation to the large follicles, which is very similar
to the relation of the bud-like diverticula and is different from
it only by the lack of communication (Fig. I, /).

152 E. UHLENHUTH AND HILDA KARNS.

The other circumstance of interest in this connection is the
frequent presence of tiny solid cell masses in thyroids of nearly
every stage. These cell masses are difficult to find in the
macerated glands and it may be for that reason that they have
been overlooked in many glands. With the method of macera-
tion a minute analysis of these cell masses is impossible. It
should be pointed out, however, that they are epithelial in
nature, distinctly different from the connective tissue as well as
from the clusters of red blood corpuscles. It is believed that they
correspond to similar cell masses found in histological thyroid
sections of A. opacum and called "reserve cell masses" in a
previous paper (n). Microscopical study of these cell masses in
the marble salamander showed that they are the source for
continuous development of small young follicles. Most probably
the same condition prevails in the spotted salamander; the
reserve cell masses split off, during a large part of the animal's
life, new primary follicles which in turn fuse with the old follicles.
The increase in the size of the smallest follicles as observed in
the glands of adult animals indicates that in these glands develop-
ment of new primary follicles has ceased and no further fusion
of the older primary follicles with the large follicles takes place.

These facts suggest that at least a certain number of diverticula
are the result of fusion of smaller follicles with larger ones.

It was observed that in A. opacum intracellular colloid globules
may enlarge within the cells of the follicular wall, come to lie
free within the epithelial wall and finally fuse with the main
mass of intrafollicular colloid (n). While they lie free in the
epithelial wall they may, sometimes, form a small, bud-like
protrusion on the outside of the follicle. Possibly some of the
very smallest diverticula in A. maculatum are formed by such
intracellular colloid droplets.

It is of interest to note that Heidenhain found separation of
buds from the mother follicles and cleavage of large into small
follicles entirely absent in the thyroid of mammals. Our findings
in the thyroid of A . maculatum corroborate entirely Heidenhain's
observation in the thyroid of mammals. As mentioned above,
Heidenhain distinguishes two types of thyroids, an association
and a dissociation type. The thyroid of Ambystoma maculatum

MORPHOLOGY OF SALAMANDER THYROID GLAND. 153

belongs to the dissociation type. The separation of the follicles,
however, is absolutely complete. The bud-like diverticula of
the large follicles, if they occur at all, are, according to Heiden-
hain, primary follicles not yet isolated. In A. maculatum the
bud-like diverticula cannot be interpreted this way; as indicated
by the perfectly smooth and spherical shape of the primary
follicles isolation takes place at an early stage.

Fusion of follicles isolated previously was not observed by
Heidenhain in the mammalian thyroid gland; it takes place,
according to this author, only between the lumina of follicles
which have not been separated before. As described in a
previous article (n) this kind of fusion is not uncommon in the
marble salamander. In the spotted salamander, it seems, fusion
is accomplished between follicles which have been really separated
before. We are in doubt, however, that in the material and
with the method used by Heidenhain a secondary fusion of
previously separated follicles could be detected, if it did exist.

Epithelial reserve cell masses from which follicles are split off
continuously were not observed by Heidenhain in the mammalian
gland. He considers the small follicles of the cattle thyroid as
the primary result of the general dissociation taking place at
a very early stage. According to his views large numbers of
the small primary follicles remain permanently unchanged, from
the time of their separation till old age. In A. punctatum
these primary follicles are absent in the adult glands indicating
that finally the primary follicles begin to grow too.

EXCEPTIONAL CASES OF THE FOLLICULAR PATTERN.
In A. opacum the follicular pattern shows a very definite
relation to metamorphosis. Two types of follicular design are
characteristic. As the follicles begin fusing with each other
shortly after the completion of the limb-development, the thyroid
consists, some time before metamorphosis takes place, for the
most part of a large, tube-like follicle distended greatly with
colloid. The other characteristic pattern is found just after the
first skin shedding. At this time the large, tube-like follicle
collapses nearly in its entire extent, as the colloid is released
from its lumen, and then forms a nearly solid and continuous
mass of cells.

154 E- UHLENHUTH AND HILDA KARNS.

As will be evident from the facts described above neither of
these conditions was found to be an integral part of the develop-
mental cycle of the thyroid of A. maculatum. Nevertheless we
found two thyroids which recall in a most conspicuous manner
the structure characteristic for the thyroid of A. opacum.

The first case is represented by an animal of series CCLXII.,
animal No. 47 in Table I.; its thyroid was examined one half
hour after the first skin shedding. Only one of the two thyroids
was macerated. It consisted of only 5 follicles (Fig. 6). Small
spherical follicles were missing entirely. Four of the five follicles
were of good size, but did not show anything unusual. The
fifth follicle, however, was excessively large, making up the
largest portion of the entire gland. It consisted of seven com-
ponent follicles of rather smooth outlines; one of these was
partly solid. On the whole this follicle resembled the large bag-
like follicle found normally in the thyroid of A. opacum and
developing there by fusion of the smaller follicles. In A.
maculatum this condition was found only once among 120
animals.

The second case is represented by the thyroid of animal No. 54
in Table I., which was killed one day after the first skin shedding.
This thyroid consisted almost in its entirety of a solid mass of
cells, devoid of any lumen and colloid. In addition u smaller
masses were isolated. But only 4 of them were follicles of
vesicular nature; the other seven were solid cell masses; it was
suspected that they had been broken off, in the process of
dissecting the macerated gland, from the main mass. This
thyroid resembles closely the condition of the thyroid found
normally in A. opacum shortly after the first skin shedding and
called in a previous paper (n) "stage of colloid release." In
A. maculatum it was found only once among 120 animals. But
in another species, A. jeffersonianum, of which only few thyroids
were examined by the maceration method, exactly the same
condition was met with in an animal which was examined one
day after it had been placed on land (CCXXXIII. 7). Animals
of the same series from which the A . maculatum described above
had been taken were examined more than one year after the
first skin shedding; none of them showed this condition of

MORPHOLOGY OF SALAMANDER THYROID GLAND. 155

almost complete continuity of the epithelium and lack of colloid.
It is therefore certain that it was not a peculiar thyroid structure
characteristic for the offspring of one particular female.

It may be said, then, that although as a rule the follicular
pattern of the thyroid of A, maculatum does not show any
striking changes corresponding to certain functional changes,
such changes may develop in rare cases. They resemble the
changes taking place regularly in A. opacum. It should be
noted especially that while the development of one particularly
large tube-like follicle is not an integral phase in the functional
cycle of the thyroid of the spotted salamander, it does occur
occasionally.

On the whole the differences of the follicular pattern of the
thyroid, existing between two so closely related species as A.
maculatum and A . opacum, are very striking. Yet the occurrence
of the two exceptional cases of follicular pattern in A. maculatum
shows, that in the thyroid of this species the ability of developing
the follicular pattern of A. opacum is potentially present.

Comparison of the thyroids of the two species of salamanders
examined furnishes, we believe, further evidence to show that
the plasticity of the endocrine system is very considerable in the
group of amphibians.

VARIABILITY OF THE NUMBER OF FOLLICLES.

Owing to the extreme variability of the number of follicles
it is difficult to determine whether or not a change in the number
of follicles can be effected by any particular experimental pro-
cedure. Nevertheless, as we have seen in experiments to be
described in a subsequent paper (13), it is possible to change so
considerably the number of follicles, that the degree of the
change exceeds the degree of variability. In the present article
we will discuss only such variations the causes of which are not
entirely understood.

As was pointed out, when reference was made to Table I.
the number of follicles varies between 5 and 72, if all of the
animals examined are included. If, however, animals of the
same parentage are compared, the variability is considerably less,
as may be seen from an inspection of Table III. All animals
11

156

E. UHLENHUTH AND HILDA KARNS.

a «. yz .2
tn

10 IT)

l-c (0 c.

41 ^ c! •

•5.. . -O E rt

CN O 1C

-t l-> CN

fe

*

°0

O O fs M MOO

•§

LO o ^f u~j O v
t^ 06 00 O\ Ov

CQ
<J

H

cfl

'S

O

<u

c/;

a

CO
W CO CO

•a -a -o

o o ^ ^ ^3
g S E E E

a .a

•° o

M C

73 -a T3
c c c

1J Qj I/ QJ CJ QJ QJ

o o o o o o o

ed

CO

a3 §
o, •
.-•a

•? o.

o
a

•2

E

0

Ov O

O O t— OO M (M •

o a a « a

MORPHOLOGY OF SALAMANDER THYROID GLAND. 157

a
s
•-

w

< u _) O a; £ ^ S
en uj c

OONCNONON^f r~- 01 ^ ^ O *T ""> O O ^ *o
N r')r*)TJ'TJ-u->r--^i-*fu:)Tfcs 10 CO ^ ^° f*5 N "T^

1-s^ S £

C O ^-. u

I 2

»o •-•*

oe c

•3 M 5

o c c

CQ <" r-

1
'2

<l)

S1

13

u

O O\00

r^iotoo^ OO ^""lO C

MM 1-1 — _ ^! „ -,

OOO

I^-M i^l O>. O O N OiOO O "-! 1-1 >O O "-i N 1"Ooo
OC OOOOO OO C O OOO O O 1^ N

00

"O T3 "O T3
o aj m o
to tn to to

r^rt rt rt rt rt ctf rt'O'n'U'O'O'O'n'C

rt

~ -

— —

15 5

W CO

~ ~
O f^

(/) a, M CS (S N ro -3-O MIHI-IMMI-I>-IN«I-I

M s! ij ij ij j iJ j j i-i j i-i i-j i-j ij j H-i >-i h-i

T3j^^'O"n'U'U"C™*C"n'Ti3"O'C'w~"n"C"n"n

C/) X) C/5 C/3 C/] X X X X

c/: c/; t/: T) x c/3 t/3 c/> c/; x x c^ :/: c/} x x x x

C ^ 1000 OOO

M N IN N N

Cl

M
rt

o

<

J.
id

•/) C

:= rt

o a

O X

^ o

•r. >

ji C

.

— ^

•~ x

rt «

158 E. UHLENHUTH AND HILDA KARNS.

included in this table are the offspring of the same female. The
average number of follicles in this brood is 42, the smallest
number was 24, the largest 72. Among 31 animals there were
seven possessing more than 50 follicles, and only three whose
thyroids contained more than 60 follicles.

In Table IV. are compared two series which were raised under
the same conditions, but were the offspring of two different
females. There is a marked difference in the follicular number
between the two series. The average numbers show a difference
of 15.

In Table V. "lot a" of the series CCLXXI. is compared with
"lot 6" of the same series. Of the thyroids of the 8 animals of
"lot b" seven (87.5 per cent.) contain more than 40 follicles,
three (37.5 per cent.) more than 65 follicles; in "lot a" the
thyroids of only two animals (8.3 per cent.) among 24, contain
more than 65 follicles, 15 (62.5 per cent.) more than 40 follicles.
The average number of follicles in "lot o" is 44, in "lot 6" 54.
Both lots are the offspring of the same female. But in "lot a"
the larvae were isolated, at an early stage, into individual finger-
bowls; individual attention was given to each animal both in
regard to food and water. The larvae of "lot b" were left
together in one large dish; the supply of oxygen was poor and
the food scarce. Only at the end of the larval period they
were placed under conditions similar to those of "lot a." The
follicular number of the animals raised under favorable con-
ditions is lower than the follicular number of those raised under
poor conditions.

In Table VI. "lot 6" of the preceding series is compared with
series CCLXII. As mentioned before, "lot 6" consisted of
animals raised under unfavorable conditions of food and oxygen.
Series CCLXII. was composed of larvae of unknown parentage,
collected in one of the ponds of the vicinity of Woods Hole and
brought to the laboratory only shortly before metamorphosis
These animals were extremely well nourished, of large size and
had lived under optimum conditions. The difference in the
follicular number between this series and "lot b" of the series
CCLXXI. is still greater than the difference between the well
nourished "lot a" of series CLXXI. and the "lot b." In series

MORPHOLOGY OF SALAMANDER THYROID GLAND.

159

K

n

*j9C|uin^

N ro *^i ^D *~* ^* oo ^*

H

O

z

d •

S

<U T3

"— ^

a.

0 g

o

= O °

w

13

1 "e^

a

E

X cC *C to tn co

u
u

'S

O Ci c3 rt cd crj

..dO • "0-OT3

CL,

'c

12 z - -o " ^ "*"

O

U

-CJ

be

cd

=== i s •- -J J-3

^^O"O-^ 'O'O'O

w

r*>

'^ ^^ i r f> r* f I"1
^J [/J C C C

0!

3

•^ .» .* oj _^ cd rt rt

z

X

*o5T3T3'wC/3 J^^J^
^uu^^j C/3C/3C/3
d Oi A co ra ,j^^;

O
H

u
u

W U U ^v H. xxx

S
z

_jj

•UIUI UI

M0 « ^- M r-OOO

o

u

9i

q^Suaq Apog

r- O w « i^- «X)OO

•

S 0

< z

•SABQ UI 3§V

^^ r*~ oo M t^« w 10 t^*
r— ooooOst^ OsOvoo

H O

1

•S9p}1JOJ

fa szqtunft

^^? — ^ 5

t; M
^H

Q

'T3OJV 'jUID

f-J W W

b>> r/} £

jad sapinoa

^* t^» M ^" !"*• \O *~^ ^* "O

t"^ "^ MH

HH < *"

jo jaquinjsi

W ^ §

hJ «

< Is

H | y

aaTsIpnToj

rOCsl^r^'-' O\O OOO

TfNM M t-t M MM

v j

jo aaquin^

s o

B fe

a

p

S ~

*~l~fgla[

N GO Oi O\ f*j OQ ^ f^ O\

0 C

fo ^aquin^i

«ON>^ f) ^ ^ ^ «n

to oi

l|

•^ D

•sABQ ui a8y

LOO«N^O ooo o
OOOxOv O\ *O O O

fcfc

t/i
u

^

•UIUI UI

O f*3 w ro r^- 10 ^~

y

ipSuo'j Xpog

O *^ ^^ w ^O w ^-O

O) C"4 CS CO (*O W *"*5

S

il

w
fe

H

J

U

2

Z

^

*

U

CO

"3

Pfi

.2;

>.

U

ta

r /^

13

* ** .

to

c\

o
>

'e

Ills

U "-" ""> nj

,*

"o

03 >

>, >.J J <;

u.

M

0

S

•o "° c c

>.

73

M O CS CO

2^

w

rt oj ~ ~ _^ _^ ^

o

> > i^ i^ 2 (/3 X C/3

o

JJUWE XX C/3

OH

•jaqumw

M (N M N CO « 10 t

H

M JM

a o o O o

i6o

E. UHLENHUTH AND HILDA KARNS.

w

X

c

Q
Z

o

u

H

Z

w

8

«

W D
O «

Q a
w h
w

3 w

«
w

H

C/3 g

W £

o «
2S
go
£ £

£ D

03
gg

Sg

D E
w H

"o

2

sS

O <

u w

-

e 5
I °
_^ w

§

-

O

H

CCLXX

a

HH

X

U

U

c/;

1

e

°3

U)

rt

•UIUJ UI

Xpog

0
•a

D.CXD.

W W U

6
-o

C3

at

d

-s

T3

ri

c;

O
-o

C

cd

d

-s

T)

c

5

-0

.c

O
od

oo
oo

ui

fo

jo

-B3JIV 5'™°

jad sapjijoj
J°

rOMOOoocM"

or—

/o

ui

•uiui ui

Apog

oooot^o\o\oooo

Os
oo

o\

ooo\oooo\o\ooo

O\OOO NOOO

ro
ro

oo
ro

2
'S

o

(U

•jaqtunjsj

_; _: -a

C W (-<

<) R

p ,

,2 OJ O c •
• O ft . U ft

ft

o

>•

cs

O. ft

oo

>•

cd

>>

a!

.ajcoKaicficnww

^» »^i cd cd cd cd cd cd cd cd

o

N

Cv

u

cti

o

<

"> •« -a

D.D.D.

.

C/} C/)

^

C/3 C/,'

MiooOO\OvOJ>*'-<wro

MORPHOLOGY OF SALAMANDER THYROID GLAND.

161

w

z
o

H

5

o
u .

II

o o
z a.

D >.

a

O w

« H
w

a ta

H o

S 3
a u

kj
J o

O v

O H

' P3

^

S W

J!

s 5
s §
s <

H O
^^

QP<

rJ rs
S B

Q a
z j-
DO

Q 3

w u

w „
- R

Sw
^

- <

_o .

Ix

ta
5<J

w

3
w

o
H

•jaquinN

N -C un O •-• r-» oo *T

c . Odd

0.15 -a-o-o

c c c
O rt rt rt

= = d"3 a a a

"rt

^JJ -od OOO

C

' ~ K C ......

'S

^4<f ^1 ^ S &

>-O<_> art rtrtrt

.... O • -o -o -o

"o

^22 -0 c, ro -r

M

•;—;._; ^ .- J tj ,J i

rt

-C

* s ¥ ST. *o —^ *^ "^ "^

_;

X

W ^* ^* "CCO CCC

X
X

* ** . . ,_• *^

rt "O T3 "cflC/) ^'^^g^i

^ O< Oi f/jf/. 'O'O'O

u
u

^ a a ^^ ^ $ ^

x

<u

•UIUJ in

N-C"-" ^t»-« r^ooc

o
X

qi8uaq Apog

S5S, ^5T ^ ^ "S

•SAEQ u; a§v

t— oo oo Ot~ OvOoo

/^:s

^R U5 ^ £ *

?

•E3JV 'jUID

T3fT c;^T~)TITO T

^~ r^* M T}~ t~*- *o ^^ c^*

s^

loU ^ 1* . 1 1 ^1

HH M M HHM »-HCS>-<

H

p jaqiun^j;

aad sapijjoj

ror^-r-ooro TfMOOt^- O 10 •^p

M

lQ

jo jaquin^j

•s»WttPd

O-^OK.^ -,>/-^-C »^ Ov

O

fo taqiunft

-, PO i»> ^. m ^ •<*•*> <r> <*> "o

"'

LO fi r^» v^ *o ^D *o ^ oo oo O

•SAEQ u; a§v

CSWWMW CS<NCNCS M (N

•uiui ui

O ro ^" f"O O ^" •—* O ^^ ^" ^A *^

X

qiSuai Apog

sO r*** t*1* x*^* N wo r^* ^O N O *~*

U

"C'hfK *** '"'

u

o g_~ ^ O C

to

?• o 5 c c

'§

^ ^ • «> rt rt |^ x

X

"rt

"^^r 71 -0 SO C "° "=

"S:

"-^ Srt>.- .- M "5
^^ s^ o X 1>» ^
.... "u a5" rt rt J

I
ec
rt

^

• • . >*, ^ »^ ^

i-

o

0

^ ' *-? . x x £r '' t-t i-« jr

>

|^| ^T^ fT^ ' ' ^^ L f • • b. (/I • y

^^

M

— — -^ ..33^_; _j rt'o

rt

"aJ IT jj" 4) o) ^22— — *o "3 -o '-^

X

X u £ ^ ^ '^ -'c. "^5 § 5 N rjj •* - "

— r _« _rt oj w -C . __

^ ° 3 '^ '!) «xSox^x^x'§ x =

rt^5rtoo 3^3J3jdU^°j3 ""^

J J J M M i->X X C/3 C/3 X X

uaqmnN

0 * 0 00 t~ .^^« r; « J

1 62 E. UHLENHUTH AND HILDA KARNS.

CCLXII. the average follicular number is only 30, while in
"lot b" it is 54; there is a difference of 24 between the average
follicular numbers of these two series. And it is again the series
reared under better conditions, in which the follicular number
is lower.

As far as we can trust these scanty figures it seems that the
follicular number of the thyroid gland is influenced by hereditary
and by environmental factors. Among the descendants of one
particular female there may be more animals with conspicuously
low or high numbers of follicles than among the descendants of
other females. Animals growing up in a favorable environment
have a lower number of follicles than those growing up in a less
favorable environment.

We do not know what role the number of follicles might
play in the function of the thyroid gland. Nevertheless it is
known that certain pathological conditions are characterized by
high or low numbers of follicles. It is of interest in this con-
nection that the inclination towards a higher or lower number
of follicles may be inherited.

SUMMARY.

1. In the species Ambystoma maculatum, as contrasted to the
species A. opacum, the number of follicles does not show, as a
rule, the characteristic changes related to metamorphosis and
consisting in the formation, by fusion of follicles, of a large,
tube-like or bag-like follicle filled with colloid. Nevertheless a
large, bag-like follicle making the largest part of the entire
organ was found in one animal among 120.

2. In the species A. maculatum the average number of follicles
is 39.7.

3. The number of follicles in each thyroid remains constant
during the entire life of the individual.

4. The growth of the thyroid is effected entirely by an increase
in the size of the follicles.

5. The primary follicles are simple spherical vesicles.

6. As the follicles grow in size the surface becomes markedly
irregular. The large follicles possess diverticula of varying
number, size and shape or may be composed of two or more
equally large and diverticulated component follicles.

MORPHOLOGY OF SALAMANDER THYROID GLAND. 163

7. Separation of the bud-like diverticula from the mother
follicles or cleavage of the larger follicles into smaller follicles
does not occur.

8. The bud-like diverticula and composite follicles are the
result of fusion between adjacent follicles.

9. The source for the continuous formation of new follicles
are probably small masses of epithelial cells found in many
thyroids.

10. Owing to the fusion of the primary follicles, which goes
on at the same rate as the formation of new follicles, the number
of follicles remains constant.

11. In adult, sex-mature animals primary follicles cease to
develop; those present do not fuse with larger follicles, but
themselves begin to enlarge in size.

It is a pleasure to the authors to express their thanks here to
Mrs. Elsie B. Uhlenhuth and Miss Elsie H. Field, who took
care of the rearing and feeding of all of the animals from which
the data reported in this article were taken.

BIBLIOGRAPHY.

1. Heidenhain, M. Ueber verschiedene Typen im Bau der Schilddriise. Verb.

d. anatom. Ges., 1921, p. 141-151 (Anat. Anz., 1921, LIV., Erganzungsheft).

2. Uhlenhuth, E. The Elaboration and Release of the Colloid of the Thyroid.

Proc. Soc. Experim. Biol. and Medic., 1923, XX., 494—496.

3. Uhlenhuth, E. The Essential Role of the Stainable Colloid in the Specific

Function of the Thyroid. Anat. Rec., 1924, XXVII. , 222.

4. Uhlenhuth, E. The Staining Affinity of the Colloid of the Thyroid. Anat.

Rec., 1924, XXVII., 222.

5. Uhlenhuth, E. The Function of the Bensley Cell in the Thyroid of the Sala-

mander, A. opacum. Anat. Rec., 1924, XXVII., 223.

6. Uhlenhuth, E. Elaboration of the Colloid in the Thyroid of the Salamander,

A. opacum. Demonstration of Histological Sections. Anat. Rec., 1924,
XXVII., 227.

7. Uhlenhuth, E. The Growth of the Thyroid and Postbranchial Body of the

Salamander, A. opacum. J. Gen. Physiol., 1924, VI., 597-602.

8. Uhlenhuth, E. Die Kolloidzelle und ihre Funktion in der Schilddriise des

Marmorsalamander&. Zeitschr. f. wiss. Zool., 1925, CXXV., 483-501.

9. Uhlenhuth, E. The Secretion Granules and the Vacucles in the Living Thyroid

Gland. Science, 1925, LXIL, 569-571.

10. Uhlenhuth, E. The Secretion Process in the Living Thyroid Gland. Anat.

Rec., 1926, XXXII., 224.

11. Uhlenhuth, E. Die Morphologic und Physiologic der Salamander-schilddruse.

i. Histologisch-embryologische untersuchung des Sekretionsprozesses in
den vetschiedenen Lebensperioden der Schilddruse des Marmorsalamandeis,
Ambystoma opacum. Roux's Arch., 1927, CIX., 611-749.

1 64 E. UHLENHUTH AND HILDA KARNS.

12. Uhlenhuth, E., and Hilda Karns. The Follicular Pattern of the Thyroid

Gland in Ambystoma maculatum. Anat. Rec., 1927, XXXV., 27.

13. Uhlenhuth, E., and S. Schwartzbach. The Morphology and Physiology of the

Salamander Thyioid. 2. The Anterior Lobe of the Hypophysis as a
Control Mechanism of the Function of the Thyroid Gland. British J.
Experim. Biol., 1927, V., 1-5.

14. Williamson, G. S., and I. H. Pearse. A System of Tubules in Secreting

Epithelia. J. Anat., 1923, LVIL, 193-198.

15. Williamson, G. S., and Pearse. The Structure of the Thyroid Organ in Man.

J. Path, and Bact., 1923, XXVI., 459-469.

16. Williamson, G. S., and I. H. Pearse. The Pathological Classification of

Goiter. J. Path, and Bact., 1925, XXVIII., 361-387.

17. Williamson, G. S. The Applied Anatomy and Physiology of the Thyroid

Apparatus. British J. Surg., 1926, XIII., 466-496.

REACTIONS OF AN ELASMOBRANCH (SQUALUS

SVCKLII) TO VARIATIONS IN THE SALINITY

OF THE SURROUNDING MEDIUM.

J. P. QUIGLEY.

From the Pacific Biological Station, Nanaimo, B. C., and the Department of
Physiology and Pharmacology, University of Alberta, Edmonton, Alberta.

A study of the changes which occur in fish when placed in a
medium other than their natural habitat has attracted the
attention of many investigators. Variations have been made
both in the concentration and in the kinds of salts used and
experiments of this nature have been conducted with teleosts
and elasmobranchs. Our knowledge of the subject is, however,
by no means complete, and a further investigation of certain
phases was thought to be of value.

It is generally appreciated that elasmobranchs are usually less
resistant to changes in salinity than are the teleosts. The
former, therefore, make excellent subjects for an investigation of
certain phases of the problem. Squalus sucklii was selected for
the following investigation. This is a species of elasmobranch
which has heretofore received little attention from physiologists.

In addition to observing the length of time this fish was able
to live in certain solutions differing in amount and kind of
salinity from its natural habitat an investigation was made of
the changes in the gill movements, the heart beat, the amount of
hemoglobin and the weight of the animal when taken from sea
water and from the experimental solutions. A comparison was
also made of the weight of the liver, spleen and pancreas taken
from fish that had died in the various media.

The fish used in this research were taken on a set line from the
Strait of Georgia near Departure Bay, Vancouver Island, and
usually from a depth of 30 yards. The fish were handled very
carefully but some injury always resulted from this method of
capture. However, good specimens usually would live for a
week when kept in cages fastened to a float in sea water despite

165

1 66 J. P. QUIGLEY.

the fact that they were not fed during this time and that the
salinity of this water was lower than that of their natural habitat.
The fish were taken during the months of June, July and August
of 1926 and a total of 219 were obtained suitable for this investi-
gation. The fish were kept in a cage fastened to a float in sea
water and were usually used within a few hours after being
taken from the set line, and always within 15 hours of the time
of capture.

For the purpose of comparison some physical characteristics
are given of the solutions with which this investigation is con-
cerned (Table I.).

METHODS.

In this investigation, the following media were used: (a)
distilled water, (6) tap water, (c) tap water which had been given
a pH of 8.4 by the addition of NaOH solution, (d) NaCl, in
distilled water, (e) NaCl, CaCl2 in distilled water, (/) NaCl,
CaCl2, KC1 in distilled water, (g) NaCl, CaCl2, KC1, MgSO4
in distilled water, (h) sea water to which NaCl, CaCl2, KC1,
MgSC>4 had been added. In each case the quantity of the salt
added was the amount required to give a concentration of the
cation approximately the same as that found in sea water.
When distilled water was used it was freshly made, cooled and
then shaken with air. The water was obtained from a copper
still but this is considered to be of negligible importance. The
various solutions were frequently agitated during the course of
the experiment and the fish were changed to a fresh medium of
the same kind if the experiment continued for more than an
hour.

Since it was desired to have conditions approximate as nearly
as possible those which a fish would encounter if it were to swim
naturally into the solution used, the fish were not "wiped.'1
Before being placed in the solution they were allowed to hang
head downward for one half minute, and then weighed. A
slight amount of bleeding was produced by introducing the
point of a scalpel having a thin blade into the caudal artery at
the base of the caudal fin. Samples of blood were taken for the
determination of the amount of hemoglobin and then the fish
were placed in the desired solution.

REACTIONS TO VARIATIONS IN SALINITY.

I67

w
ij
m

'D

Cti

O

8

Sfl

U

oJ
U

Bj

2

c-S

a; *j *>

M C •-

>. D H-3

H a >-

"3 a;

w

E £ o

<L» D ^

ffi
a

3

i/-, u",

O O O O C C

o
• ee —

o o o o o o o

o o o -^ o

o o M « « -5 «
-a

o <

ro M
to m

M 00

t> q

od o 6

ts

VO M «

q M q\ •*
t^ oc i 06 iA
M O) «-) ro

00 ^f O ro

M oo O O

N M O O

o o c o

-1 OOOOOOOOOOOOOOOO
*O ^D ^3 *O *O *O ^C ^

•o
SH

ft^

Si ?-

§<5 -

- ^. s

M- ^

'

K

—

•c
•o
rt

0

c

3

E

o
o

"a

o
2

S.

o

U
U

168 J. P. QUIGLEY.

It is well recognized that the effectiveness of a stimulus is
partly dependent on the rate at which it is applied, therefore the
sudden immersion of the fish in the solution should result in a
maximum effect. Also, the injuries which the fish received
incident to capture might alter slightly the results of an ex-
periment.

After a variable length of time, the respiratory movements
became slower and weaker. As the paralysis of the respira-
tory movements became more marked some struggling usually
occurred and finally respiration completely failed. When respira-
tion ceased, as was evidenced by cessation of movements of the
spiracle, the time was noted, the fish was removed from the
water, suspended head downward for one half minute as before,
and reweighed. Blood was again taken for the determination
of hemoglobin. The fish was placed on a tray, the heart exposed
and if found beating, as was practically always the case, the
time when the cardiac movements ceased was noted. The
liver, spleen and pancreas were then exposed, freed from adjacent
tissue and weighed.

Observations were made of the respiratory movements of all
the fish used while in the various solutions but a series of thirteen
fish were used in making a more complete study of the effect of
tap water on the nature of the respiratory and cardiac move-
ments. Freshly caught fish were weighed and then fastened to
a wooden frame. The frame was submerged in a tank of sea
water in such a manner that all of the fish, with the exception
of the ventral surface directly over the heart, was covered by
the water. A median incision was made in the exposed area,
the tissue at the sides of the incision was retracted in such a
manner that the heart was exposed, but the entrance of water
to the wound was prevented. By means of a thread, the apex
of the ventricle was attached to a heart lever arranged to write
on a kymograph drum. A second heart lever was attached to
one of the gill slits and arranged to record respiratory move-
ments on the same drum. A record was made of the respira-
tory and cardiac movements of the fish in sea water. This
water was then quickly drawn out of the tank and replaced
by tap water and a second record was made. After the fish

REACTIONS TO VARIATIONS IN SALINITY. 169

had been in tap water for thirty minutes, a third record was
made and then the water was replaced by fresh tap water. A
fourth record was made after the fish had been in the tap water
for fifty-five minutes. After the fish had been in the tap water
for sixty minutes, it was removed from the water and killed by
destruction of the central nervous system.

In determining the normal weight of the liver, spleen and
pancreas 103 fish (34.9 per cent, females) of various sizes were
used. These were fish that died in sea water either shortly
before lifting the set line or soon after being placed in the cages.
The fish were held head downward for one half minute to drain
and then weighed. The abdominal cavity was opened by a
median incision. The liver, spleen and pancreas were each
dissected from the adjacent tissue and immediately weighed.
From the figures so obtained, the weight of each organ per
gram of fish was calculated.

RESULTS AND DISCUSSION.

Toxicity of Media. — As a result of the studies of many investi-
gators [for a review of this subject see Carrey (i, 2), Scott (3, 4),
Macallum (5)] it has been shown that the osmotic pressure of
the blood of a fish is rarely the same as that of its natural habitat.
This is made possible by the relative impermeability of the
integument, alimentary tract mucosa and of the gill membranes.
The kidneys also play an important role in keeping the blood
composition constant. However, when the fish is transferred
to a medium having a different osmotic pressure the imperme-
ability of the gill membranes is reduced and a passage of water
and salts occurs which can not be entirely counter-balanced by
the activity of the kidneys. This leads to a partial equalization
of the osmotic pressures of the solutions separated by the gill
membranes. In general, this equalization occurs more rapidly
and more completely in elasmobranchs than in teleosts. Toxicit y
of a solution appears to some extent to be related to the degree
to which it differs in osmotic pressure from that of the natural
habitat of the fish. Therefore, teleosts will frequently survive
changes in osmotic pressure of the external medium which would
be fatal to elasmobranchs.

1 70 J. P. QUIGLEY.

Distilled water has usually been found to be very toxic for
fish, Ringer (6), Wells (7), but Carrey (2) found that fish can
live in this medium for weeks and he quotes a similar observation
made by Loeb. Garrey (2) has likewise corroborated the findings
of Ringer (6) (8) that the toxicity of distilled water is reduced
by the addition of a small amount of salts; for example, fresh
water is less toxic than distilled water.

Many physiologists Loeb (9), Ringer (6), Garrey (2) have
demonstrated with a number of species of marine and fresh
water teleosts the mutual antagonism and progressive decrease
in toxicity as Ca, then K, and finally Mg ions have been added
in definite proportions to NaCl solutions. The usual explanation
offered is that these ions exert their protective action by reducing
the injurious effect which the individual ions have on the perme-
ability of the gill membranes to salts and water.

In so far as I have been able to ascertain no previous investi-
gation has been made on the effects of physiologically un-
balanced solutions, with the possible exception of fresh water,
on any elasmobranch or the effect of physiologically balanced
solutions on Squalus sucklii.

In determining the effect of unbalanced, partially balanced
and completely balanced solutions on dogfish it was found, that
although the differences in the toxicity of the solutions used
were not very great (Table II.), the results in general agreed
rather well with those obtained by the other investigators.
Using the time required for failure of the respiration as a guide
to the toxicity, the solutions in order of their decreasing toxicity
are Na > distilled water > Na, Ca, K, Mg, > Na, Ca, > Na,
Ca, K. Since the solution containing the Na, Ca, K, and Mg
ions was in substance an artificial sea water, it is difficult to
explain its apparently high toxicity.

Using failure of the heart as a guide to the toxicity, the results
are distilled water > Na > Na, Ca > Na, Ca, K, Mg > Na,
Ca, K. The solution made by adding NaCl, CaCl2, KC1, and
MgSO4 to sea water so that the cations would be present in
approximately twice their normal concentration was a balanced
solution. Therefore it is interesting to note that this solution
as regards its effect on respiratory or cardiac movements was the

REACTIONS TO VARIATIONS IN SALINITY.

TABU-; II.

Solution Used.

Number
of Fish
Used.1

Ave. Duration of
Movements (in
Minutes).

Ave.
Change
in Hemo-

L'Jobin
Gm. per

IOO CC.

Blood.

Ave.
Per-
centage
Change in
Weight.

Respira-
tory.

Car-
diac.

Distilled water. . . .

24
23

4

10

5

7

6

6

73
113.2

87-5
7i
89

125

88.3
53-3

IOO

144.7
128
137.7
138

165

153-3
83-3

-0.613

-0.323
-0.487
— 0.046
-0.035

— 0.0057

+0.108
+0-345

+ I-75
+3-66
+6.7
— 1.09
— 0.132

-2.379

—0.032
+3-46

Tap water

Tap water pH 8.4

NaCl solution

NaCl, CaCh solution. .

NaCl, CaCh, KC1 solu-
tion

NaCl, CaCh, KC1, MKs< >,
solution

Sea water with added salts .

1 Does not apply to determination of hemoglobin.
+ Indicates increase, — indicates decrease.

most toxic solution used. Experiments of Loeb and Wasteney-
(10), Carrey (2) and Portier and Duval (n) also indicate that
a balanced solution may be toxic when the concentration exceeds
a certain limit.

The water referred to as tap water was a ground water caught
in a small private reservoir and supplied to the Pacific Biological
Station. It may be considered as similar to that which fish
would encounter if they were to swim to a point above tide
water in the streams around Departure Bay. This water had
a very low salinity but it is quite possible that the cations Na,
Ca, K and Mg were all present.

Tap water was found to require a longer time to produce
cessation of respiration or cardiac failure than any of the experi-
mental solutions with the exception of the Na, Ca, K solution,
an observation which would strengthen the conclusion that
toxicity is not a simple question of osmotic pressure. Since the
dogfish continued to breath for an average of 113 minutes in
tap water and remained active during most of this time, tlu-y
probably could escape from a fresh water stream even if they
were to swim into it above tide water level. As in the case of
the salmon, Greene (12), the resistance of the dogfish to fresh
water might be increased if it were to ascend the stream very

gradually.

12

172 J. P. QUIGLEY.

Scott (3), however, has shown that dogfish may be permanently
injured by solutions whose osmotic pressure is markedly different
from that of the normal habitat. This investigator has shown
that when specimens of Mustelus were placed in either a hypo-
tonic or hypertonic solution, the freezing point of the blood
returned to normal either slowly or incompletely after being
returned to sea water depending on the length of time the fish
remained in the abnormal solution and the degree to which this
solution differed from normal sea water. A similar conclusion
may be drawn from the results obtained in this investigation for
it was observed that after being left in distilled water or tap
water until a disturbance of the respiration was observed, the
condition of Squalus sucklii did not improve when transferred
to sea water. The specimens of Mustelus used by Scott (3)
in nearly every case died in less than 100 minutes when immersed
in fresh water. Although it is likely that the tap water used in
the experiments here described contained less salts than the
fresh water used by Scott, Squalus sucklii (Table II.) continued
to breath for an average ot 113.2 minutes and cardiac failure
occurred after 144.7 minutes.

Chidester (13) quotes the work of many investigators, from
which it is to be concluded that fish are very sensitive to changes
in the pH of the medium and, moreover, that the toxicity of a
solution is less, the nearer it approaches the pH of the natural
habitat. It was therefore considered of interest to determine
whether the toxicity of tap water (pH 6.9) would be decreased
by making the concentration of the hydrogen ion correspond
with that of the normal sea water. For this purpose, NaOH
was used rather than Na2COs so that the variation in the salinity
would be slight. Contrary to expectations it was found (Table
II.) that the water at pH 8.4 was more toxic than at 6.9. As
will be shown later, the toxicity of abnormal media appears to be
related to a depression of the respiratory center. It is therefore
possible that the fish lived longer in the tap water pH 6.9 because
this solution, being of an acid nature, was less depressant to the
respiratory center than the tap water pH 8.4. It is also possible
that the NaOH exerted a toxic action in some manner other than
through an alteration of the pH. As will be shown later, the

REACTIONS TO VARIATIONS IN SALINITY. 1 73

fish in this solution underwent a marked gain in weight, a fact
which would indicate that the solution was very injurious to
the gill membranes.

It was impossible to demonstrate any relation between the
size of dogfish and their resistance to changes in salinity. Similar
reports have been made for teleosts by Young (14) but Bert
(15) maintained that the larger fish are more resistant. The
fact that most of the large females used by the author were
pregnant may be used in explaining why these fish were not
more resistant.

Bert (16) has reported that when fresh water fish were trans-
ferred to sea water, circulatory changes occurred in the blood
vessels of the gills so that the appearance of these structures
was decidedly changed. In my comparison of the gills of dogfish
that had died in sea water with those of fish that had died in the
experimental solutions no constant variation in the appearance
was observed.

When the dogfish were placed in the experimental solutions
their behavior was in all cases, much the same. They usually
remained quiet, giving no indication that the medium was
obnoxious to them, but they moved energetically when disturbed.
A white slimy material collected in the water and on the bodies
of the fish after they had been in the solution for some time, and
in some cases opisthotonos, most marked in the region of the
neck, was observed. Shortly before and after the respiratory
movements ceased, struggling movements frequently occurred.
The eyes were usually opaque by the time respiration ceased.
The significance of the previously mentioned slime is not known,
but it may indicate that the solutions had an action on the
integument of the fish, perhaps stimulating secretion from the
gland cells of the integument. The formation of slime in
teleosts associated with changes in the environment has been
reported by Young (14) and by Bert (16).

Many pregnant fish were used in the experiments but in no
case did abortion occur as the result of introducing the fish into
the experimental solutions. On opening the abdominal cavity
of these fish, after cessation of the heart beat, the embryos were
in all cases found to be dead. Several pregnant fish taken

174 J' P' QUIGLEY.

directly from sea water were killed by destruction of the central
nervous system. They were allowed to remain undisturbed for
an hour or more and then the embryos were examined. In
most cases they were found to be alive and active. The cause
of death in embryos following the introduction of the mother
into an abnormal medium is not known. Asphyxia and dilution
of the blood probably play minor roles in this phenomenon.

Gill Movements. — A gradual decrease in rate and amplitude of
the respiratory movements followed the introduction of Squalus
sucklii into a medium of abnormal salinity. Until the fish had
been in the medium for about half the time they survived the
change this decrease was usually slight and in some cases was
absent, but during the latter half of the experiment the decrease
in both rate and amplitude was more rapid.

For thirteen fish in sea water, following exposure of the heart,
the respiratory rate varied between 18 and 73 but averaged 41.6
to the minute. The fish gave little or no evidence of injury as
the result of the operation and the results obtained are believed
to approximate rather closely those to be observed in the normal
intact animal. One hour after placing the fish in tap water,
the respiration of five had ceased and the respiratory rate of the
remainder averaged 31.7 per minute.

Lyon (17) reports that the normal respiratory rate of the
shark under experimental conditions varies between 18 and 30
per minute with an average rate of 23. A specimen of Mustelus
examined by Scott (3) had a respiratory rate of 59 per minute
just as the change from sea water to fresh water was being made.
After sixty-seven minutes in fresh water this fish made only 8
very feeble respiratory movements per minute. A specimen of
Squalus acanthias observed by the same investigator breathed
while in sea water at the rate of 14 times per minute. Greene
(12) states that the respiratory rate of the salmon varies between
60 and 1 20 per minute. Parker (18) records 35 to 40 respiratory
movements in the normal specimen of Mustelus resting in sea
water and 50 to 55 per minute in the fish swimming slowly.

Scott (3) noted that the respiratory and the heart rate in
Mustelus were at times equal but they appeared to be little
correlated. Lyon (17) found evidence to indicate that the heart

REACTIONS TO VARIATIONS IN SALINITY. 175

of the sand shark normally takes its rate from the respiration.
In the present investigation it was observed that in Squalus
sucklii the respiratory rate and cardiac rate seemed to be un-
related (except as considered later, when struggling or gill
cleaning movements occurred). They were rarely equal and
usually the respiratory rate was between two and three times as
rapid as the heart rate, except shortly before the death of the
animal when the respiratory rate was markedly decreased while
the heart rate had changed but little.

Hyde (19, 20) has noted slightly convulsive movements of the
gills occurring in the normal skate and suggests that these may
be for the purpose of forcing more water through the gill apertures
and thus removing foreign matter. Lyon (17) observed that
such movements occur in the shark when any foreign body or
solution enters the mouth or when a manipulation of the body
occurs. Scott (3) found that such movements became very
marked in Mustelus some time after being transferred to fresh
water but decreased in intensity and frequency before the death
of the fish. He intimates that the injurious action of the fresh
water on the gill membranes may be the cause of the increase in
intensity and frequency of these movements. The gill cleaning
movements did not occur according to Scott (3) when similar
experiments were carried out on Squalus acanthias.

Movements similar to those described were noted in Squalus
sucklii. These appear to be a normal movement and were
observed while the fish were in sea water or in one of the experi-
mental solutions. A more detailed study was made of these
movements as they occurred when the fish were in tap water.
With Squalus sucklii in tap water the gill cleaning movement
may begin during any phase of the respiratory movement and
corisists of a single or more rarely two or three vigorous move-
ments of the gills; the normal rhythm and amplitude is then
regained. There is no preliminary movement preceding the
gill cleaning movement. In some fish the gill cleaning move-
ments may not be observed for some time but usually they occur
at fairly regular intervals of one or two minutes. With certain
fish, the intervals between these movements decreased to reach
a minimum about one half hour after immersion in tap water

176 J. P. QUIGLEY.

and then decreased in frequency until respiration ceased. This
was not of constant occurrence, for cases were observed where
the movements disappeared after the immersion of the fish in
tap water. The amplitude of the gill cleaning movements was
not increased by immersion of the fish in fresh water. In fact
it could not be definitely stated that immersion of the fish in
fresh water produced any constant variation in these movements.

During the ventricular diastole following the gill cleaning
movement, or more rarely during the second ventricular diastole,
the heart apparently loses tone and marked dilatation occurs,
the heart loses a beat and then quickly resumes the normal tone
and rate. Scott (3) observed a similar change in Mustelus and
suggested that "the cardiac spasm" is an instance of reflex
inhibition of the heart beat due to the cardiac inhibitory center
being stimulated by impulses from the sensory nerves. Reflex
cardio-inhibition does readily occur in fish, Lyon (17), Greene
(12), but because of its rhythmic nature and for other reasons
the explanation of Scott does not appear to apply to the move-
ments which I observed in Squalus sucklii. I am, however,
in accord with his conclusion that the respiratory convulsions
do not produce the peculiar cardiac movements, but that the
two processes have the same cause.

Distinction should be made between the movements described
above and a somewhat similar movement observed in Squalus
sucklii. This consisted of a spasm or series of struggling move-
ments of the skeletal muscles but also involved the heart and
gills. They appeared to occur during any part of the interval
during which the fish were in the solutions and to originate in
any external stimulus; occurring usually when the fish was
touched or the water agitated. Several fish in a tank would
remain quiet for a long interval, then a sudden movement of
one individual would usually result in struggling movements
among the others. These movements were more easily initiated
after the fish had been in the solution for some time and especially
just before and after respiration had ceased, that is, when the
oxygen want might be expected to be greatest. The effect of
these movements on the cardiac and respiratory movements was
more pronounced but otherwise similar to those described as gill
cleaning movements.

REACTIONS TO VARIATIONS IN SALINITY.

177

Although these movements would be likely to have a gill
cleaning effect, this probably was not their primary object.
They appeared to be the response given by the fish to a sensory

Simultaneous Records of Respiration (upper line) and Heart Beat (middle
line).

Fish in sea water, (a) gill cleaning
movement without a tonus wave in the
record of the heart beat, (b) record on rapid

drum.

After fish had been in
tap water for 54 min-
utes, (b) record on rapid
drum.

Record from fish after 6 minutes in tap
water, (a) gill cleaning movements with (b)
tonus waves.

stimulus after an increase in excitability had occurred. This
increase in excitability may have been the result of the accumu-
lation of CO2 in the tissues due to the asphyxia which followed

178 J- P. QUIGLEY.

the introduction of the fish into an abnormal medium. Bert (16)
has described the agitation and irregular respiration occurring
in fresh water fish transferred to sea water and Young (14) has
observed that when fresh water teleosts were placed in a solution
whose salinity differed from that of their natural habitat the
fish were . restless and easily excited before death. Wells (21)
has noted a somewhat similar condition in fish dying in a solution
containing a high concentration of carbon dioxide and a low
concentration of oxygen. This apparently is the same condition
as that described above for the dogfish. No other forms of
irregularities in the respiration of the dogfish were noted.
Cheyne-Stokes rhythm is at times observed in cases of anoxemia,
I therefore expected that it might occur while the fish were
in the experimental solutions but it was not observed. Lyon
(17) has noted this type of respiration occurring in the shark
shortly before death.

Cause of Death of Fish in Experimental Solutions. — Perhaps the
most noticeable change which occurred in the dogfish when placed
in the experimental solutions was a gradual but progressive
slowing of the respiration terminating finally in a complete
cessation of the respiratory movements. Since this change
occurred with all of the solutions used, it does not appear to be
the result of the presence or absence of any of the salts used nor
of the pH of the solution. The rate at which failure of respiration
becomes complete as shown in Table II. does vary with the salt
or combination of salts used. This has previously been con-
sidered.

Many investigators studying the effect of variations in salinity
have noted the injurious effect on the respiration and Lyon (17)
has noted that the respiration ceases before the heart beat when
sharks die in sea water, while Greene (12) from a study of the
salmon concludes that the spawning act produces death through
some more vulnerable channel than the heart and blood vessels.
It may be a rule of more or less general application for fish that
the respiratory apparatus succumbs more readily than the
circulatory.

Assuming that the fish used in our experiments died from
asphyxia, it is interesting to note that the solutions used contained

REACTIONS TO VARIATIONS IN SALINITY. 179

about as much oxygen at the beginning of the experiment as
did the sea water from which the fish were captured. The
solutions were agitated frequently and it is not likely that any
marked decrease in the oxygen content occurred during the
experiment. Moreover Powers (22) has noted that the oxygen
content of the normal medium in which fresh water or marine
fish live can be decreased to a low level (between 1.7-0.4 cc. per
liter) before the fish will exhibit oxygen want.

That anoxemia was an important factor in the death of the
fish placed in experimental solutions has been the conclusion of
many investigators. Indeed, it has been shown by Backman
(23) that if the sea water in which dogfish live, with A = • 1.88°
is diluted to A = - 0.5°, fifteen minutes after being placed in
this solution, the tension of oxygen in the blood is diminished
from 19.1 to 3.7 per cent. This anoxemia was believed to be a
result of injury to the gill membranes and to the change in the
water and salt content of the blood with the attendant change
in size and the destruction of the blood corpuscles, thus decreasing
the gas carrying power of the blood.

However, if fish were dying from asphyxia due to an injury of
the gill membranes while the irritability of the respiratory center
remained normal, hyperemia and dyspnoea might be expected.
They did not occur in Squalus sucklii. The changes observed
correspond more with those to be expected from a depression
of the respiratory center.

As a result of this investigation and the contributions to this
problem made by others, the author is inclined to explain the
death of fish in media having an abnormal salinity to a progressive
depression of the respiratory center. Although a change in the
composition of the surrounding medium frequently alters the
permeability of the limiting membranes of the body, the changes
in the osmotic pressure of the blood and the injury to the gills,
provided they occur, are considered to be only a contributary
cause of death. The action of the abnormal medium on the
respiratory center and perhaps on the motor nerves supplying
the gills appears to be the primary cause of death. Changes in
permeability of the limiting membranes are of importance in
this regard insofar as they alter the activity of the structures

ISO J. P. QUIGLEY.

mentioned by (a) making it possible for salts to diffuse into or
from the inner ear and brain cavity, (6) initiating a reflex over
some cutaneous nervous structure (lateral line?), (c) altering
the composition of the blood supplying these structures. The
relative importance of these factors may, of course, be subject
to variations. According to this concept, teleosts are more
resistant to changes in salinity, either because the changes
mentioned above occur less readily or because the fish are more
resistant to the changes after they do occur. In support of
mechanism (a) we may consider the suggestion of Loeb and
Wasteneys (10) that toxic salts (NaNO3, NaBr, KC1) in the
external medium exert their action in this manner. Further-
more, it is to be expected that a disturbance of equilibrium
might be concomitant with changes in the inner ear. This
phenomenon has been observed shortly before death when
fish were placed in solutions of abnormal salinity, Loeb and
Wasteneys (10) and Young (14), in solutions of glycerine,
Siedlecki (24), and in solutions of low oxygen and high carbon
dioxide content, Wells (21). Maxwell (25) states that in most
selachians the lymph of the vestibule is in free communication
with the exterior sea water through the ductus endolymphaticus,
and that it is reasonable to suppose that the density of the lymph
would be practically equal to that of sea water. Since this
condition is never encountered in teleosts it may be that we have
here an important factor in the different susceptibility of these
two groups of fish to variation in salinity.

Cardiac Movements. — Observations of the heart beat of Sqtialus
sucklii were confined to fish in sea water, tap water and fish that
were removed from one of the other experimental solutions
following cessation of respiration. In general the heart is much
more resistant to changes in the external media than is the
respiration.

The heart rate of fish in sea water, taken shortly after exposure
of the heart varied between 14 and 31 beats per minute with
an average of 18.9.

In the experiments where records were made of the cardiac
movements while the fish were in tap water very little change in
the beat occurred. The fish were removed from the tap water

REACTIONS TO VARIATIONS IN SALINITY. l8l

at the end of an hour, or earlier in those cases where the respira-
tion ceased in less than an hour. No failure of the cardiac
movements occurred during this interval. No irregularities ncr
loss of tone of the heart muscle occurred other than those changes
associated with the gill cleaning and struggling movements.
A slight decrease in the amplitude of the contractions did occur
and the number of beats per minute at the end of the hour
varied between 4 and 28 with an average of 14.6.

Scott (3) records a heart rate of 50 per minute in a pithed
specimen of Mustelus canis immersed in sea water. After the
fish had been in fresh water for an hour the amplitude of the
beat was about the same as at the beginning but the rate was
only one fourth as great. When the respiration failed the heart
beat was forcible and strong but it failed rapidly after this.

This investigator also records that in a pithed specimen of
Squalus acanthias placed in the same medium from which it was
captured (brackish water) the rate was 16 per minute. One
specimen was placed in fresh water and observed for five and one
half hours. At the end of this time, the heart rate was 8 per
minute.

Scott (3) observed Traube-Herring waves in blood pressure
records made from pithed specimens of Mustelus and thought
that they might be the result of the destruction of the spinal
cord. They ceased when the animal was placed in fresh water.
Greene (12) observed somewhat similar waves in the tracings
made from the Chinook salmon but considered them to result
from the rhythmical effects of the respiration on the blood
pressure. Lyon (17) observed something similar to Traube-
Herring waves in blood pressure tracings made from sharks in
sea water.

Rhythmical variations in cardiac tone (tonus waves) were
observed in Squalus sucklii but they were more prevalent while
the fish were in tap water than with fish in sea water. They
appeared to begin and sometimes to end with the gill cleaning
movements. Respiratory waves were not observed in the heart
records. (See figures, page 177.)

Changes in the Amount of Hemoglobin. — We determined the
amount of hemoglobin in 44 fish before and after immersion in

1 82 J- P. QUIGLEY.

the experimental solutions. Meischer's modification of Fleischl's
hsemoglobinometer was employed in making these determina-
tions. This apparatus was calibrated for human blood and
since it was not checked against any other method, the absolute
amount of hemoglobin recorded may not be correct. However,
the same method was employed for all the determinations and
the relationship between them is therefore of value.

The results obtained from normal fish varied between 0.68
and 5.14 grams of hemoglobin per 100 cc. of blood with an
average value of 2.84. We were unable to demonstrate any
definite relationship between sex or weight of fish and the amount
of hemoglobin. In an examination of the blood of the skate,
Harris (26) found that the percentage of HbO2 in the blood of
this animal varied between 3.1 per cent, and 6.2 per cent. He
considers that the average value for the normal skate is between
3.5 per cent, and 3.8 per cent. By the use of Oliver's tintometer,
v. Fleischl's hsemometer or the spectrocolorimetric method of
Rollet he obtained results which were in close agreement.

The average changes in amount of hemoglobin shown in Table
II. are in close agreement with the differences in osmotic pressure
which the fish encountered when transferred to the abnormal
media. They appear to result largely from the dilution or con-
centration of the blood which followed the change in osmotic
pressure of the external medium. However, it has been observed
by Hall, Grey and Lepkovsky (27) that in the menhaden asphyxia
leads to an increase in the hemoglobin value. A similar change
may have been operating in Squalus sucklii.

Changes in Weight. — When the impermeability of the gills is
decreased as the result of introduction of a fish into a solution
having a salinity to which it is unaccustomed, water and salts
would enter or leave the blood and tissue of the fish depending
on the direction of the difference in the osmotic pressure. The
direction and extent of the exchange of water might be indicated
by a study of the weights of fish before and after the change in
the external media had been made. It has frequently been
observed that teleosts gain in weight when in a medium having
a lower osmotic pressure than normal and lose weight in a
hypertonic solution, Greene (12), Portier and Duval (28).

REACTIONS TO VARIATIONS IN SALINITY. 183

However, it was found by Gueylard and Portier (29) that
sticklebacks "unlike any other fish" gain in weight in hyperton'c
and lose weight in hypotonic solutions, while Sumner (30) and
Scott (31) observed that fish frequently gained but later lost in
weight in the same solution and Scott especially noted that
the changes in weight varied greatly for individual fish.

Regarding the average changes in weight of Squalus sucklii
recorded in Table II.: fish in distilled water, tap water and tap
water with a pH of 8.4 made a gain in weight which may be
ascribed to the greater osmotic pressure of the blood than of the
environment. The fish in tap water gained 3.66 per cent., a
result which agrees well with the observation of Scott (3) that
the average gain in water in a number of the tissues of Mustelus
after immersion in fresh water was 3.1 per cent. The fish in
distilled water gained less than those in tap water a result which
may be explained by the fact that the latter fish lived on an
average of 40 minutes longer in the tap water than did the
fish in distilled water. The fish in fresh water pH 8.4 made a
much greater gain in weight than did the others although they
remained in the medium only a few minutes longer than the
fish in distilled water. The large gain may be taken as indicative
of an extensive injury to the gill membranes by the alkaline
solution. The osmotic pressure of the solution containing NaCl,
CaCl2, KC1 and MgSO4 would be about that of sea water.
Practically no change in weight resulted in Squalus placed in
this solution. The osmotic pressure of the solution of NaCl,
CaCl2, KC1; of NaCl, CaCl2; and of NaCl is each progressively
lower than sea water. A loss of weight was observed in fish
placed in each of these solutions. Fish placed in sea water to
which salts had been added gained in weight, although the
osmotic pressure of this solution was greater than that of the
blood.

In general, the length of time fish were able to live in a solution
wras not related to the change in the weight of the fish. These
results as well as those of other investigators previously mentioned
indicate that a study of the changes in weight gives results
which are difficult to interpret. The amount of water and salts
passing through the gill membranes and the amount of these

184 J- P- QUIGLEY.

substances eliminated by the kidneys should largely determine
the change in weight of a fish placed in an abnormal medium.
Marked individual variations of these two factors may explain
the peculiar results noted.

Variation in Weights of Liver, Spleen and Pancreas with the
Weight of the Fish. — The weight per gram of fish of the liver, the
spleen and the pancreas taken from fish that had died in sea
water varies with the weight of the fish. Although individual
fish frequently show considerable variations in the relative
weights of these organs, when an average is taken of the results
from a number of fish with approximately the same body weight,
a rather definite relationship becomes evident (Table III.).

In fish taken dead from sea water the average weight of the
liver per gram of fish rises rather rapidly from 0.066 for fish
weighing between 300 and 999 grams to 0.115 f°r nsn with
weights between 3,000 and 3,999 grams. As larger fish are
considered, the figure tends to remain comparatively constant;
that for fish weighing between 6,000 and 6,999 grams being
0.117.

With the spleen, the greatest weight per gram of fish (0.0040)
occurs in fish with a body weight between 300 and 999 grams.
In larger fish, the figure falls in a rather regular manner until
with fish having a total weight between 6,000 and 6,999 grams,
the weight of the spleen per gram of fish is 0.0015.

The pancreases taken from fish that had died in sea water
show a variation similar to that noted in the case of the spleens.
With fish whose weight lay between 300 and 999 grams, the
weight of the pancreas per gram of fish was 0.00286, and for
fish weighing between 6,000 and 6,999 grams, the figure is
0.00142.

Pregnancy or sex did not appear to result in an alteration in
the weight of the organs and therefore in this connection, these
factors may be neglected. Insufficient data, however, is at
hand to permit the formulation of an opinion as to whether the
increased weight of the fish due to the presence of embryos was
balanced by a compensatory decrease in the weight of the body
of the fish or an increase in the weight of the organs. Since the
weights recorded were obtained from fish captured only during

REACTIONS TO VARIATIONS IN SALINITY.

185

l/l

I— I

(I,

te

O

O

<
o

as
O

o

°g
§§
««

>O Tt

t- 0\

•H 00
M O

c 6

To "^5
10 O

— >-l
0 O

o o
d d

4 »

§ a

0 Q

10 ">

(O C4 00 _

10O Tf *

ro N <N —

M oo ao oo
MOO O

odd :'

«> c* w —

t~- O> *-* O

^C 00 ro O

« « <S M

COO O

000 0

d d d d

E

M

c

i^
§1

4 ^r

"?te. "r- _ ~^-

OC r^ O O 1^

O\ O 00 i- Cv
O M O - O

d d d d d

to o» ^- -• •—

ro O\ •* oo O
O oo tj- ro •^-
es oi oj r-] «
O O O O O
O O O O O

d d d d d

<L»

-j

JS

4->

bf

° g
§?

ro ^

go — —
O c-j - OO T}-

i/-. — -r . y.

M 01 «; -t r-

M M « O O

d d d d d

00 t-: — - —

-1 TJ- « o o\

c— 1-1 >o t— oo

N M M M 1-1
O O O O O

o o o o o
d d d d d

1
>.

-r
pq

§1

tN M

Cl Tl C« — — -« N

t ro — •^t'tOO LO
O >-i •& 1-1 O *f *f LO

Mp.|rt«O\CSM M

f! n ^- ci — — — tl

«", O w m t^ ^f oi C
t^-Onojioiyjro O\
(sro^OrOrOMN N
OOOOOOO 0
OOOOOOO 0

•5

OOOOOOO O

OOOOOOO O

j

W

X

4 g;

£§

M "

«5 M _- -t

O re t^ O t—
^- rC M OO ro

O\ M 00 VC
0 M M 0 O

d c d do

<c n ?> — —

LO >O O\ O O

Tf M Tj- Tf O

00 O 00

0 O C 00

do d do

o g

§t

_^ M

^ 'i "rf <; f-

O "> >O »A> cq

t^ r^ O 10 o\

o q o q q
odd do

O 00 lO O ^ M
1H IH <3 O 0) 0)

^- T)- ro O fO ^o

o o o o o o
o o o o o o

odd d do

§&

2><*

S ^J* e* . •& ua c» n
O M M O •* 00 O\
•O M O Tf O 10 t-«

\O <^ \o t^- r*- O ^J*

O O O O O O O

-T '-2
-T — O — . -* J3 TIM

t^- >o o\ oj 10 O O\ 'S'
O t— ro 1-1 10 O\ rot^.
^J-^-Tfroro ro ^ro

ooooo o c;

OOOOO 0 00

odd d d do

OOOOO 0 00

i

• ,

3 ' '

0 • •

o • •

c/3

to

—

•*

• ^

'^o <2

. c O 2

*O

0 00 -^

0 or —

u

Ti bfl &

TJ 5£ ^

' — Jr M

p

• — s —

c c

Co

0 BJ - _=

Ox- ^i

• •* • -nu -5

: T : -nu "3

tr*

£

to

i00 i-3^^ ^
n : a c m . - 5

9, IP..OJ5J3.2 •?

iiffi-y

^'•5 c. Q.UUUU.2 "

S--rtc«n!rtaJrt*J™

c^QHHZZ^Z cJJ

00 :l^k^ 2
o SC c m . - ~
Q c..2^-r.'r^' '%

Kuu^^U-^ ^
r --^rfojrt ij

Hlll5^-«l

^ '-^ - - ^ U U U 2 "*
? •- rt rt rt rt rt rt - g

(KQt-«H^Z^55 en

cS

M

6

u
O

c

0

o

a
/.

186

J. P. QUIGLEY.

8

•^i

•*<>

8

W
.J
ffl
<H

H

s

G

<u

<u

0
.0

5
M

1

O

CQ

03
fe

0

0

q

Os

OS

O

O.OOI424

H
CO

Os
O
O
0

d

0

o

o

Os

OS

IO

CO

Os

o

0

d

'-r

M

O

r
d

oo
N

-f
H

O

o
d

LO

vO

M

O

o
d

o

o
o

rf-

Os

Os

OS

10

0

M

0

0

d

O.OOI732

10

O
q

d

O.OOI7O1

0.00140'

o

0

0

Os
Os

q\

DO
O

O

d

CN

oo

1— I

o

o

d

00 'O
0 0

o o
d d

O.OOI5O1

-ooo'z

Os

Os

a

CN

°Os
OS

h-4

C

o

002II2

o o o

OS SO I"}
CN (M CN

O O O

o o o

s.

0
O

c—

0

X

O
0

d

0

o

0

o o

o

o

i
o
o

h- 1

Os
Os
OS

to

CN
IN

O

o
d

o

M

0

q
d

O.OO2662

0.00242'

b
o

d

0

o

o

OS

Os

en c*J

i-O OO

CN CN

O O

o o

d d

';

CN

O

o
d

0

ro
O

O

d

0.00244'

0.00227'

0

o

1*3

Cs
Os
Os

>o

00
CN
O

O

to

"-.

o
o

oo

-r

CN

O
O

0 0
0 0

"os
Cs

CN

o

0

o

n
O
O

to

CN

0

0

0

o

o

o o

0

o

o

G

T3

OJ

Q

4-)

rt

£
£

£
ta

3

03

M

C

U!

U
rt
U

U

(rt

c

'fj

"rt

•a
cu
"C

rt
u

i

%

NaCl, CaCh, KCI solution. . .

NaCl, CaCh solution.

rf •

J-,

a.

1
rt

ti

Distilled H2O. .

Tap water
Tap water pH 8
NaCl solution. .

rt

Sf

Q

§

9

rt

Os

Cs
O

o
o

CO

CO
(43

a
+j

Os

Os .
ex-;

li

— *~~1

C/3

REACTIONS TO VARIATIONS IN SALINITY. 1 87

the summer months, nothing can be said regarding the occurrence
of seasonal variations in the weights of the organs.

With fish that had died in the experimental solutions, the
weight of the fish after cessation of respiration was used in
calculating the weight per gram of fish for the organs examined.
Although possible sources of error in the method of determining
the weights of the organs per gram of fish may be pointed out,
the methods used in determining these values for fish that had
died in the experimental solutions are so nearly the same, that
a comparison of the figures is certainly of value. In the case of
fish dying in sea water and also in the case of fish dying in the
experimental solutions the results would probably have been
more uniform if a much larger series of fish had been available.
In general the weights of organs taken from the two classes of
fish were in close agreement. Weights of organs taken from
fish that had died in distilled water usually did run slightly
higher than those taken from fish that had died in sea water,
while these latter weights in turn were usually higher than those
obtained from fish that had died in sea water with added salts.

Gueylard (32) working with a fresh water teleost, the stickle-
back, found the normal weight of the liver to be 0.0435 grams per
gram of fish but after a 24-hour sojourn in water containing 20
grams per liter of NaCl, the weight had decreased to 0.0420.

A marine fish such as Squalus might be expected to show an
increase in the weight of the liver after being placed in distilled
water and in tap water but the results obtained show that this
is not the case.

It is the contention of Gueylard that a relationship exists
between the weight of the spleen and the resistance of fish to
changes in salinity. This investigator has found the weight of
the spleen per gram of fish in four species of marine teleosts to
average from 0.0005 to 0.00073, while the spleen of five species of
fresh water teleosts ranges from 0.00065 to 0.0024. Since the
figures for the dogfish which had died in sea water range from
0.00407 to 0.00153, this fish appears to have a weight of spleen
corresponding to that of the fresh water teleosts.

In a study of the stickleback, a fish readily adaptable to
changes in salinity, Gueylard (32) found the weight of the spleen

13

1 88 J. P. QUIGLEY.

per gram of fish to be 0.00535 while the average figure for five
less adaptable species of fresh water teleosts was 0.00065-0.0027.
She has further stated that if the stickleback were placed in
water of its natural habitat to which 20 grams of NaCl per liter
has been added, the weight of the spleen per gram of fish at the
end of 15 minutes was 0.00356 and at the end of two hours
0.00266. Somewhat similar changes were observed when eels
were transferred from fresh water to sea water (33).

Introduction of a marine fish into fresh water might therefore,
be expected to produce an increase in the weight of the spleen.
The fact that this change was not observed with Squalus sucklii
may be taken as an indication that a relationship exists between
the inability of the dogfish to make a change in the weight of
the spleen and its lack of adaptability to changes in salinity.
A second hypothesis is that in Squalus, the change in weight of
the spleen occurs slowly and has not progressed sufficiently before
death takes place to make itself apparent. Neither of these
explanations appears satisfactory. Gueylard (32) also reports
that when sticklebacks were placed in water to which NaCl
had been added, the spleen became flabby and the color which
normally was brownish red had changed to reddish yellow. No
changes in the color or consistency were observed by the author
when the spleens taken from dogfish that had died in sea water
were compared with those taken from the experimental solutions.

It is a pleasure to express my appreciation to the Biological
Board of Canada by whom a portion of the expenses incurred in
this work has been defrayed. I am also grateful to Dr. W. A.
Clemens of the Pacific Biological Station for the many facilities
extended, to Professor A. W. Downs and Professor William
Rowan of the University of Alberta for helpful suggestions, and
to Viola W. Quigley for assistance with the practical work.

SUMMARY.

i. An investigation has been made of the effect on Squalus
sucklii of transference to the following media : (a) distilled water,
(6) tap water, (c) tap water pH 8.4, (d) NaCl solution, (e) NaCl,
CaCl2 solution, (/) NaCl, CaCl2, KC1 solution, (g) NaCl, CaCl2,
KC1, MgSO4 solution, (h) sea water to which NaCl, CaCl2, KC1,
MgSO4 had been added.

REACTIONS TO VARIATIONS IN SALINITY. 189

2. The most toxic solution was sea water with added salts,
the solution of NaCl, CaCl2 and KC1 was the least toxic.

3. Tap water had a relatively low toxicity, but tap water
which had been given the same pH as sea water by the addition
of NaOH was more toxic.

4. Cessation of respiration invariably occurred more readily
than did failure of the heart.

5. It appears likely that respiratory failure is the cause of
death of fish in abnormal media. This is believed to result from
a depression of the respiratory center.

6. No variation which could be ascribed to pregnancy, size or
sex was observed in the duration of respiration or heart beat,
the change in amount of hemoglobin, weight of fish or of organs
while in the experimental solutions.

7. Changes in the amount of hemoglobin closely paralleled
the changes in the osmotic pressure of the external media.

8. An increase in weight usually but not invariably resulted
from introduction of fish into hypotonic media.

9. It could not be shown that a change in the comparative
weight of the liver, spleen or pancreas followed the introduction
of the fish into an abnormal medium.

REFERENCES.

1. Carrey, W. E. BIOL. BULL., 1915, XXVIII., 77-

2. Carrey, W. E. Amer. J. Physiol., 1916, XXXIX., 313.

3. Scott, G. G. Ann. New York Acad. Sc., 1913, XXIII.. i.

4. Scott, G. G. Amer. Naturalist, 1916, L., 641.

5. Macallum, A. B. Physiol. Rev., 1926, VI., 316.

6. Ringer, S. Jour. Physiol., 1883, V., 98.

7. Wells, M. M. BIOL. BULL., 1915, XXIX., 221.

8. Ringer, S. and Buxton, D. W. Jour. Physiol., 1885, VI., 15.4.

9. Loeb, J. Science, 1915, XLL, No. 1064, 757.

10. Loeb, J. and Wasteneys, H. Jour. Biol. Chem., 1915, XXI., 223.

11. Portier, P. and Duval, M. Compt. rend. d. 1'Acad. d. Sci., 1922, CLXXIV.,

1493-

12. Greene, C. W. Physiol. Rev., 1926, VI., 201.

13. Chidester, F. E. Brit. J. Exptl. Biol., 1924- H., 96-

14. Young, R. T. Amer. J. Physiol., 1923, LXIII., 373.

15. Bert, P. Compt. rend. d. 1'Acad. d. Sci., 1883, XCVIL, 133.

16. Bert, P. Compt. rend. d. 1'Acad. d. Sci., 1871, LXIII., 382, 464.

17. Lyon, E. P. J. Gen. Physiol., 1926, VIII., 279.

18. Parker, G. H. Bull. U. S. Bureau Fish., 1909. XXIX., 45.

19. Hyde, I. H. Amer. J. Physiol., 1909, XXIII., 201.

IQO J. P. QUIGLEY.

20. Hyde, I. H. Amer. J. Physiol., 1904, X., 236.

21. Wells, M. M. BIOL. BULL., 1913, XXV., 323.

22. Powers, E. B. Publ. Puget Sound Biol. Sta., 1921, III., i.

23. Backman, E. L. Zentralbl. f. Physiol., 1914, XXVIII., 495.

24. Siedleicki, M. Compt. rend. d. 1'Acad. d. ScL, 1903, CXXVIL, 469.

25. Maxwell, S. S. Jour. Gen. Physiol., 1920, III., 158.

26. Harris, D. F. Jour. Physiol., 1903, XXX., 319.

27. Hall, F. G., Grey, I. E. and Lepkovsky, S. J. Biol. Chem., 1926, LXVIL,

549-

28. Portier, P. and Duval, M. Compt. rend. d. 1'Acad. d. ScL, 1922, CLXXIV.,

1366.

29. Gueylard, F. and Portier, P. Compt. rend. soc. d. Biol., 1917, LXXX., 683.

30. Sumner, F. B. Bull. U. S. Bureau Fish., 1906, XXV., 53.

31. Scott, G. G. BIOL. BULL., 1913, XXV., 121.

32. Gueylard, F. Compt. rend. d. 1'Acad. d. Sci., 1923, CLXXVI., 91 7-

33. Gueylard, F. Compt. rend. d. 1'Acad. d. Sci., 1924, CLXXIX., 1292.

THE LUMINOUS MATERIAL OF MICROSCOLEX
PHOSPHOREUS DUG.

STANISLAW SKOWRON.
From the Biological Laboratory of the University, Cracow, Poland.

In my former paper ('26) I have described the luminescence of
Microscolex phosphoreus and some of its chief characteristics,
which seem to exclude the possibility of infection with luminous
microorganisms in this species. The present note deals with the
luminous material alone, and with its changes during the emission
of light. The changes accompanying the light production have
never been adequately studied in living forms, owing to technical
difficulties or unsuitable material. The results of Hickling ('26)
who investigated the luminous slime of a fish species, Malaco-
cephahis Ixvis, differ in many points from my own observations
on Microscolex.

The luminous slime of Microscolex,1 usually discharged in small
quantities and examined in a drop of tap water to keep it moist,
consists of not very numerous small granules, which are situated
within large, round cells, or set free into the surrounding fluid.
Their round shape may be somewhat modified, owing to the
close contact with other granules, and their diameter although

1 Microscolex phosphoreus, of rather common occurrence in Southern Europe,
has never been as yet reported from Poland. Through the courtesy of Mr. Z.
Wasyliszyn I received some specimens found in one of the coal-mines situated near
Cracow. Visiting this place I observed in one passage not used for over two years,
about 230 m. below the surface, great quantities of Microscolex which in these
special conditions of constant temperature and moisture had propagated very
rapidly. Walking in darkness, hundreds^of luminous points were seen, glowing
brilliantly after every step. This species seems to need a higher temperature and
that explains why it can live in mines, where it was occasionally introduced.

Contrary to the individuals I have examined in Naples and others which were
recently sent to me from Naples, the coal-mine forms do not show the day-night
rhythm in luminescence, a fact probably connected with the constant darkness in
which they are living. The degree of irritability of the nervous system, regulating
the ejaculation of the luminous slime, seems to be the same during the night and day
time, owing to these exceptional conditions.

191

STANISLAW SKOWRON.

very variable does not exceed a few microns. In many cases the
granules form compact aggregates within the cells and show some-
times Brownian movement which does not cease at higher tem-
perature of 80° C. and in 90 per cent, alcohol. They are of
greenish color, strongly refract the light and do not show any
optical activity. Under high magnification one can easily see
that every granule is composed of two substances, one of which
forms the central part, while the second lies around the periphery,
an observation which appears to be similar to those of Forster
('12) on Pholas and of Hickling on Malacocephalus .

The granules seem to be highly resistant to cytolytic reagents
(chloroform, saponin, xylol, etc.) and do not change much after
a long time. In hydrochloric acid, however, they swell to some
extent and the central part of each granule dissolves at once,
while the outer layer can be preserved for a few minutes without
any visible changes. The next step is the slow dissolution of the
latter. The granules are also quickly destroyed by pepsin solu-
tion at a certain pH and about 40° C. but the changes are mark-
edly different from those produced by hydrochloric acid.

The picture of the slime if examined immediately after ejacula-
tion and without any addition of water is very different. Besides
the cells and granules mentioned above, there are great masses of
large cells filled completely with granules, the other cells being
already broken and the granules set free. If now a drop of water
be put near the edge of the cover-glass, it is possible to observe
how in the moment of contact with diffusing water the bright
flash of light is produced and the granules dissolve instantaneously
leaving sometimes but a very transparent sheath similar to the
shadows of red blood corpuscles in hypotonic solution. During
this process many cells are breaking up, but some of them remain
intact and the water passing through the permeable cell mem-
brane dissolves the granules situated in the cell. In this case one
can see afterwards a fine network within the cell which represents
the rest of the dissolved granules. As far as I was able to observe,
the granules dissolving in water look quite as the formerly de-
scribed ones, except that their size is usually greater and it seems
to me that they are but fully developed stages of the same kind
of structures. This is further corroborated by the fact that the

LUMINOUS MATERIAL OF MICROSCOLEX PHOSPIIOREUS.

undeveloped granules are never as numerous in the single cell
as the others.

It is necessary to mention that every cell contains only unde-
veloped or developed granules alone, and in the slime one can
always find both kinds of these cells, sometimes the cells with
undeveloped granules being more numerous than usual but never
as abundant as the others. Studying under the cover-glass the
dissolution of the granules and the luminescence which lasts as
long as granules are present, one must regard the dissolving
granules as the real luminous material, a conclusion which dis-
proves entirely the possibility of infection with luminous bacteria.

In order to find out the necessary conditions for luminescence
I have first tried the effect of osmotic pressure of the surrounding
medium, which, as pointed out by Hickling, plays a chief role in
the diffusion of luminous substances from the granules of Malfico-
cephalus. The experiments were, however, without result, as
the luminous slime glowed in saturated KC1 solution almost as
bright as in distilled or tap water, the granules dissolving rapidly.
Many similar observations showed that the osmotic pressure and
different ions have no effect whatever on dissolution and lumi-
nescence of the luminous material. Glucose solutions gave differ-
ent results. In higher concentrations both glucose and glycerin
suppressed luminescence of the slime, which returned, however,
upon addition of water. It seems that the amount of water
regulates the dissolution of the granules and thereby the lumi-
nescence. Filter paper saturated with luminous slime ceases to
glow in alcohol absolute and the granules remain intact, but the
luminescence returns if some water is added. Taking for granted
the importance of water, it will be easy to explain why the lumi-
nous slime, when ejaculated, glows longer in a moist atmosphere
than in pure water, its quantity being the same in both cases.
Larger drops of slime can be luminescent even in pure water for
a few minutes because they form viscous clusters through which
the water diffuses slowly.

I am not quite certain what prevents the luminescence of the
granules inside the body of the worm. The amount of oxygen is
ruled out of the question because the slime can be made luminous
in water where the oxygen is in such small quantity that colorless

194 STANISLAW SKOWRON.

reduced methylene blue does not turn blue. The amount of
oxygen necessary for luminescence must be very small indeed,
contrary to the observations on Malacocephalus, and without any
doubt smaller than its supply in the ccelomic fluid of the living
animal. Greater quantities are necessary for the ejaculation of
the slime only, which function however is under control of the
nervous system. The amount of water in the secreted fluid is
also sufficient because the slime gives off the light for some time
even in dryness, not as long however as to use up all the luminous
material, and here lies the explanation why the dried worms or
slime are still luminous upon moistening. The process of break-
ing up the cells always occurs when the luminescence proceeds and
the amount of water is not too abundant, because in this case
the dissolution of the granules may take place within the cell
membrane. It seems to me that the changes occurring normally
during or after the ejaculation of the slime in the cells themselves,
probably in their cortical layer, lead to the contact of granules
with surrounding water and this causes their luminescence.

As I have already pointed out, the highly concentrated glucose
solution, alcohol abs., and also pure ether and chloroform stop
the luminescence but the process is reversible if water is added
and the action of these reagents is not too prolonged. Quite
different is the effect of HgCl2, concentrated solution or diluted
to one half with water. The power of luminescence is in this
case irreversibly destroyed. The same effect can be obtained
with alcohol abs. acting for over half an hour.

The granules are luminescent at pH from 3 to u. More acid
or alkaline solution extinguishes the luminescence irreversibly.
It was impossible to show quantitatively the duration of lu-
minescence in relation to different temperatures, owing to the
difficulty in getting exactly the same quantities of slime in all
cases. Nevertheless it was found that the light is fainter and of
shorter duration both in higher (40-50° C.) and lower ( — 5-10°
C.) temperatures.

DISCUSSION.

It seems more than probable that the two substances which
compose the luminous granules are both necessary for lumi-
nescence. We may suppose that these two substances represent

LUMINOUS MATERIAL OF MICROSCOLEX PHOSPHOREUS. 195

luciferin and luciferase separated from each other by a film as
suggested by E. N. Harvey, while discussing the luminescence in
other animals. When the cell breaks up and water is absorbed
by the substances of the granules, the film may be destroyed and
the fluid mixture glows till the whole material is used. It is
possible that this destruction can be accelerated by the small
addition of ether or chloroform to the water because of the
effect of these solutions upon the intensity of luminescence, as I
have formerly described. If the water is present in a very small
quantity, the structure of many granules is not changed and the
power of luminescence can be retained. The same result can be
obtained when strong alcohol is added. Other substances, e.g.,
HgCl2, change chemically or physically the composition of the
luminous material and these changes are irreversible. Osmotic
pressure of the surrounding fluid does not play any role, contrary
to the observations of Hickling on Malacocephalus who did not
see the dissolution of the granules because he thinks that "in the
artificial conditions of experiment the granules never become
exhausted, since conditions automatically become unsuitable
long before exhaustion occurs" owing to the too low pH for
luminescence. In Microscolex the material was suitable enough
to observe the dissolution of the granules and further investiga-
tions may show if this process is of more universal occurrence,
because, as far as I know, it has only been mentioned by E. N.
Harvey in jellyfishes by the action of saponin, sodium gly-
cocholate and of fresh water. Besides, it is possible that different
types of changes in luminous material may be met with, of which
Microscolex and Malacocephalus are but two examples.

I am greatly indebted to Professor E. N. Harvey for criticizing
my results.

BIBLIOGRAPHY.
Fbrster, J.

'12 Uber die Leuchtorgane, etc., von Pholas dactylus. Zeitschr. f. Wiss.

Zool., 109.
Harvey, E. N.

'22 Studies on Bioluminescence. XIV. J. Gen. Physiol., Vol. IV., No. 3.
'24 Recent Advances in Bioluminescence. Physiol. Reviews, Vol. IV.
Hickling, C. F.

'26 A New Type of Luminescence in Fishes. 2. Journ. of the Mar. Biol.

Association, Vol. XIV.
Skowron, S.

'26 On the Luminescence of Microscolex phosphoreus Dug. BIOL. BULL., Vol.
LI.

FUNCTIONLESS MALES IN TWO SPECIES
OF NELROTERUS.1

*

J. T. PATTERSON.

In connection with a proposed cytological investigation of the
Cynipidae, I have found it necessary to make a rather extensive
collection and study of the different species present in the Austin
area. Many new species and varieties have been discovered.
These have been turned over to Dr. A. C. Kinsey for identification
and description. Dr. Kinsey has already included a number of
the Austin species in his important publications on the Cynipidae,
and others will be dealt with in future papers. I wish to take this
opportunity to express my appreciation of his fine cooperation in
this work.

During the course of this study many interesting facts con-
cerning the biology of the Cynipidae have been brought to light.
Not the least among these is the discovery of functionless males
in the (so-called) agamic generation of two species of Neuroterus.
The species concerned are TV. c^itortus (Weld) and N. rileyi
(Bassett). Dr. Kinsey has recently redescribed these species and
their varieties.

Neuroterus contortus has two generations a year, one called
"TV. contortus agamic form contortus," the other "TV. contortus
bisexual form principalis" (Kinsey, '23, pp. 91-92). Here at
Austin this species is restricted to a single locality. Indeed, it is
found in small clumps of sprouts around the base of two trees of
Quercus breviloba. The galls of the agamic generation are found
on the roots of the young shoots, just below the surface of the
soil. The adults of this generation emerge during the last week
of January and the first two weeks of February (1923). The
female cynipid oviposits immediately in the tissues of the newly
formed sprouts, just above the old galls, and usually above the

•

level of the surface of the ground. By the middle of February

1 From the Zoological Laboratory of the University of Texas. Contribution
No. 209.

196

FUNCTIONLESS MALES IN NEUROTERUS. 1 97

the new galls begin to appear and are fully formed by the middle
of March. The adults of the bisexual generation emerge during
the first two weeks in April (April 3-15, 1922). New galls of the
next agamic generation were first recognized on May 19, 1922.

There is an interesting difference in the behavior of the females
of the alternating generations. Those of the agamic generation
react positively to light, while those of the bisexual generation
react negatively. This was determined by tests in the laboratory,
as well as by observations in the field. This difference in be-
havior may account for the relative positions (the agamic gall on
the root, the bisexual on the stem) of the galls of the two genera-
tions.

I have bred out from the galls of the bisexual generation a
total of 576 females and 1,076 males. This apparent dispropor-
tion of males to females is not due to any real difference in sex
ratios, but is to be accounted for on an entirely different basis.
I have found in this, and in certain other species of Cynipidae,
that all of the adults emerging from a single, isolated gall are of the
same sex. This means that the agamic females must be of two
kinds, one female- and the other male-egg-producing. If one
collected galls a majority of which happened to contain adult
wasps of one sex, this would give a larger number of individuals
of that sex in the totals.

From galls of the agamic generation I bred out 231 females and
10 males. It is this interesting fact that forms the basis for
the present note. I suspected from the first that these males
might be functionless, and consequently decided to test out their
sex behavior. On February 5, 1923, I bred out 62 females and
6 males from one lot of galls. As soon as they emerged the males
were isolated and kept in a vial for several hours. Females were
then placed with these males, and the culture kept under observa-
tion for an hour. During this period no matings took place.
The females did not display any sex reactions, but some of the
males did exhibit certain of the "courtship" reactions, such as
approaching the female and vibrating the elevated wings. A
single male made a very weak effort to mate with a female, but
copulation did not occur. The experiment was repeated in the
afternoon, but no attempts to mate were observed.

198 J. T. PATTERSON.

In order further to test out the behavior of these wasps, I
collected a number of galls of the agamic generation and placed
them in the ice box, on February 25, 1923. These were left there
until the bisexual generation emerged in April. On April 14
the galls were taken from the ice box and kept at room tempera-
ture, and on the following day four agamic females came out.
These were then tested with males of the bisexual generation.
Before making the tests, however, the behavior of the males of
the bisexual generation toward the females of that generation was
studied. The normal sex reaction is as follows : As soon as virgin
females are introduced into the dish containing the males, the
latter become exceedingly active, and race rapidly about the dish.
The male then approaches to within two or three millimeters of
the female, stops, raises the wings above the body, and then
rapidly vibrates the elevated wings. If the female responds,
mating follows immediately. During this courtship on the part
of the male, the female assumes a passive attitude, by becoming
quiet, and then lowering the antennae until the tips rest on the
bottom of the dish. She remains perfectly motionless during the
process of mating.

The four agamic females were introduced to a group of newly
emerged males of the bisexual generation. The behavior of these
males in the presence of the agamic females is identical with that
described above. They show the normal sex reactions. On the
other hand, these females do not give the slightest evidence of
response. In fact, they make every effort to escape from the
aggressive males. During the course of an hour about fifty
attempts were made by the males to mate with these females,
but there was not a single successful copulation. It would have
been interesting to study the behavior of these males in the pres-
ence of the sexually active females of the bisexual generation,
but unfortunately no males emerged from the refrigerated ma-
terial.

It is evident from these observations that the agamic females
have entirely lost the mating instinct. It is also clear that the
aberrant males have nearly lost this same instinct and, so far as
reproduction is concerned, are functionless.

The second species in which similar functionless males have

FUNCTIONLESS MALES IN NEUROTERUS. 199

been found is Neuroterus rileyi variety mutatus (Kinsey, 1923, p.
1 1 8). Only the agamic generation of this variety is known. It
produces an enlargement of the stems of Quercus Aluhlenbergii.
From April 4 to n, 1922, I reared 94 females and 3 males from
three galls. The sex behavior of these males is similar to but
weaker than that of the functionless males of contortus.

The presence of a few aberrant males in the agamic generation
of species showing the typical alternation of generations is of
considerable interest, because it indicates that such males repre-
sent a remnant of a more primitive, bisexual condition in the
generation which has now become agamic. As Kinsey points out,
these observations are of considerable importance in connection
with our understanding of the development of agamy and heter-
ogeny in the Cynipidae. I have no doubt that careful observa-
tions will reveal the presence of similar males in the agamic
generation of other species. In most species, however, such
males are evidently not present.

Another point of interest is the one concerning the origin of
these males. Our cytological studies indicate that males of all
species of the Cynipidae are haploid, while the females are diploid.
Males arise from unfertilized eggs that have undergone chromatin
reduction. The females of the bisexual generation come from
eggs which have failed to reduce the number of chromosomes,
while those of the agamic generation develop from fertilized eggs.
I have accumulated evidence which offers a possible explanation
of the fact that the bisexual generation produces an agamic
generation of females. In several species examined, the sexual
females usually do not oviposit until after fertilization takes place.
Eggs laid by these females should be fertilized and consequently
would produce females. The few functionless males appearing in
this generation might arise in one of two ways: either certain un-
impregnated females occasionally lay eggs that develop into males
or else an inseminated female lays a few unfertilized eggs, after
the manner of the queen bee. Such eggs would produce males.

The suggestion that certain males are functionless in the
Cynipidae is not entirely new. Thus Adler ('81) has pointed out
that in the genus "Rhodites" the few males found in certain
species must be superfluous, but so far as I am aware the two cases

2OO J. T. PATTERSON.

cited above are the first to be recorded for the agamic generation
of heterogenous species.

REFERENCES.
Adler, Hermann

'81 Alternating Generations. A Biological Study of Oak Galls and Gall Flies.

Straton's translation. Oxford, 1894.
Kinsey, Alfred C.

'22 Studies of Some New and Described Cynipidae. Indiana University Studies,

Study No. 53.
Kinsey, Alfred C.

'23 The Gall Wasp Genus Neuroterus (Hymenoptera). Indiana University
Studies, Study No. 58.

Vol. LIV.

March, 1928

No. 3.

BIOLOGICAL BULLETIN

SEXES IN THE CYNIPID/E AND MALE-PRODUCING
AND FEMALE-PRODUCING LINES.1

J. T. PATTERSON.

i. INTRODUCTION.

In certain groups of insects there occurs an alternation of
generations in which a bisexual generation alternates with one or
more parthenogenetic or agamic generations. Thus in Aphids
and Phylloxerans such an alternation occurs, and in many of the
gall wasps, or Cynipidae, a similar condition is found. In the
case of Phylloxerans, Morgan has pointed out that in some
species (e.g., P. fallax) a single agamic female lays (probably)
both male-producing and female-producing eggs, while in other
species (P. caryxcaulis} a given agamic female lays but one kind
of egg. Her progeny therefore will be either all females or else
all males. I have found similar conditions to exist among the
Cynipidae, and it is the purpose of this paper to give the essential
facts.

The cytological basis of sex-determination in the Aphids and
Phylloxerans is very well worked out, but the same can not be
said for the Cynipidae. However, the facts already known are
sufficient to indicate that the males of all species are haploid,
while all females are diploid. Males develop from unfertilized
eggs in which the number of chromosomes has been reduced
during maturation. Females develop either from fertilized eggs
or from unfertilized eggs in which reduction has not taken place
(Doncaster).

This means that the primitive species, which usually have a
single bisexual generation a year, develop by facultative partheno-
genesis. In such species all eggs are alike and all undergo the
typical maturation divisions with reduction. If the egg is ferti-

1 Contributions from the Zoological Laboratory, University of Texas, No. 210.
14 201

202 J- T. PATTERSON.

lized it produces a female (diploid), but if unfertilized it develops
into a male (haploid). It is not known whether the unimpreg-
nated female can lay both fertilized and unfertilized eggs, as is
the case in the queen bee, or whether the unimpregnated female
alone lays unfertilized eggs.

In species showing alternating generations the agamic females
are the product of the fertilized eggs of the bisexual generation,
but all of the individuals, both males and females, of this genera-
tion develop from unfertilized eggs laid by agamic females. As
we shall show, in some species the female lays both male-produc-
ing and female-producing eggs, while in other species a given
female lays but one kind of egg. Finally, in some of the more
highly specialized Cynipidas males are never found, and develop-
ment is exclusively by parthenogenesis.

In breeding out various species of Cynipidae for other purposes,
a large number of data on sex ratios has incidentally been ac-
cumulated. From this we may select such cases as have a bearing
on the questions outlined above. The plan has been followed of
collecting single, isolated galls and breeding out all of the indi-
viduals from each gall. In this way it is possible largely to
eliminate the danger of breeding from a gall which has been
produced from the eggs of two or more females.

2. SEXES IN THE PRIMITIVE GENERA.

Two genera of the primitive Cynipidae have been investigated
here — the genera Aulacidea and Aylax. Two species belonging
to the first genus have been bred out. Both of these species in-
fest the wild lettuce, Lactuca ludoviciana. These are Aulacidea
bicolor radialis and A. tumida conjuncta (Kinsey, MS.).2 Bicolor
does not produce a gall but inhabits the pith of the stem. Tumida
produces a distinct gall on the lettuce, usually in the form of a
twisted swelling at the top of the stalk.

I have reared 20 females and 1 5 males of A . bicolor, and since
it is one of the most primitive species, males and females are
expected to appear in about equal numbers.

I have bred out 231 individuals from the galls of A. tumida.

* I am indebted to Dr. A. C. Kinsey for the identification or description of the
species of Cynipidae mentioned in this paper.

SEXES IN THE CYNIPID/E.

203

Of this number 51 only, or about 23 per cent., are males. In
breeding from single isolated galls, it is found that each gall
yields both sexes (Table I.).

TABLE I.

Aulacidea tumida conjuncta.

Galls.

Females.

Males.

Galls.

Females.

Males.

A

ii

6
5

i

i
4

D

2

5
151

2

i
42

B

E

c

Mass

Totals

180

Si

Per cent, of males, 22.51

The genus Aylax is represented by a single species, Aylax
laciniata compacta (Kinsey, MS.). It produces clusters of
galls on the receptacles of the flowers of the rosinweed, Silphium
albiflorum. The galls are formed in the sterile flowers toward the
center of the disc. A total of 405 individuals of this variety have
been bred out. About 45 per cent, of these are males, indicating
that the sexes occur in about equal numbers.

TABLE II.

Aylax laciniata compacta.

Galls.

Females.

Males.

Galls.

Females.

Males.

A

2
2
4

3

5

2

3

0
0
0

3

4

G

2

4
7

2

I
192

o

i
o

2
I
165

B

H

c

/

D

J

E

K

p

Mass

Totals

226

179

Per cent, of males, 44.19

The rearings from isolated galls are shown in Table II. Of the
eleven galls bred from, six yielded individuals of both sexes and
five those of one sex. On account of the numerous parasites, it is
difficult to secure more than a few cynipids from each flower head.
Thus 64 parasites and only 48 cynipids were bred out from the
eleven galls.

The genus Rhodites is generally regarded as one of the primitive,
in that none of its species is known to have an alternating genera-

2O4 J- T. PATTERSON.

tion. However, in some respects it gives evidence of high special-
ization, especially in its concentration on the genus Rosa. Its
interest in this connection is found in the fact that in many
species the males are much fewer in numbers than the females.
In certain species the males do not exceed two per cent., and are,
according to Adler, superfluous. Kinsey ('20) believes that on
the whole most of the species reproduce agamically, "with the
males still existent but not usually abundant enough to fertilize
many, if any, of the females." I have not been able to find any
species of Rhodites at Austin.

3. SEX RATIOS IN SPECIES WITH ALTERNATING GENERATIONS.

Among the most primitive species with alternating generations
are those found in the genus Neuroterus, but there are several
other genera exhibiting the same condition. The few cases given
below are selected with the view of bringing out the main points
concerning the female- and male-producing lines of the agamic
females.

a. Andricus operator austrior form austrior Kinsey.

This variety has been described by Dr. A. C. Kinsey ('22) from
material collected here at Austin. It is a bisexual generation,
and what is in all probability the alternate generation is an acorn
gall on Quercus marylandica. From the acorn galls I have bred
out six females. The bisexual galls occur on Q. schneckii. The
gall is a large, compact mass of wool, within which are found
clusters of seed-like capsules containing the larval cells. It in-
volves the young stems, young leaf clusters, and flower clusters.

A total of 1,235 individuals have been reared. About 46 per
cent, of these are males. The results of the breeding experiments
are shown in Table III. The point of chief interest is the fact
that a large majority of the galls yielded individuals of both
sexes. Thirty-four of the forty-five galls bred from gave cynipids
of mixed sexes, eight yielded females only, and three gave males
only. There is no doubt that the majority of agamic females are
capable of laying both male-producing and female-producing eggs.
However, some of them lay but one kind of egg (Table III.,
galls i-s}.

SEXES IN THE CYNIPID/E.

205

TABLE III.

Andricus operator austrior form auslrior.

Galls.

Females.

Males.

Galls.

Females.

Males.

A

2

25
103
69
16

4

21
II

4
7i
26

S

34
35

2

8

4
8
26
3

2
2

4

4
4
IS
33
8

2

6

2
2
2
8

4
1 1
I

3
i
i

2

4
i

2
I
2

X

2
2

8
18

2
21
15

I?

5

10
2

8
3
5
IS

2

7

21

4
o
o

0

19

i

2

3

2

I
I
I

5

2
I
I
0
0
0
0
0
0
0
0

51

118

138

146

B

F

c

Z

D ....

a

E

&

F ....

c

G

d

H

e

I

/

J

P

K

h

L

i ....

M

;

N

k

O

i

P

o

m

n

R

o

s

P .

T

a

u

r

v

5

W

Mass

Totals

671

564

Per cent, of males, 45.66

b. Andricus maxwelli Bassett.

This species forms polythalamous galls at the ends of twigs of
Quercus stdlata. It is an ideal gall from which to test out the
kind of eggs laid by the cynipid. The galls are scattered and
frequently a single gall will be found on one tree. Under such
conditions it is highly improbable that a single isolated gall would
be the product of the layings of more than one female. The
number of individuals emerging from a single gall varies from one
to as high as 27.

I reared 474 individuals of this species. About forty-two per
cent, of these are males. The results of the breeding tests are
tabulated in Table IV. Sixteen galls yielded males, eighteen
yielded females, and seven gave individuals of both sexes. As
compared with the preceding, maxwelli shows a distinct increase in
the number of agamic females which lay one kind of egg.

2O6

J. T. PATTERSON.

TABLE IV.

Andricus maxwelli.

Galls.

Females.

Males.

Galls.

Females.

Males.

A

4
13
4
ii

3

i

i
i
25

10
20

8

14

6
8

25
8
8
7
4
13

23

14
7
4

i

3

i
o

0

o
o
o
o
o

0

o
o
o
o
o

0

V

6

2

4
3

0

o

0

o
o
o
o

0
0
0
0

o
o
o
o
o
68

o
o
o
o

12

4

i

5
4

22

15
II

4

2

4
9

5

2

3

2

39

B

W

c

X

D

Y

E

z

F

a

G

b

H

c

I

d

J

e

K

f

L

0

M

h

N

i

O

/

P

fe

Q. .

I

R

m
n

5

T

o

U

Mass

Totals

277

197

Per cent, of males, 41.56

c. Andricus maxwelli var.

A distinct variety of the preceding makes a somewhat similar
gall on the twigs of Q. breviloba. This variety has not as yet
been described. It is No. 47 of my collection of Cynipidae. I
have been able to breed out 138 wasps of this species. The results
are shown in Table V. Of the fifteen individual galls bred from,
all but two gave individuals of one sex. This variety, therefore,

TABLE V.

Andricus maxwelli (variety).

Galls.

Females

Males.

Galls.

Females.

Males.

A

4

c

I

o

8

B

2

i

f

o

c

C

8

o

K'

o

It

D

1 1

o

L

o

8

E

•2

o

M

o

F

2

o

N

o

2

G

•3

o

o

o

2

H

o

7

A/ass

7

A1

Totals

AO

O8

Q J

SEXES IN THE CYNIPID^E.

207

shows a further tendency for the agamic female to lay but one
kind of egg.

d. Eumayria floridana texana (Kinsey, MS.).
This species produces a polythalamous gall at the end of roots
of Qtiercus schneckii. Table VI. gives the results of the rearing

TABLE VI.

Eumayria floridana texana.

Galls.

Females.

Males.

Galls.

Females.

Males.

A

124
8
267

0

o

I

88
o
138
149

p

o
o

0

7
406

69
53
6
225
729

B
C
D

G .

H

Ma<^

E

Total^

Total reared, 1135

Per cent, of males, 64.22

tests. Of the 1,135 individuals bred out, 729 or 64 per cent, are
males. Two of the eight isolated galls give both sexes, one
yielding 124 females and a single male, the other, 8 females and
88 males. Each of the other six galls gave individuals of one sex.
This species must have an agamic generation, for it has been ob-
served that the females begin to oviposit in the leaves of the same
tree immediately after emerging.

e. Belenocnema treatce. kinseyi form parallela (Kinsey, MS.).

We may now consider four species in which the individuals
emerging from one gall are always of the same sex. This means
that the agamic female must be either a male-producer or a
female-producer.

The polythalamous galls of the bisexual generation of Belenoc-
nema treatx variety kinseyi occur on the rootlets of the common
live oak, Q. virginiana. The galls of the agamic generation (B.
treattz kinseyi form kinseyi) are monothalamous and are found on
the under side of the leaves of the same oak. The galls of the
bisexual generation invariably yield cynipids of the same sex
(Table VII.). The sexes occur in about equal numbers. Of the
677 individuals reared, 329 are males. This gives about 49 per
cent, of males.

208

J. T. PATTERSON.

TABLE VII.

Belenocnema treatee kinseyi Jorm parallela.

Galls.

Females.

Males.

Galls.

Females.

Males.

A ....

7

o

K

o

7

B ...

o

8

L

c

o

C

o

QT

M

6

o

D

IO

o

N

22

o

E

2

o

O .

8

o

G

7

o

P

2^

o

H

o

16

0

o

/

8

o

Mass

2J.I

2^7

J

o

12

Totals

?48

^2O

Total reared, 677

Per cent, of males, 48.59

/. Neuroterus irregularis albipleurx Kinsey.

This variety of Neuroterus forms galls on the young leaves of
Q. breviloba. It is a typical bisexual generation, but the alternate
or agamic generation is not known. The individual galls always
yield wasps of one sex (Table VIII.). The sexes probably occur in
equal numbers, although but 40 per cent, of the 902 wasps bred

out are males.

TABLE VIII.

Neuroterus irregularis albipleurce.

Galls.

Females.

Males.

Galls.

Females.

Males.

A

o
o

0

5

0
10

4i
o
o

20

19
9
o

9
o

0

8
34

J

18

7

0
0
0

4

2

454
541

o

0

3

12

4

0

o

243
361

B

K

C

L

D

M

E

N

F

O

G

P

H

Mass

I

Totals

Total reared, 902

Per cent, of males, 40.02

g. Neuroterus vernus evanescens Kinsey.

This variety has galls in the form of swellings on the ament
stem of Quercus breviloba. The results of the rearings are seen in
Table IX. Perhaps not enough galls were bred out to show
conclusively that individuals of one sex always emerge from a
single gall. However, in addition to the five galls listed in the
table, my field notes state that mixed sexes have never been seen
emerging from one gall.

SEXES IN THE CYNIPID/E.

2O9

TABLE IX.

Neuroterus vernus evanescens.

Galls.

Females.

Males.

Galls.

Females.

Males.

A

i

o

D

o

4

B

o

c

E

3

o

C

o

4

Mass

44

46

Totals

48

SQ

Per cc

nt. of males, 5

5-14

h. Plagiotrochus candidus (Kinsey MS.}.

This species forms galls on the leaf petiole of Quercus schneckii.
As a rule, a single cynipid emerges from one gall, but whenever
there are two or more, they are of the same sex (Table X.).

TABLE X.

Plagiotrochus candidus.

Galls.

Females.

Males.

Galls.

Females.

Males.

A

2

i

3
i

i

2
O
O

3

0
0
0
0
0
0

i
i

0

J

i
o
i
i

4
i
i
1 1
33

o

2
0
0
0
0
0

72
76

B

K

c

L

D

M

J7

N

p

O

G

P

ji

Mass

I

Totals

Total reared, 109

Per cent, of males, 69.72

i. Neuroterus niger patter soni form patter soni (Kinsey).
In all of the forms considered above the galls are polythala-
mous, or else clusters. It is interesting to know that in a species
producing scattered monothalamous galls it is possible to test out
the agamic female as to the kind of egg she lays. The oppor-
tunity to do this is found in Neuroterus niger pattersoni. Both
generations are known. The galls of the agamic generation are
found as small swellings on the leaves of Q. breviloba, and probably
Q. stellata also, while those of the bisexual generation occur on the
leaves of Q. stellata. The agamic females emerge in March and
immediately oviposit their eggs on the underside of the post
oak leaves. It was noted that once a female began laying her

210

J. T. PATTERSON.

eggs in a given leaf, she never moved to another leaf, but continued
until all of her eggs were laid. This suggested the possibility of
breeding out the bisexual individuals from separate leaves. The
results of these tests are shown in Table XI. Of the 26 leaves
thus tested, 22 gave cynipids of one sex, and four gave mixed sex
(leaves W-Z).

TABLE XI.

Neuroterus niger pattersoni form patter soni.

Leaf.

Females.

Males.

Leaf.

Females.

Males.

A

55
3i
o

14
o

0
0

4
o

4

0
0

o

21

o
o

19
o

21

9
6

0

16
o
ii

19
6
o

o

52
o
o
o
o

0
0

o
9

12

8
I
468
683

o
9
9

10
12

24

4

2

58
i
53

21
465

775

B . . . .

P

C .

O

D

R

E

s

F

r

G

u

H . . .

v

I .

w

J

x

K

Y .

L

z.

M

Mass

N

Totals

Total reared, 1,458

Per cent, of males, 53.15

It is highly probable that the four exceptions are due to the
fact that more than one agamic female laid in each leaf. That
this is so, is indicated by breeding out ten galls recently. These
galls were carefully selected from a region in which the galls were
very scarce. Care was taken to secure a single isolated leaf from
each of ten trees. Each of these ten leaves gave wasps of one
sex (leaves M- V, Table XI . ) . From this it may be concluded that
the agamic female of this species lays but one kind of egg.

4. DISCUSSION.

The species listed above are a few of a rather large number in
which the breeding tests indicate that the agamic females are of
two types. The evidence shows clearly that among heterogenous
species the function of laying one kind of egg by the agamic
female has evolved from a condition in which such a female laid
both male- and female-producing eggs.

The chief point of interest is whether it is possible to find

SEXES IN THE CYNIPID/E. 211

cytological evidence that would be adequate to explain this con-
dition in the Cynipidae. Doncaster ('16) has attempted to solve
the question of the nature of the difference between the female-
producing and male-producing females of the agamic generation
of Neuroterus lenticularis, one of the heterogenous species found
in Europe. He attacked the problem both by breeding experi-
ments and by cytological investigations. He concludes from the
results of his breeding tests that the two types of agamic females
could not be due to a dimorphism of sperms produced by one
male. He suggests two other possibilities: (i) That the two
types of female may be due to two kinds of eggs laid by different
sexual females; (2) or, if each sexual female mates with only one
male, they may be due to two kinds of males which produce
different spermatozoa.

Doncaster found no evidence in the spermatogenesis which
would support the last suggestion. In studying the maturation
of the eggs laid by the sexual females, he did find differences indi-
cating that there might be two classes of eggs. However, he did
not regard these differences sufficient "to correlate them with
the sex-phenomena with any confidence." The question of the
nature of the difference between the two types of agamic
females in the Cynipidse is therefore unanswered, and its solution
awaits further study.

LITERATURE CITED
Adler, Hermann

'81 Alternating Generations. A Biological Study of Oak Galls and Gall Flies.

Straton's translation. Oxford, 1894.
Doncaster, L.

'09 Gametogenesis of the Gall-fly, Neuroterus lenticular is. Parti. Biol. Proc.

Royal Society, Vol. 82.
'10 Gametogenesis of the Gall-fly, Neuroterus lenticularis. Part II. Ibid., Vol.

83-
'16 Gametogenesis and Sex-determination in the Gall-fly, Neuroterus lenticularis.

Part III. Ibid., Vol. 89.
Morgan, T. H.

'09 A Biological and Cytological Study of Sex-determination in Phylloxerans

and Aphids. Jour. Exp. Zool., Vol. ?•
'10 The Chromosomes in the Parthenogenetic and Sexual Eggs of Phylloxerans

and Aphids. Proc. Soc. Exp. Biol. and Med., Vol. 7.
Kinsey, A. C.

'22 Studies of Some New and Described Cynipidae. Indiana Univ. Studies, No.

53-
'23 The Gall Wasp Genus Neuroterus. Ibid., No. 58.

FACTORS INFLUENCING THE RATE OF METABOLISM
OF &SHNA VMBROSA NYMPHS.

MARY HONORA SAYLE.

Introduction 212

Historical 212

Material and Methods 213

Observations 217

A. The Carbon Dioxide Output of Control Insects 217

B. The Effect of Starvation on the Carbon Dioxide Output 219

C. The Effect of Darkness on Carbon Dioxide Output 221

D. Carbon Dioxide Output with Rising Temperature 223

E. The Effect of Low Temperature on Carbon Dioxide Output 226

Summary and Conclusions 228

Literature Cited - 228

INTRODUCTION.

The respiratory exchange of an animal is of special physiological
interest, because it affords a means of securing quantitative
evidence regarding the metabolic processes taking place inside
the living animal. It may be measured by ascertaining either
the oxygen consumption or the carbon dioxide output. At
present, methods for the determination of the latter have been
greatly improved and are especially favorable for work on lower
forms, such as insects. The literature dealing with metabolic
rate as influenced by feeding, starvation and change of tempera-
ture and light, includes a number of papers on the metabolism of
nsects, but little or nothing in the way of direct quantitative
measurements of metabolism in dragon fly nymphs.

HISTORICAL.

Spallanzani (22) and Plateau (18) determined the vital limit
of tolerance to high temperature of many species of insects.
Spallanzani (23), Treviranus (25), von Linden (27), Weinland
(28), and Fink (9) studied the gaseous exchange of various in-
sects. Newport (15) and Sosnowski (21) determined the carbon
dioxide output of insects in various stages of development. In
the classical work of Regnault and Reiset (19), the carbon dioxide

212

METABOLISM OF /ESHNA UMBROSA NYMPHS. 213

output of May-beetles was determined. Biitschli (6) and Vernon
(26) studied the effect of temperature upon the gaseous exchange
of Blatta orientalis fed on different substances. Slowtzoff (20),
Batelli and Stern (2), Parhon (17), and Dirken (7) also observed
the influence of temperature on the respiratory exchange of such
insects as ants, bees, bettles, silkworms, and flies. Many ex-
periments have been performed on the gaseous exchange of the
silkworm. Of these the most outstanding ones are those of
Luciani and Piutte (14), Luciani and Lo Monaco (13), and Farkas
(8).

Loeb and Northrop (12), Bodine and Orr (5), and Northrop (16)
studied the respiratory metabolism of Drosophila cultures.
Bodine (3) (4) studied the water content and the rate of metab-
olism of active and hibernating grasshoppers. Studies on the
influence of light on the metabolic rate were made by Loeb (n)
on the chrysalids of certain butterflies and moths and by Alice
and Stein (i) on the May- fly nymphs.

Perhaps the most useful compilation of data on metabolism
is to be found in Krogh's (10) monograph on "The Respiratory
Exchange of Animals and Man." This work includes the results
of several original experiments on insect pupse, particularly on
chrysalids of Tenebrio molitor.

The purpose of the present study was to determine the effect
of starvation, darkness, and temperature upon the carbon dioxide
output of Mshna umbrosa nymphs. These nymphs are easily
obtained and live successfully under general laboratory condi-
tions. They offer especially desirable material for the study of
the respiration in a truly aquatic form. Acknowledgment is made
to Prof. A. S. Pearse for suggesting the problem and to Prof. L. E.
Noland and Dr. Samuel Lepkovsky for advice and criticism
during the course of the work.

MATERIAL AND METHODS.

The insects studied were the nymphs of JEshna umbrosa E.
Walker, and were collected during September 1925 and 1926 in-
Lake Forest Creek, near Madison, Wis. The material used was
collected at the same time and from the same place and hence
was all of approximately the same age and had been under fairly

214

MARY HONORA SAYLE.

uniform environmental conditions. The creek is a spring-fed,
woodland stream about two feet wide and varying in depth from
two inches to two feet. The stream is choked here and there
with water cress to which the nymphs cling. The pH of the
water on collecting days was around 8.

When the nymphs were not being used for experimentation,
they were kept separately in finger bowls containing lake water,
the pH of which ranged around 8.6. A bit of water plant was
placed in the bowl with the insect, except in the case of insects
subjected to starvation. These were supplied with a pebble on
which to cling. The temperature of the water ranged from 20°
to 22° C. The nymphs were fed pieces of mealworms three times
a week and the water changed an equal number of times. Care-
ful controls were run. When the nymphs were fed and changed
to fresh water, the controls were also fed and changed. The
nymphs used in the study were of about the same age, usually in
the penultimate stage at the start of the experimental work.
Three sets of ten insects each were used in each type of experi-
ment and an effort was made to arrange the insects so that the
weight of each set was about the same.

Fig I

•

The experimental apparatus consisted of a closed system of
chambers and tubes, through which a circulation of air was main-
tained by means of a small air pump operated by an electric

METABOLISM OF ^ESHNA UMBROSA NYMPHS. 215

motor (Fig. i, M and P). The airtight glass jar (/) in which
the animals were placed, rested in a constant temperature bath
(.B), which was heated by an electric coil (C) with thermostat
control (7"). To provide constancy of illumination, light was
supplied, when required in the experiment, by a 75-watt electric
light bulb which was hung three feet above the jar.

At the beginning of each experiment, 300 cc. of physiological
salt solution was placed in the jar. This solution consisted of
7 gr. NaCl, .3 gr. KC1, and .25 gr. CaCl2 per liter of glass distilled
water or a total of 7,550 parts per million of dissolved salts.
The solution had a specific gravity of 1.003 at 24° C. ar>d a pH
near 7.4. Preliminary experiments showed that distilled water,
which would naturally be preferred, was quite unsuitable; the
nymphs died after a five-hour stay in the water. That the
injurious effect of the distilled water was of an osmotic nature
was shown by the fact that the addition to the same water of the
neutral salts mentioned above made it a satisfactory medium.

In each experiment, a set of ten animals was used and each
nymph was placed in a paraffin-dipped wire cage, just large
enough to hold the nymph, so that body movement was reduced
to a great extent. The cages were closed and tied together in
sets of five, with a long thread attached to each set. Experiments
were done at the same time each day and always forty-eight
hours after feeding, so that excreta had been passed. At the
beginning of each experiment, the nymphs were hung above the
liquid in the clamp top jar and the threads issued from under
the jar cover. The insects were placed in the jar before the CO2
was removed from the circuit in order to avoid opening the jar
after the preparatory run.

After the nymphs were thus arranged in the jar, the motor was
started and the air was forced from this jar through a bottle
(D) containing a solution of sodium hydroxide, which absorbed
the carbon dioxide from the circuit and also from the liquid in
the experimental jar. A test bottle (£) containing a solution of
barium hydroxide was then opened into the circuit to determine
whether or not all the carbon dioxide had been removed. Fifteen
minutes were found to be sufficient time to remove the carbon
dioxide. At the end of this time, the clamp top on the experiment

2l6 MARY HONORA SAYLE.

jar was loosened slightly for a moment, releasing the string at-
tached to the cages and allowing them to slip down, immersing
the nymphs in the liquid. Bottles D and E were then cut off
from the circuit by the glass stopcock and the experiment was
actually started. The air in the circuit was bubbled through the
Pyrex test tube (F) containing 25 cc. of N/2O Ba(OH)2 for five
hours. At the end of this period, the clamp top was loosened
momentarily and the cages were drawn up to a point above the
liquid, and then for fifteen minutes more the air was circulated
to remove any carbon dioxide not already absorbed from the
water.

The test tube containing the carbon dioxide precipitated as
barium carbonate was detached from the circuit and carefully
protected from air until titrated. After the precipitate had
settled, the excess of barium hydroxide was titrated with N/2O
HC1, using phenolphthalein as indicator. To check the accuracy
of titration, two aliquot portions (5 cc.) of the clear supernatant
liquid were used and the average of the two, multiplied by five,
was taken as the true value. This amount was then subtracted
from the amount of N/2O HC1 required to neutralize 25 cc. of the
original barium hydroxide solution and the difference was taken
as equivalent to the quantity of carbon dioxide produced by the
organisms during the period of the experiment. This method of
CO2 determination is based upon the assumption that the only
volatile acid produced by the nymphs is carbon dioxide, or that
other acids, if produced, are negligible in quantity.

Since the rates of respiratory metabolism are expressed per
gram of body weight per unit of time, it was necessary to weigh
the insects used in the experiments. Weighing excited the insects
greatly, so the process was performed immediately following
rather than immediately preceding each experiment. Since the
nymphs live in the water and alternately draw water into the
rectum and expel it in the respiratory act, it was necessary to
remove surplus water from the surface of the body and to see that
the nymphs discharged water from the rectum before each weigh-
ing. The excess moisture was removed by placing the nymphs
between pieces of filter paper. Even though the greatest care
was used, it is probable that the amount of moisture remaining
in the rectum varied a little at different weighings.

METABOLISM OF yESHNA UMBROSA NYMPHS.

217

OBSERVATIONS.

A . The Carbon Dioxide Output of Control Insects.
The carbon dioxide output of the nymphs under ordinary
laboratory conditions was taken as representing the standard
output, to which comparison could be made with the data secured

.70
.60

.50

>

v

X

y

\

I92&

-\

\

—

I1Z.

r

X

\

^'

A

^

V

\

Y

^

\

\

L
Q

\.\

V

C

••

\

\
\

E

V

s

a

^

\\

i

1

^

>^

20

2

^

^

a

5;

^

Oo<

/K

We

ion

i

6^

^

s

./50 .22$ .300 .3^0

GRAPH i. Curves show rate of COz output over period of 3 months. Time
interval 10 days. Temperature 22° C. Abscissas, gram body weight per insect.
Ordinates, milligrams COz per gram insect per hour. For further explanation see
description in text.

•

under experimental conditions. The temperature of the water
was uniformly between 20° and 22° C. Two sets of insects were
used each year (1925 and 1926) and the records of their CO2 out-
put are given in Table I., and the average record in Graph i.
15

218

MARY HONORA SAYLE.

TABLE I.

SHOWING THE ACTUAL COa OUTPUT IN MILLIGRAMS PER GRAM BODY WEIGHT

PER HOUR, DURING A PERIOD OF THREE MONTHS AT 22° C.

Time interval — 10 days.

1925.

Weight of i Insect
in Grams.

CCh Output in Milligrams per Gram
Body Weight.

Set i.

Set 2.

Average.

Set i.

Set 2.

Average.

.160

.171

.165

.6630

• 7391

.7010

.185

.183

.184

.6950

•6505

.6727

.220

.214

.217

.6501

.6499

.6500

.250

.238

• 244

.6402

.5989

.6195

.274

.260

.267

.5218

.5076

•5147

•293

.288

.290

.4129

•4333

•4231

.319

.313

.316

•3343

.3101

.3222

•346

-334

•340

.2697

•2454

• 2575

1926.

Weight of i Insect
in Grams.

COz Output in Milligrams per Gram
Body Weight.

Set i.

Set 2.

Average.

Set i.

Set 2.

Average.

.172

.166

.169

•6555

.6900

.6727

.191

• 183

.187

.6968

•6731

.6849

.217

.214

.215

•6336

.6504

.6420

• 237

.230

• 233

• 5737

•5903

.5820

.267

• 255

.261

.5482

• 5545

• 5513

.291

.279

•285

. -4350

.4791

•4570

•328

.301

•314

•3018

•3508

•3263

•359

•331

•345

• 2757

.2884

.2820

The determinations were made every ten days and the data taken
over a period of almost three months. The animals were still
in good condition at the end of this period.

The results show a relatively smooth curve; the rate of CO2
output is higher for animals lighter in weight and decreases
progressively as the animals increase in body weight. Differences
in body weight, especially in nymphs, are closely correlated with
differences in age and one is led to assume that younger indi-
viduals have the higher rate of respiratory output. Smaller
animals are generally more active and are immature and growing.

At 22° C. the saturation point of oxygen at the stated degree
of salinity of the chloride solution is about 8. Since each experi-

METABOLISM OF ^ESHNA UMBROSA NYMPHS.

219

ment was run over a period of five hours, a determination of the
dissolved oxygen content of the water at the beginning and at the
end of one of the experiments was made, using theWinkler method
as described in "Standard Methods for the Examination of Water
and Sewage" (24). The physiological salt solution was found to
contain 6.959 parts per million at the start of the experiment and
7.145 parts after five hours. It was thus shown that there was
plenty of oxygen in the circuit to meet the needs of the insects for
the period of the experiment. The fact that there is more dis-
solved oxygen in the solution at the end of this experiment than
at the beginning is accounted for by the fact that the oxygen of
the air in the circuit previous to starting the experiment dissolved
in the solution during the five hours of circulation.

B. The Effect of Starvation on the Carbon Dioxide Output.

Three sets of nymphs were used in the starvation experiments,
carried on at 22° C. Their carbon dioxide output in a well-fed
condition was determined twice at the beginning of the experi-
ment, and then the animals entered starvation. The water was
changed regularly, but no food was given. Carbon dioxide
determinations were made once a week up to the fifth week. In
each set, one or more animals died during the fifth week, which
necessarily ended the experiment.

An examination of Table II. and Graph 2 shows that the car-
bon dioxide output decreased during the first week, but in each

TABLE II.

SHOWING THE ACTUAL CO2 OUTPUT IN MILLIGRAMS PER GRAM BODY WEIGHT

PER HOUR, DURING STARVATION.

Time interval — 7 days.

Average Weight

Milligrams CO2

of i Insect

Output per Hour per

in Grams.

Average

Gram Body Weight.

Average

Weight.

COa.

Set i.

Set 2.

Set 3-

Set i.

Set 2.

Set 3.

.194
• 195

.181
.183

.185
.190

.188

.4252
•4512

.6077
.5809

•5747
• 5632

.5338

Normal.

• 195

.197

.192

•195

•3333

.4187

.4988

.4169

Starved i week.

.194

.220

.201

.205

.4670

.5200

•5073

.4981

Starved 2 weeks.

.199

• 231

.205

.212

.4208

•3238

•4033

.3826

Starved 3 weeks.

.200

.229

.205

.211

.2200

.2882

.2671

.2584

Starved 4 weeks.

220

MARY HONORA SAYLE.

set the output increased during the second week. The explana-
tion of this increase is questionable. It fnay be that at this stage
the nymph draws heavily on its reserve food or on its tissues.
During the third and fourth weeks, the animals showed a steady
decrease in metabolic rate. Nearly all the starved nymphs died
before the seventh week, although two lived two months without
food.

\

20

aoo'v

iVe

t c-,

GRAPH 2. Curve shows rate of CCh output during four weeks of starvation.
Time interval 7 days. Temp. 22° C. Abscissas, gram body weight per insect.
Ordinates, milligrams COz per gram insect per hour. For further explanation see
description in text.

METABOLISM OF ^SHNA UMBROSA NYMPHS. 221

In comparing this graph with the graph of controls (Graph i),
it is noticed that the carbon dioxide output of a starving insect
decreases in one week at an amount comparable to the decrease in
twenty days in a younger insect and to ten days in the older
animals. At the end of four weeks of starvation, the carbon
dioxide output is about the same as the output of the controls at
the end of three months. It is interesting also to note an increase
in weight in the starving insects. Since the nymphs are aquatic,
this increase is probably due to the replacement of reserve fat by
water during starvation.

C. The Effect of Darkness on Carbon Dioxide Output.

The carbon dioxide output of three sets of nymphs was de-
termined under normal conditions of light and then the nymphs
were placed in darkness. Darkness was maintained by placing a
cap of heavy black paper over the finger bowls. The paper was
of the type used to protect photographic paper from the light
and was admirably suited for the present purpose. Air could
pass under the edge of the bag, which was fastened here and there
to the table. The temperature remained around 22° C. and the
animals were fed as usual, but in a dark room. Sticks were used
as the objects upon which the nymphs could cling. Water plants
would decay in the dark and offer a possibility of toxic materials
being added to the water. During experimentation the room was
in darkness and the experimental jar and the water bath were
covered with black paper. Once a week the carbon dioxide
output was determined and the results are recorded in Table III.
and Graph 3.

The metabolic rate decreased progressively with the weeks of
darkness up to the sixth week, when four individuals died.
Although the experiment was then brought to a close, the re-
maining nymphs were kept in the dark until all died, which
occurred within a period of two weeks. That the metabolic rate
of these animals is decreased by darkness can readily be seen
by comparing the data here given with the data obtained from
the controls. Per unit of weight the CO2 output of the nymphs
was much less than that of the control animals. A control
insect weighing .214 gram gave off .6499 milligram of CO% per

222

MARY HONORA SAYLE.

70.

do

.4-°

s

-<J

\

X

Gr

.Zto

GRAPH 3. Curve shows rate of CO2 output in darkness. Time interval 7 days.
Temp. 22° C. Abscissas, gram body weight per insect. Ordinates, milligrams
COa per gram insect per hour. For further explanation see description in text.

METABOLISM OF /ESHNA UMBROSA NYMPHS.

223

TABLE III.

SHOWING THE ACTUAL COj OUTPUT IN MILLIGRAMS PER GRAM BODY WEIGHT

PER HOUR, DURING DARKNESS.

Time interval — 7 days.

Average Weight of i Insect

Milligrams COa per Hour per

in Grams.

Gram Body Weight.

Set i.

Set 2.

Set 3.

Average.

Set i.

Set 2.

Set 3.

Average.

.158

.165

• 173

.165

.6613

.7650

• 5999

•6754

Light.

.164

.170

.181

.172

•4359

• 5931

.4071

• 4787

i week in dark.

.182

.181

•193

.185

.3626

.4207

.3404

.3746

2 weeks in dark.

.211

.199

.207

.206

• 2345

.3093

• 2303

.2580

3 weeks in dark.

.218

.219

.214

.217

• 2523

.2641

.2298

.2487

4 weeks in dark.

•233

.225

.229

.229

•231?

.2207

.2177

.2234

5 weeks in dark.

hour, while an insect in darkness weighing .211 gram produced
only .2345 milligram.

Since most of the animals lived but six or seven weeks, the
decreased CO2 production must have been accompanied by a
pathological condition resulting in early death. It is evident
that light is necessary for the normal physiological processes,
and that a lack of it produces an abnormal condition bringing
about death. It would be interesting in further research to
determine what part of the spectrum is of greatest importance
in the normal life of these nymphs.

D. Carbon Dioxide Output with Rising Temperature.

Table IV. gives the actual amounts of CO^ given off by three
sets of insects as the temperature of the water in which they were
living was gradually raised. The average figures are shown in
Graph 4.

Most of the authors already cited determined the CO2 output
at each increase in temperature without making a study of the
metabolic rate after a longer exposure to the same temperature.
Instead of making the usual successive determinations at different
temperatures, three determinations were made at each tempera-
ture. These records were made after twenty-four, forty-eight,
and seventy-two hours' exposure to the given temperature. In
following this plan, a different curve was secured than is usually
drawn. Reference to the graph shows that the CO2 output goes

224

MARY HONORA SAYLE.

up for each temperature after the twenty-four hours of exposure,
but goes down again after the longer periods of exposure. The
output is about normal after the seventy-two hours at the given
temperature. These results were found to be true in the case of
each rise in temperature. If one connected with a line the twenty-

TABLE IV.

SHOWING THE ACTUAL CO2 OUTPUT IN MILLIGRAMS PER GRAM TOTAL BODY
WEIGHT PER HOUR, DURING PERIODS OF HIGH TEMPERATURE.

Weight of i Insect

Milligrams COs per Gram Body

Time

in Grams.

Weight per Hour for I Insect.

Temp.

in

Hours.

Set i.

Set 2.

Set 3.

Average.

Set i.

Set 2.

Set 3-

Average.

22

.148

.165

.161

.164

.6087

•5579

.6091

• 5499

22

.163

.170

.177

• ****fi

.4723

•5341

.5173

25

24

.163

.170

.177

.170

.6518

.6742

.7155

.6805

25

48

.163

• 173

.177

.171

.4538

• 5404

.6496

•5483

25

72

.165

•173

.182

• 173

•4320

• 5511

• 5774

.5202

28

24

.171

.174

.182

.176

.7655

-7915

.7910

.7827

28

48

.171

.177

.182

.177

.4875

•6731

.6101

.5902

28

72

.177

.179

.184

.180

.3946

.5070

.4978

.4665

3i

24

.178

.181

.184

.181

.7746

-7999

.7899

.7881

3i

48

.178

.181

.184

.181

.6637

•6359

.6449

.6482

3i

72

.178

.181

.186

.182

.4400

• 5252

.5088

•4913

34

24

•175

.183

.186

.181

.4914

.5100

.6666

.5560

four-hour points on this graph, the usual rising curve could be
secured. However, according to the method of determination
followed here, the insects show a mechanism for adjustment or
acclimatization as evidenced by their carbon dioxide output.

The greatest output is between 28° and 31° C., this representing
for the nymphs the maximum in catabolic reactions. Above 31°
the temperature has a deleterious effect upon the organism, for
the insects respond but slightly to a further rise in temperature.
At this high temperature, the decreased carbon dioxide produc-
tion is no doubt accompanied by a physiological injury, for the
animals died after twenty-four hours in water at 34°.

After twenty-four hours at thirty-one degrees, the oxygen con-
tent of the water was determined. It was found that the water
still contained around 5.9 parts per million.

METABOLISM OF ^SHNA UMBROSA NYMPHS.

225

GRAPH 4. Curves show rate of COs output at high temperatures. Abscissas,
time in hours. Ordinates, at top, temperature degrees Centigrade; at bottom,
milligrams COa per gram body weight per hour. For further explanation see
description in text.

226

MARY HONORA SAYLE.

E. The Effect of Low Temperature on Carbon Dioxide Output.

The insects used in the experiments to show the effect of low
temperature on CO2 output were somewhat larger than the
insects used in the previous experiments. The former experi-
ments were carried on in the fall months, while the present ex-
periments were carried on in the winter, since it was thought
advisable to use running water from the lake at a time when it
maintained a constant temperature of 6°. Comparisons may be
made, however, with insects of similar weight in the controls,
to determine the relative amounts of carbon dioxide given off.

TABLE V.

SHOWING THE ACTUAL CO2 OUTPUT IN MILLIGRAMS PER GRAM TOTAL BODY
WEIGHT PER HOUR, DURING PERIODS OF Low TEMPERATURE.

Weight of i Insect

Milligrams CO2 per Gram Body

Time

in Grams.

Weight per Hour for i Insect.

Temp.

in

Hours.

Set. i

Set 2.

Set 3-

Average.

Set I.

Set 2.

Set 3.

Average.

22°

22°

.258
.266

• 243
.246

.211
.217

.240

.4263
.4135

•4750
.4519

• 5001
•4722

•4548

17°

24

.260

.246

.217

.241

.2538

.2891

.2906

.2778

17°

48

.266

.246

.217

• 243

• 3501

•3477

.3404

.3460

17°

72

.266

.250

.221

.246

•3721

-3655

•3494

•3623

13°

24

.278

.250

.221

.250

.1978

.1861

.2054

.1964

13°

48

.278

.250

.221

.250

.3079

•2737

.2977

.2931

13°

72

.286

• 253

.227

•255

•3165

.2999

•3147

.3104

9°

24

.288

-253

.228

.256

• 0593

•0577

.0694

.0621

9°

48

.288

•253

.228

.256

•1337

.1272

• 1399

.1336

9°

72

.289

.260

•237

.262

.1582

.1488

• 1577

• 1549

3°

24

.289

.260

•237

.262

.0380

.0299

.0376

•0352

3°

48

.289

.260

•237

.262

.0425

.0471

•0517

.0471

3°

72

.289

.264

•237

.264

.0570

.0441

•0597

• 0536

Table V. shows the carbon dioxide output of the nymphs as
the water temperature was lowered gradually from 22° to 3° C.
The same scheme of duration of time at each temperature was
used as previously employed in the records with rising tempera-
ture. Graph 5 shows that the carbon dioxide output decreased
progressively with falling temperature after twenty-four-hour
durations, but also shows an acclimatization to a low temperature
after being exposed to it for a longer period of time. The carbon
dioxide output drops for each drop in temperature, but the out-
put increases again in forty-eight and seventy-two hours at a

METABOLISM OF .-ESHNA UMBROSA NYMPHS.

227

30.

GRAPH 5. Cuives show rate of COt output at low temperatures. Abscissas,
time in hours. Ordinates, at top, temperature degrees Centigrade; at bottom,
milligrams COa per gram body weight per hour. For further explanation see
description in text.

228 MARY HONORA SAYLE.

given temperature. At 3°, the metabolic processes are going on
at a very slow rate and acclimatization at this temperature was
only slight. Low temperature did not affect the nymphs in any
permanent way, for they lived a normal length of time after the
experiment was completed.

SUMMARY AND CONCLUSIONS.

Although a great deal of work has been done on the metabolism
of insects, very little has been published on the metabolic rate in
truly aquatic forms. As far as I know, nothing has appeared
relative to the carbon dioxide output of /Eshnid nymphs. Their
respiratory mechanism is an interesting one and the present work
is only an introduction to the many related research problems
that present themselves.

1. During the nymphal stage of jEshna umbrosa, the carbon
dioxide output decreases progressively as the nymphs grow older.
The younger, more active individuals have the higher rate of
respiratory output per gram body weight.

2. The CO2 output of starved animals decreases during the
period of starvation, except during the second week, when the
metabolic rate is higher than the normal.

3. Darkness has a marked effect on the metabolism of these
animals. The CO2 output decreases steadily with the weeks in
darkness and the physiological condition is so affected, that the
nymphs die.

4. Higher temperatures cause increased rates of CO2 output
and lower temperatures tend to have the reverse effect. How-
ever, after an initial increase in CO2 output following twenty-four
hours of exposure to the high temperature, the output decreases
again after longer exposure to the same temperature. The same
effect is evidenced by the nymphs exposed to low temperature,
i.e., decrease in CO2 output is followed by an increase after a
longer period at the same temperature. The insects indicate a
mechanism for acclimatization.

LITERATURE CITED.

i. Alice, W. C., and Stein, E. R.

'18 Light Reactions and Metabolism in May-fly Nymphs. Jour. Exp. Zool.,
26:423-458.

METABOLISM OF .ESHNA UMBROSA NYMPHS. 22Q

2. Battelli, F., und Stern, L.

'13 Intensitat des respiratorischen Gaswechsels der Insekten. Biochem.
Zeitschr., 56: 50-58.

3. Bodine, J. H.

'21 Factors Influencing the Water Content and the Rate of Metabolism of
Certain Orthoptera. Jour. Exp. Zool., 32: 137-164.

4. Bodine, J. H.

'23 Hibernation in Orthoptera. Jour. Exp. Zool., 37: 457-488.

5. Bodine, J. H., and Orr, R. R.

'25 Respiratory Metabolism. BIOL. BULL., 48: 1-14.

6. Biitschli, O.

'74 Ein Beitrag zur Kenntnis der Stoffwechsels, insbesondere der Respiration
bei den Insekten. Arch. f. Anat. u. Physiol., 348-361.

7. Dirken, M. N. J.

'22 La relation entre les changements de temperatuie et la consommation
d'oxygene par les animaux a sang froid. Archives Neerland. de Physiologic
7: 126-131.

8. Farkas, K.

'03 Uber den Energieumsatz des Seidenspinners wahrend der Entwicklung im
Ei und wahrend der Metamorphose. Pfliiger's Arch. f. d. ges. Physiol.,
98: 470-546.

9. Fink, D. E.

'25 Metabolism during Embryonic and Metamorphic Development of Insects.

Jour. Gen. Physiol., 7: 527-545.
10. Krogh, A.

'16 The Respiratory Exchange of Animals and Man. London and New York.
u. Loeb, J.

'88 Der Einfluss des Lichtes auf die Oxidationsvorgiinge in thierischen Organis-
men. Pfliiger's Archiv. f. d. ges. Physiol., 42: 393-407.

12. Loeb, J., and Northrop, J. H.

'17 On the Influence of Food and Temperature upon the Duration of Life.
Jour. Biol. Chem., 32: 103-121.

13. Luciani, L., et Lo Monaco, D.

'93 Sur les phenomenes respiratoires de la chrysalide du bombyx du murier.
Arch. ital. de Biol., 19: 274-283.

14. Luciani, L., et Piutte, A.

'88 Sur les phenomenes respiratoires des oeufs du bombyx du murier. Arch,
ital. de Biol., 9: 319-358.

15. Newport, J.

'36 On the Respiration of Insects. Philosophical Transactions of the Royal
Soc. London: 529-566.

16. Northrop, J. H.

'26 Carbon Dioxide Production and Duration of Life of Drosophila Cultures.
Jour. Gen. Physiol., 9: 319-324.

17. Parhon, M.

'09 Les echanges nutritifs chez les abeilles pendant les quatre saisons. Ann. des
Sc. natui. Zool., Ser. 9, 9: 1-58.

18. Plateau, F.

'72 Recherches physico-chimiques sur les articules aquatiques. Bull. Acad.
Roy. Sci. Belg., 34: 274-321.

230 MARY HONORA SAYLE.

19. Regnault, V., et Reiset, J.

'49 Recherches chimiques sur la respiration des animaux des diverses classes.
Annales de Chimie et de Physique, 26: 299—519.

20. Slowtzoff, B.

'09 Uber den Gaswechsel der Insekten und dessen Beziehung zur Temperatur
der Luft. Biochem. Zeitschr., 29: 497-503.

21. Sosnowski.

'02 Contribution a 1'etude de la physiologic du developpement des mouches
Bull, intern, de 1'Acad. des Sciences de Cracovie: 568.

22. Spallanzani, L.

1787 Opuscules de Physique, animate et vegetale. Trans. Jean Senebier,
8:56-58-

23. Spallanzani, L.

1807 Sur la respiration des insects. Geneve.

24. Standard Methods for the Examination of Water and Sewage. American Pub-

lic Health Association. New York. 1925.

25. Treviranus, G. R.

'31 Versuche iiber das Atemholen der niederen Tiere. Zeitschr. f. Physiol. von
Tiedmann und Treviranus, 4: 1-39.

26. Vernon, H. M.

'97 The Relation of the Respiratory Exchange of Cold-blooded Animals to
Temperature. Jour. Physiol., 21: 443-496.

27. von Linden, G.

'12 Die Assimilationstatigkeit bei Schmetteilingspuppen. Leipzig.

28. Weinland, E.

'05 tiber die Stoffumsetzungen warn end der Metamorphose der Fleischfliege.
Zeitschr. f. Biol., 47: 186-231.

FURTHER OBSERVATIONS AND EXPERIMENTS ON

THE SYMBIOSIS BETWEEN TERMITES AND

THEIR INTESTINAL PROTOZOA.

L. R. CLEVELAND,

DEPARTMENT OF TROPICAL MEDICINE,

HARVARD MEDICAL SCHOOL,

BOSTON, MASSACHUSETTS.

It has been possible through a grant from the Bache Fund of
the National Academy of Sciences and the collecting ability of
Dr. Harold Kirby, Jr., to extend somewhat my observations on
the symbiosis between termites and their intestinal protozoa.

Hitherto the investigations (Cleveland, '24, '250, '256) have
been confined principally to two genera, Termopsis and Reticuli-
termes; similar observations have now been made on six more
genera (subgenera according to some investigators), with frag-
mentary observations on several other genera. All the termites
in this investigation except Reticulitermes came from Panama
and Costa Rica and were identified by Dr. T. E. Snyder.

In most of the experiments the protozoa were removed by
confining the termites in oxygen at a pressure of 60 pounds for
one hour ; in a few instances higher pressures for a shorter period
and lower pressures for a. longer period wrere used. It was never
possible, regardless of how high a pressure was used, to remove all
the protozoa in less than ten minutes. The termites, as in previ-
ous experiments (Cleveland, '250, '256), showed no ill effects from
the oxygen even if confined in it for a much longer period than
that required to kill all their intestinal protozoa.

Large numbers of termites were used when available and after
the removal of the protozoa they were kept in large Petri plates
and were fed either cellulose (filter paper) or wood. Moisture
was supplied in all except the earlier experiments by placing in
each Petri plate three small pieces of absorbent cotton dipped in
water and then mashed between the fingers until as much water
as could be gotten out in this way was removed. Every third

232 L. R. CLEVELAND.

day the pieces of cotton were remoistened. This appears to be
the best way to keep termites in the laboratory so that they may
be constantly observed with little effort and without being greatly
disturbed. When fed wood, they were not given enough to hide
themselves by burrowing into it. They prefer the wood and
paper and do not eat the cotton unless the supply of wood or
paper is exhausted. Death due to molds — the bane of poorly
kept laboratory termites — may be avoided entirely when ter-
mites are kept in this manner. Controls of all the experiments
were kept in the same manner and are in excellent condition now,
ten months after the defaunated termites were all dead.

The principal results lend themselves rather easily to tabulation
and are set forth in Table I . These data are not as complete as
they should be. They show, however, that none of these termites
are able to live indefinitely after their protozoa have been re-
moved by oxygen treatment. But, in most cases, they do not
show why some termites lived longer than others. This may be
a difference inherent in the termites themselves, some being more
dependent on protozoa than others. Further investigations, in
fact, may even show that certain protozoa-harboring termites are
able to live for a long time, if not indefinitely, on a normal diet
of wood after their protozoa have been removed. The reason
why the termites of some genera lived so much longer than those
of other genera may also, in part, be a matter of food. Too few
termites were available in some instances to properly control this
factor. Some of the termites were strikingly different in ap-
pearance and behavior from others, and we may well expect con-
siderable differences in their ability to maintain themselves on
the same diet after the removal of their protozoa, but it will be
necessary to carry out more extensive experiments before this
can be definitely demonstrated. However, a large number of
experiments were carried out with Reticulitermes under ideal
conditions with the result that this termite of the family Rhino-
termitidse was not able, on the average, to live nearly so long as
termites of three genera of the family Kalotermitidae (Neotermes,
Kalotermes, Cryptotermes). In the case of Reticulitermes, the
reason why some of the experiments ran so much longer than
others is quite clear. Those experiments which ran so long

SYMBIOSIS BETWEEN TERMITES AND PROTOZOA.

233

•O

'•^

S

QJ -J
'/ "''

a

: -

>.

— ?

c

—

*-M

CO ^

M

-

•/ •/. ™ x •/'

^

w >

- i w 3 -j

c3

_s —

-*-» •»-• ^J *^ *_>

ctf CO CQ *^ rt

e

~~

D* £•« Z* « — t/

•

(2

'C 5

•/ '- •/ 'E -;/ 'E '^ I'

•S S?

~ *^ " *~L ~~ ~ *^j ""
.S — .S — 2L — — -

.

^ TO

S «

^^^w-^~.^

C/)

W

^1, ^

S'S'S'S'S'S'S*—

H

^

i2 u; ^ 'ji Ld L^ ^

5

f-t

tA

•

U

o

H

o

D
W

•o

^

-d

<;

u

0

£2

0)

^

<!

"O

•d

to
(d

Q

•O

O
O

T3

O

O

ta

O

in

73

•d -a -d -r ~

o

C

o o o os

d

O O O O D

D-

C

fe fe "4 ^* •*

<

tc

^ . ^

*•*

<u

*£ c c S "c •— ' S

to

a

^ ^ ^3 O- ^ --H Q-

1— 1

0

CU

Cfl C/2 C/3 OH t/3 to P-c

[/>

W

>

•saiiuuay

J

ffl

Q

pajBunBjaQ jo ajiq jo

o

0 xO H t~ «- JO «>

<

z

qiSua^j jo riuauiuadxg

M M

r*

K

jo uopBjriQ aSBJ3Ay

to

hH

to

•pEaQ 3-iaA\ s}uam
-uadxg nv ui sajiui

00

o

O r» N <~O oo O >o

M t- !N M M OO O

M M

C

•J3X IIV a-ioj^q SABQ

B

o

•pB3Q 3J3AA. 1U3UI

-uadxg auQ ui sa^iui

00

O N O "*O « O

W

1

"J3X IIV 3JOJ3(1 SABQ

>— 1
W

•luauiuadx'j ipBTi ui

o

o o o o o o o

H

*-* *

o

\T) IO O *O WO O WS

saiiuijaj^ jo jaquin^

M

M

g

o

•}11() pOUJB ) S1U31U

O

^ ^ N ^. ^ „ N

0,

-i lodxii 10 j.xnunvr

0)

PL,

*^

r*

z

CO

[>

^ ' —

o

_rt

— x '!• ./'_•; —

K

1

«"| §"c i'f.^

CO

£

~ •' "? f- "~

o

•0.

o •5-2 -~ b 5. -r

'._ *"^' ~~™ 2r ^ sS

1

g

8 ^ 8 K -^ ^ §,

kl

t*

S*. "** '*** ^* ci

o
H

2

•^ ~ ^^ -c 5 •*

1

fe fe 1 1 i 1 1

~ «. g S ci v ^

"a

**• -2 ^ ^* ^ "^ d

<0

•*A

2 •^•'t: ~ ^.^5

16

234 L- R- CLEVELAND.

contained a fairly large number of second form young adults.
These individuals are fed and supported by the workers and
nymphs (Cleveland, '256). They had few, if any, protozoa to
lose (Cleveland, '256) and were able to exist in a fairly normal
manner on the salivary secretions of the workers and nymphs
for three weeks or more, or as long as the workers and nymphs
lived and were able to feed them, and then came their period of
starvation followed by death three to four weeks after the workers
and nymphs who supported them died.

The conditions under which Lobitermes, Calcaritermes , and
Glyptotermes were kept were perhaps not so favorable as the con-
ditions of the other genera of this family (Kalotermitidae),
Kalotermes, Cryptotermes, and Neotermes, and perhaps for this
reason they did not live so long. These experiments were carried
out before the Petri plate method of keeping termites was dis-
covered, and all termites of these genera were used up.

Some material of two other genera, Leucotermes and Copto-
termes, of the family Rhinotermitidae was procured, but these
termites behaved so differently from Reticulitermes and the
genera of the Kalotermitidae that it was impossible to carry out
the same experiments with them. One or two days in closed
jars or test tubes usually resulted in their death. Almost any
kind of handling, especially confinement in a closed jar or, some-
times, shipment for eight to ten days in a wooden box, resulted
in the loss of their intestinal protozoa. Owing to the fact that
material of these genera was not plentiful, no attempt to study
them in aerated containers was made. It will be very interesting
indeed for comparative purposes, and possibly for gaining some
definite information on the origin of termite symbiosis and para-
sitism, to carry out a large number of experiments on these and
other genera of the Rhinotermitidae. Valuable information in
this same direction may also be obtained by a more critical study
of the ability of several genera of the Kalotermitidee to live on
various diets after the removal of their protozoa. It is also very
important to study the single genus Mastotermes of the most
primitive family, the Mastotermitidae.

Several genera of the most highly specialized family, the
Termitidae, were obtained. This family as a rule (Cleveland,

SYMBIOSIS BETWEEN TERMITES AND PROTOZOA. 235

'230, '26) does not harbor protozoa, although a few small flag-
ellates and fairly large xylophagous amoebae l occur in some species,
but perhaps not in sufficient numbers, for the most part, to be of
any decided advantage to their hosts. Many species in this
family do not feed on wood at all, and those that do, feed on more
decayed wood than most termites of the other three families.
Numerous attempts have been made to feed cellulose (filter
paper) to several genera of the Termitidse (Anoplotermes, Miro-
termes, Microcerotermes, Cornitermes, Orthognathotermes, Armi-
termes, Convexitermes), but with no very definite results. These
termites eat moistened filter paper readily and it will be an easy
matter to determine whether or not they can digest it as soon as
a convenient method for keeping them in the laboratory is worked
out. In most of the experiments that were carried out the con-
trols, or those termites fed very much decayed wood and fresh
soil or humus, lived but little longer than those fed paper. When
the termites were fed paper and allowed to remain in the nest or a
portion of it, they lived much longer than when placed in jars
with paper only. This was perhaps due to the fact that the
fungi in the nest aided them in some way in the digestion of food.
However, the laboratory conditions at best were too different
from those of nature to lend much value to the experiments.
When better means of keeping the Termitidae in the laboratory
are obtained, this subject will be considered again.

Termites of all four families harbof very many spirochaetes,
which, like the protozoa, may play a role in the digestion of
cellulose and hemicellulose. Considerable time has been given
to an attempt to remove the protozoa from several genera of the
Kalotermitidae without removing the spirochaetes, but with no
very definite results, because practically every treatment that
removes the protozoa removes the spirochaetes also. It has been
possible, however, in a few experiments to remove the protozoa by
heat treatment without killing the spirochaetes, but they were all
dead three or four days later. It will be necessary to repeat these
experiments many times under easily controlled conditions be-
fore any definite conclusions are drawn regarding the relation of
the spirochaetes to the termites and to the protozoa. They may

1 These amoebae will be described by Dr. Harold Kirby, Jr.

236 L. R- CLEVELAND.

live in some sort of a symbiotic relation to the protozoa — we do
not know. They are attached to the protozoa — a fact overlooked
by Damon ('26) and by Hollande ('22), who have recently given
brief descriptions of termite spirochaetes. Very few of the
spirochsetes, indeed, except those of the large form similar to the
Cristispira of oysters, are free in the intestinal fluid. Countless
millions are often attached to a single protozoon, which seems
scarcely able to move. Some protozoa have no attached spiro-
chaetes, while others of the same species may have from one
individual to several millions attached to them. Sometimes half
of the protozoa in a termite are completely covered with spiro-
chaetes, which, of course, may be mistaken for flagella unless
examined by the dark field method. Spirochaetes may also at
times be seen inside the protozoa when the termites have been
fed paper for a long period. All attempts to grow the spiro-
chaetes have failed so far, and all animal inoculations have been
negative. The subject is being pursued further and will be re-
ported on in more detail later.

Somewhere in the 1,500 species of termites there is a key to the
unique and most interesting animal association throughout all
nature. There have probably been many gradatory steps in the
formation of the association, but perhaps enough of the steps
remain today to enable us, after many years of investigation, to
trace the order of events. The fact that in the Termitidae we
find a few protozoa in a species here and there suggests that in
this family we are now viewing either the beginning or the end of
a widespread symbiotic association. The ease with which those
genera of the Rhinotermitidae resembling the Termitidae closely
in behavior lose their protozoa when handled in the labo:atory
or when packed and shipped, indicates that possibly the Termi-
tidae are losing their protozoa.

ADDENDUM.

Since this paper went to press termites have been fed cellulose
thoroughly moistened with a five per cent, aqueous solution of
acid fuchsin and it has been possible in this way to remove all
spirochaetes from them without doing any damage whatever to
the protozoa or to the termites. Some of the termites have been

SYMBIOSIS BETWEEN TERMITES AND PROTOZOA. 237

spirochjeteless now for three months and it is impossible to de-
tect any deviation from their normal appearance and activity,
even though they have been fed continuously on the diet which
removed their spirochaetes. Also no change has occurred in the
number and activity of the protozoa. Evidently the spirochsetes
play little, if any, role in the digestion of wood and cellulose.
These results will be reported in detail later.

REFERENCES.
Cleveland, L. R.

'233 Correlation between the Food and Morphology of Termites and the

Presence of their Intestinal Protozoa. Amer. Journ. Hyg., 4: 444-61.
'24 The Physiological and Symbiotic Relationships between the Intestinal

Protozoa of Termites and their Host, with Special Reference to Reticulitermes

flavipes Kollar. BIOL. BULL., 46: 177-225.
'253 The Feeding Habit of Termite Castes and its Relation to their Intestinal

Flagellates. BIOL. BULL., 48: 295-308.
*25b The Effects of Oxygenation and Starvation on the Symbiosis between the

Termite, Termopsis, and its Intestinal Flagellates. BIOL. BULL., 48: 309-27.
'256 Toxicity of Oxygen for Protozoa in Vivo and in Vitro : Animals Def aunated

without Injury. BIOL. BULL., 48: 455-68.
'26 Symbiosis among Animals with Special Reference to Termites and their

Intestinal Flagellates. Quart. Rev. Biol., i: 51-60.
Damon, S. R.

"26 A Note on the Spirochaetes of Termites. Journ. Bact., n: 31-36-
Hollande, A. Ch.

'22 Les spirochetes des teimites; processus de division; formation du schizo-

plaste. Arch. Zool. Exper. et Gen., Notes et Revue, 61: 23-28.

NOTES ON A 32 MILLIMETER FREEMARTIN.

THOMAS HUME BISSONNETTE,

TRINITY COLLEGE, HARTFORD, CONN., AND MARINE BIOLOGICAL LABORATORY,

WOODS HOLE, MASSACHUSETTS.

In an earlier paper, Bissonnette ('24) described the develop-
ment of the reproductive ducts and canals in cattle embryos from
28.0 mm. crown-rump length to 20.0 cm., and compared with that
series one of 5 freemartins from 9.0 cm. to 20.0 cm. In that paper
the gonad of a 9.0 cm. freemartin was described which showed
unmistakable development of ovarian cortex or secondary sexcord
region. The material described gave support to the theory of
Lillie ('16 and '17) and Keller and Tandler ('16) that the free-
martin is a female modified in utero in the male direction by the
action of a male hormone from the testis of her twin brother. It
showed that she may go so far in the female direction as to have a
secondary sexcord region and even to the condition with an
intermittent wolfnan duct before undergoing regression of the
mullerian duct derivatives and added growth and differentiation
of the so-called "male" structures. The one there described was
the first showing such sexcords. Since earlier writers had agreed
that such cords were never found in freemartins (Chapin, '17,
Willier, '21, Lillie, '23) the question remained as to whether this
was not a very abnormal situation among freemartins. Further
study of younger stages of the freemartin was needed to answer
this question.

Lillie ('23) described and figured a 3.75 cm. freemartin with no
such cords but with a germinal epithelium more than one cell
thick — in some places even 3 or 4 cells thick — but never more than
about a third to a sixth of the thickness of the corresponding
layer in the single-born female. He found the ducts and canals
in no way different from those of the female of corresponding
size. This was the smallest stage of the freemartin described and
it showed definite modification of the gonad in the male direction.

The subject of the study described in this paper is a freemartin

238

NOTES ON A 32-MILLIMETER FREEMARTIN. 239

of 32 mm. crown-rump length which is of interest because (i) it
is the smallest stage so far studied, (2) it has a minimum of union
of the chorions compatible with any blood transfusion, (3) it is
another case where ovarian cortex is found outside the tunica
albuginea, (4) it appears to be as near as possible to the stage of
distinguishable sex-differentiation in which modifications of the
freemartin type can be distinguished, (5) it shows slight free-
martin modification and forms the initial end of a series of inter-
grades between the postnatal freemartin types and the normal

female.

MATERIAL AND METHODS.

This 32 mm. freemartin, twin to a 34 mm. male, was taken July
28, 1923, at the killing floor of Swift & Co., Chicago Stockyards.
Since it was the smallest freemartin in over 50 taken up to that
time by the writer, it was especially carefully treated. The uterus
of the mother was slit open throughout both horns and the mem-
branes and twins carefully lifted out. The embryos were far up
the opposite horns of the uterus and the chorions were but slightly
united and showed no bloodvessel union visible to the naked eye.
The placentae pulled apart with a small amount of bleeding only,
when they were handled after removal, though they held together
during that operation. Chorions and embryos entire were
placed at once into Bouin's fluid which was changed once about
two hours after they were put in. No injection was attempted
since the vessels were judged too small for the apparatus at hand
on the killing floor and the help available, and the internal organs
were wanted for microscopic work and early fixation was desired.
The membranes and embryos were washed in 50 and 70 per cent,
alcohols, to the latter of which at its second addition a few drops
of a saturated aqueous solution of lithium carbonate was added
to help remove the picric acid. They were then stored in 80
per cent, alcohol for about a year and a half, when the embryos
were removed and the posterior parts of both embryos were
stained in bulk in alum cochineal and sectioned 30 micra thick.
The series was counterstained on the slide in orange G and
mounted in balsam.

Both maternal ovaries had been taken by the boy who collects
those with large corpora lutea for extraction, so it is supposed each
had a corpus luteum.

240

THOMAS HUME BISSONNETTE.

Cloweru/ws.
Wolff, bo^.

Rete.

'•••'^-jf'^ .

nuryJ ex-Cords
u,nica aiDaflinei

FIG. i. Transverse section through gonad, mesonephros, and ducts of right side of
normal female, 32 mm. crown-rump length. X 38.3.

FIG. 2. Transverse section through right gonad, mesonephros, and ducts of
normal male, 32 mm. crown-rump length. X 38.3.

NOTES ON A 32-MILLIMETER FREEMARTIN.

241

Glome

Wff.pt

h left gonad, mesonephros, and ducts of free-
martin, 32 mm. crown-rump length. X 34.3.

gonad, mesonephros, and ducts of normal
female, 32 mm. crown-rump length. X 34.3.

FIG. 3. Transverse section through left gonad, mesonephros,
martin, 32 mm. crown-rump length. X 34-3-
FIG. 4. Transverse section through left gonad, mesonephros, ai

— , crown-rump length. X 34-3-

242 THOMAS HUME BISSONNETTE.

On washing out, no vascular anastomosis could be made out
with the binocular dissecting microscope, low power, but the
bleeding on separation was evidence that there was capillary
connection at least.

Previous work had shown that thick sections have some ad-
vantages over thin in showing the zones in the gonads and regions
of even slight degeneration with consequent pigment production.
They also save time in reconstruction of the ducts and canals.
Such reconstruction was made for this freemartin and for a normal
female of the same size for a control, after the manner of Bis-
sonnette ('24). The notochord was used as reference line for
dorso-ventral relations of ducts, canals, mesonephros, and anlagen
of the round ligament or gubernaculum, and gonads (Figs. 5 and
6). Typical cross-sections through the gonad, mesonephros,
miillerian and wolffian ducts, in the normal female, male, and
freemartin, are shown in Figs. I, 2, 3, and 4.

DESCRIPTION AND COMPARISON WITH NORMALS.

The Placental Situation. — The twins were far apart in discrete
but united chorions with cotyledons beginning to appear as
opaque patches. One chorion was telescoped slightly into the
end of the other and they were so united that bleeding occurred
on separation. Otherwise the chorions were normal. Both
older and younger cases have been taken by the writer in which
the chorions were in contact but not so adherent as to cause
bleeding on separation. In all such cases so far investigated the
female has proved normal. This situation was much the same as
that in the 9.0 cm. freemartin No. I, Fig. 2, of Bissonnette's ('24)
paper, where a slight though definite amount of modification in
the male direction was demonstrated and secondary sexcord re-
gions or cords of Pfliiger were found. It is another case permit-
ting a small blood interchange between the twins and a relatively
small concentration of hormone would be expected on the female
side and so a longer period might elapse before the effective
minimum concentration of the hormone is reached. The am-
nions were separated widely and were normal.

The External Genitalia. — Externally this freemartin was female
in type as shown by Figs. 5 and 6 in sagittal section. The

NOTES ON A 32-MILLIMETER FREEMARTIN.

243

clitoris was slightly more recurved than that of the normal con-
trol. However there are variations in this respect among single
females due to the difference in relative curvatures of the embryos
at the time of killing and fixing. It resembles the one figured

tori

W.Dt

le sotte/m

^^

Osi

iflttllDt. i
W.Dt.

l-s«%.

t-Bl

I

FIG. 5. Reconstruction of the urinogenital system of the freemartin of 32 mm.

crown-rump length. X 9-5-
FIG. 6. Reconstruction of the urinogenital system of the normal female of 32 mm.

crown-rump length. X 9-5-

by Lillie ('23) for his 3.75 cm. one (Fig. 5) externally as well as
in gross internal dissection. So it is not figured here in these
aspects. The urogenital aperture is open although in the male
of the same size and in the male twin it is closed and has a raphe

244 THOMAS HUME BISSONNETTE.

to mark the midline in place of a slit. The perineum is not
expanded ventrally as in the male.

The vulva, vestibule, and bulbo-vestibular glands are female in
type and proportions (Figs. 5 and 6). The ureters also come off
the urethra or urogenital sinus as in the female from a ridge in
the dorsal wall.

The wolffian ducts are not much affected though their lumen is
greater just posterior to the mesonephros than in the control. It
is not to be expected that they would be affected greatly since they
are still the ducts of the functional mesonephros.

The miillerian ducts are not so large in section as the control
and lack a lumen in some regions where there is one in the control.
They appear to be intermediate between the male and the female
and, as in the male, show some signs of regression. Their staining
reaction is like that of the male (Figs, i, 2, 3, 4). This difference
in size is most apparent in the region just posterior to the mesone-
phros where the wolffian ducts are larger. The slight difference
in the type of entry into the vestibule shown in Figs. 5 and 6 is
probably not significant as such differences are found in single
females. The ostia tubae abdominale at the anterior end of the
miillerian ducts open out flat with lips on the ventral aspect of
the anterior end of the mesonephros and show a tendency to be
associated with the rete there. This rete appears to lead over
into the mesonephros and to join a glomerulus as in the normals
of both sexes at this age.

The Gonads. — Figs, i and 2 are from sections through near the
middle of the right gonads of 32 mm. female and male respectively,
and figures 5 and 6 are similar sections of left gonads, etc., of the
same sized freemartin and female, drawn with camera lucida to
the same scale. As stated above, these thick sections (25 and 30
micra) show up the regions of the gonad better than thin sections,
and cortical regions stand out better. The gonads of the free-
martin have the same relations to the mesonephros as those of
the female. Rete ovarii are similar in that they do not extend
into the middle of the gonad as they do in the male, but are
found only near the mesenterial border of the gland (Willier, '21 ;
Bissonnette, '24). The type of tunica albuginea is female but
the medulla of the ovary does not show signs of as great activity

NOTES ON A 32-MILLIMETER FREEMARTIN. 245

as in the control. A cortex is present outside the tunica albuginea
and is continuous with the germinal epithelium or serosa coat
from which it appears to develop, though organization into defi-
nite sexcords is not evident in the preparations of either female or
freemartin. The secondary sexcord region is not quite so thick as
in the control nor is the gonad quite so large in cross-section. It
is about 21/23 of the normal in diameter. The length of the
freemartin gonad is 1,740 micra while that of the control is 2,150.
This may or may not be within the limits of normal variation at
this age for no series of measurements of normal ovaries has been
made at this stage to determine these limits. The difference in
size is apparent from the sections drawn to the same scale.
This would be even greater were both sections taken from exactly
the same relative level of the gonad. The freemartin was cut
62/100 of the length from the front end and the female 58/100 of
the corresponding distance. As shown in the reconstructions,
the gonads are larger in diameter posterior to their middle regions
but near them. The blood supply of the freemartin and control
gonads is distributed through the tunica albuginea inside the
cortex, while, in the corresponding male, since there is no cortex
outside the tunica, the blood supply is superficial in the definitive
tunica albuginea, present at this stage. In the testis at this stage
there is marked cellular growth in the medulla or primary sexcord
region ; in the ovary this activity is less marked ; and even less so
in the freemartin gonad.

The anlagen of the gubernacula or round ligaments are similar
to those of both sexes at this stage. No sex differences in this
organ appear till later stages as described by Bissonnette ('24).

DISCUSSION.

Lillie ('16 and '17) and Keller and Tandler ('16) came to the
conclusion that the freemartin is a modified female, that sex-
determination as a female was consummated at fertilization, and
that the abnormalities are due to the action of a male hormone
passed to her through the vascular anastomosis in the united
placental circulations. They believed that this was a case of
normal zygotic sex-determination followed by abnormal sex-
differentiation in the female twin. This material supports such

246 THOMAS HUME BISSONNETTE.

a theory by showing a case near the initiation of the differentia-
tion process where the embryo is essentially female in type and
condition of the organs, but is slightly underdeveloped in respect
to the secondary sexcord region and miillerian ducts. This case,
taken in connection with other described cases, particularly where
the placentae were in contact but not united by either capillaries
or arteries or veins, as described by Lillie ('17) and Keller and
Tandler ('16) and verified by the writer, appears to show that a
blood-carried agent or hormone is at work here and not some
agent like the "organizer" which Spemann ('25) shows so un-
questionably is operative in the early development of the am-
phibian larva. If such an agent were operating here, mere
tissue cohesion or connection would be sufficient to condition full
effect, and the great differences between freemartins of the same
embryonic age would be difficult to account for except on the
basis of variations in zygotic sex-balance of the two individuals of
the twins or differences in time of contact of the chorions. Now,
in cattle, mere contact is usually attained early by reason of the
great elongation of the trophoblastic vesicle; but vascular con-
nection is not always so early, and the effect appears to depend
on blood transfusion rather than on tissue continuity.

However, this case and others such as the one described by
Keller and Tandler ('16), in which there was slight vascular inter-
change accompanied by a scarlike welt across the place of junc-
tion of the chorions, reducing what they believed was once a
greater vascular connection to proportions too small to account
for the degree of modification, appear to show that even a capil-
lary anastomosis or an anastomosis of very small bloodvessels is
sufficient to condition the modification. The state of the organs
appears to depend on the time elapsed since the effective thresh-
old of hormone action is reached, rather than on a varying
amount of effective hormone above such threshold. This free-
martin and the 9.0 cm. one of Bissonnette ('24) both showed a
minimum vascular connection accompanied by a type of modifi-
cation, or degree of it, which suggests a marked development as a
normal female, probably for some time before the hormone be-
came active, followed by regression of the peculiarly female
organs and a delayed acceleration of the growth of the essentially

NOTES ON A 32-MILLIMETER FREEMARTIN. 247

male parts. They do not support the assumption -that there
has been a slower than normal development of the female struc-
tures with a slightly faster development of the male ones over the
whole period of union, as will appear from what follows.

Lillie's Fig. n, page 68, of a gonad section from a 3.75 cm.

freemartin, in which he found demonstrable vessels connecting

the circulations of the twins with no clearly distinguishable place

of chorionic union, shows a germinal epithelium together with

three or even four layers of cells outside the tunica albuginea in

some regions. These cells of the extra layers are irregularly

arranged and not in definite cords. These extra cell layers over

and above the single layer of coelomic epithelium homologous to

that in the bull of similar size (Fig. 12) he does not interpret as a

sexcord region and cords are not possible with so few cells in any

case. After inspection of the 32 mm. freemartin in this study

and all the freemartins younger than the 9.0 cm. stages examined

by the writer microscopically, all of which show the secondary

sexcord region, it appears that these extra cells in Lillie's specimen

may well be secondary sexcord primordia in which cords have not

yet organized. Brambell ('27) finds that in the mouse cords do

not organize till later and that in males the germinal epithelium

becomes several cell-layers thick after the tunica albuginea forms

but later thins out to a single layer. This latter situation does not

develop in males in cattle as seen in Fig. 2 of this paper and Fig.

12 of Lillie ('23). Chapin ('17) in some of her gonad sections

shows regions which she does not interpret as derivatives from

secondary sexcord anlagen, but wThich appear in the light of this

more recent younger material and from an inspection of her slides

themselves (kindly loaned by Professor F. R. Lillie) to be derived

from such source with possible later proliferation. In any case

the secondary sexcord region is not a layer of uniform thickness

all over the gonad but is thicker in lateral regions than in ventral.

The layer also appears to be resorbed at different rates in different

regions and so to remain in some regions longer than in others

and possibly permanently. This appears to be correlated with or

dependent on pressure by adjacent organs in some way. It is

possible for the tunica albuginea to reach the outside of the gonad

in some regions and not in others, leaving bands of secondarv

248 THOMAS HUME BISSONNETTE.

sexcord derivatives extending lengthwise in the gonad parallel to
the medullary cord derivatives. This situation is figured by
other workers. These remnants were difficult or impossible
to interpret without comparison with the earlier stages or with
stages which had undoubted secondary sexcords. Such a series
was not available for study by other students of the freemartin.
This will be discussed more fully in a later paper on the basis of
more material.

From a study of the above mentioned freemartins and of a
6.75 cm. and 7.5 cm. pair from a set of triplets with one bull only
(to be discussed in a later paper) where cords of Pfliiger are also
present but show signs of regression, one is led to conclude that
the rate of resorption of these cords is comparatively slow as well
as that they may continue to grow in some cases for a short time
before regressing. This also leads one to believe that secondary
sexcord regions were never present in Lillie's specimen mentioned
above, in any greater amount than is shown in his figure (Lillie,
'23, Fig. n, p. 68). This also suggests that the initiation of the
male hormone action as inhibitor of the secondary sexcord region
was longer delayed in this 32 mm. one than in Lillie's 3.75 cm.
one. This may be explained on the basis of the small amount of
hormone passing to the female at first through the small vascular
connection. This would delay the effective minimum concentra-
tion of the hormone. That there is such an effective minimum,
in some mammals at least, is shown by the partial castration and
grafting experiments of Moore ('21), Sand ('19), Steinach ('20),
and many others for postnatal stages. No prenatal testis grafts
have yet been possible in mammals so far as the writer can learn.

It is possible that the vascular anastomosis was just becoming
established in these twins and that, as the pregnancy progressed,

s

it would have become more effective by the pressing together of
the chorions. This would insure a continuous larger supply of
hormone on the female side and the rate of modification in this
case would be as fast as in those with earlier complete anastomosis,
though the condition reached at any time would lag behind that
of the others and this freemartin would fall in the group showing
small degree of modification as classified by Willier ('21).

Willier ('21) classifies freemartins into three classes on the

NOTES ON A 32-MILLIMETER FREEMARTIN. 249

basis of degree of modification — those with small, medium, and
large amount of abnormality — and Lillie states that the free-
martins "form an exceedingly well-defined group without inter-
grades to normal females or normal males." This may be true
for postnatal freemartins; but, from the series discussed in this
and other papers cited, it appears that a series of intergrades
connects the female with the freemartin group, or that the free-
martin group moves steadily away from the female group of
types, as we pass from younger to older stages. No such series is
seen between the freemartin and male. The rare cases of modifi-
cation externally in the male direction are probably due to upset
of some kind in genetic or zygotic sex balance and not to the
freemartin type of hormonic modification alone. Such cases
occur without twinning in cattle, and in as great a percentage of
cases as in freemartins. It is this series in the younger stages
leading back to the normal female type among freemartins that so
conclusively shows that the freemartin is a modified female and
not a modified male. Taken with Lillie's ('17) statistics as to
sex-ratios in cattle twins, it furnishes most conclusive proof of
the theory of the freemartin of Lillie and of Keller and Tandler.
Lillie ('23) suggests that probably there are individual differ-
ences in zygotic balance of sex factors in cattle even in single
born calves, and those who have been much associated with
cattle breeding and judging, and have seen the differences in
potency of males and females inter se, and their differing degrees
of male and female "conformation," will agree. This appears to
be a fundamental factor in the condition of all freemartins and
doubtless works from both sides, male as well as female, in the
pair. The hormone, even, may not be equally potent in all cases.
Its appearance may be delayed in some males for this reason.
So the initial action may be delayed even in cases with early
complete intercommunication of blood vessels. The females
also may differ in their resistance to the masculinizing action of
the male hormone: they may differ already in femaleness.
Comparison between this case and Lillie's and other small free-
martins illustrates how these factors may lead to variations, among
the early stages at least. It is also just possible that the differ-
ences from the control female found in this specimen may be
17

250 THOMAS HUME BISSONNETTE.

within the limits of individual variation resulting from these
differences in zygotic sex balance. Only a careful study by bio-
metric methods of a large number of single female embryos at
this stage can settle this point. Such a study has not been made
because of lack of material and the long time involved in search.
In any case, should this case fall inside the limits of individual
variation discussed in the preceding paragraph, it will show all
the more conclusively that the first stage of the freemartin is, at
least, in some cases, a female, normal for that stage of develop-
ment.

CONCLUSIONS.

1. On the basis of this 32 mm. freemartin and comparison with
others described by the writer and by others, where the free-
martin modifications are conditioned by vascular unions between
the placentae and not by mere contact and adhesion of one to the
other, it is concluded that a blood-carried agent in the nature of
a hormone is effective in producing the freemartin rather than an
"organizer" of the type found in Spemann's work on early am-
phibian development.

2. Small vascular or capillary anastomosis is sufficient to con-
dition the freemartin effects in some cases at least ; but such small
amounts of transfusion as are possible in such cases at first ap-
pear to delay the modifications and permit longer female develop-
ment before the effective minimal concentration of hormone is
reached on the female side. Early stages in such cases show a
more markedly female embryo than do those with strong anasto-
mosis, particularly in respect to gonad and mullerian ducts.
Study of such borderline cases as these lend support to the "all-
or-none" conception of the hormone action.

3. Cords of Pfliiger or secondary sexcord regions develop
frequently in freemartins and are resorbed later at unequal rates
in different parts of the gonad, so that some of the peculiarly
complicated interiors of gonads figured by previous students of
the freemartin gonad can be interpreted in the light of these
earlier stages as the results of survivals of isolated parts of the
secondary sexcord region, separated from the medulla of the
gland by tunica albuginea. Organization into sexcords may not
be possible from the first and some individuals may not develop

NOTES ON A 32-MILLIMETER FREEMARTIN. 25!

cords though they may have the cell proliferations from which
the cords would later develop but for the action of the hormone.
This unequal development of sexcord region and unequal re-
sorption seem to be related in some way to pressure by surround-
ing organs.

4. There is no intergrading in freemartins between the free-
martin types and male types except in rare cases which resemble
the anomalies found among single born males, where the sex
balance appears to be upset. And in postnatal freemartins there
is no intergrading to the female type. But in prenatal free-
martins there is a definite series of intergrades to the female con-
dition as the younger stages are studied till practically the female
condition is reached. This furnishes the strongest of proof for
the theory of the freemartin maintained by Lillie and by Keller
and Tandler.

5. There are probably individual differences in zygotic sex
balance among both male and female cattle. This is a factor in
the variations of condition among freemartins and doubtless is
operative from the bull's side as well as on the freemartin side.
The hormone may not be equally potent in all cases and it may
possibly differ in potency at different times in the same male, or
be delayed in its appearance. In this way the hormone action
may be delayed even when early complete anastomosis is effected.
So, too, it is possible that the differences between this freemartin
and the control female may be within the limits of individual
variation correlated with zygotic differences in sex balance, for
the exact limits of such variation have not yet been determined.

6. If so, it is even stronger proof for the theory of the freemartin
of Lillie and of Keller and Tandler, since it would show that the
first stage of the freemartin is a female, normal in sexual condition

for that age.

SUMMARY.

1. A 32 mm. freemartin twin to a 34 mm. male is compared
with a normal female of like size as to reproductive organs, and
with normal male as to gonad.

2. It differs from the female control in having a thinner second-
ary sexcord region and thinner miillerian duct, lacking lumen in
places where the control has one. Nevertheless it is more like

252 THOMAS HUME BISSONNETTE.

the female than any other freemartins with vascular anastomosis
so far studied.

3. Placental vascular anastomosis between the twins was very
slight — hardly more than capillary — and the placentae pulled
apart with slight bleeding. So it permitted very limited trans-
fusion of male hormone, a situation correlated with slight modifi-
cation from female type shown in small gonad and thin ovarian
cortex and mullerian duct.

4. Comparison with other freemartins as to placental and
sexual conditions indicates a blood-borne hormone as the agent
rather than an "organizer" found effective in early amphibian
differentiation. Mere tissue cohesion fails to condition the modi-
fications as it does in Spemann's transplants.

5. Comparison with other gonads figured for freemartins and
with new material indicates that some formerly anomalous regions
in older freemartin gonads are derived from secondary sexcord
remnants which failed to be resorbed and that these cortical
regions frequently develop in freemartins to some degree and may
or may not disappear.

6. This study adds support to the theory of the freemartin of
Lillie and Keller and Tandler, that it is female, normally deter-
mined zygotically, but abnormally differentiated sexually by the
intervention of a male hormone received from her twin brother
in utero.

REFERENCES.
Bissonnette, T. H.

'24 The Development of the Reproductive Ducts and Canals in the Freemartin

with Comparison of the Normal. Am. Jour. Anat., 33: 267-345.
Brambell, F. W. R.

'27 The Development and Morphology of the Gonads of the Mouse. Part I.
The Morphogenesis of the Indifferent Gonad and of the Ovary. Proc. Roy.
Soc., Series B, Vol. 101, No. 6711, pp. 391-408.
Chapin, Catherine L.

'17 A Microscopic Study of the Reproductive System of Foetal Freemartin?.

Jour. Exp. Zool., 23: 453-482.
Huxley, Julian S.

'22 Some Recent Advances in Biological Theory of Sex. Jour. Roy. Soc. Arts,

70: 187-202, 206-220.
Keller, Karl, und Tandler, Julius.

'16 Ueber des Verhalten der Eihaute bei der Zwillingstrachtigkeit des Rindes,
Untersuchungen ueber die Entstehungsursache der geschlechtlichen
Unterentwicklung von weiblichen Zwillingskalbern, welche einen mann-
lichen Kalbe zur Entwicklung gelangen. Weiner Tierarztl. Wochenschrift,
III. Jahrg., pp. 513-526.

NOTES ON A 32-MILLIMETER FREEMARTIN. 253

Lillie, F. R.

*

'16 The Theory of the Freemartin. Science, 43: 611-613.

'17 The Freemartin: A Study of the Action of Sex Hormones in the Foetal Life

of Cattle. Jour. Exp. Zool., 23: 371-451.

'23 Supplementary Notes on Twins in Cattle. BIOL. BULL., 44: 47-77.
Moore, C. R.

'21 et seq. On the Physiological Properties of the Gonads as Controllers of the
Somatic and Psychical Characters. III. Artificial Hermaphroditism in
Rats. Jour. Exp. Zool., 33: 129-171.
Pezard, A.

'20 Secondary Sexual Characters and Endocrinology. Endociin., 4: S37-54O.
Sand, Knud.

'19 Experiments on the Internal Secretion of the Sexual Glands, Especially on

Experimental Hermaphroditism. Jour. Physiol., 53: 257-263.
'23 Experiments on the Endocrinology of the Sexual Glands. Endocrinology,

7- 273-301.
Spemann. H.

'25 Some Factors of Animal Development. Brit. Jour. Exptl. Biol., 2: 493.
Steinach, E.

'20 Kunstliche und Natiirliche Zwitterdriisen und ihre analogen Wirkungen.

Archiv. f. Entw.-Mech., Bd. 46, S. 12-25.
Willier, B. H.

'21 Structures and Homologies of Freemartin Gonads. Jour. Exp. Zool., 33:

63-127-

ABBREVIATIONS.

Bl., bladder.

B. V. GL, bulbovestibular or Bartholin's glands.

Clit., clitoris.

COM., gonad.

Gr. Ph. Un., undercut groove in phallus.

Gub., gubernaculum.

Mesoneph., mesonephros.

Mill. Dt., miillerian duct.

Ost., ostium tubae abdominale.

Reel., rectum.

Sac. Vag., saccus vaginalis.

Scr., scrotum.

Ura., urethra.

Vest., vestibule.

Vul., vulva.

W. Dt., Wolffian duct.

THE DIGESTIVE SYSTEM AND ITS FUNCTION IN
FUNDULUS HETEROCLITUS.

B. P. BABKIN AND D. J. BOWIE.

(From the Atlantic Biological Station, St. Andrews, N. B. and Department
of Physiology, University of Toronto.)

In the course of investigation of the function of the alimentary
canal in Fundulus heteroditus (North American common killifish,
mudfish, mummichog1), it was found that this animal does not
possess a stomach, i.e., an organ secreting the pepsin-hydrochloric
acid. Although several species of fishes do not possess peptic
glands (see in Oppel's ' Lehrbuch der Vergleichenden Mikro-
skopischen Anatomic der Wirbeltiere," i Teil, Jena 1896, S. 33,
the corresponding table), we have never seen any reference to any
investigation concerning the gastric digestion in Fundulus hctcro-
clitus.1- From the anatomical as well as from the physiological
point of view, the digestive system of the Fundulus presents great
peculiarities, which are worth attention. The present paper is to
be regarded as a preliminary communication. A series of investi-
gations were made by one of us (B. P. B.) during the summer of
1926 at the Atlantic Biological Station, St. Andrews, N. B., and
subsequently one of us (D. J. B.) made histological examinations
at the Department of Physiology, University of Toronto.

^H ANATOMICAL DATA.

The whole family of Poeciliidse are comparatively small fishes.
The largest specimens of Fundulus taken at Birch Cove on the
Bay of Fundy, near St. Andrews, N. B., were from 65 to 83 mm.

1 It is interesting to note that the whole family of Cyprinidae (Cyprinus
carpio, etc.) are deprived of peptic glands. Fundulus belongs to the family
of Cyprinodontidae, and the Germans call it " Amerikanische Zahnkarpf en "
(Brunning (2)). According to Gill (3) "the Cyprinodonts or Poecilids
are really related to the Esocids and Umbrids, and to them they should be
approximated in the sub-order Haplomi." The results of the present in-
vestigation, especially the absence of the glands secreting the pepsin-hydro-
chloric acid, show that Fundulus has some features in common with the
Cyprinidae.

254

DIGESTIVE SYSTEM IN FUNDULUS IIETEROCLITUS.

255

in length, measured from the snout to the insertion of the caudal
fin.

The alimentary canal in this species is comparatively short,
being about equal to the length of the body, excluding the caudal
fin. On opening the abdomen and carefully dissecting away the
liver, we see a short oesophagus directly connected to the intestine.

FIG. i. Diagram of the alimentary tract of Fundidus. (E, oesophagus;
D, duodenum; GB, gall-bladder; P. I., principal islands of Langerhans ; 5",
spleen ; L, liver ; P ', pancreas ; A, anus.

Longitudinal sections through the junction of oesophagus and in-
testine show that, at this point, there is a well-developed sphincter
quite comparable to the pyloric sphincter in mammals. The in-
testine is bent upon itself ventrally and to the right to form three
portions ; a first portion, descending ; a second portion, ascending ;
and a third portion, descending to the rectum (Fig. i). Pyloric
caecse are not present.

The first part of the intestine, especially when distended with
food, has a somewhat pear-shaped form. The second and third
parts of the intestine do not display any such unusual degree of
dilatation when food is present in them.

In Fig. 2 is represented a scheme of the alimentary tract of

256

B. P. BABKIN AND D. J. BOWIE.

5 C a. I C.

FIG. 2. Scheme of the alimentary tract of Fundulus. The intestinal tract

is fully relaxed.

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS.

257

Fimdulus. The actual measurements are taken from a female
Fund ul us whose body length (caudal fin excluded) was equal to
78 mm.

The capacity of the first part of the intestine is four times greater
than the capacity of the second part. Without pressure in the first
part, four drops of water enter ; in the second part only one drop
enters. Under a moderate pressure (filling from a pipette) the
cardiac part retains 7 drops, and 8 drops more may be introduced
into it by using a greater pressure. Thus the arrangement of the
initial part of the alimentary canal of Fimdulus is similar to the
so-called " syphonal ): stomach of many other fishes. From a
physiological point of view the first part of the intestine of the
Fimdulus fulfils one of the duties of the stomach, being a receptacle
and container for the food.

In order to avoid repetition of cumbersome phraseology we shall
frequently designate the first part of the intestine as the duodenum,
since, as will be shown, it is into this part that the bile and pan-
creatic juice are poured. Both the physiological and the anatom-
ical observations show that a stomach is really absent in Fimdulus.

The liver is comparatively large, as it is in most fishes, and lies
in a left ventral position, closely applied to the oesophagus and to
the loop formed by the first and second portions of the intestine.
In the upper part of the longitudinal space bounded by the liver
and the three portions of the intestine, we find the gall-bladder, the
spleen and the " principal " islets of Langerhans, along with a con-
siderable portion of the diffuse pancreas. The pancreas has a
wide distribution along the three parts of the intestine and is so
diffuse that in fresh specimens it cannot readily be differentiated
with the naked eye. The gall-bladder is comparatively large in
these fish ; in a specimen' 78 mm. in body-length it measured 6 mm.
by 3.5 mm., and had a capacity of 2.5 drops. The duct which
drains the gall-bladder arises from the posterior end of that vesicle,
but soon turns in an anterior direction and, after passing along the
first part of the intestine to within 3 or 4 mm. of the oesophagus,
it pierces the intestinal wall.

It is important to note that the pancreatic duct, though it runs
parallel to the bile duct, does not unite with the latter to form a
common duct, but has its own separate opening into the intestine

258 B. P. BABKIN AND D. J. BOWIE.

beside the orifice of the bile channel. The hepatic ducts com-
municate with the cystic duct to form the common bile duct, though
one or more of them may empty directly into the neck of the gall-
bladder. The spleen is a small, flattened, bright red organ, and
serves as a convenient landmark for locating the so-called principal
islets of Langerhans, which are situated in the region between the
neck of the gall-bladder and the duodenum. The largest of them,
the " chief " principal islet, appears as a whitish, spherical nodule
about i mm. in diameter. The pancreas is best studied in a series
of stained sections. These show that slender processes extend
off from the larger portions of the gland to invade the liver,
particularly along the blood vessels and less so along the hepatic
ducts. Sections show also that a thin layer of pancreatic tissue
completely envelopes the gall-bladder. Serial sections made with
the aim to study the duct connections of the pancreas which sur-
rounds the gall-bladder showed that the ducts are leading away
from this pancreatic tissue and communicating with the larger
pancreatic ducts. It is important to note, however, that in no
case were we able to find any evidence of communication between
any of the channels for bile and any of those for pancreatic juice.

METHODS.

Since the fistula method is not practical for the study of the
digestive processes in fishes, after a suitable interval following
natural or forced feeding the contents of the duodenum were
sucked out by means of a pipette with a rubber bulb on the end.
This method had the well known disadvantage that the intestinal
secretions were mixed with food masses and also that removal of
the partially digested food caused more or less disturbance of the
normal process of digestion. Fiinduli are exceptionally hardy
fishes and did not show any ill effects from the manipulations con-
nected with the investigations of their intestinal contents.

The fish were kept in small aquaria or in glass jars containing
running sea-water, or else in jars in which the water was kept cool
(below 15° C.) and was well aerated. When kept in non-aerated
sea-water which is allowed to reach a temperature of 15° to 20° C.
(a comparatively high temperature for these Funduli), the fish
very often eject the food through the mouth and some of them
may die under such unfavorable conditions.

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS. 259

The reaction of the intestinal contents was first tested by means
of litmus paper. Since the indications of litmus are not very
reliable, whenever possible the hydrogen ion concentration was de-
termined by the "spot" method elaborated by Felton (4). The
values of the pH found in the present investigation must be looked
upon as approximate only, because in most cases the intestinal
fluids were mixed with food masses, a fact which would introduce
the "protein" and "salt" errors described by Clark (5). The
presence of bile in the intestinal contents was determined by
Gmelin's test and by the surface tension test with flowers of sul-
phur. The methods employed for the detection of different enzy-
matic actions will be given later on.

As a preliminary measure all of the fish were deprived of food
for one or more days prior to the experiments.

EXPERIMENTS ON FASTING Funduli.

Usually in fishes which have fasted for one or more days the
duodenum is practically empty, it being difficult to obtain even as
much as one drop of fluid in the pipette and what is obtained is
composed partly of mucous. In some cases one may obtain two
or three drops of a yellowish fluid which gives quite positive results
in tests for bile, and occasionally the remains of copepods and
other small animals may be found when microscopic examination
of the contents is made. After 7 to 10 days of starvation, it is
hard to get anything at all from the duodenum. Water is very
seldom present in this part in fasting Funduhis.

The contents of the duodenum of fasting Fund ul us is always
alkaline to litmus, and its hydrogen ion concentration, as deter-
mined in more than 50 fishes ranges from pH 8.0 to pH 9.2. The
variations in alkalinity depend chiefly on the presence or absence
of bile, since bile taken directly from the gall-bladder has a hydro-
gen ion concentration of only pH 7.0 to 7.2. When mucus alone
was present, the values were pH = = 8.8 to pH = = 9.2, whereas when
bile also was present the values recorded were from pH = = 8.0 to
pH = 8.6. Regarding the possible presence and influence of sea
water, we may say that samples taken from the sea-water pipe
system at St. Andrews Biological Station were found in several
determinations to have pH values of from 8.O to 8.1. The con-

260

B. P. BABKIN AND D. J. BOWIE.

tents of the second and third parts of the intestine of freshly
killed fasting fish were also always alkaline to litmus.

EXPERIMENTS ON FED Funduli.

Although Funduli prefer animal food such as clam and the in-
ternal organs of dead fishes, even of their own species, they are
practically omnivorous and will eat fish muscles, raw meat, bread,

TABLE I.
DIGESTION OF CLAM IN Fundulus.

Date.

Fish
No.

Contents of Intestine.

Reaction of the
Contents.

pH.

Litmus.

July 17:

i

Yellowish fluid

8.8

Strong alkaline

9.30 A.M.

2

ft it

8.6

** **

to

3

mucus

8.8

tt n

10. oo A.M.

4

Mucus and water

8.6

Alkaline

5

Mucus

8.8

Strong alkaline

10.15 A.M.

12 pieces of clam were put in the

aquarium; 6 pieces were eaten

at once. At 12.00 noon two

pieces were left, which were

removed from the aquarium

12.15 P.M.

3

Pieces of clam ; no bile

8.4

Alkaline

4

Scant; no bile

8.8

Strong alkaline

2.15 P.M.

i

Clam; bile

8.2

Alkaline

2

** **

8.8

"

3

Scant; "

8.6

* t

4

Nil.

8.6

* <

5

Clam; bile

8.4

« *

7.15 P.M.

I

Nil.

8.4

ii

2

Clam; bile

8.2

t *

3

Mucus only

8.8

Strong alkaline

4

i t t t

9.0

* * * *

5

Clam; a little bile

8.8

ti it

July 18:

9.00 A.M.

5

Yellowish mucus

9.2

tt it

hardboiled egg (yolk and white) and even watermelon. They
digested perfectly well milk introduced by a pipette into the duo-
denum. The alimentary tract of freshly caught Funduli may con-
tain mud, and small animals, insects, copepods, etc., but no matter
what might be the nature of the food eaten or the stage of its

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS. 26l

digestion, the reaction in the duodenum as well as in the second
and third parts of the intestine was always alkaline, both in fish
that were caught and investigated at once, on the spot, and also in
those kept in aquaria and fed naturally or forcibly. However, the
degree of alkalinity varied somewhat, according to the nature of
the contents. As stated above, the reaction in the duodenum of
fasting Funduli is decidedly alkaline, pH = - 8.0 to 9.2. The pres-
ence of foods rich in protein, such as meat and clam, which are
acid ; of milk, which is neutral or only very slightly alkaline ; of
sea-water (pH = = 8.2) ; and particularly the presence of consider-
able amounts of bile, which is also nearly neutral (pH = 7.0 to
7.2), all tend to reduce the degree of alkalinity.

Table I shows the course of digestion in Fundulus after natural
feeding with clam. The contents of the duodenum were sucked
out at certain intervals for each of five fishes. Prior to feeding
they had been fasting for five days, and no excrements were found
in the aquarium during the last two days of this period. Shortly
before the fast was terminated, the pipette was introduced into the
duodenum to withdraw any fluids which might be present.

Results analogous to those shown in the table were obtained
with all the other kinds of food ; the reaction in the duodenum was
always alkaline.

The following experiment shows that there is an active discharge
of bile into the intestine after a meal has been swallowed.

EXPERIMENT OF AUGUST 4.

In an aquarium in which nine fishes were kept for several days without
food, at 8.30 A.M. there were placed several pieces of clams' flesh, but the
amount was not sufficient to feed all of them. At 8.30 P.M. all of these
fishes were killed and the contents of the duodenum and of the gall-bladder
were investigated. In five fishes the duodenum was filled with clam and
bile, while the gall-bladder was quite empty and collapsed. In the other
four fishes the duodenum was empty, while the bladder was distended with
bile.

In the case of injection of milk into the duodenum, the digestion
was also always alkaline, the degree being somewhat less in the
first hours (pH = = 7.8) than in the later hours (pH = = 8.8). Bi!e
was present in the contents. The pH of the milk itself was
PH = 7.2.

262 B. P. BABKIN AND D. J. BOWIE.

Alkaline reaction was also always found during the digestion of
raw beef introduced into the duodenum. Six and one half hours
after forcible feeding the contents were sucked out and the reddish-
yellow fluid obtained gave an alkaline reaction, pH==8.2.

REACTION TO VARIOUS STIMULI.
Mechanical Stimulation.

Experiment of July 10. — As a means of mechanical stimulation,
a few small pieces (2 mm. cubes) of cork or rubber were intro-
duced into the duodenum in three Funduli, which were starved for
at least three days. Just prior to the experiment the contents of
the duodenum, consisting of a very little amount of mucus and
some bile, were tested and found to be alkaline, pH = 8.6 to 9.0.

At 10.15 A.M., three pieces of cork were introduced into the duodenum
of one fish; two pieces into that of the second; and two pieces of rubber
into that of the third fish. The duodenal contents were sucked out at 2.15
P.M. and at 6.15 P.M. At both times, in each case, one could remove from
2 to 4 drops of water, with some mucus, but in only one fish was bile pres-
ent, as indicated by the slightly yellowish colour of the contents. The re-
action was in all cases alkaline, pH = 8.2 to 8.6. The next morning the
corks were found in the aquarium, stuck together with mucus and prob-
ably had been ejected through the anal opening.

It is interesting to note that towards the end of the experiment
water was found in the duodenum of these fishes, since its presence
is quite unusual. From these experiments it appears that purely
mechanical stimulation of the duodenum does not provide the
necessary stimulus to provoke the pouring of bile into the intestine
but that it does cause the fish to " drink water."

Hydrochloric Acid.

Table II, gives the results of introducing a few drops of 0.36
per cent, hydrochloric acid into the duodenum in two fishes that
had been starved for 10 days. Its introduction did not provoke a
flow of bile into the gut.

The table shows that in 6 hours after the introduction of 0.36
per cent, hydrochloric acid the duodenal contents had again become
alkaline. It is interesting to note that in warm blooded animals
the introduction of hydrochloric acid in the concentration in which

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS.

263

TABLE II.
INTRODUCTION OF HC1 INTO DUODENUM.

Date and
Time.

Fish
No.

Contents of Intestine.

pH.

Litmus.

August 13:
9.00 A.M.

i

2

Almost nil.

—

Alkaline

9.10 A.M.

Introduced 5 to 6 drops
0.36 %HC1

10.10 A.M.

I
2

Mucus and fluid

4-8
5-0

Acid

ii. 10 A.M.

I
2

Mucus; very little fluid l

S-o

5-4

1 *

12.10 P.M.

I
2

Mucus only l
" " i

S-o
5-4

1 4

3.10 P.M.

I
2

Mucus and water
Watery fluid

7-8
7-6

Very weak alkaline

it is in the gastric juice, i.e., 0.5 to 0.36 per cent, practically does
not provoke the entry of bile into the duodenum.

Alcohol.

From 4 to 6 drops of 5 per cent, pure ethyl-alcohol was intro-
duced into the duodenum by means of a pipette 3 to 4 times per
day. Four sets of analogous experiments were performed. In
one of them, 2 fishes received alcohol regularly during 6 days.
In the earlier experiments, in order to make sure that the alcohol
reached the duodenum, it was mixed with a little finely ground
carmin and in every case the red granules were afterwards found
in the duodenal contents and also in the excrements.

The results of one of these experiments are shown in Table
III. The fish had been starved for 2 or 3 days before the ex-
periment, and the duodenum was almost empty. The very small
amount of fluid present was alkaline to litmus. When the duo-
denum was evacuated 2 or more hours after injection of alcohol,
it was found to contain only a very small amount of fluid, usually
mucus. Except occasionally in the morning tests after a night

1 To determine the reaction in these cases a few drops of distilled water
were added to extracted material.

264

B. P. BABKIN AND D. J. BOWIE.

when the fish did not receive injections of alcohol, there was no
bile present in the duodenum. The alkalinity of the contents was
very high, pH = = 8.8 to 9.2, but it was not determined whether this
was due to the presence of pancreatic juice or of intestinal juice.
The alcohol injected was slightly acid, pH = = 6.8.

TABLE III.
INTRODUCTION OF ALCOHOL INTO THE DUODENUM.

Date and Time.

Contents of
Intestine.

pH.

Litmus.

5 % Alcohol
Injected.

July 19:

4.30 P.M.

— •

—

4 drops

7.00 P.M.

I drop of fluid

8.6

Alkaline

41 * *

July 20:

9.30 A.M.

Mucus

9.0

"

* *

2.45 P.M.

2 drops

8.8

Strongly alkaline

t i

7.00 P.M.

2 drops

8.8

Alkaline

* *

July 21 :

11.25 A.M.

i drop and mucus

8.2

Alkaline

* *

12.25 P.M.

3 drops

8.8

Strongly alkaline

* (

3-30 P.M.

2 drops

9.0

n .*

< i

7.15 P.M.

2 drops

9.0

14 ti

None

July 22:

11.50 A.M.

Mucus only l

9-4

it t 1

4 drops

July 23:

9.45 A.M.

Nil.

— •

<f «

11.50 A.M.

3 drops

8.8

Strongly alkaline

1 1 i i

2.45 P.M.

i drop and mucus

9.0

** * *

" * '

July 24:

9.15 A.M.

Mucus only l

9.2

« 4 4

fi it

12.15 P-M.

i to 2 drops

9.2

It 11

None

N. B. Bile was not found in the contents during this experiment.

It is important to note that the experiments with alcohol afford
further evidence that gastric glands are absent in Fundulus. Al-
though alcohol is one of the strongest stimuli for the secretion of
gastric juices, in our experiments the intestinal contents never
had an acid reaction, but, as stated above, were always highly
alkaline.

Pilocarpin.

Table IV. gives the results of one of eight sets of experiments
with pilocarpin. Five or 6 drops of i per cent, solution of pilo-
carpin was injected into the duodenum of two fishes which had

1 To determine the reaction in these cases a few drops of distilled water
was added.

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS.

been starved for 7 days. At the end of this period the duodenum
contained in each case a small amount of mucus and fluid which
was strongly alkaline, pH = = 9.2 and 9.0 respectively.

The interesting feature of the experiments is that pilocarpin
provoked chiefly the motor phenomena of the alimentary canal.
Its introduction into the duodenum excited evacuation of the
bowel and this was accompanied by the entry of the bile into the
intestine. The contents of the duodenum, extracted one hour
after injection ot the pilocarpin, had a yellowish or greenish
colour and gave a positive test for bile, with flowers of sulphur.
Following the injection, the alkalinity of contents of the duodenum
was comparatively low, pH between 7.8 and 8.6.

TABLE IV.
INTRODUCTION OF PILOCARPIN INTO THE DUODENUM.

Date and
Time.

Contents of Intestine.

pH.

Litmus.

August 10 :
8.35A.M.

i. Mucus and some fluid
2. '

9.2
9.0

Alkaline

8.45 A.M.

Injected 5 to 6 drops of i % sol. of
pilcarpin into each fish

9.45 A.M.

i. 2 to 3 drops of emerald green bile
2. 8 drops of green fluid; re-injected
6 drops of it

S.o

7.8

Weakly alkaline

10.45 A.M.

i. 2 to 3 drops of light green fluid
2. 10 drops of green fluid; re-injected
8 drops of it

8.4

8.2

14 (4

11.45 A.M.

i. 2 to 3 drops of light green fluid
2. 8 drops of green fluid

8.4

8.2

Alkaline
Weakly alkaline

3.15 P.M.

i. 4 to 5 drops of a very light green
fluid with mucus
2. 3 to 4 drops of clear, colorless fluid

8.6
8.6

Alkaline

Atropin.

The injection of i per cent, atropin sulphate into the duodenum,
as one would expect, inhibited the intestinal secretions (it was
hard to get anything but mucus from the intestine) and also pre-
vented the entry of bile into the gut. Table V. gives the results
of experiments on two fishes which had been starved for 10 days.
18

266

B. P. BABKIN AND D. T- BOWIE.

The reaction of the duodenal contents was usually strongly alka-
line.

It is interesting to note that one hour after the introduction of
atropin into the intestine, the skin of the fish became dark, almost
black. In five hours the normal colour of the skin had returned.
These changes were not due to light effects, as the whole day was
grey and cloudy.

From the foregoing experiments it may be seen that the first
part of the alimentary canal of Fundulus may properly be desig-
nated the duodenum, since it corresponds to the duodenum of
vertebrates possessing a stomach.

TABLE V.
INTRODUCTION OF ATROPIN INTO THE DUODENUM.

Date and Time.

Contents of Intestine.

pH.

Litmus.

August 13:

9.10 A.M.

i. Nil.

— •

Alkaline

2. "

—

"

9.20 A.M.

Injected 4 to 5 drops of i % sol. of

atropin sulphate into each fish

10. 20 A.M.

i. 2 to 3 drops of fluid, with mucus

9.0

Strongly alkaline

2. i drop of fluid, with mucus

8.8

Alkaline

11.20 A.M.

i. Very little fluid and mucus; difficult

to obtain it

8.8

"

2. Very little fluid and mucus; difficult

to obtain it

8.8

« t

12.20 P.M.

i. Almost nil.; a little mucus

8.8

Strongly alkaline

2. Almost nil.; a little mucus

8.8

Alkaline

3.20 P.M.

i. Mucus and i to 2 drops of water

9.0

Strongly alkaline

2. Mucus and 3 drops of water

8.8

N. B. There was no bile present in the intestinal contents during the
whole experiment.

ENZYMES OF THE Mucous MEMBRANE OF THE FIRST PART OF

THE INTESTINE.

The only method which we applied to the study of the enzymes
of this mucous membrane was the reaction of its extract on various
media. The imperfections of this method are very well known.
A certain possibility that we might have to consider autolytic
enzymes, as well as digestive enzymes, was suggested by the ob-

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS. 267

servations of Bradley (1922"), on autolysis of liver and kidney.
He produced evidence to show that in autolysis there are two auto-
lytic enzyme-complexes, one of which is analogous to trypsin, but
has its optimum activity at pH 4 to 4.5, whereas true trypsin has
its optimum activity at pH 8. Beyond the range from pH 7 to
pH 3, the action of this autolytic enzyme-complex is insignificant,
while for trypsin the range of activity is pH 9 to pH 4.5. The
other autolytic enzyme-complex is analogous to pepsin, but has
its optimum reaction at pH 4.5, whereas pepsin has its optimum
at pH 1.5. It is destroyed at pH 2.6, while pepsin is very active
at that degree of acidity. The range of activity for pepsin is from
pH 0.5 to pH 6.5 ( McFarlane, et al.7)-

By adjusting the reaction of the media to hydrogen ion concen-
trations which are beyond the limits of activity for any autolytic
enzyme-complexes which might be present, one could expect to
obtain reactions due solely to digestive enzymes. To minimize
the possibilities of contamination all extracts of the intestinal mu-
cous membrane were kept under toluene and all test tubes and
pipettes were sterilized before use.

The extracts from the mucous membrane usually of the duo-
denum were prepared in the following way. The intestine, cut
out from the body, was cleaned from all surrounding tissues, split
longitudinally, and washed several times in tap water. Only the
mucous membrane was scraped by a knife and mixed with the
corresponding fluid. The temperature in the incubator was 35° C.

For use in the various experiments extracts of the mucous mem-
brane of the intestine were made with 30 per cent, alcohol, 0.9 per
cent. NaCl, glycerine, and 0.36 per cent. HC1. These were used
at various intervals after their preparation and were kept under
various conditions of temperature.

In preparing the extracts for use they were filtered, or other-
wise separated from the tissue cells, and were then diluted with 3
or 4 volumes of distilled water. The hydrogen ion concentration
of the mixture was adjusted to the desired level by adding either
0.36 per cent. HC1 or dry sodium carbonate. All tests of enzy-
matic action reported here must be looked upon as qualitative only.

268 B. P. BABKIN AND D. J. BOWIE.

Action of the Extracts on Protein (Fibrin').

For the experiments on the digestion of protein, fibrin was ex-
tracted from fresh blood and was then kept in glycerol. Before
use it was thoroughly washed ; dried between leaves of blotting
paper; and then minced and small portions of it were placed in
several test tubes along with the extract to be tested and a few
drops of toluene. The whole mixture was incubated at 35° C.

In order to rule out the problem of autolysis the reaction of the
media during these tests was adjusted to H ion concentrations of
pH - - 2 ; pH = = 8.0 ; pH = 8.4 and pH = 9.0 respectively.

No trace of digestion of fibrin was manifested in periods rang-
ing from 48 to 62 hours respectively, a fact which indicates that
the extract of intestinal mucous membrane contains neither pepsin
nor trypsin.

In one case, on the sixth day the fibrin was found to be partly
dissolved. The initial pH of the mixture in this case was 8.4, but
it may have been that during such a long period of incubation the
medium may have lost its alkalinity to such an extent that the
autolytic enzyme-complexes had started to act.

Action of the Extracts on Peptone.

In one set of experiments a very gradual but not complete dis-
coloration of a mixture of purified casein, phenol red, and saline
extract was observed during 24 hours in the incubator. The
initial pH of the mixture was 8.0.

In another set of experiments a 2 per cent, solution of Witte's
peptone was mixed with boiled, and with unboiled, saline intestinal
extract, respectively, and was adjusted to pH = 7.8. After 3 days
in the incubator biuret tests showed that there was less peptone in
the mixture with fresh extract, than in the control, in which boiled
extract was used, indicating that the fresh extract had a weak
ereptic action.

Action of the Extracts on Starch.

Table VI. gives the results of two experiments in which the
extract of the intestinal mucous membrane on normal saline was
allowed to act on starch. In one case fresh extract was used and
in the other the extract was boiled. In each case the mixture was

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS. 269

adjusted to pH 6.7 and 2 drops of toluene were added, after which
it was placed in the incubator. In the subsequent test for starch a
weak solution of iodine in potassium iodide was employed. The
table shows that, whereas the tests with boiled extract were nega-
tive, those with fresh extract gave a progressive diminution in the
amount of starch present and a corresponding increase in the sugar
content of the mixture, indicating a decided amylolytic action.

That the amylolytic action of the intestinal extract was due to
the enzyme contained in the mucous membrane and not to the pan-
creatic juice, which could be fixed on the intestinal mucous mem-
brane, in spite of its thorough washing, is seen from the fact that
proteolytic action was absent in the same extracts.

TABLE VI.

AMYLOLYTIC ACTION OF EXTRACT OF THE INTESTINAL Mucous MEMBRANE.

Date. Experiment I. Experiment 2.

August 4 3 cc. of 2% soluble starch 3 cc. of 2% soluble starch

10 drops boiled extract 10 drops fresh extract

pH = 6.7 pH = 6.7

Iodine reaction — blue Iodine reaction — blue

3.30 P.M Placed in incubator Placed in incubator

August 5 :

8.30 A.M Iodine — no change Iodine — light purple

Fehling — negative Fehling — positive

pH = 6.7 pH = 6.7

7.30 P.M Iodine — no change Iodine — light purple

August 6 :

10.15 A.M Iodine — no change Iodine — very light purple

Fehling — no change Fehling— strongly positive

pH = 6.6 pH = 6.7

7.30 P.M —no change — progressive change

August 7 :

7.30 PM Iodine — no change Iodine — almost colorless

Fehling — no change Fehling — strongly positive

pH = 6.6 pH = 6.7

Action of the Extracts on Lipoids. '

It was found that unboiled extract of the intestinal mucous
membrane had a weak lypolytic action on olive oil and a some-
what more pronounced effect on cream. The reaction on cream is
shown in Table X., experiments 5 and 6.

270

B. P. BABKIN AND D. J. BOWIE.

Enterokinasc of the Extract.

In addition to its action as a ferment acting directly on peptones,
starch and fats, it was shown that the unboiled intestinal extract
had also the power of greatly accelerating the tryptic action of
liver extract of Fundulus and of the bile : to a lesser degree, this
extract increased the action of lipase on the digestion of fat,
whereas boiled extracts had no such effects.

Activation of Trypsinogen.

Table VII. shows the results of adding unboiled saline intestinal
extract to a mixture containing minced fibrin and glycerin extract

TABLE VII.

ACTIVATION OF TRYPSINOGEN BY EXTRACT OF INTESTINAL Mucous

MEMBRANE.

Date.

Exp. i.

Exp. 2.

Exp. 3.

August 13:

12.00 Noon

1. 15 P.M.
3-30 P.M.
5.30 P.M.
6.30 P.M.

August 14:

9.00 A.M.
August 15:

9.00 A.M.
August 16:

8.30 A.M.

i. oo P.M.

3.00 P.M.

Minced fibrin

8 dr. glycerin ex-
tract of hepatic
pancreas

12 dr. water
No intestinal ex-
tract (control)
pH = 8.2

No change

Minced fibrin

8 dr. glycerin ex-
tract of hepatic
pancreas

10 dr. water
2 dr. intestinal ex-
tract (in saline)
pH = 8.2

No change
Digestion begun
Half digested
Completely digest-
ed

Minced fibrin

8 dr. glycerin extract
of hepatic pancreas;
and intestinal mu-
cous membrane pre-
pared a day before

12 dr. water

2 dr. intestinal ex-
tract (in saline)

pH = 8.2

Completely digested

Marked digestion
Half digested
Completely digest-
ed

N.B. In this experi-
ment (3) the ex-
tracts of hepatic
pancreas and of in-
testine were mixed
together on the pre-
vious day

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS. 271

of the liver of Fundulns. (As stated above, the liver of these
fishes is invaded by pancreatic tissue, so that extracts might con-
tain considerable trypsinogen.) In Experiment No. 3, the fibrin
was very quickly digested. The probable cause was that the liver
extract and the extract of the intestine had been mixed together
on the previous day and that in the interval the trypsinogen had
already become active trypsin. In Experiment No. 2, the activa-
tion of protrypsin required a latent period of about 3 hours. The
presence of glycerin in the mixture probably inhibited the process
of activation. In Experiment No. I, intestinal extract was not
used, and, therefore, digestion did not take place until the third
day. The most probable explanation of this phenomenon is that
it required such an interval for the spontaneous activation of the
protrypsin.

Other experiments which demonstrate the activation of tryp-
sinogen by intestinal extracts are recorded in Table IX.

Acceleration of Lipolysis.

The experiments in which the digestion of fat was accelerated
by the addition of unboiled intestinal extract to a mixture contain-
ing bile and cream are recorded in Table X. and in Table XI.

Enzymes in Gall-bladder Bile.

At the time of these experiments, the histological investigations
had not been made and, therefore, it was not then known that the
gall-bladder of Fundulns is surrounded by a thin layer of pancreatic
tissue, the presence of which introduces the possibility that a small
amount of pancreatic enzyme may have become mixed with the
bile during the manipulations for emptying the gall-bladder.

The gall-bladder was dissected out and removed from the body
with sterile instruments, the bile was pressed out of the viscus,
and the contents were collected in a sterile container. The bile
secured in this manner contained a very active amylase, trypsino-
gen and prolipase. The last two substances could be activated by
the unboiled extract of intestinal mucous membrane.

Table VIII. gives the results of two experiments to demonstrate
the amylolytic action of fresh bile.

272

B. P. BABKIN AND D. J. BOWIE.

August 7 :
Time.

9.30 A.M.

9.40 A.M.
10.40 A.M.
11.40 A.M..

2.40 P.M.

TABLE VIII.
AMYLOLYTIC ACTION OF GALL-BLADDER BILE.

Experiment I.

2 cc. of 2% soluble starch

2 drops boiled bile

pH = 6.6

Placed in water bath at

35° C.

Iodine — dark blue
Fehling — negative
No change

No change

PH = 6.6

Experiment 2.

2 cc. of 2% soluble starch

2 drops fresh bile

pH = 6.6

Placed in water bath at

35° C.

Iodine — purple co1or
Fehling — positive
Iodine — light purple
Fehling — strongly positive
Iodine — very light purple

almost colorless
Fehling — strongly positive
pH = 6.6

TABLE IX.

TRYPSINOGEN OF THE BILE AND ITS ACTIVATION BY THE EXTRACT OF THE
INTESTINAL Mucous MEMBRANE.

Date and Time.

Experiment i.

Experiment 2.

Experiment 3.

August 5 :

Minced fibrin

Minced fibrin

Minced fibrin

2 cc. bile

2 cc. bile

2 cc. bile

8 cc. dist. water

8 cc. dist. water

8 cc. dist. water

No extract

i drop boiled in-

i drop fresh intestinal

(control)

testinal extract

extract

pH = 8.3

pH = 8.3

pH = 8.2

10.20 A.M.

Put in incubator

Put in incubator

Put in incubator

12.05 P.M.

No change

No change

Digestion begun

1. 15 P.M.

tt n

tt tt

Half digested

2.00 P.M.

n tt

tt tt

Completely digested

7.30 P.M.

it tt

a tt

— •

August 6:

10.20 A.M.

tt n

tt tt

—

Added i drop fresh

Added i drop fresh

intestinal extract

intestinal extract

11.30 A.M.

Digestion begun

Digestion begun

12.05 P.M.

Half digested

Half digested

i.oo P.M.

Completely digest-

Completely digest-

ed

ed

Table IX. shows that the bile alone or in the presence of boiled
intestinal extract has no digestive action on fibrin, but when fresh

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS.

273

X

H

fc

<

OS

8

J

H
c/i

U

o

</l

W
fe

w
t/}

W

Q
fe

W
en

J

3

w

w

Q
O

o

cd

_a

CO

•O

IO

cy

o

-t-J

ti

0>

CU "*~*

2 'S

IH

rt

73
•o

cd

"C

1 1

3
cu

M

^ ffi

w

C rz3 . ^

3

1 :
o

'5, >5

^ cu *^ O.

u O cl

3

O

[M IH -o c -a

UZ -t

Z

J > H _ S

1

,4.

CO

CU S

•u
C

2 -s

0

4-)

cu

cd

a

to 73

1

to o *;

QJ o ^

0

cu
QO
C
cJ

,

w

1 5 ^ i

c3

O

S o 73 X

ul^1'

-*->
Cu

O

: : : g

1

73 73

73 73 73

PO

CO

cu cu

cu cu cu

£ -2

M

73 73

73 73 73

£

CU

a

H -s

•M

1

73 73

cd cd

c 53 c 53

'O *"O *O
cd cd cd

u

"^ ?

cu

CU Q CU Q

O QI ^

cu

a

M ^ ^ -"-*

C

Si cd Si cd

^ cd cd ^ cd

X

W

£2^2

_C

•*-* •*-*

5 i: >• c ^ 8

-^1-^ -^- U

U 73 uj

3

,£?73 .£? 73

.2? 73 IH "O .2 73 73

U TftN

Cu

,-J <N >_) ro

J M ON CU w 0

73

73 73 73

CS

CU

CU CU CU

riment

CU

4J

2

4->

M

c3

4J

1

CJ

73
73
cd

73 73 73

*o ^o *o
cd cd cd

&

ZM

3 cd

£ cd Si cd 3 cd

X

W

H-5
^

_C

4J

3

Z cu

J3 ^ D.

1^ 1^ |^ |

U TI-

Cu

J M CU

^ _j M i j r*"5 \^ HH \£i

H

•^J

cu

^

cd

i>

cu

u

cd

a

-t-1

to

3

-n
a

X

u

3l

cd o

u
C

M

cd -

o

: : :

Cu o
t* N^ ^

•4-J

3

O

*7^

u

Cu

Z

: : £

1 CO

73 U

73 3

cd O

(H "^

^J

jj ^

cu

^

^ ^

S S S -2 «

-4->

c3

Cu

CU Cu

cu < cu J2 c3

Q

o

o o

O •• O O § 'u

M

q

ro O

O W O O 3

M

CO

*^* O

OO 06 ro

bi)

M ° 7

3

3 CO V

< Z

274

B. P. BABKIN AND D. J. BOWIE.

X

w
j
«
<
H

(J

w

H
(«

a
H

55

K

H
>

n

Z
o

H

t/3

H

a

i—
PQ

o

Q

CQ

o

H
Z

en

PH
&

•a .i

d -g

•

^ 1

•d

*d *O *O *O "O

C

2^1

o
rt

•o

T>
C3

"T33 HZJ *O *G *O
*O 'O *O *O H3
re" rt cd cd cd

Experim

to S 2J *J

S? °l

C . ^ ^

Q) rj-\ rf

C ^ <N C

J3
1

4J

1

|0 CD 0 ^ |C>

l-a §£ N s-o' 1-5 *

U N

OH

12 M

jMO-OroOrojN M

•a .i

§ «

,

ra <u

4->

"*

IH

4-1

4_) '^? ""

O

C

4J

0)

*H ^

C3

E

m "S s t3

3

fl)

IH

CJ

&

g o -0 i;

C

-

^

rt

W

1 u "0 ~

•2

0

- ts C

3

0

— '

U 0)

OH

Z

: : : : : z

•a

ro

4->
U

«J^

2s

O

ta

•d
•3

1 1

Experim

4-> J3

CO (A

.0 <U

S-

**H

E °
^^

Uc ~

J2
O

tc
_c

4->

3

1

o aj i-pj

& 3 & "e-o

" b " £?M £

030 c T3 -a

U ro

OH

>2 M

*7 »^^ *7 « r h
A (> f-* W M N

_0)

^

S S

O

a

(U

2-g

•|

-^ O

O

OJ

M

w

E"°

i

CS
<J

8-a'

-*->

3

0

^;

U *r>

OH

z

: : : : : ^

M

cu

IH

-4-J

O

c

I

CO

1

i

X

W

If

— — •

11

U

M
CS

CJ

: : : : :

£i -3*

3

o

r^

u

OH

£

Z

1 CO

-0 *-

'O g

c

; ; " 4rj

g

S

o
o

S S J§ S ^ J J^

HVJ

<I<

2

PH OH PH OH "^ ti OC

cd

Q

O
O

o
o

O O O O O "3 .S
O O O fO"O OT*-

... fN. . * jj

M

M

<N

M M m oo M QO £*o

si

M 9 "2

3 cd ^

<;

< z

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS. 275

extract is used, tryptic digestion takes place in a short time. Ap-
parently trypsinogen was present in the bile, and was activated by
the fresh intestinal extract.

To these findings might also be added the fact that when toluene
is not used, self activation of the tryptic enzyme in the bile takes
place after one or two days at room temperature, and much more
quickly in the incubator.

Tables X. and XI. show that the gall-bladder bile contains a
prolipase, which may be activated by the extract of the intestinal
mucous membrane. From the same tables we learn too that the
intestinal extract possesses a weak lipolytic action and contains an
activator for the bile lipase.

In all of the experiments recorded in Tables X. and XI. boiled
cream, diluted with one volume of distilled water was used as the
substrate, and to this was added one drop of i per cent, phenol-
phthalein and sufficient 0.09 per cent NaOH solution (2 to 4
drops) to produce in each case a distinct pink color (pH = = 8.3).
This alkaline pink mixture of diluted boiled cream is referred to
in the table as the cream substrate. After decoloration of the
mixture in the incubator, a sufficient amount of 0.09 per cent.
NaOH solution was added to restore its initial pink color.

From Tables X. and XI. one may see that : the bile alone pos-
sesses a weak lypolitic action which develops slowly (Exp. 2,
Table X. and Exp. 3, Table XL) ; this action is due to an enzyme,
because boiled bile loses its lypolitic action (Exp. 2, Table XI.) ;
the intestinal extract alone has a weak lipolytic action (Exp. 5,
Table X.) ; boiled intestinal extract loses its enzymatic action
(Exp. 4, Table X.) ; when bile and intestinal extract are com-
bined the lipolysis is much accelerated (Exp. 3, Table X. and Exp.
5, Table XI.) ; boiling destroys the lipolytic action and the activa-
tion (Exp. 4, Table XL).

It is evident that the result of combined action of the bile and
of the intestinal extract is not merely a simple summation of their
activities, but indicates rather that the intestinal extract possesses
a substance which activates the bile viz. pancreatic lipase. There
is also a possibility that the bile increased the action of the intes-
tinal lipase.

Since the presence of pancreatic tissue on the gall-bladder makes

276 B. P. BABKIN AND D. J. BOWIE.

it possible that enzymes of the pancreas may have become mixed
with the bile used in these experiments, the latter must necessarily
be repeated with bile obtained by more adequate methods. If it
can be demonstrated that the enzymes in question were admixed
to the bile with the pancreatic cells surrounding the gall-bladder in
Fundulus, these experiments will offer a proof that the pancreas
of this animal elaborates the same enzymes as the pancreas of
warm-blooded animals, two of them in form of zymogen (pro-
trypsin and prolipase), and one in active form (amylase).

DISCUSSION.

The data reported in this investigation show that Fundulus
heteroclitus lacks a stomach, the intestine joining directly to the
oesophagus. Since there are several species of fishes with an
analogous structure of the alimentary canal, from the point of view
of their digestive system, all fishes may be divided into two main
groups, namely, (i) those possessing a pepsin-hydrochloric acid
digestion, and (2) those lacking it. This makes it possible to dif-
ferentiate two types of digestion in fishes ; namely, acid-alkaline,
and exclusively alkaline. There is very little doubt that these
decidedly different types of digestion greatly affect the internal
chemisms of the body, in the two cases ; because, in the one case,
important changes in the chemical composition of the body fluids
are provoked following the secretion of acid into the stomach.
Similar changes must necessarily be quite absent in fishes deprived
of a stomach and must be greatly attenuated in the pathological
syndrome, known as achylia gastrica, or in animals with experi-
mentally removed stomachs. Therefore, the study of the animals
of a type of Fundulus presents an interesting problem.

SUMMARY.

i. That Fundulus heteroclitus does not possess a stomach is
shown by the following evidence :

(a) There is absolute absence of pepsin and hydrochloric acid
in the digestive juices.

(b) Every phase of digestion takes place in an alkaline medium.

(c) The first part of the intestine (duodenum) is joined di-
rectly to the oesophagus.

DIGESTIVE SYSTEM IN FUNDULUS HETEROCLITUS. 277

The bile and pancreatic juice are poured into the gut only
a few millimeters (3 or 4) below the oesophagus.

(e) The duodenum is capable of dilating to form a container
for food.

2. Extracts of the intestinal mucous membrane manifest an
amylolytic, a lipolytic and an ereptic digestive action. They prob-
ably contain also an enterokinase.

3. The gall-bladder bile contains a trypsinogen, a prolipase and
an amylase. The origin of these enzymes must be investigated
further, because they could be admixed to the bile from the layer
of pancreatic tissue surrounding the gall-bladder, during its collec-
tion.

REFERENCES.

1. Jordan, D. S., and Evermann, B. W. Bull, of the U. S. Nat. Mus.,

Washington, 1896, Part I, p. 640.

2. Brunning, C. Wochenschr. Aquar. — Terrar. Kunde 1910, Jg. 7, p.

161.

3. Gill, T. N. Proc. U. S. Nat. Mus., 1895, Vol. 17, p. 115; and Ibidem,

1896, Vol. 18, p. 221.

4. Felton, L. D. Journ. Biol. Chem., 1921, Vol. 46, p. 229.

5. Clark, W. M. " The Determination of Hydrogen Ions," 2d ed. Bal-

timore, 1922, p. 118.

6. Bradley, H. C. Journ. Biol. Chem., 1922, Vol. 52, p. 467.

7. McFarlane, J., et al. Journ. Gen. Physiol., 1927, Vol. 10, p. 437.

The authors wish to acknowledge their indebtedness to the
Biological Board of Canada for the permission to use the facili-
ties of the Atlantic Biological Station, St. Andrews, N. B., and
to Professor A. G. Huntsman and his staff for valuable assistance
rendered in the course of the study.

Vol. LI V. April, 1928. No. 4.

BIOLOGICAL BULLETIN

GRAFTING AND REINCORPORATION IN ACTINO-
SPH/ERIUM EICHHORNTI EHR.

RUTH B. ROWLAND,
WASHINGTON SQUARE COLLEGE, NEW YORK CITY.

It is a well known fact that contiguous individuals l in a cul-
ture of Actinospliccrium cichhornii may fuse along the surfaces
in contact. Cytoplasmic union of this type, or plastogamy,2 has
been frequently described, and has often been considered as an
important initial step in the evolution of nuclear union.

In the closely allied form Actiuophrys sol, plastogamy is a pre-
liminary step toward nuclear fusion. A striking photomicrograph
of this process, taken from living material, is shown by Belar, '22,
(Taf. 3, Fig. 24). Fusion of a number of these individuals
around a large food mass, has recently been described by Looper,
'25. This union is temporary, for separation occurs as soon as
digestion is completed.

Attempts to produce plastogamous fusion artificially have met
with varying degrees of success. Cienkowski, '65, was able to
fuse two vegetative masses of Actinosphcerium if he brought to-
gether two cut surfaces, and held them together with paper strips.
Johnson, '94, could not bring about artificial coalescence in a
single instance. Greff, '67, and Penard, '04, observed that a
single vegetative mass, if fragmented into a large number of por-
tions under a cover slip, might eventually fuse into the original
form. In some cases, this may have been a simple contraction
phenomenon, for no attempt was made to break apart the cyto-
plasmic threads connecting the crushed fragments.

1 In behavior, the single sphere commonly termed an individual Actino-
sphterium, shows all the characteristics of a colonial aggregate. This would
seem the more accurate terminology to apply to each vegetative mass.

- Also spelled plasmogamy. Belaf, '22.
19 270

28O RUTH B. HOWLAND.

In the course of an extended physiological study of Actino-
sphserium eichhornii, it was observed that artificial plastogamy
could invariably be induced by suspending two or more animals
in a hanging drop small enough to exert moderate compression.
Various types of simple grafts were found to be successful when
sufficient pressure was exerted by this method to enforce tem-
porary contact.

Also, the interesting behavior of axopods 3 which were occa-
sionally severed from the main body during the transfer of
animals to hanging drops, seemed worthy of critical attention.
Reincorporation of pseudopodial fragments has been reported in
a few instances. Jensen, '96, and '01, describes this phenomenon
in Orbit olites and Amphistegina. Kepner and Reynolds, '23, ob-
served in Difflngia the recovery of separated pseudopodial frag-
ments by their cell bodies.

The results derived from these accidental and experimental
graftings, and from the severing of axopods, in Actinosphacerium

are presented in the present paper.

i

MATERIAL AND METHODS.

Actinosphcerium eichhornii was reared in large numbers in a
medium containing ciliates, rotifers and round worms. Sub-
cultures were made in Great Bear Spring Water, to which had
been added concentrations of Colpoda, Paramecium, Noteus and
Philodina.

Gross merotomy was performed under a binocular microscope.
To aid in differentiating between the cut portions of two indi-
viduals, half of the animals to be grafted were stained in a very
dilute solution of neutral red, made up in the original medium.
The neutral red penetrated easily, and stained the cytoplasmic
granules a deep red. A stained (red) and an unstained (white)
animal were placed in a Syracuse watch glass, and cut into halves
or quarters with a glass hair. Before normal contraction could
cause rounding of these portions, two or more were transferred to

3 Greeff (loc. cit.) saw " unipolar, pear-shaped cells," among the frag-
ments resulting from his crushing experiments. In all probability, these
were derived from axopodia, but their origin could not be accurately ob-
served when so drastic a method of separation was applied.

REINCORPORATION IN ACTINOSPH^ERIUM. 28 1

a hanging drop, the size of which was reduced until the animals
came into contact at the desired angle. These were transferred
to a moist chamber for observation under high power (Leitz oc.
8x, obj. /a).

Severing of individual axopodia was accomplished by means
of the Chambers micrurgical apparatus, and the subsequent be-
havior followed under high power.

RESULTS.
(a) Grafting of Cut Portions.

Grafted animals fused perfectly in all cases. The portions
taken from each of the two contributing animals remained dis-
tinctly marked off. The surfaces held in contact varied in posi-
tion and extent.

Record 3, (Text-figures i to 7), covers briefly the range of pos-
sible variations. In Figs. I and 2, two cut medullary surfaces
(M-M) were brought into contact. Two quadrants of a stained
individual brought into contact, medulla to medulla, (M-M), and
medulla to cortex, (m-C), are shown in Figs. 3, 4 and 5. Figs.
6 and 7 represent a case in which an entire stained (red) animal
was grafted on to half of an unstained (white) individual.

The time taken for complete fusion of surfaces varied from
five to twenty minutes. As the animals were returned to the
deeper water in a Syracuse watch glass, they resumed their normal
spherical shape.

The extent and position of the grafted portions was still recog-
nizable at the end of thirty hours.

(&) Merotomy and Reincorporation of Axopodia.

Axopodia were severed from the main body, carried away by
the needle for some distance to break all connecting strands of
protoplasm, and returned to the immediate region of the animal.
Freed from the microneedle, they exhibited slow, erratic, swing-
ing movements. If these movements brought them into chance
contact with an unsevered axopod, or with the cortical surface
of the main body, they were incorporated (Records i, 2 and 5).

282

RUTH B. HOWLAND.

The first step after severance is usually the rounding up of the
broad basal end of the axopod as an irregular granular mass, into
which the extended portion is gradually withdrawn (Record 4;
Figs, i, 2, 3), and from which new pseudopodial extensions of

1-2:00

7- 2:15 8

10-2:18 I!

12-2:24

14-2:31 15

1 8 - 2:40

RECORD I, Figs. 1-18. Changes of form exhibited by a severed axopod
before reincorporation. Time lapse, 2 : 00-2 : 40. P, pseudopodial exten-
sions. V , newly formed vacuoles. U. A., axopod still unsevered from the
main body.

great length are thrown out (Fig. 4). During its migration, this
body becomes increasingly vacuolated (Record i, Figs. 10-18).
No axial cores were ever observed in its pseudopodia (filopodia).
A number of severed axopodia may unite to form a loose net-
work (Record 2, Figs, i to 9), which behaves as a single unit.

REINCORPORATION IN ACTINOSPH^ERIUM.

The first sign of protoplasmic union may occur at either the
tip or the side of an axopod. Fragments reincorporated near the

I - 3:36

4-3:45

5
B-

, ; S?

/ — • /
•• •• •

-—-r: •
-

: •-".

8-3:56

.

V

•i - r

— '.r

&—c/. vac.

RECORD 2. Figs. 1-9. Protoplasmic network formed from a number of
severed axopodia. x, y, s, individual axopodia, gradually fusing, and finally
reincorporated as a single mass. Time lapse, 3 : 35 to 4 : oo. A and B, the
two unsevered axopodia chiefly concerned in the reincorporation process.
Ct. vac., vacuoles of the cortical layer of the main body.

tip, increase the size of an axopod at that point to several times
the original diameter, and have the appearance of a large lateral

284

RUTH B. ROWLAND.

swelling (Record 5, Fig. 6). This mass travels centripetally,
flattening as it goes, and fuses completely at the level of the cor-
tical surface (Record 5, Fig. 7; Record i, Fig. 18).

RECORD 3, Figs. 1-7. Diagrammatic representations of different types of
grafting between stained and unstained vegetative masses of Actinospharium
cichhornii. Stippled areas represent portions of animals stained with dilute
neutral red. Clear areas represent portions of unstained animals (white).
M-M, medulla grafted to medulla. C-M , cortex grafted to medulla.

Axopodia cut from one animal and carried on the needle point
to the immediate region of another, will be incorporated if the
protoplasm is still capable of locomotion. Dead (motionless)
axopodial masses are refused.

REINCORPORATION IN ACTIXOSPH/ERIUM.

Frequently, severed axopodia were observed which approached
very closely to an animal without fusing, and subsequently moved
away. If fusion had not occurred at the end of 40 to 50 minutes,
the pieces became spherical and quiescent, and showed signs of
hyalinization and disintegration (Record 4, Fig. 6).

I

RECORD 4. Figure i represents an axopod just severed from the main
body. Figs. 2. and 3, stages during resorption of the original axopodial
extension. Fig. 4, newly formed pseudopod, without axial core. Fig. 5,
twisting movement which results in a clumsy locomotion. Fig. 6, hyaliniza-
tion and disintegration after failure to become reincorporated.

It is demonstrated in the foregoing experiments that the highly
vacuolated cytoplasm of the vegetative masses (colonial aggre-
gates) of Actinosphcerium is very adhesive. The ease with which
artificial plastogamy or grafting is accomplished under compres-
sion, strongly suggests the possibility that the failure of Johnson,
'94, to fuse vegetative individuals was due to the lack of some
similar device. Also, the essential factor in the experiments of
Cienkowski, '65, would seem to be the compression which he ap-
plied by the use of paper strips, rather than the approximation of
two cut surfaces.

Reincorporation of severed axopodia is, in reality, the same
process on an infinitely smaller scale. The fusion, in these cases,
is unlike reincorporation in Difflugia in that it may occur at the
tip or side of an unsevered axopod, or at the cortical surface of
the main body.

286

RUTH B. ROWLAND.

1-2:16

• "•

'

.'

2- 2:20 •;

:20 ~,/ , A?

, • - '

m \ •

o-a is

2345
ti-

\ - ''6

7-2:30

10-2*0

RECORD 5. Figs. i-io. Diagrammatic representation of the process of
reincorporation of six severed axopodia, numerically labelled. Time lapse
for reincorporation of numbers 2, 3, 4 and 5, was four minutes ; for num-
ber I, fourteen minutes. Number 6 did not fuse with unsevered axopodia,
but moved slowly toward the cortex, fusing there in thirty-four minutes.

REINCORPORATION IN ACTINOSPH^ERIUM. 287

SUMMARY.
Grafting.

1. Merotomized portions of Actinosphceriwm eichhornii may
be grafted together under slight compression.

2. Fusion occurs between two medullary, or between medullary
and cortical surfaces.

3. Grafted animals regain a spherical shape, and show no sign
of separation at the end of four days.

Rcincorporatlon of Axopodia.

1. Axopodia completely severed from the protoplasmic mass
may be reincorporated, either by the same or by another animal.

2. Severed axopodia often become vacuolated, and develop long
pseudopodia.

3. Reincorporation may occur at either the tip or the sides of
unsevered axopodia, or on the cortical surface of the main body.

4. Severed axopodia which have not been reincorporated at
the end of 40 to 50 minutes, round up, become quiescent, and show
signs of disintegration.

5. Motionless axopodial fragments are not incorporated.

LITERATURE CITED.

Belaf, Karl.

'22 Untersuchungen an Actinophrys sol Ehrenberg. I. Die Morpholo-
gic des Formwechsels. Archiv. f. Protistenkunde, Bd. 46, Heft I.
Cienkowski, L.

'65 Beitrage zur Kenntniss der Monaden. Archiv. f. Mikr. Anat.,

Bd. I, 201-232.
Greff, R.

'67 Ueber Actinophrys Eichhornii und einen neuen Stisswasserhi-
zopoden, besonders in Riicksicht auf Theilbarkeit derselben
resp. Vermehrung durch kiinstliche Theilung. Arch. f. Mikr.
Anat., Bd. 3, 396-403.
Jensen, Paul.

'96 Ueber individuelle physiologische Unterschiede zwichen Zellen

der gleichen Art. Pfliiger's Archiv., Bd. 62, 172-200.
'01 Untersuchungen uber Protoplasmamechanik. Pfliiger's Archiv,
Bd. 87, 361-417.

288 RUTH B. ROWLAND.

Johnson, H. P.

'94 Plastogamy of Actinosphaerium. Jour, of Morphology (Whit-
man), Vol. 9, No. 2, 269-276.
Kepner, W. A. and Reynolds, B. D.
'23 Reactions of Cell-Bodies and Pseudopodial Fragments of

Difflugia. BIOL. BULL., Vol. 44, 23-39.
Looper, J. B.

'25 Observations on the Food Reactions of Actinophrys sol. Anat.

Record, Vol. 31 (abstract only, pp. 333).
Penard, Eugene.

'04 Les Heliozoaires d'eau Douce. Geneva.

MOSAICISM AND MUTATION IN HABROBRACON.

P. W. WHITING,
BUSSEY INSTITUTION, HARVARD UNIVERSITY.1

In a recent publication (Whiting and Whiting, 1927) the hy-
pothesis was advanced that certain irregular types in the para-
sitic wasp, Habrobracon, were caused by incomplete maturation
of the egg. Impaternate females were assumed to originate from
eggs in which there was failure of second maturation division.
Mosaic males from unfertilized eggs of heterozygous mothers
were thought to be due to failure of extrusion of second polar
body, two ootids with different genes taking part in maturation ;
gynandromorphs to have originated in a manner similar to that
of the mosaic males, except that one ootid was fertilized. A
single heterozygous mosaic female was explained by mitotic ir-
regularity in somatogenesis.

In the present paper are presented data giving additional evi-
dence for the hypotheses previously advanced although no more
impaternate females have been found. Factors considered here
are the allelomorphs black, O, light, o1, orange, o, and ivory, o1
(eye color), (Whiting and Burton, 1926) and normal, R, and
reduced, r, (wings), (Whiting, 1926). Materials involved are
stock IT, (type) from Iowa City, Iowa, which when previously
used in crosses with Lancaster material gave the irregular types
already discussed; stock 17 (ivory) and stock 20 (ivory reduced),
both of mixed origin from the Lancaster and Iowa City materials ;
stock 24 (type) from Lowell, Massachusetts; and orange-eyed
males from various stocks.

Stock 24 has not been previously described. A female, cap-
tured September 7, 1926 in Lowell, produced 24 males and 73
females. Thirteen virgin females produced 651 males. Thirty-

1 The present work, begun at the University of Maine, has been carried
on at the Bussey Institute under a grant from the Committee for Investi-
gation of Problems of Sex of the National Research Council. The writer
wishes to express his indebtedness to this committee for support and to the
Bussey Institution for courtesy in extending space and facilities.

289

290

P. W. WHITING.

K

Q
W
05

C/3

O

«

CJ

W fc

w
to

u

tt

"b
O

1

CM

.

o

rb<-

O

ro

•b*

ro

O

Ov

CM

^

g

OO

fi

•A

1

o

c S'b

9|-b

c>

00
CM

•O
oo

ir>

O

0

o

ro

_»-.

oo

'fr

oo

rt

Ov
CM

CM

to

Ov

c

»

•

CM

S "o

CM

Ov

0>

o\

pt-

g.'b

CM

r-j

a

ro

3

O

ro

Ov
CM

CM

M

<u 73

bo Q^

oo

^-

C ^

CM

O

fe Ov

r-»

c3 ~^

i_l

r-

O oo

oo

^H 'O

CM

o 5;

il

M
M

K

O\

ro
vO

CM

o

T3

o

U

o

OO

M

*^

>. Ov

l_l

c

Ov

Ov

O

o

M

M

CS

CM

CM

M

5 ^

ro

C

6

1—1

O.

t£

0

g

CM
CM

M

0

O

Ov

0
0

t^

ro

\_,

ro

HI

CM

I-H

ro

"3

H

H

M
OJ

o" i a

ro

^t-

3

0

OO
Ov

t?

00
Ov

o

u->

PH

O.

•*

IN

rj-

ro

I-H -O

CM

•6

0

^

CM

0

t^

^.

r-

0

ro

U ^

Tj-

M

O

o

t->.

H

o

ro

O^ 3

o\

'"t

CM

**

M

ro

'b

T3

CM"

'b

>,

Ov

^

CO

0

in

10

ro

CM

^

t-,

O

OO
Ov

Ov

t^

*"•

ro

'§•

oo

t—t

CM

•

^

OO

^

M

oo

O

O

to

0

a

O^

sO

0

OO

0

ro

N

IT}

CM

o

^}"

ro

^J1

M

IO

H

CM

1

O

CO

j

V

M

2

6

Orange
Reduced

00

C

2

o

Orange
Reduced

Ivory
Reduced

hi
G

+j

C

O

a;

00

ca

p-

M

O

0

0

Tf O

•* o

't 0

H

^

3_i

CJ

M

CM

CM

IN

CM CM

CM CM

CM M

M

cd

0 C

•1

rT O

Ui

3
Cfl O

X

0
CM

X

M

X

M
I-H

X

>H
M

XX

O T

CM CM

XX

CM CM

XX

o •*

CM CM

X

0
CM

OO _I

2i o

o

,-f.

^.

Ov

t^

O O

LO (N

rOvO

CO

cd c^

^

IN

M

CM M

H

o
o

«

CL.

MOSAICISM AND MUTATION IN HABROBRACON. 2QI

six mated females produced 1417 males and 1129 females. These
wasps were of normal appearance, like Lancaster " type " stock
i, but mesosternum was very " sooty " and 9 males and 6 females
had breaks in wing vein r4.

Reciprocal crosses were made between stocks 20 and n and
between stocks 20 and 24. The Fx females were in some cases
isolated as virgin, in others bred to orange males (stocks 3, 12,
13) or orange reduced males (stock 19). Results in the segregat-
ing generation (from females Oo'Rr virgin or mated to oR or
to or males) are given in Table I in which are also included three
fraternities from F1 females which had mated to their ivory re-
duced, o'r, brothers.

Disregarding for the moment the 21 freaks which are desig-
nated by numbers at the right of the table, it may be seen that
offspring fall into the expected classes on basis of independent
segregation of ivory and reduced and dominance of orange over
ivory. Males segregate in all cases while females segregate only
when their fathers are recessive. None of the regular males are
orange while ivory is concealed in all females except those with
ivory fathers. Females segregating into the four classes are
summarized as are all regular males at bottom of table. Reduced
fall below expectation both in males, 47.7 per cent., and in females,
45.9 per cent. Recombinations have taken place in 50.1 per cent, of
males and in 47.9 per cent, of females segregating for both factors.
There is thus no evidence of linkage. In order to determine
whether there might be linkage at some period during the mother's
life or some effect of age of mother upon recombination of
factors, segregating progenies of all cultures were added according
to vials through which mothers were passed. No significant dif-
ference from average was found.

The infrequent types, " freaks," may now be considered.

BlPARENTAL MALES.

Biparental males, formerly called anomalous or, because they
are to be distinguished from other males by possessing paternal
traits, patroclinous males, resemble their sisters in showing domi-
nant characters inherited from either or both parents (Whiting,
Anna R., in press). They are almost or quite sterile. Their few

292 P- W. WHITING.

daughters tend toward morphological abnormality and are usu-
ally sterile.

The nine males classed as biparentals in Table I are distin-
guished from the regular males by possessing orange eyes of pa-
ternal origin. They occur only in bisexual fraternities and only
from vials containing females, in other words, before their moth-
er's supply of sperm has become exhausted. Freaks 293 and
2930 were not tested but are probably biparental. The remaining
seven were almost or quite sterile. The four with long-winged
father had long wings as expected. Of the five with reduced
father, only one (290) had long wings received from its Rr
mother.

MUTANT AND MOSAIC MALES.

Mutant Male 289. — An Fx female from ivory reduced female
by type male produced in addition to males, — type 75, ivory 61,
reduced 69 and ivory reduced 60, an orange reduced male (289)
(in vial /) with compound eyes rather dark and ocelli of typical
orange color. He was mated with eight different females over a
period of forty-four days producing daughters by each of them.
A subsequent mating after ten days resulted in males only (58)
indicating sterility due to age. Of the eight females producing
daughters, two, orange reduced, produced 60 orange reduced
daughters; five, ivory reduced, produced 172 orange reduced
daughters, and one, type carrying reduced, produced 13 type and
4 reduced daughters. The eight mothers also produced 160
normal matroclinous males. Nine of the orange reduced daugh-
ters with orange reduced mothers were tested and produced only
orange reduced offspring. Freak 289 therefore bred as he ap-
peared somatically transmitting reduced in at least 236 cases and
non-black (orange or ivory) in at least 232 cases of which at least
181 were certainly orange.

Mutant Male 319. — An Fl female from ivory reduced female
by type male produced in addition to males, — type 60, ivory 75,
reduced 64 and ivory, reduced 92, an orange reduced male (319)
(in vial g} with compound eyes rather dark and ocelli typical for
orange. He was mated over a period of fourteen days to eleven
ivory reduced females. There were produced 490 ivory reduced
males, 2 orange reduced (biparental) males, and 640 orange re-

MOSAICISM AND MUTATION IN HABROBRACON. 293

duced females. Two subsequent matings resulted in males only,
207. Nine orange reduced daughters produced 218 orange re-
duced males and 210 ivory reduced males. No new types ap-
peared.

Mutant Male 308. — Freak 305 (see below) had eyes which ap-
peared mosaic for black and orange, gonads mosaic for black and
ivory. All of his offspring had long wings. Four of his type
(di-heterozygous Oo'Rr) daughters from an ivory reduced mother
(two had mated to ivory reduced brothers) produced males — type
129, ivory 139, reduced 134, and ivory reduced 121 and females-
type 37, ivory 29, reduced 20, and ivory reduced 19. In one of
these four fraternities which included females in vials a, b, and
c, but males only in vial d there occurred, in vial d, an orange
(deep red) long male (Freak 308).

Freak 308 was tested by mating to twelve ivory reduced females
over a period of fifteen days. There were produced males — ivory
reduced 701, orange long (biparental) 13, and females orange long
244. Two of these orange biparental males were tested by mat-
ing, twice each at an interval of eight days. Males of maternal
type only appeared, 67 and 104 respectively, showing that bipa-
rental males from Freak 308 tend like other biparental males to
be sterile.

Seven orange virgin daughters of Freak 308 segregated males-
orange, ivory, orange reduced and ivory reduced as expected.
Selection of two deep red of these " orange " males and crossing
with ivory reduced females gave orange males in F2 that were of
the usual variability. No effect of selection immediately following
mutation appeared.

Freaks 289, 319, and 308 may be regarded as mutants to orange
among the expected four classes of males from Oo'Rr mothers.

Mosaic Male 286. — An Fx female virgin from type female by
ivory reduced male produced in addition to males — type 40, ivory
48, reduced 35, and ivory reduced 31, a black-eyed Tnale with left
wing reduced, right long (Freak 286), occurring in vial d (Fig. i).
He was set with seven different females over a period of eight
days. He made prolonged and vigorous attempts to mate which
appeared at times successful. No daughters were produced among
427 sons. Examination showed external genitalia to be abnormal

294 p- w- WHITING-

in the presence of an extra right external clasper, as also a curious
chitinous structure suggesting an extra malformed penis (Fig. 2).
Sterility of this male may be attributed to his inability to complete
copulation. His potential breeding capacity is therefore unknown.
It is postulated that the first oocyte division was equational for Rr
and that two ootids containing R and r respectively took part in
cleavage.

Mosaic Mole 287. — An F1 female virgin from type female by
ivory reduced male produced in addition to males — type 37, ivory
38, reduced 36, and ivory reduced 42, an ivory-eyed male (Freak
287) with right wings reduced, left long, occurring in vial d (Fig.
10). Eyes and ocelli were typical for ivory.

Freak 287 was mated to each of thirteen females over a period
of fifty-two days. Daughters were produced by all of them.
Three subsequent matings resulted in males only, 276, indicating
sterility due to age. In addition to the 270 males of maternal
type produced by the first thirteen females there were the follow-
ing biparental offspring :

Six ivory reduced females produced 113 black reduced, 8 males
and 105 females.

Three orange/ivory reduced females produced 120 black re-
duced, 5 males and 115 females, and I ivory long female.

Four homozygous orange reduced females produced 26 black
reduced, 4 males and 22 females.

One homozygous ivory long female produced 24 type, 2 males
and 22 females.

Freak 287 therefore had the capacity to produce biparental males,
19 to 265 females. Biparental males from recessive mothers indi-
cate as do females what genetic factors are in the sperm. Black
biparental offspring number 284 of which those with reduced
-mothers, 260, were all reduced. The single ivory long daughter
from orange/ivory reduced mother represents the combination
visible in the eyes and left wing of Freak 287.

Each of two of the reduced biparental males with ivory reduced
mother was mated once. Ivory reduced males only resulted, indi-
cating that biparental sons of mosaic males tend, like other bi-
parental males, to be sterile.

The origin of Freak 287 may be expressed

MOSAICISM AND MTTATIoX IX 1 1 AI;R( iKKACOX. 2<)5

First }>olar body Oo'Rr

Cleavage nuclei Or olR

Italics are used to designate origin of gonads.

Mosaic Mole 507. — Five type females, F: from ivory female
(stock 17) by type male (stock 24) produced males only — type
412 and ivory 408 and in vial h a male (Freak 307) with pale
orange eyes and ocelli typical for orange. He was mated with
fourteen ivory females over a period of twenty-five days. Be-
sides the 939 ivory sons there were 470 biparentals, all black, 9
males and 461 females.

Two of these biparental males were tested by mating, each to
two females at an interval of eight days. Males only resulted, 66
and 104 respectively. Biparental males from Freak 307 are
therefore of the usual sterile type.

Three daughters of Freak 307, produced males only, type 139,
ivory 154.

Reduced is not concerned in the production of Freak 307. His
origin may be expressed

First polar body Go'

Cleavage nuclei O o'

If it be assumed that o1 mutated to o. Freak 307 is not only a
mosaic but also a mutant. This question is discussed below.

Mosaic Male 321. — A black reduced male was mated to an ivory
reduced female. 2 ivory reduced males and 28 black reduced fe-
males resulted. One of these daughters produced ivory reduced
males 2, and a reduced male (321) with ocelli and left eye black,
right eye black with red or orange area ventrally (Fig. 9). Long
wings are not involved in this experiment. Freak 321 was mated
to twenty-five ivory females over a period of twelve days. Ivory
males 1816, black (biparental) males 66, and black females 806,
resulted.

Freak 321 may be regarded either as a mosaic of black and ivory
in which the ivory has taken on a reddish appearance due to
proximity of " black " tissue or as a simultaneous mosaic and
mutant. Its origin may be from a binucleate egg, or from a
normal egg with subsequent mutation of O to o.

Mosaic Male 323. — A type male was mated with an ivory re-
20

296

P. W. WHITING.

duced female. Type females only (26) resulted. One of these
produced 44 males — type, ivory, reduced, ivory reduced as expected
and in vial d a male, Freak 323, with long wings and mosaic eyes
and ocelli (Figs. 5 and 6). He was mated with sixteen ivory re-
duced females over a period of eight days. Ivory reduced males
158, type (biparental) males 16, and type females 280, resulted.
Freak 323 is comparable with Freak 321 in that light areas of
eyes are orange rather than ivory. They are, however, much more
extensive. His origin on basis of two ootids may be expressed:

First polar body Oo'rr

Cleavage nuclei OR o'R

o' may have mutated to o or the orange color may be purely
somatic.

Mosaic Male 283. — An F1 female from ivory reduced female
by type male produced in addition to type males 50, ivory males
36, reduced males 38, and ivory reduced males 41, an orange-eyed
male (Freak 283) with long wings occurring in vial c.

Two matings of Freak 283 with ivory long females (stock 17)
resulted in 133 ivory males and 161 type females. The production
of black-eyed daughters indicates that Freak 283 was a mosaic
involving black. Fifteen of these type daughters isolated as virgin
produced males — type 321, ivory 277, reduced 252, and ivory re-
duced 278. The production of reduced males proves that Freak
283 transmitted reduced and was therefore mosaic of long and re-
duced, as well as of black and ivory (or orange).

A mating of Freak 283 with orange reduced (stock 19) female
resulted in orange reduced males 41, reduced females 56, showing
directly that Freak 283 produced reduced only. Five of these re-
duced females produced reduced males 238, and orange reduced
males 208, as expected. Freak 283 was also mated to type female
(stock i). This resulted in type males 93, and type females 30.
Six of these females produced males — type 99 and reduced 83,
and one which was mated to Freak 283 produced females — type 14,
and reduced 15.

Freak 283 was also mated with three ivory reduced females.
There resulted ivory reduced males 28, and reduced females 139,
as well as a gynandromorph (Freak 304) with reduced wings and

MOSAIC'ISM AND MUTATION IN HABROBRACON. 297

eyes mosaic of black and orange discussed below. Eight of these
females which had mated with their ivory reduced brothers pro-
duced males — reduced 30, ivory reduced 34, and females — reduced
93, and ivory reduced 92, as expected.

Summarizing the results of the eight matings of Freak 283 ex-
tending over a period of fourteen days, we find that he transmitted
black, O, in at least 363 cases and reduced, r, in at least 232 cases.
In other words he bred as a black reduced and no new factors or
unusual types appeared except the gynandromorph, 304, discussed
below. His origin on the basis of egg binuclearity may be ex-
pressed.

First polar body Oo'Rr

Cleavage nuclei Or \ o'R

Freaks 283 and 307 differ from freaks 321 and 323, in that
there is no trace of black in the eyes. Orange appearance is
therefore not due to proximity of black facets, but to some
physiological influence from " black " tissue or to mutation.

Mosaic Male 306. — An F3 female from type female by ivory
reduced male was mated with an orange long male (stock 12).
From vials a-d there appeared males — type 9, ivory 13, reduced 6,
ivory reduced 9, and females — type 26, and orange 27, as expected.
Subsequently (vials e-h} there appeared males — type 26, ivory 26,
reduced 34, ivory reduced 17, and a long-winged male with light
orange eyes and mosaic ocelli appearing in vial h (Freak 306, Figs.
7 and 8). As may be seen from the figure the right ocellus is
without pigment, even the slight amount characteristic of " orange "
being absent. Anterior and possibly also left ocelli are mosaic.
That the thorax is also possibly mosaic is indicated by its lighter
color on right side.

Freak 306 was mated over a period of thirty-eight days with
nine ivory reduced females and with five ivory long carrying re-
duced. In addition to the 814 males of maternal type there were
601 ivory long daughters. 464 of these (from ivory reduced
mothers) prove that Freak 306 transmitted long. Of the re-
mainder from Rr mothers, about half must have been long because
of factor R in the sperm.

Freak 306 therefore breeds like an ivory long. The ocelli cer-

298 P. W. WHITING.

tainly contain black and either ivory or orange. The compound
eyes are of definite orange appearance. Origin may be expressed

First polar body Oo'rr

Cleavage nuclei OR | o*R

" Black " tissue is regarded as present because of the ocelli. If
orange appearance be clue to mutation of o1 to o then there are
three types of tissue here, O, o, and o'.

Mosaic Male 505. — An F1 female from type female by ivory re-
duced male was mated with an orange long male (stock 3). There
were produced males — type 23, ivory 24, reduced 18, and ivory re-
duced 28, and females — type 94, and orange 74 as expected.
There also appeared in vial g along with type and orange sisters,
a male, Freak 305, with long wings, orange ocelli, and eyes mosaic
of black and orange (Figs, n and 12).

Freak 305 was mated with eleven females over a period of
twenty-nine days. Three subsequent matings resulted in males
only, 297, indicating sterility due to age. The eleven females pro-
ducing daughters gave in addition to 529 matroclinous sons, the
following biparentals :

Three homozygous ivory females produced type females 6, ivory
females, 122.

Eight ivory reduced females produced type females 19, ivory
long male I, ivory long females 243, ivory females with wings un-
expanded 8.

Freak 305 therefore transmits long wings, 263 cases, both in
association with black, 19 cases, and ivory, 244 cases.

Table 2 shows distribution of the 25 black and 374 ivory progeny
according to age of Freak 305.

Origin of Freak 305 may be expressed

First polar body Oo'rr

Cleavage nuclei OR \ oiR

Gonads are evidently mosaic for O and o1, eyes for O and o' or
O and o.

DIPLOID MOSAICS.

Mosaic Female 5/5. — An orange long female (stock 10) mated
with an ivory reduced male (stock 20) produced orange long —

MOSAICISM AND MUTATION IN HABROBRACON.

299

males 20 and females 6 and an orange female (Freak 313) with
left wing reduced, right long (Figs. 22 and 23) and fifteen joints
in each antenna, typical for female.

TABLE II.

DISTRIBUTION OF BLACK AND IVORY PROGENY ACCORDING TO AGE OF MOSAIC

MALE 305.

Progeny.

Age in Days since Eclosion.

Total.

4

5

7

12

13

16

18

26

28

30

32

Black
Ivorv

9

4
56

IO

I

49

5
64

2
48

2

39

i
23

IO

74

i

i

25

374

Freak 313 which had mated with her orange brothers produced
males — orange 13, ivory 9, orange reduced 13, ivory reduced 10,
and females — orange 39. Two of these females, when tested, pro-
duced reduced sons while five gave only long. Eggs of Freak 313
bearing R therefore numbered 27 while eggs bearing r numbered

25-

Freak 313 may be regarded as a heterozygous female, oo'Rr, in

which some somatic mitotic irregularity occurred eliminating R in
the development of the left wing.

It may be noted that this mosaic female, unlike the mosaic males
thus far discussed, had a homozygous mother.

Mosaic Biparental Male 318. — An ivory reduced female mated
with an ivory long male produced ivory reduced males, ivory long
(biparental) males, and ivory long females and ivory male (Freak
318) with left wing reduced, right wing long and terminal joints
of right antenna fused (Figs. 3 and 4).

Freak 318 differs from mosaic males previously described in
having a homozygous mother. It is evidently biparental, receiv-
ing R from its father. Its mosaicism was regarded as comparable
in origin to that of female, 313 (somatic mitotic irregularity
eliminating R). Being biparental, it was expected to be sterile or
to produce a few sterile daughters. It was mated with six females
over a period of twelve days. Nothing but males of maternal
type appeared, totalling 946.

300

P. W. WHITING.

GYNANDROMORPHS.

Gynandromorph 296. — An Female from type female by ivory
reduced male was mated with an orange reduced male (stock 19).
There were produced males — type 23, ivory 29, reduced 28, ivory
reduced 18, and females — type 34, orange 25, reduced 26, orange
reduced 35, and in vial c a gynandromorph (Freak 296) with ivory
eyes, right wing long, left reduced, sixteen joints in right antenna,
24 in left, and female abdomen (Figs. 18, 21). Measurements
showed ocelli of male type. Responses resembled those of the
male.

Since the father was orange the male parts (head with ivory
ocelli and compound eyes) were of maternal origin. Difference
in primaries indicates that one ootid contained R, the other r.

Gynandromorph 302. — An Fx female from type female by ivory
reduced male was mated to an orange reduced male (stock 19).
There were produced males — type 14, ivory 17, reduced 10, ivory
reduced n, and females — type 39, orange 32, reduced 33, orange
reduced 34, and in vial / a gynandromorph (Freak 302) with black
eyes and ocelli, male head (right antenna with 20 joints, left with
21 and ocelli of male size), long wings and female abdomen. Re-
sponses were characteristic of the male.

Head and wings show maternal traits. Antennae and ocelli and
therefore presumably compound eyes are male. Whether wings
are male like head (haploid) or female like abdomen (diploid) is
uncertain.

Gynandromorph 303. — An Fx female from type female by ivory
reduced male was mated to an orange long male (stock 12).
There were produced males — type 10, ivory 14, reduced n, and
ivory reduced 8 and females — type 36, and orange 46, and a
gynandromorph (Freak 303) in vial b with ivory eyes and ocelli,
long wings, male head, antennae with 23 joints) and female ab-
domen (Fig. 24). Responses were characteristic of the male.

Eye color indicates maternal origin of male parts (head). Wing
character may have been derived from either parent and may be
either haploid (male) or diploid (female).

Gynandromorph 288. — An ivory reduced female mated to a
light (black eyes, dilute ocelli), o1, long male produced 4 ivory re-

MdSAICIS.M AND MUTATION IN HABROBRACON. 30!

duced males, 3 light long males, (biparentals), 23 light long fe-
males and a gynandromorph (Freak 288) in vial c with ivory eyes,
male head (23 joints in each antenna), left wing long, right re-
duced, and abdomen female (Figs. 19 and 20). Pupal membrane
adherred to tip of abdomen and specimen was nearly dead when
found, perhaps diseased. Darker color of right side of sternum
indicated that right side of thorax, with wings reduced (matro-
clinous) was probably male. This is consistent with fact that
head, also matroclinous, is male.

Gynandromorph 122. — An ivory reduced female mated to a
black reduced male produced 10 ivory reduced males, 46 black
reduced females and in vial b an ivory reduced gynandromorph
(Freak 322) with male head (20 joints in each antenna) and a
female abdomen. Reactions were male but weak. Long wings
are not concerned in the cross. Eye color shows male parts to be
maternal in origin.

Gynandromorph 304. — An ivory reduced female mated to Freak
283, orange long male breeding as black reduced (see above) pro-
duced ivory reduced males 12, black reduced females 71, and in
vial b a gynandromorph (Freak 304) with reduced wings and eyes
mosaic for black and orange (Figs. 13-17). Left antenna is
clearly female (15 joints), right is male but with terminal joints
fused. Ocelli are mosaic, the right and anterior with considerable
black pigment, predominantly female ; the left with very little pig-
ment clearly larger than the right and exclusively or predomi-
nantly male. Compound eyes are mosaic for black and orange.
Abdomen is entirely female, sting split, and right gonapophysis,
in agreement with right antenna, deficient. The left secondary is
considerably longer than the right indicating femaleness.

Responses proved to be exclusively male. Repeated and vigor-
ous attempts were made to mate with females.

Freak 304 is evidently an intricate mosaic of male and female
parts, very different in this respect from any gynandromorphs
previously found in this species.

The size of the left ocellus indicates that it is male and its re-
duced amount of pigment shows its maternal origin. Black re-
gions of eyes indicate paternal influence and are assumed to be
female (diploid). Orange regions of eyes are presumably male

302

P. W. WHITING.

(haploid) and of maternal origin, ivory being changed to orange
at least in appearance.

Gynandromorph 325. — An ivory reduced female mated to a
black reduced male produced 7 black reduced females and in vial a
a small-sized gynandromorph (Freak 322) with male head and fe-
male abdomen. Eyes and ocelli were entirely black, the latter of
male size. Wings were wrinkled with skin adhering. Abdomen
was collapsed, genitalia immature. Sternites were of female type.
Antennae were deficient. The left had seventeen joints, the right
eighteen with the two terminal fused. These antennae are inter-
preted as male since counts of joints of over eighteen hundred fe-
male antennas previously made showed none above sixteen and this
number occurred in the larger individuals only.

The specimen was weak when found. Reactions could not be
tested.

Freak 322 is the first gynandromorph found with male parts
patroclinous. Male biparentalism is indicated by the general ab-
normality and weakness of the specimen. As explanation for
origin, either egg binuclearity and dispermy or somatic mitotic
irregularity may be suggested.

FREQUENCY OF OCCURRENCE OF MUTANTS AND MOSAICS.

Of the 14,023 wasps recorded in Table I., 9,787 (normal males
9,780, mutant males 3, and mosaic males 4) may be regarded as
coming from unfertilized eggs and 4.236 (normal females 4,224,
biparental males 9, and gynandromorphs 3) from fertilized eggs
laid by diheterozygous mothers, Oo'Rr. These mothers came in
all cases from crosses of stock 20, ivory reduced, and a type stock,
either n or 24.

Mosaic and mutant males and gynandromorphs number 9 among
the 7,90*8 descended from stock 1 1 , while there is but one mosaic
male among the 6,115 descendents of stock 24. This may indi-
cate a difference in hereditary tendency toward binuclearity in
favor of stock n.

Among the eleven haploid freak males above discussed, three
(289, 308, 319) were mutants but not obviously mosaics, two
(286, 287) were mosaics but not obviously mutants, and six (283,
305, 306, 307. 321, 323) were mosaics and at least apparently

MOSAICISM AND MUTATION IN HABROBRACON. 303

mutants for they showed orange color in eyes. Since two (283,
307) at least of the latter would not have been tested and detected
as mosaics had they not been orange, it was thought that other
males apparently segregating in normal manner might be mosaic
also. Gonads might differ from soma.

Accordingly F., males of normal appearance from a cross of
ivory reduced (20) by type (n) were tested by mating to ivory
reduced females. Thirty-five type bred as type (offspring total-
ing ivory reduced males 1/2, type males 24, type females 750),
forty-two ivory bred as ivory (offspring totalling ivory reduced
males 218, ivory males 28, ivory females 911), thirty-eight re-
duced bred as reduced (offspring totalling ivory reduced males
254, reduced males 21, reduced females 1,217) an<^ twenty-nine
ivory reduced bred as ivory reduced (offspring totalling 77 males
and 285 females. Thus of 144 normally segregating males from
Oo'Rr females none showed gonads different from soma.

HYPOTHESES SUGGESTED.

The data above presented show that orange eye color may ap-
pear in males from eggs laid by mothers heterozygous for black
and ivory, Oo1, in at least two or possibly three different ways.
Biparental males obviously receive the factor O from their father.
Males with eyes mosaic for black and orange have thus far bred as
black (321, 323), as ivory (306), or as black and ivory mosaic
(305). In the case of three of these it may be supposed that
proximity of black facets may cause the " ivory " to take on an
orange appearance, but number 306 had compound eyes entirely
orange with dark pigment appearing in the ocelli only. Moreover
since two orange males (283, 470), breeding as black showed no
trace of black pigment, it must be supposed that influence of
" black " tissue in changing ivory to orange is not merely an optical
effect from proximity of black facets, but is due to some physio-
logical (enzymatic?) influence which actually transforms ivory to
orange. If this be the case, it must be supposed that in number
287 (ivory breeding as black-ivory mosaic) the " black " tissue did
not involve the " enzyme-producing gland."

An alternative to this " somatic " explanation is to suppose that
actual genetic change has taken place. In the mosaic males gonads

P. W. WHITING.

have not included this orange tissue. It may be suggested that
the three orange mutants (289, 308, 319), although not obviously
mosaics, may have been derived from binucleate eggs carrying O
and o' for their mothers were Oo'.

The hypothesis of mutation of o' to o caused by association with
O tissue brings the various phenomena into harmony. Evidence
both that it is o' rather than O which has changed and that this
change has taken place in somatogenesis rather than in ovogenesis
is afforded by the mosaic-eyed (black-orange) gynandromorph
304 in which O must have been derived from the male parent for
the mother was o'o'.

Evidence decisive between the somatic and the genetic hypothe-
ses has not yet appeared. A mosaic male from Oo' female pro-
ducing o gametes or a male with eyes mosaic of orange and ivory
would be of much interest in this connection.

SUMMARY.

1. Independent segregation of ivory (eye color) and reduced
(wings) is shown in progeny of Oo'Rr females.

2. When such females are mated to orange, o, males there occur
among the normal offspring, orange (biparental) males with char-
acteristic sterility.

3. One female from RR female by r male and one sterile bi-
parental male from rr female by R male were each characterized
by possessing one long and one reduced wing. Elimination of
chromosome bearing R is suggested as explanation.

4. Eight mosaic males were produced by Oo1 females. Of
these — one was sterile, four produced black daughters only, one
ivory only, and two both black and ivory. It is suggested that
these males arise from binucleate eggs and that gynandromorphs
are produced when one nucleus of such eggs is fertilized.

5. Of seven gynandromorphs six showed male parts matro-
clinous while in one they appeared patroclinous. In the latter
case they may have been diploid, comparable to ordinary male bi-
parentalism.

6. Two mosaic males had eyes completely orange while four
mosaic males and one gynandromorph had eyes mosaic of black
and orange. Since presumably only black or ivory entered into

MOSAICISM AND MUTATION IN HABROBRACON. 305

the gametes producing these mosaics, the orange color must have
been due to some somatic physiological effect or to mutation.

7. Three mutant orange males showing no mosaicism and breed-
ing as orange were produced by Oo'Rr females.

LITERATURE CITED.
Whiting, Anna R.

Genetic Evidence for Diploid Males in Hymenoptera. In
Press.
Whiting, Anna R. and Burton, R. H.

'26 Quadruple Allelomorphs Affecting Eye-Color in Habrobracon.

Am. Nat., LX, May-June. 286-290.
Whiting, P. W.

'26 Two Wing Mutations in Habrobracon and their Method of Inheri-
tance. Am. Nat., LX., Sep.-Oct., 443, 454.
Whiting, P. W. and Whiting, Anna R.

'27 Gynandromorphs and other Irregular Types in Habrobracon. BIOL.
BULL., LII., 2, Feb.

306 P- W. WHITING.

EXPLANATION OF PLATE I.

FIGS. 1-24. Camera lucida drawings of various parts of mosaics oi
Habrobracon juglandis.

Figs, i, 4, 10, 13, 1 8, 19, 22. < 10.

Figs. 3, 11, 12, 14, 15, 17, 20, 21. 23, 24. X 22.

Figs. 2, 6, 9. X 30.

Figs. 5, 7, 8, 1 6. X 63.

Figs. 1-12 are from mosaic males.

Figs. 13-21 and 24 are from gynandromorphs.

.Figs. 22 and 23 are from a mosaic female.

Figs, i and 2, wings and ventral view of genitalia of mosaic male 286.

Figs. 3 and 4, right antennae and wings of mosaic male 318.

Figs. 5 and 6, dorsal view of ocelli and anterior view of head of mosaic-
eyed male 323.

Figs. 7 and 8, dorsal and anterior view of ocelli of mosaic male 306.

Fig. 9, right view of head of mosaic-eyed male 321.

Fig. 10, wings of mosaic male 287.

Figs, ii and 12, dorsal and ventral views of head of mosaic-eyed male

305.

Fig. 13, wings, Figs. 14 and 15, right and dorso-sinistral views of head.
Fig. 16, dorsal view of ocelli and Fig. 17, ventral view of abdomen of
mosaic-eyed gynandromorph 304.

Figs. 18 and 21, wings and abdominal sternites of gynandromorph 296.

Figs. 19 and 20, wings and ventral view of abdomen of gynandromorph
288.

Figs. 22 and 23, wings and abdominal sternites of mosaic female 313.

Fig. 24, abdominal sternites of gynandromorph 303.

BIOLOGICAL BULLETIN, VOL. IIV.

PLATE I.

o o

0 Q

P. W. WHITING.

OVULATION IN THE FOUR-TOED SALAMANDER,
HEMIDACTYLIUM SCUTATUM, AND THE EX-
TERNAL FEATURES OF CLEAVAGE AND
GASTRULATION.

K. R. HUMPHREY,
DEPARTMENT OF ANATOMY, SCHOOL OF MEDICINE, UNIVERSITY OF BUFFALO.

INTRODUCTION.

Incidental to the collection of material for a study on the germ
cells of Hemidactylium scut at um the writer brought to the labora-
tory a female of this species which subsequently deposited a
number of ova while under observation. The eggs laid proving
fertile, the progress of cleavage and gastrulation was carefully
noted. Though the observations made cover but a very limited
material, they serve to supplement the work of Bishop ('18) who
begins his account of the development of Hemidactylium with the
neural plate stage ; moreover, ovulation in this species seems not
to have been previously observed and reported. For these reasons
the observations made are here reported, even though the limited
material prevents consideration of but little more than the super-
ficial features of cleavage and gastrulation.

OVULATION IN Hemidactylium.

The female on which these observations were made was cap-
tured in the sphagnum bog on the shores of Mud Pond near
Ithaca,1 the site of Bishop's discovery of the egg-laying habits of
this species some years before. This particlar female was found
in a sphagnum hillock together with two eggs, on the morning of
May u, 1923. She was brought to the laboratory the afternoon
of that day, and placed in a small glass dish of sphagnum, to-

1 This study was carried on in the laboratories of the Department of
Histology and Embryology of Cornell University. The writer wishes to
acknowledge the kindness of Dr. B. F. Kingsbury at whnsr suggestion the
observations herein reported were undertaken.

307

308 R- R- HUMPHREY.

gather with the eggs assumed to be hers. About 9 P.M. of the
same day it was noted that eight eggs were present, all but three
of which were scattered over the bottom of the dish. The female
had been partially concealed by the sphagnum, and was of course
necessarily disturbed by the attempt to locate her and her eggs for
observation. It would seem, however, that her position when first
accurately observed was probably but little changed, since she was
in contact with two ova which later observations would indicate
had been but recently extruded. She lay with her back turned
against the bottom of the dish, her abdomen upturned against the
overlying sphagnum. Two eggs lay above her, still in contact
with her ventral body wall immediately cephalad of the cloacal
opening, but adherent to the overlying sphagnum ; by gently pull-
ing upon this sphagnum the eggs could readily be pulled away
from the body of the female, who was not at all disturbed by this
procedure, but remained quietly in position. Her body was some-
what curved, so that the middle part of her abdomen was in con-
tact with the side of the dish ; the end of her tail also lay in con-
tact with the side of the dish, elevated along it and not in contact
with the bottom.

With the body of the female in this position, it was noted at
9: 14 P.M. that an egg was beginning to appear at the cloacal
opening. The female made no unusual movements, but lay quietly
while the egg continued to move slowly outward. In about two
or three minutes it was practically expelled from the cloaca, but
the female lay motionless for several minutes before a slight shift
of her body left the egg attached to the overlying sphagnum.

The process of ovulation was not repeated until about 10: 55
P.M. During the expulsion of this second egg the female twisted
the front half of her body in such fashion that her left side was
toward the bottom of the dish ; the caudal part of the body how-
ever, was kept in its former position, with the dorsum against the
bottom of the dish and the vent towards the overlying sphagnum.
This position was retained for several minutes following the ex-
pulsion of the egg. After about twelve minutes the female was
brought into a brighter light for more careful observations, and
thereupon moved slowly away from the position she had main-
tained throughout the expulsion of the two ova. She now lay

OVULATION IN T11K F< )l'R-T( >KI> SALAMANDER. 309

with the abdomen pressed against the side of the dish, the cephalic
third of her body in a vertical position, the remainder, except for
the end of the tail, extending in a horizontal direction along the
junction of the side and bottom of the dish. In this position a
third egg was expelled. As in previous ovulations, no unusual
movements of the female were noted during the process. Pre-
ceding the appearance of the egg at the cloacal opening the female
was seen to straighten her body and move the hind limbs slightly-
movements perhaps necessary to facilitate the passage of the egg
through the pelvis. The female then lay motionless with the
abdomen pressed against the side of the dish while the egg was
slowly forced from the cloacal orifice and against the glass. After
the female had remained for about eight minutes with the egg in
this position she was gently pushed away from the side of the
dish, to which the egg remained adherent. As she showed no
indication of further attempts at ovulation, the observations were
discontinued. This female subsequently deposited a few other
ova, but at times when she was not under observation.

The position of the female during the expulsion of the first
two ova is of considerable interest, since it may possibly represent
an approach to the normal. A horizontal position of the body with
the ventral surface uppermost at the beginning of ovulation would
prevent the eggs from dropping down into the spaces of the
sphagnum, and would insure the contact and adherence of eggs
successively expelled from the cloaca, since each egg as it leaves
the cloacal opening would be supported upon the body of the
female in contact with the egg previously extruded. \Yith the
accumulation of a number of eggs, the size of the egg mass, to-
gether with the viscid character of its surface would effectively
prevent its dropping downward in the spaces of the sphagnum.
Females taken in sphagnum bogs after ovulation ordinarily have
their eggs loosely arranged in a mass around or over which the
body of the animal is coiled.

The eggs when first expelled from the cloaca show only an ex-
tremely thin gelatinuous capsule ; after several minutes this is
noticeably thicker, due to the absorption of water from the moist
environment. The outermost layer of this jelly is of a semi-fluid,
viscid nature, which causes the eggs to adhere readily to each
21

3IO R- R- HUMPHREY.

other or to sphagnum, glass, or other gelatinous egg envelopes,
while a thin, membrane (vitelline membrane) closely surrounds
the egg or embryo and separates it from the more fluid gelatinous
envelope; a similar membrane of the egg of Cryptobranchus is
stated by Smith ('26) to be identical with the zona pellucida of
ovarian eggs.

Fertilization in this species is internal, as in other Urodeles.
The breeding habits of the animals are quite unknown to the
writer, but since males have not been found with the females
near ponds at the time of ovulation (Bishop, '18; Blanchard, '23)
it may be concluded that mating occurs at some previous time, and
very probably in a terrestrial environment, the transfer of sperm
from male to female possibly being effected by contact of the
ventral surfaces of the animals. Bishop quotes C. and H. Thomp-
son ('12) and Moesel ('18) to the effect that males and females
have been taken from beneath the same log or stone during the
spring months.

The female possesses the sperm-storing organ or spermatheca
characteristic of Urodeles. According to Dieckmann ('27) the
spermatheca of this species is unique in that it consists of but a
small number of spermathecal tubules opening directly into the
cloacal chamber ; the latter feature is regarded as a primitive one,
while the former (reduction of tubules) is characteristic of the
most highly developed spermathecse, in which the tubules open to
a single duct-like tube (common tube, — a modified portion of the
cloacal chamber). Dieckmann's descriptions of the cloaca and
spermatheca of this species are based upon two specimens cap-
tured in August near Buffalo, New York. In one of these, sper-
matozoa are very abundant, in the other absent ; on the assump-
tion that Heinidactylium males are incapable of mating during
the summer months, as are practically all Urodeles, these spermato-
zoa must be interpreted as having been stored for at least several
weeks after a mating in the spring months. Their large number
and whorl-like arrangement in the spermatheca suggest however,
that no ovulation had occurred after they had been received into
the spermathecal tubules.

OVULATION* IN TTIE FOUR-TOED SALAMANDER. 31 1

CLEAVAGE.

The eggs laid under observation in the laboratory, together
with others deposited by the same female, proved to be fertile,
making possible a study of the early cleavage divisions. While
the limited number of eggs necessarily prevented fixation and sec-
tioning of successive cleavage stages, the external features of seg-
mentation were carefully noted over a period of several hours, or
until gastrulation was in progress.

In order to follow the progress of cleavage in individual eggs
the ova were removed singly to numbered Syracuse watch glasses
each containing a few drops of water ; each egg was left in con-
tact with a small bit of sphagnum rather than in the water. The
watch glasses were transferred to the stage of a binocular micro-
scope for observations ; the position of the egg was changed as
desired, either by moving the sphagnum to which it was attached,
or by direct manipulation with needles. Though it was realized
desirable to handle the eggs as little as possible, the difficulty of
keeping them in position for viewing the vegetal hemisphere neces-
sitated considerable manipulation. Though possibly this and other
environmental factors may have contributed to certain of the
atypical features of cleavage observed in some of the ova, it is
nevertheless certain that eggs showing such features in some cases
gave rise to normal embryos. The eggs were of course kept under
observation without removal of their protective membranes.

Some of the eggs, when first viewed under the microscope, still
showed the polar bodies adherent to their surfaces ; these as a rule
became indistinguishable after the first two or three cleavage divi-
sions had been completed. The stage of maturation at the time
of laying was not determined.

Since in none of the eggs laid in the laboratory during the night
of May 11-12 did the first cleavage furrow make its appearance
later than 10: 30 A.M. of the following day, cleavage may be as-
sumed to begin within from ten to fifteen hours after the egg is
deposited. The egg of Hemidactylium apparently agrees with that
of Spelcrpcs (Eiirycca} in this respect (see Goodale, '11).

The pattern of cleavage will first be described in an egg in
which the divisions proceeded in a relatively orderly, symmetrical
fashion. In this egg the first cleavage furrow was first noted at

312

R. R. HUMPHREY.

9:30 A.M.; it then extended over about a sixth of the egg cir-
cumference (see Fig. i, a). From its middle point at the animal
pole a short broad furrow extended at right angles to it on either
side, the two furrows together forming a rather distinct cross.
The short transverse furrow subsequently disappeared, being prac-
tically invisible by the time the first cleavage furrow had reached
the egg equator. A similar furrow was observed in only one
other egg. In another several short furrows radiating from the
animal pole were formed preceding the appearance of the first
cleavage furrow and disappeared after the latter had extended
over a fourth of the egg circumference.

FIG. i. Egg showing the most regular or symmetrical cleavage pattern
observed. Viewed from animal pole except in c; cleavage furrows as
numbered, a, 9:30 A.M. 5-12-23; b, 12 M. ; r, 2 P.M.; d, 3:40 P.M.; e,
vegetal hemisphere at 5 P.M.; /, 8: 30 P.M. The origin of the micromeres
in d is indicated by the lettering.

The progress of the first cleavage furrow through the more
heavily-yolked vegetal hemisphere of the egg was at a much
slower rate than its formation between animal pole and equator.
From this egg and others observed we may conclude that the fur-
row is usually completed (on the egg surface) within about two
hours after its first appearance.

OVULATION IN THE FOUR-TOED SALAMANDER. 313

The second cleavage furrows made their appearance before the
first had been completed. In Fig-, i, b they are shown extending
practically to the equator of the egg, 2% hours later than the
stage of Fig. i, a. The first cleavage furrow in this egg is now
complete. The second cleavage furrow apparently reaches the
vegetal pole of the egg in from two to three hours.

The third set of cleavage furrows appeared about an hour after
the stage of Fig. i, b, and were completed within an hour, the egg
at 2 P.M. having the appearance shown in Fig. i, c. These fur-
rows were roughly parallel with the egg equator and lay well up
in the animal hemisphere. The four micromeres thus formed
rotated a short distance in a clockwise direction to the position
they occupy in Fig. i, c.

The cleavage furrows of the fourth set were completed in the
micromeres at 3 : 40 P.M., though just beginning to appear at that
time in the macromeres. The direction of these furrows in B
and D was roughly parallel with the first cleavage furrow ; in A
it was at right angles to this furrow, and in C it was meridional.
With the completion of the division the daughter micromeres
underwent readjustments of form and position such that the egg
at 3:40 P.M. presented the appearance of Fig. i, d. The fourth
cleavage furrows of the macromeres are shown in this figure at
an early stage of their formation ; they had not been completed
when the egg was sketched at 5 P.M. (Fig. i, <?). Their direc-
tion, while approximately meridional, was such that they would
end in contact with the second cleavage furrow a short distance
from the vegetal pole.

No further observations on this egg were made until 8 : 30 P.M.,
at which time it presented the appearance indicated in Fig. i, /.
It was impossible at this time to determine accurately the relation
of any of the micromeres to those of Fig. i, d.

It would appear from the above that cleavage in Hemidactyliwm
is similar to that in the egg of Eurycea (Goodale) and is of the
holoblastic unequal type characteristic of the moderately-yolked
eggs of Urodeles. As in Eurycea and other species, however, the
cleavage divisions conform to a symmetrical pattern for but few
divisions if at all, and the blastomeres soon become of such ir-
regular form and arrangement that their descent from any par-
ticular cell of the four- or eight-celled stage can not be determined.

314

R. R. HUMPHREY.

Though irregularity of the cleavage pattern may be expected to
appear during the third or fourth division in any egg of this
species, as shown in Fig. i, d, it is of earlier origin in a great many
cases. Of ten eggs in which the furrow pattern could be deter-
mined for a stage equivalent to that of Fig. i, b, in only four did
it conform to this plan, and of these only one egg showed in the
third division the regularity of pattern illustrated in Fig. i, c.

FIG. 2. Egg showing cleavage pattern resulting from inequality in the
first division. Viewed from animal pole except in c; furrows numbered
and origin of cells in d shown by lettering, a, i : 15 P.M., 5-12-23; b and
c, 3 : 25 P.M. ; d, 5 P.M.

Irregularity or asymmetry of cleavage in two of the ten eggs
began with the first division. In one case it consisted merely in
an inequality in the division of the egg (see Fig. 2, a) ; in the
second division the larger of the two cells was divided more
slowly than the smaller, the cleavage furrow in it reaching only
thirty degrees below the egg equator when division in the smaller
cell was completed. The third cleavage furrows were horizontal
in three cells ; in the fourth the furrow took a meridional direc-
tion, as shown in Fig. 2, b, in which it reaches almost to the
egg equator. The vegetal hemisphere of the egg now appeared
as in Fig. 2, c. A cleavage furrow along the dotted line of Fig.
2, b, subsequently cut off a micromere from cell B ; this, with the
division of micromeres A, C, and D, formed a group of seven
cells of very unequal sizes arranged as in Fig. 2, d. The de-
scendents of micromeres C and D formed a group of cells still
distinguishable by their smaller size several hours later, in the
fine-celled blastula stage ; they then lay just above the egg equator.
This egg subsequently completed gastrulation in the normal man-
ner.

OVULATION IN THE FOUR-TOED SALAMANDER.

315

In the second egg to show irregularity beginning with the first
division, the first furrow to appear had the position illustrated
in Fig. 3, a. Within two hours other furrows of irregular direc-
tion had developed, giving the animal hemisphere the appearance
of Fig. 3, b. At this time no furrows extended below the equator.
Later several furrows extended into the vegetal hemisphere (Fig.

FIG. 3. Egg showing extreme! y irregular cleavage pattern beginning
with the first division. Viewed from animal pole except in c. a, 2 : 45
P.M., 5-12-23; b, 4:45 P.M.; c, 8 P.M.; d, 10:25 P-M-

3, c}. Fig. 3, d illustrates the appearance of the animal hemis-
phere in this egg about eight hours after the beginning of cleavage.
Though continuing development for 24 hours longer and reach-
ing a fine-celled blastula stage this egg finally died without begin-
ning gastrulation. Its markedly atypical cleavage, therefore, may

FIG. 4. Egg showing irregular cleavage pattern. Viewed from animal
pole; egg in b is tilted to show full extent of atypical furrows, a, 10: -'5
A.M., 5-12-23; b, 12 M.; c, i : 45 P.M.; d, 10: 45 P.M.

well have been the result of injury from drying, handling, or
other causes.

Another egg with almost as pronounced irregularity of cleavage
as that illustrated in Fig. 3 nevertheless gave rise to a normal
embryo. In this egg the first cleavage furrow developed norm-
ally, and had completely encircled the egg within three hours.

316

R. R. HUMPHREY.

Before the usual second furrows made their appearance, a fur-
row was noted leading from the first a short distance above the
egg equator. About a half-hour later the appearance was as in
Fig. 4, a, with the second furrows now present. Later the egg
showed other furrows in the region in which the first atypical one
had made its appearance (Fig. 4, b) and about five hours after the
onset of cleavage the animal hemisphere displayed the atypical pat-
tern of Fig. 4, c. In the somewhat later blastula stage illustrated
in Fig. 4, d this egg showed a group of distinctly smaller jnicro-
meres at and below the equator in the region of the first atypical
furrow. Though this peculiarity persisted until late in cleavage
this egg completed gastrulation in the normal fashion.

FIG. 5. Egg showing irregular cleavage pattern starting with retarda-
tion of one of second furrows. Viewed from animal pole except in a;
furrows as numbered, a and b, 10:30 A.M., 5-12-23; c, 12 M. Origin of
cells in c indicated by lettering.

Probably the most common type of irregularity initiated during
the second cleavage division is that resulting from retarded or
atypical formation of one of the two furrows of the set, so that
one reaches the vegetal pole long in advance of the other, as illus-
trated in Fig. 5, a. The animal hemisphere of this egg pictured
in Fig. 5, b shows that the third cleavage furrows were likewise
retarded in the same half of the egg, they being completed in the
half on the right while only one had appeared in the opposite
hemisphere, assuming that the furrow marked 3 represents a fur-
row of the third set. In a stage corresponding to that of Fig. i,
d the polar cap of micromeres consists of but six cells (Fig. 5, c)
instead of the eight found in the regular type of cleavage. This
egg subsequently developed into a typical fine-celled blastula in
which stage it was fixed for sectioning.

OVULATION IX THE FOUR-TOED SA I. A M A XDER.

317

In many eggs the cleavage pattern becomes irregular in the
course of the third division, after the first two divisions have
proceeded in the regular or symmetrical fashion illustrated in
Fig. i. This irregularity may originate through retardation of
division in one or more of the blastomeres, as illustrated in Fig.
6, a, in which only two micromeres are shown completely separated
from the corresponding macromeres. Often, however, the irregu-

1 1

d e f

FIG. 6. Eggs showing irregularity of ck-avage pattern beginning with
third division, after first two have proceeded as in Fig. I, b. Viewed from
animal pole except in e; cleavage furrows as numbered, a, cell with all
furrows of the third set horizontal, but furrows delayed in two of the blas-
tomeres. b and c, egg with two furrows of third set atypical in direction ;
b at II : 45 A.M., c at 12: 50 P.M. d and e, egg with two furrows of third
set horizontal and the other two parallel to the second furrows in the animal
hemisphere (d) but meeting the second furrows below the equator (r).
/, egg with furrows marked 3 and 3' atypical in direction.

larity results from the atypical direction of one or more of the
cleavage furrows of the third set. In the egg shown in Fig. 6, /'.
for example, the line marked 3 represents a furrow of the third
set which takes a direction parallel with the second furrow, and
hence does not cut off a polar micromere in the usual fashion ; the
third furrow around micromere A is also atypical in direction,
since it curves back almost to the animal pole instead of joining the
first cleavage furrow nearer the equator. In the fourth cleavage

318 R- R. HUMPHREY.

division of this egg micromeres A and B were divided transversely
(see Fig. 6, c) and micromeres were cut off from cells D and F,
giving a polar cap of six micromeres, to which a seventh was
added later by division of blastomere C. The plane of this divi-
sion is indicated by a dotted line in Fig. 6, c. A somewhat more
symmetrical pattern was exhibited by the egg shown in Fig. 6, d;
in this egg two furrows of the third set take the usual position
parallel to the equator, while the other two run parallel to the
second furrows in the animal hemisphere of the egg but join these
furrows in the vegetal hemisphere (Fig. 6, e~) ; cells A and D are
thus not typical polar micromeres but extend well toward the
vegetal pole of the egg. Still another egg with atypical cleavage
furrows of the third set is shown in Fig. 6, /. In this egg two
furrows of somewhat typical direction had cut off the micromeres
C and D, while the furrow marked 3, meeting the second furrow
below the equator, had cut off a much larger cell (5). The fur-
row 3' ultimately joined the first cleavage furrow some distance
below the equator, to cut off another large cell (A}. This egg
thus resembles that of Fig. 6, d in possessing only two micro-
meres of the usual size in the eight-cell stage.

Although only a limited number of eggs were followed through
early cleavage, it would seem that irregularity or asymmetry of
cleavage pattern in Hemidactylium probably begins with the third
division in the majority of cases, but with even the first or second
in some cases.

Though cleavage in Hemidactylium follows the general plan
described by Goodale ('n) for Spclerpes (Eurycea) the vegetal
hemisphere of the egg is earlier divided into small cells than in
the latter species. According to Goodale the cleavage furrows of
the fourth set in Spclerpes " do not even reach the lower pole,
but usually join the earlier furrows less than 45 degrees below the
equator." As a result of this mode of division, the egg of
Spelerpes in a later stage may show 130 cells visible from above
and only 7 from beneath, or 400 above and 14 beneath. In Hcmi-
dactylium the egg is less heavily yolked than in Spelerpes and less
markedly telolecitha5. The cleavage furrows of the fourth set,
though not necessarily reaching the vegetal pole, nevertheless end
farther below the equator (see Fig. i, r), while eggs showing 7

OVULATION IN THE FOUR-TOED SALAMANDER. 319

or 8 cells when viewed from the vegetal pole show only 25 to 30
when viewed from above. These facts indicate that the egg of
Hemidactyttum is intermediate between that of Amblystoma and
Spelerpes, and far less markedly unequal in its cleavage than the
relatively large eggs of Desmognathus (Hilton, '09) or Crypto-
branchus (Smith, '12). In the latter forms the tendency is for
the third cleavage planes to be radial or meridional in direction,
and the relatively small micromeres of the animal pole arise by
later divisions ; in Hcmidactylium, on the other hand, the third
cleavage furrows are more frequently equatorial (latitudinal) in
direction, and cut off micromeres of relatively larger size, the egg
in this respect resembling that of Amblystoma.

FIG. /. Early blastula of Hcmidactylium in vortical section. Outlined
with camera lucida. < 31.

So few eggs were available that it was impossible to make a
careful study of the internal changes of cleavage. Blastulae of
but two stages were sectioned. The earlier of these, represented
in Fig. 7, shows a distinct segmentation cavity roofed by a single
layer of cells. It is probable that the blastocoele makes its ap-
pearance early, as in the eggs of Cryptobranchus (Smith, '26) or
Desmogtiathus (Hilton, '09), and as a distinct cavity rather than
a collection of intercellular spaces such as Goodale states is fre-
quently the condition in Spelerpes. Older blastulse show a rather
characteristic loose arrangement of the internal blastomeres, but
a distinct blastocoele is nevertheless present.

320

R. R. HUMPHREY.

GASTRULATION AND CLOSURE OF THE BLASTOPORE.
The earliest indication of gastrulation appeared about 90 hours
after ovulation in the form of a crescentic depression well down
in the vegetal hemisphere (Fig. 8, a). This crescent blastopore
is bordered by a dorsal lip frequently showing a greater pigmen-
tation than the neighboring surface. The crescent blastopore is
gradually extended laterally through further invagination of cells

e f g h

FIG. 8. Diagrams illustrating the formation and " closure " of the blas-
topore in Hemidactylium. No attempt is made to indicate exact relation
of blastopore to egg equator. Margin of neural plate indicated by broken
lines, a, early crescent blastopore; b, later crescent blastopore (from an-
other egg) ; c, blastopore of b about 24 hours later; d, blastopore about n
hours later than in c; c, later blastopore from another egg; /, blastopore of
c about 9 hours later ; g, smaller blastopore from another egg ; h, blastopore
of g 10 hours later, reduced to short vertical slit which persists as anus.

at its lateral margins (Fig. 8, &) until it assumes the form of a
half circumference; sometimes it may become more extensive, as
in Fig. 8, c, but in no case was the line of invagination or over-
growth observed to become a complete circle surrounding a yolk
plug. Following its period of greatest ventral extent the blas-
topore shortens and assumes the form of a crescent with laterally
directed margins (Fig. 8, d} ; by further narrowing and elevation
of its lateral portions it assumes the form shown in Fig. 8, e, and
finally the form of an inverted crescent (Fig. 8, /). As the
neural folds close the blastopore gradually shortens (Fig. 8, g)

OVIM.ATION IN TIIM FOUR-TOED SALAMANDER. 32!

and finally assumes the form of a slit elongated in the direction
of the body axis ( Fi,^. 8. // ) . It does not appear that the blasto-
poric opening is lost but rather that it persists as the anal opening.
From the above brief account of the changes in form of the
blastopore it may be seen that gastrulation in Hemidactylium
closely resembles that in Spelcrpes (Goodale, '11) so far as ex-
ternal features are concerned. Presumably the egg would show
the same movements of surface material that Goodale was able
to demonstrate in Spelerpes by the use of Nile Blue Sulphate ; the
egg is so lightly pigmented that it, like the egg of Spelerpes, might
readily be used for studies of this sort. The conclusions of
Goodale as to the formation of the germ layers in Spelerpes are
probably likewise applicable to Hemidactylium.

Hemidactylium and Spelerpes are possibly unique among Amer-
ican Urodeles in lacking the " yolk-plug stage " characteristic of
the development of Amblystoma and the Aimra. In the larger,
yolk-rich eggs of Necturus and Cryptobranchus the blastopore as-
sumes the form of a complete circle (Smith, '12). Although
Hilton claims this to be true also for Desmognathus, his Fig. 30
shows a late blastopore of the inverted crescentic form charac-
teristic of the same stage in Spelerpes (Goodale, PI. i) or Hemi-
dactylium (Fig. 8, /and g). Unfortunately observations on this
stage of development of the eggs of Gyr'mophilus and the Pletho-
dons are not available ; possibly in these species gastrulation would
follow the pattern found in Hemidactylium and Spelerpes.

In his account of the development of Hemidactylium, Bishop
('18) estimates that his youngest specimen, an embryo showing
well-elevated neural folds, is not more than 72 hours old. This
estimate would appear to be much too low, assuming that the
eggs timed by the writer had not been retarded by the conditions
to which they were subjected. In these eggs the early crescent
blastopore was first observed about 90 hours after ovulation. and
the neural folds did not close until from 60 to 80 hours following
this stage, or at from 150 to 175 hours after ovulation. Since
temperature determines to a great extent the rate of the develop-
mental processes, and since the laboratory temperature averaged
higher than that of the normal environment, it is probable that
the eggs observed developed at least as rapidly as would eggs in
the field.

322

R. R. HUMPHREY.

For stages of development later than those of Fig. 8 the reader
is referred to the excellent sketches published by Bishop ('18).
Transections of older embryos of this species are figured in a
previous report dealing with the primordial germ cells (Hum-
phrey, '25).

SUMMARY AND CONCLUSIONS.

1. The egg complement of the H etnidactyliuin female probably
requires a period of several hours for its deposition, since the
expulsion of a single egg occupies several minutes.

2. The observed position of the body of the female during
ovulation (ventral surface upmost) may be of value in preventing
scattering of the eggs, since they are thus supported upon the
body of the female as well as by adherence to neighboring sphag-
num or other eggs.

3. Under laboratory conditions cleavage begins in from ten to
fifteen hours after ovulation ; it is of the unequal holoblastic type.
The first and second cleavage furrows tend to be meridional in
direction. The furrows of the third set are frequently horizontal
(latitudinal) and lie well above the egg equator, but in many cases
they take a direction parallel with the first or second furrows in
the animal hemisphere and meet these furrows at some point below
the equator. A well developed blastocoele is very early present.

4. Irregularity in cleavage is to be expected in all eggs after
the third division. Only occasionally do the eggs show symmeri-
cal eight-cell stages, due to the fact that irregularity or asymmetry
in cleavage may begin with any of the first three divisions.

5. The vegetal hemisphere of the egg undergoes more division
than in the eggs of Spclerpes, Dcsmognathus, or Cryptobranchus.
In this respect it resembles the egg of Amblystoma.

6. Gastrulation begins about 90 hours after ovulation with the
formation of the usual crescent blastopore. No ventral lip de-
velops and no yolk plug is formed. The crescent blastopore
shortens and becomes inverted, the horns of the crescent pointing
upward, and finally closes to a vertical slit which becomes the
anus.

/. Closure of the neural folds occurs in from 150 to 175 hours
after ovulation. The further course of development has been
described by Bishop ('18).

OVULATIOX IN THE FOUR-TOED SALAMANDER. 323

LITERATURE CITED.
Bishop, S. C.

'18 Notes on the Habits and Development of the Four-toed Salamander,
Hcmidactylimn scut a turn (Schlegel). New York State Museum
Bulletins, 219, 220 — Fifteenth Report of the Director. Albany,
N. Y. Pp. 251-282.
Blanchard, F. N.

'23 The Life History of the Four-toed Salamander. Am. Nat., Vol.

57, pp. 262-268.
Dieckmann, Johanna M.
'27 The Cloaca and Spermatheca of H emidactylium scutatum. BIOL.

BULL., Vol. 53, pp. 281-285
Goodale, H. D.
'n The Early Development of Spelerpes bilineatus (Green). Am. Jour.

Anat, Vol. 12, pp. 173-247.
Hilton, W. A.
'09 General Features of the Early Development of Desnioi/nntliu.'; fusca.

Jour. Morph., Vol. 20, pp. 533-559-
Humphrey, R. R.
'25 The Primordial Germ Cells of H emidactylium and other Amphibia.

Jour. Morph. and Physiol., Vol. 41, pp. i-43-
Moesel, Julia
'18 A Study of the Caudata of the Cayuga Lake Basin. M. S. Thesis;

Library of Cornell University.
Smith, B. G.

'12 The Embroyology of Cryptobranchus alleghemensis. Part II. Gen-
eral Embryonic and Larval Development with Special Reference
to External Features. Jour. Morph., Vol. 23, pp. 455-579-
'26 The embryology of Cryptobranchus alleghcnicnsis. Part III. For-
mation of the Blastula. Jour. Morph. and Physiol., Vol. 42, pp.
197-252.
Thompson, Crystal and Thompson, Helen.

'12 The Amphibians of Michigan. Mich. Geol. and Biol. Surv., Pub.
10, Biol. Ser. 3, pp. 34~36.

FERTILE TERMITE SOLDIERS.

HAROLD HEATH,

HOPKINS MARINE STATION,
PACIFIC GROVE, CALIFORNIA.

In two earlier studies (Heath, '03, '27) attention was called to
the fact that in the termite genus Termopsis, represented in Cali-
fornia by two species (T. nevadensis and T. angusticollis) , occa-
sional individuals occur which externally resemble a soldier and
yet are sexually mature. In the typical soldier the head is dis-
tinctly longer than broad, whereas in the fertile type the two di-
ameters are the same or nearly so. Furthermore the jaws of both
project far beyond the labrum, and are built upon essentially the
same plan. Hence, unless it is decided that they belong to a spe-
cial caste, these unusual insects must be looked upon merely as
soldiers with fully developed sex organs.

Up to the time when the manuscript of the later paper went to
press only four of this type of termite had been discovered. Since
then fourteen additional individuals have been collected, and it
now appears that they may prove to be of rather frequent occur-
ence. All of these later acquisitions were living in essentially the
same conditions. A limb of a Monterey pine (Finns radiata). ten
or fifteen feet in length, and evidently occupied by a colony of T.
nevadensis throughout its entire extent, had broken off near its
base, and in falling had broken into several fragments. When the
weathering of these, and the depth they had sunk into the leaf mold
indicated a sojourn of several months each fragment was carefully
opened. A large number of such colonies were examined, and
while in most instances complemental royal forms were present,
those with soldier-like appearance existed in approximately twelve
per cent, of the communities.

In addition to their relatively small heads the behavior of these
fertile soldiers is unmistakable. Unlike the typical soldier, which
moves about from place to place in the observation nest, the fertile
type remains in close association with the normal complemental in-

324

FERTILE TERMITE SOLDIERS.

325

sects. Also the females are readily distinguished by their distended
abdomens. Conjugation was witnessed on six occasions where
the soldier paired with an individual of the usual neoteinic type.
No especial attempt was made to observe the actual egg-laying

FIG. i.

process, since this had been observed on two former occasions.
Dissections, however, of six specimens showed the sex organs to
be well developed. In the text figure, for example, a specimen of
female complemental soldier is compared with a primary queen
from a colony of T. ncvadcnsis two and one half years old. In
the first named individual (A) there were seven fully-formed ova,
including one in the oviduct (not represented), while in the other
22

326

HAROLD HEATH.

specimen the ovary contained eight fully-developed eggs. In ad-
dition there were approximately the same number of developing
ova in both insects.

In a former account (Heath, '27) some observations were re-
corded which indicate that the complemental royal forms, at least
in the genus Termopsis, are fertile soldier nymphs. The fertile
soldier also points to the same conclusion. Under normal circum-
stances it appears that in an orphaned colony insects of this type
appear as a rule to be in the penultimate instar, and yet there are
occasional specimens which give evidence of belonging to the ante-
penultimate stage. Also it appears to be true that in some in-
stances the sex gland stimulus may be applied to individuals which
have progressed so far along the path to complete soldier develop-
ment that with the functional activity of the gonad there is a slight
suppression only of those characters which normally belong to the
fully-developed soldier. This seems to be a reasonable explana-
tion of the origin of these individuals, which are thus seen to com-
bine the activities of the usual complemental royal insect together
with many of the structural features of the typical soldier.

Up to the present time no essential differences have been dis-
covered between the offspring of the primary royal pair and those
of the complemental royal forms including the fertile soldier. The
data relating to this last named category, however, are based upon
the study of two colonies only. A report upon a more extensive
series of experiments, now being conducted, will be made at a
later date.

REACTIONS OF THE CYPRID LARV^ OF BARNACLES
AT THE TIME OF ATTACHMENT.1

J. PAUL VISSCHER,

WESTERN RESERVE UNIVERSITY.

Knowledge of the actual process of attachment and of the re-
action of the organism immediately preceding this act is very
meager for many of the sessile marine organisms. As a large
number of these forms cause fouling on ship's bottoms, a careful
study of some has been made by the author as part of a study to
determine the nature and extent of fouling on ship's bottoms.

The activities of the tadpole larvae of several species of tunicates
have been carefully studied by Caswell Grave (1921-1923). He
found that after a free-swimming period which varied from a few
minutes to several hours, the larvae invariably became negative to
light and attached by means of adhesive papillae. The factors
which determined when the organisms attached appeared to be
related to the physiological state of the organism, while the place
of attachment is probably a matter of chance, although the re-
actions to shadows may determine this in some degree.

Nelson (1924) found "that in the oyster the larvae 'test out.'
by means of their feet, an appreciable area of solid surface before
actually attaching, which is preceded by circling movements." It
had long been supposed that these larvae had simply settled and
the lucky ones survived.

The only record of attachment of larval barnacles of which the
author is aware is found in Darwin (1851). He states that "the
larva fixes itself with its sternal surface parallel and close to the
surface of attachment, and the antennae become cemented to it; if
the cirripede, after its metamorphosis had remained in this posi-
tion, the cirri could not have been exserted, or only against the
surface of attachment; but there is a special provision that the
young cirripede shall immediately assume its proper position at

1 Published by permission of the U. S. Commissioner of Fisheries. Ex-
periments were performed at the U. S. Biological Station at Beaufort, N. C.

328 J- PAUL VISSCHER.

right angles to the position which it held whilst within the larva,
namely with its posterior end upwards. This is effected in a
singular manner by the exuviation of the great compound eyes,
which we have seen are fastened to the outer arms of the double
UU-like, sternal apodemes ; these together with the eyes stretch
transversely across, and internally far up into the body of the
larva; and, as the whole has to be rejected or moulted, the mem-
brane of the peduncle of the young cirripede has necessarily to be
formed with a wide and deep inward fold, extending transversely
across it; this when stretched open, after the exuviation of the
larval carapace and apodemes, necessarily cause the sternal side of
the peduncle to be longer than the dorsal, and, as a consequence,
gives to the young cirripede its normal position, at right angles to
that of the larva when first attached."

Barnacles attach to a great many materials and under very
diverse environmental conditions. Darwin (1853) stated that
" sessile cirripedes adhere to all sorts of objects, floating and fixed,
animal and vegetable, living and dead, organic and inorganic," and
that " living mollusca are the most frequent objects of attachment."
The following list of materials will serve to give some idea of the
varied array of materials on which the writer has seen barnacles
attached : wood, glass, rubber, beeswax, shells of more than thirty
varieties, rocks of many sorts, rope, cloth, leather, metals of many
kinds including iron, copper, zinc, lead, tin, aluminum, and to many
alloys as brass and bronze, turtles, crabs, lobsters, fucus fronds,
bamboo, zostera, sponges, corals, echinoderms, bryozoa, tunicates,
and several types of cetacea. It is thus apparent that barnacles
attach to almost every type of material submerged in sea-water,
which is sufficiently large and permanent to afford a place of
growth.

But although barnacles attach to all these materials, there is
nevertheless a high degree of specialization among them. Thus
Bdanus galeatus is found only on certain gorgonian corals;
Coronula diadema only on the humpback whales (Megaptera) ;
Cryptolepas rachianecti only on the Calif ornian grey whale (Rha-
chlanectis glaucus). Octolasmis mulleri is found in the gill
chambers of several crabs, while Octolasmis geryonphila is found
only in the gill chamber of a single type of crab (Geryon quin-

CYPRID LARV^: OF BARNACLES. 329

quedens). Chclonibia patula grows only on the carapace of cer-
tain crabs (and on Limuhis) while Chclonibia tcstubinaria grows
only on the carapaces of turtles.

Although many thus show a high degree of specialization as re-
gards their habitat, yet within a single genus one often finds wide
extremes. Balanus amphitrite is perhaps the most widely dis-
tributed barnacle existing today, with a world wide distribution,
being found in all tropical and warm temperate waters. It attaches
to wood, to rocks, to fucus fronds, to rubber, to metals of many
sorts and to shells as well as living molluscs of many kinds. In
contrast to this, Balanus galeatus attaches only to a limited num-
ber of gorgonian corals and is typically a West Indian species,
although a few specimens have recently been found from Southern
California.

The larvae of all barnacles are hatched as nauplei, the eggs being
carried within the mantle cavity of the adult until hatched. These
nauplei moult from one to three times during the course of the
first week or ten days, changing slightly with each moult into forms
known as metanauplei and then metamorphose into an entirely dif-
ferent type of organism called the cypris larvse or cyprid. The
cyprids of all barnacles are bilaterally symmetrical organisms
about four times as long as high and often ten times as long as
broad. These proportions vary depending upon the genus of the
barnacles but the species of any given genus seem to vary but little
in this regard.

This free-swimming period as cyprids appears to vary greatly
with different forms of barnacles. Balanus amphitrite has been
observed to attach within seven days after hatching, while Chtha-
malus fragilis probably attaches within four days. Balanus ba-
lanoides, on the other hand, has a much more extended larval life
and apparently does not attach for from 10 to 12 weeks after
liberation.

These cyprid larvae propel themselves through water by means
of sudden backward movements of their appendages. These
movements occur with great rapidity so that it is quite impossible
to see this reaction unless the organisms are swimming very slowly.
They are aided during their pelagic life by the fact that they
possess great globules of a fatty substance in the anterior portion

33O J. PAUL VISSCHER.

of their bodies. That this is clearly a lipoid in its character has
been demonstrated by its reaction with Congo red and with osmic
acid. During much of this period this mass of liquid substance
seems to act as a buoy holding the larvse near the surface film.
But as the free swimming period draws to a close this substance
disappears. It seems probable accordingly that the length of the
free-swimming period is dependent upon the amount of this sub-
stance originally stored and its rate of disappearance.

Throughout this period of pelagic life the cyprids often swim on
their sides, that is, with their oral surfaces on a horizontal plane
with their dorsal surfaces. This holds good especially if the or-
ganisms are in the surface film at the surface of the water or if
they rest on some object at the bottom. They will frequently
propel themselves for some time through a considerable distance
while in this position. It has also been observed that they usually
are found when in this position to have their morphologically right
side turned downward while their left side is uppermost.

The actual process of attachment occurs, as suggested above,
after a free-swimming period of from three days to two weeks
(longer for some species). The cyprids of Balanus improvisus
and Balanus amphitrite have been kept under observation for ten
or eleven days at the end of which time some were found to have
attached but many were still active.

When the internal physiological conditions necessary for at-
tachment are present, conditions apparently correlated with the
" lipoid " content of the organisms, the larvse have been observed
on many occasions to " walk " on the substratum apparently
" hunting " a place for attachment. This remarkable performance
is accomplished by alternate attachment and release of the ad-
hesive tips of the antennae combined with a " push " from the set
of appendages.

By reference to Fig. I, it will be seen that when one of the
antennas (0-2) is released the appendages (a/>) straighten out,
throwing the larva forward. The released antenna (0-2) then
attaches, while the appendages are withdrawn, and with the next
move the antenna (a-i) is released.

In this manner these organisms have been observed to traverse
a distance of more than twelve mm. (about ~*/2 inch) and have been

CYPRID LARV.E OF BARNACLES.

331

e e

N, V

B

FIG. i. To illustrate the movements of barnacle larvae at the time of
attaching. A : ventral view to show antennae, a and a1 ; median eye, e, and
the paired lateral eyes, e'. B-D, lateral views of cyprid larva, showing
method of locomotion at time of " selecting " a place for attachment.

332 J- PAUL VISSCHER.

seen to " test " different areas for a period of more than one hour
before finally attaching. Frequently it has been observed that the
antennae appear to adhere so firmly that the larva " kicks " for
eight or ten times before releasing its hold.

On several occasions the writer has been fortunate in seeing
the actual process of metamorphosis while observing through a
microscope giving a magnification of about one hundred and fifty
diameters. It was then seen that after attachment by means of
the antennae, the organisms " kick vigorously " for some time, but
without effecting release. The animal then appears to settle
down as soon as attachment was effected, in the region of the ap-
pendages. Metamorphosis now occurred. The two-valved shell
of the cyprid stage was thrown off, including the exoskeleton of
the appendages (often also the paired eyes). From the resulting
almost amorphous mass the young barnacle is soon made out. A
secretion is continually laid down on the formerly ventral surface,
and the rudiments of a coating (the future shell) appear around
the sides of the mass. After one hour the " head " and appendages
of the animal can be seen in an inverted condition. Whereas when
attached the appendages extend downward, they now extend up-
ward and the mouth parts too have changed their position ; this
reversal in position occurring as described by Darwin as quoted
above.

No opercular valves appear until after six hours, while the
separate plates of the shell do not appear until more than twelve
hours after metamorphosis, in any of the observed cases of Bal.
amphltrite, Bal. hnprovisus, or Chthainalus fragilis.

In these types the plates appear as four units taking the place
of the original apparently chitinous material which protected the
organism during the early stages of metamorphosis.

The cyprid larvae of barnacles are also sensitive to light and it
has been shown ( Visscher and Luce, 1928) that they are stimulated
to a much greater degree by a given amount of energy in the field
of green than by like amounts of energy in other fields of the
spectrum. These larvae are usually positive in their reactions but
in the later stages they are very erratic and may not show either
positive or negative reactions for a considerable period of time.
However, at the time of attachment, although still sensitive to

CYPRIU LARV/E OF BARNACLES.

333

light and especially to wave-lengths of 4/0-500 ^ they become de-
cidedly negative and move away from the source of stimulation.

A

FIG. 2. To illustrate effect of light at time of attachment of larval
barnacles. X. glass dish, 4 cm. square, by 1.5 cm. in height, with concave
depression 2^ cm. in diameter and i cm. in depth, covered with heavy
black paper on five of its six sides as indicated. A-C , and D-F, two series
of tests, contained barnacle larvae, in which each container was covered as
in X, and with the exposed side toward the window as indicated by the
arrow. Only those larvae which had attached and metamorphosed are
indicated.

It has been impossible to select cyprid larvae of known age, but
the following experiment will demonstrate this reaction. A num-
ber of cyprid larva were isolated from a collection of tow made
in the tidal channel in front of the laboratory island. From the
original lot those with the least amount of " oil droplets " were
selected and placed in small glass dishes resembling salt cellars
which were surrounded on five of the six sides with a dull black
paper (Fig. 2). These were then set on a table in front of a large

334 J- PAUL VISSCHER.

north window with the uncovered side of the dish toward the
window.

In the first series of three experiments twelve cyprid larvae of
Bal. improvisus were placed in A, three cyprids of Chthamalus
fragilis were placed in B, and between twenty and thirty mixed
forms were placed in C.

By referring to Fig. 2, it will be seen that attachment occurred
almost without exception on the side of the dish away from the
source of light. It was observed at this time that many of these
young barnacles were definitely oriented with their anterior ends
(containing their eyes) decidedly away from the light side of the
dish. In a second series of experiments D, E, and F (Fig. 2)
this feature is even more clearly seen. Although these experi-
ments were repeated some two weeks later with large numbers of
individuals only a very few were observed to attach, and the re-
sults were no more conclusive than those shown in the first two
series of experiments.

It can be clearly seen from these experiments that for the two
types of barnacles which were tested, light is an important factor
in determining the point of attachment, and that they orient with
the anterior end directed away from the source of light.

CONCLUSIONS.

Barnacle larvse are sensitive to light at the time of attachment
and for the three species tested, are negative to light at this time.
It has also been demonstrated that barnacle larvse do not attach at
random, i.e., merely by chance, but that they apparently " test out "
the surface before attaching, in which process they have been ob-
served to spend a very considerable period. The antennae are evi-
dently the " feelers " as well as the final " hold-fasts " and once
definitely attached a secretion is deposited which protects the
young barnacle from the beginning of its life as a sessile organism.

BIBLIOGRAPHY.
Darwin, Charles
1851 A Monograph on the Cirripedia. A. The Lepadidae. Publications

of the Ray Society, 400 pp. London.

1853 A Monograph on the Cirripedia. B. The Balanidse. Publications
of the Ray Society, 648 pp. London.

CYPRID LARVAE OF BARNACLES. 335

v

Grave, Caswell
1920 Amaroucinm pellucidum. I. The Activities and Reactions of the

Tadpole Larvae. Jour. Exper. Zool., Vol. 30, pp. 239-257.
Grave, Caswell, and McCosh, Gladys K.

1923 Perophora viridis. The activities and structure of the free-swim-

ming larvae. Wash. Univer. Studies, Vol. XI., pp. 89-116.
Grave, Caswell, and Woodward, Helen.

1924 Botryllus schlosscri. The Behavior and Morphology of the Free-

swimming larvae. Jour. Exper. Zool., Vol. 14, pp. 467-513.
Nelson, T. C.

1924 The Attachment of the Oyster Larvae. BIOL. BULL., Vol. 46, pp.

I43-I5L
Visscher, J. Paul, and Luce, Robert H.

1928 Reactions of the Cyprid Larvae of Barnacles to Spectral Colors.
Biol. Bull., Vol. 44, pp. 336-350.

REACTIONS OF THE CYPRID LARV^ OF BARNACLES

TO LIGHT WITH SPECIAL REFERENCE TO

SPECTRAL COLORS.1

J. PAUL VISSCHER AND ROBERT H. LUCE.

TABLE OF CONTENTS.

1 . Introduction 336

2. Material and Methods 337

3. Behavior of Cyprids 344

4. Results • • 345

5. Conclusions 348

6. Bibliography 349

INTRODUCTION.

It is well known that the larvae of barnacles as well as the larvae
of almost all sessile marine organisms are, at times, sensitive to
light and are either photopositive or photonegative. This fact
has led to their use for experimental purposes, but the use of the
larval forms of barnacles in such experiments has been confined
to the earliest free swimming stage call the Nauplius. No records
have been found of the use of the later larval stage, the cyprid
larva in experiments with light.

Since barnacles are the most important of the various types of
organisms which cause fouling (cf. Visscher, 1928) and since
barnacles attach while in the cyprid stage and subsequently meta-
morphose into the adult form and since it had been observed on
ship's bottoms that barnacles were frequently most numerous on
the shaded area of the hull, it seemed important to ascertain the
stimulating efficiency of light of different wave-lengths in regard
to the cyprid larvae of barnacles.

In previous work on the reactions of various organisms to color,
much confusion has arisen from the fact that the apparatus used
was neither standardized nor calibrated. Where spectral trans-
missions were known, there was frequently no attempt to balance

1 Published by permission of the U. S. Commissioner of Fisheries. Ex-
periments were performed at the U. S. Biological Station at Beaufort, N. C.

336

CYPRID LARVJE OF BARNACLES. 337

the radiant energies. The only occurrences of the use of different
spectral colors with balanced radiant energies are those of Pergens
(1899), Blauuw (1908), Laurens (1911), Day (1911), and Gross

(I9I3)-

The most common error in the earlier work was that of ascrib-
ing the effect of colored light to its wave-length, and ignoring the
effect of intensity. Pergens was the first to recognize the differ-
ence between wave-length and intensity. Working with a spec-
trometer he found that the energy of the blue end of the spectrum
was too small to be measured. He then used light filters of
colored glass, and measured the light intensity with a Ritchie pho-
tometer. Pergens failed, however, to eliminate the effect of the
infra red transmitted by the filters.

Laurens, Day and Gross used apparatus which was essentially
alike. The light was diffracted and balanced and the band of
transmission was either narrowed or widened until the deflection
of the galvanometer was the same for all the colors. While this
work was done with colors of high purity and balanced radiant en-
ergies, it cannot give more than a relative idea as to the stimulat-
ing efficiency of parts of the spectrum, as only four limited regions
were employed.

Mast (1917) used a Hilger spectrometer and a Lummer-Brod-
hun rotating sector and was able to determine with great accuracy
the relative stimulating efficiency of the various parts of the
spectrum. While this method allows a direct study of the effec-
tiveness of spectral colors, it appeared to be limited in its appli-
cation to organisms which, in a single beam of light travel parallel
to the rays and do not deflect to either side and consequently could
not be used with cyprid larvae. The intensity of light obtainable
by this method was believed to be too limited for adequate work
on the cyprid larvae of barnacles. For these reasons and others,
to be discussed later, it was considered inadvisable to use this
method in the following work.

MATERIALS AND METHODS.

The larval life of the barnacle is divided, with a very few ex-
ceptions, into two distinct stages, the nauplius, and the cyprid.
The cyprid stage is characterized by the presence of a large bi-

338 J- PAUL VISSCHER AND ROBERT H. LUCE.

valve shell, giving it a general resemblance to certain ostracods.
This resemblance, however, is superficial and does not extend to
the limb structure or internal anatomy. All of the appendages
of the adult are present ; the antennules protrude from between
the valves of the carapace ; there are six pairs of thoracic limbs, but
no abdominal appendages. The unpaired eye of the nauplius is
usually still present, and in addition there is a pair of large com-
pound eyes.

Cyprids of various species of barnacles differ from each other
in the size and proportions of the carapace, and in the presence
of certain characteristic bodies within the carapace. During the
time of year in which this work was done, three main types of
cyprids predominated. These types were of two distinct sizes.
The smaller size was .51 mm. long, .225 mm. high, and .20 mm.
wide, and was found by subsequent metamorphosis to be the cyprid
of Chthamalus fragilis. The larger type was made up of two
groups which were outwardly very similar, being .65 mm. long,
.24 mm. high, and .20 mm. wide. These groups could only be
distinguished by the presence of certain granules in the carapace.
One of these was determined by subsequent metamorphosis to be
Balanus improvisus, and the second was the cyprid of B. am phi-
trite.

Balanus improinsus and B. amphitrite are known to breed most
abundantly at Beaufort during June and July, and large numbers
of their nauplii and cyprids are to be found at that time. Chtha-
malus fragilis has been observed to have a more extended breed-
ing period, apparently running from May into late September.
Balanus eburnens is abundant at Beaufort, but as far as is known,
no cyprids of this species were used in the experiments.

Cyprids were obtained by towing with a number 12 mesh net.
They were picked out of the tow by means of a capillary pipette,
then washed in clear sea water and placed in the aquarium. Dur-
ing the summer of 1924 cyprids were not numerous but during
the summer of 1925 cyprids were very abundant, comprising at
times in late June and early July a large part of the plankton.
This abundance allowed a selection of material, and in consequence
most of the work was carried out with the cyprids of B. impro-
visus and B. amphitrite. Cyprids appear to be most abundant

CYPRID LARVM: OF BARNACLES. 339

on an incoming tide and at flood tide. This is also true for the
nauplii during June when they were most abundant. There is a
marked diurnal migration, so that tows during the day must be
made with the net about eight feet below the surface, this being
slightly less than the minimum depth of the channel. The time
of exposure to sunlight exerts a marked influence on the activity
of those cyprids which remain at the surface. Cyprids taken
from a surface tow during the morning are fairly active, but those
taken from a surface tow during the late afternoon are inactive,
and could not be used.

The apparatus used in the following experiments consisted of
100 watt, gas-filled, tungsten lamps, two copper sulphate cells,
and thirteen monochromatic light filters.

In selecting the filters, an attempt was made to use only those
having a narrow transmission band. This is somewhat difficult
to obtain for the yellow and blue-violet. Wratten filters were
available for most of the needed colors, but these filters are not
wholly stable under the climatic conditions found at Beaufort.
At the suggestion of Dr. K. S. Gibson of the Bureau of Standards,
we were able to obtain a number of filters from glasses manu-
factured by the Corning Glass Works, which company very kindly
supplied them for our use. These glasses were ground to thick-
nesses giving transmissions in accordance with the curves shown
in Bureau of Standard Technologic Paper 148. The glasses
are more satisfactory than Wratten filters because of their per-
manency under all climatic conditions. However, Corning glasses
were not available for certain regions of the spectrum, so four
Wratten filters were used, making a series of thirteen filters cov-
ering the whole visible spectrum. By referring to figure No. I a
list of all " Corning "' filters with their curves of transmission
can be found. Table I. is a summarization of the curves for all
filters used giving the limits of transmission and the dominant
wave-length of each filter.

The difficulty encountered in obtaining a deep blue or ultra-
violet arises from the fact that all the filters in this region transmit
some of the red end of the spectrum. No filter was available to
block out the red without also cutting off some of the blue or
ultraviolet. Since red is relatively feeble in stimulating effect, it

340

J. PAUL VISSCHER AND ROBERT H. LUCE.

was judged best to use the blue and untraviolet filters as they
were, and to attempt a correction for the presence of the red later.
We were not able to obtain a yellow filter with a sharply limited
transmission band, since ordinary yellow filters merely cut off a
portion of the blue end of the spectrum. Perhaps the best yellow
would be produced by a combination of Corning glasses 036

i.o

0.9

o.o-

.0.0

300

Ufctra-violet

400 500

Violet Blue

600

700

Green Yellow Orange Red.

Y/ave-length in milli-mlcra.

FIG. i. Curves to illustrate the range of light energy transmitted by the
"Corning" glass niters used to test the relative stimulating efficiency of
spectral colors. (From the U. S. Bureau of Standards Technologic Papers
No. 148.)

(curve 38) and 40 iZ (curve 59) as given in Bureau Standards
Paper No. 148. Unfortunately, however, this combination was
not available to us at the time. The balancing of the radiant
energies for all filters was done by the Bureau of Standards, by
the use of an apparatus based upon the radiomicrometer of Boys.1

1 The authors are grateful to the Director of the Bureau of Standards
and also to Dr. W. W. Coblentz for their kindness in calibrating all filters
and lamps used for these experiments.

CYPRID LARV.E OF BARNACLES. 34!

TABLE I.

LIST OF FILTERS USED IN EXPERIMENTS, SHOWING THEIR TOTAL SPECTRAL
TRANSMISSION, AND THEIR DOMINANT WAVE-LENGTHS.

Filter. Total Dominant

Transmission Wave Length.

Ultra C 83 315-428 MM 355 MM

6o9-Red End

Purple C 69 310-485 MM 370 MM

6oo-Red End

Purple W 35 300-475 MM 420 MM

65O-7OOMM

Blue W 49 400-510 MM 440 MM

Blue C 60 335-640 MM 460 MM

Blue C 59 335-690 MM 480 MM

Blue Green C 56 340-700 MM 505 MM

Green €52 425-670 MM 530 MM

Green W 58 485-635 MM 540 MM

Yellow W 15 500-700 MM 590 MM

Orange W 22 545-700 MM 620 MM

Orange C 38 54O-Red End 640 MM

Red C 19 620-Red End 700 MM

The letter " C " after a filter denotes a Corning glass filter. The num-
bers after the Corning glasses refer to the transmission curves shown in
Bureau of Standards Technologic Paper 148. The letter " W " denotes a
Wratten filter, and the number refers to the transmission curves found in
the booklet " Wratten Filters " published by the Eastman Kodak Company.

Each lamp to be calibrated was set up with the center of the
curved filament facing a radiometer. The radiometer was placed
on a scale graduated in centimeters, and could be moved toward
or away from the light. Connected with the radiometer was a
Coblentz modification of Thompson's galvanometer. This was
hung on the wall, fifteen feet from the radiometer, and at right
angles to it. The lamps used were 100 watt, 115 volt, gas-filled,
Mazda lamps and during the test each lamp carried 109.3 volts
and 0.87 amperes. In these experiments the infra-red was ab-
sorbed by copper sulphate cells. Each cell was two centimeters
thick, and consisted of two pieces of optical glass cemented to a
glass ring, five centimeters in diameter. The cells were filled with
a solution consisting of 57 grams CuSO4 in two liters of water.
One of these cells was placed in front of each lamp tested. The
radiometer was then moved back and forth until the galvanometer
showed a deflection of one centimeter. This was taken as a
23

342 J. PAUL VISSCHER AND ROBERT H. LUCE.

standard deflection. A filter was then placed before the lamp
and the galvanometer again brought to a deflection of one cm.
Such balancing of the radiant energies, therefore, eliminates the
necessity of correcting for the unequal distribution of energy in
the spectrum of the lamp. The result of any experiment con-
ducted with these balanced filters represents, therefore, the actual
stimulating efficiency of the specific wave-length of the light, and
is not in any way influenced by a difference in light intensities.

The apparatus was set up so that two beams of light crossed at
right angles (Fig. 2). The lamps were enclosed in light-proof
boxes each having an opening five centimeters square. A copper
sulphate cell was placed in front of each of these openings. Each
box was readily moved, and the center of the crossing of the two
beams was used as the zero point from which to measure the posi-
tion of the light. The aquarium in which the organisms were
placed was made of the best quality slide-glass, and was 35 mm.
square and 10 mm. deep. This aquarium was placed on a stand
at the center of the two cross beams. Underneath this aquarium
was placed a small microscope lamp which could be lighted
momentarily so that the cyprids could be counted more readily.
Two screens were placed just outside of the aquarium stand, one
of these had an opening admitting white light, and the other had
a slot in which the filters were placed. In practice, the lamp
giving the white beam was placed permanently with the center of
the filament at a point 1 19.2 cm. distant from the center of the two
beams of light while the lamp used with the filters was moved to
such distances as to compensate for the energy transmission value
as previously determined for each filter.

Ten cyprids were placed in each aquarium in carrying out an
experiment. They were allowed to become dark adapted, and
were then given a one minute exposure to light. At the end of
this time the cyprids in the white light, in the filter light, and in
the corner between the two light beams, were counted. These
numbers thus obtained were recorded in a table of four columns
under the heads of White, Corner, Filter, and Non-reacting. A
dark period of about twenty seconds was given between trials.
Three trials were made with each filter, after which the filters

B

FIG. 2. To illustrate apparatus used to test the relative stimulating effi-
ciency of equal amounts of light energy, in known and limited fields of the
spectrum, upon the cyprid larvae of barnacles. A, Track for boxes (J5)
containing the electric bulbs (L). C, copper sulphate filters in front of
square opening in light box (B). D, screen holding filters (JE) and sur-
rounding aquarium stand (F). S, switches for control of lights. G, feed
line for electric current.

344 J- I'AUL VISSCHER AND ROBERT H. LUCE.

were changed. The same cyprids as a rule could be used for the
whole series of filters, though certain factors modified their be-
havior.

BEHAVIOR OF CYPRIDS.

Cyprids give definite reactions to light for two to three hours,
but after that the number of non-reactants increases rapidly.
They must also be dark adapted in order to secure definite move-
ments toward the light, but apparently there are limits to this dark
adaptation. It was noted during the experiments that if the cy-
prids were left in total darkness for periods of thirty to sixty
minutes, that on resuming the experiments there were no reactants
during the first exposure. In the second exposure there were a
few reactants, but several exposures were required before the
usual number of reactants were obtained. This, however, can
be demonstrated more strikingly with nauplii, where, in several
instances, ten exposures were required before all of the nauplii
were reacting. The effect of a shorter dark period is also well
marked. A dark period of one minute was used during the earlier
experiments. By varying this rest period, however, and counting
the number of reactants and non-reactants, it was found that a
rest period of about twenty seconds gave the smallest number of
non-reactants. Such a period varies somewhat as the experiment
progresses, for a shorter period can be used in the first part of
the experiment, and a slightly longer period is necessary toward
the end.

Cyprids usually lie upon either side when not in motion. Move-
ment is usually on the side, and is brought about by rapid and
powerful kicks of the thoracic appendages. The movement is
rapid and usually in a wide arc, but sometimes in a straight line.
It is not at all unusual for them to turn about in a complete circle
or series of circles before shooting off toward the light. The posi-
tion of the antennae apparently has much to do in determining the
direction of movement, the movement being in a wide arc or fairly
straight line when the antennas are withdrawn into the carapace,
and circular when they are extended. The cyprids may either
sink down and lie quietly on the bottom after striking the side of
the aquarium, or they may move back and forth along the side of

CYPRID LARVAE OF BARNACLES. 345

the dish in a series of jerky movements. This uncertainty in their
movements makes it necessary to have a definite time of exposure,
and also necessitates a rapid counting of those collected in each
light.

The presence of small oil droplets in the anterior end of the
cyprids modifies their behavior to a considerable extent. These
droplets are usually present, and are found either in the anterior
end or along the dorsal side. Their size may apparently be taken
as an indication of the age of the cyprid. Cyprids about to at-
tach have very small drops, are sluggish, and move about on the
bottom of the dish with a walking movement of the antennae and
a push or slow kick of the thoracic appendages. Those in which
the oil droplets are large, often become caught in the surface
film. When this occurs it is difficult to get them out, and while
they may be very active, they cannot react definitely toward the
light. Their reaction consists of spasmodic kicks and short move-
ments toward the light, but these movements soon cease and they
drop back passively toward the center of the dish. If two or
more cyprids get into the surface film, they cluster together and
are of no further value in the experiment. In such cases they
were removed and replaced by fresh organisms. The cyprids
which are the most favorable for study are those which are too
old to get in the surface film, and which are not ready for at-
tachment.

RESULTS.

In Table II. are shown the results of twenty series of experi-
ments as described above. These results are summarized in figure
3 showing a curve which clearly indicates that there is a fairly wide
region in the spectrum from 505 to 590 /A/A or from light blue to
yellow, where the stimulating efficiency is equal to, or more than
50 per cent, of white light. The maximum of this region is in
the light green from 530 to 545 /A/A where the stimulating efficiency
is between 90 and 94 per cent., or practically the equivalent of
white light.

In the region of the spectrum from the yellow-orange through
the red end, or from 590 to 700 /A/A, there is a very sharp drop in
the effectiveness of light, until at 700 /A/A this is less than 5 per cent.

346

J. PAUL VISSCIIER AND ROBERT H. LUCE.

of white light. On the other side of the maximum there is a fall
in value from 480 to 440 //,/A, but this is not as sharp as on the red
end. From 420 into the ultra-violet there is a flattening of the
curve which indicates strongly that ultra-violet light has a dis-
tinct stimulating effect. This effect, however, is very limited in
comparison with the maximum in the field of green.

Percent

of

cyprlds
reacting

filter
whits

mllll-mlcra
(mu-mu)

FIG. 3. A curve showing the distribution of the stimulating efficiency of
equal energy values among the various parts of the spectrum, for the cyprid
larvae of Balanus amphitritc and Balanus improvisus.

Due to the fact that light from a tungsten filament does not
penetrate far into the ultra-violet, we were unable to test this
region as carefully as desired, but in all tests the relative energy
values were carefully calculated.

No filter transmitting ultra-violet was available which did not at
the same time transmit some light in the far-red end of the spec-
trum. (Compare No. 69 in Fig. i.) The efficiency values for
these filters were therefore corrected for the presence of red by

CVI'KII) I.AKY.K OF BARNACLES.

347

•o 2

1

•c T:

i QJ r*

ja'i'd

00

N

s

04 u

N

3^!MA\

»H c«i ^3" ro f^ r*- LO o ro N lo O O O\ i^ o\ ^* ^* O ^^

N
N

-f

Oj

C r^

a.

•o "5

1 <U C

jajji^I

N,Q M t^-coo ^O r^iooi row lOPO^fO^ro^^cO

oo

00

-f

ou

otfW

a;iqA\

NN NNN NP^NPON NN NNNN°NNPO

IO

O

O

be N

C N

^

i,!

aa;!!^

^N &\ r** oo r** 1O ^3 ^^ ^o ^^ t~^ t"^ oo ^3 ^O t** ^^ co oo -^

M

-f

-

0^

N

O

*^" O

3,iqA\

•^SM^'SNN'N^N'NN'NN'N'N^NNNN'N

IO

-t

-.

is

a.

a.

j^jn 4

*-* -t IO O PO 'O OC PO ^ *t ON ^ PO O >O -t rf r^ rj IO

o

.2 ^

a.

\ :i

< Jl.-d

%

OJ >

ON

o o

a

PO t ON PO - rON 0 ONONt^O I^VOP-I 0 ON« -t ON

I-

oc

pH f>

10

10 r-

'.MAY

ro

C

C cc

g.

g_

io 10 10 ^t 10 »o 10 o *^ "1" *o O ^f ON to — < ^t ON C N

;

oi "^

a-

\ ^

41. Jj

c

X ^

§.

OO PO

•* 0 oo ON ON ON 0 i-ONr^«ooo t^- 10 N N PON PO

O-

*t

u P»

10

^i-o

>F1IA\

po

o>

c

a.

a_

j

PO O ON O r-- O N 10 O *o ON *o vo t^* co f^- oo r~- r^ 10

l>

: -

o! ^

a.

\ *

+l.d

2>

cX O

PO

N t^

91IU AA

^" ON M f*» u(J PO OJ N IO *^t ""< O OO N OO M O O O 00

t-

u

10

•*«O

t.MAV

po

ON

a.

a_

J3

QcsMNOMTfMroN 10 10 10 o vo r-> o 't O N

IO

KC

3 10

a.

\ 3-

'••**

N

V

CQ U

o

' O O
•* 0

ONONO ON^IOP; POtOO M t-i t-t OOO ^tO ON'OOO

M

6

IO

PO t~

MMA\

PO

3^

a.

^OC N •-< O r^O O "T"N --fcj OOOt^NOO POO IO

-t

. ^

(U ON
3 10

*.

\ a-

N

-f

PM ^.

O
00

PO ON
PO O

^!MA\

^^'^NM'MN'N^'^'SN'N^SNN'NN^N

0
N

\r.

Cj O
3 0

1

a.
i 3-

-I O^ I I >-J

^HM MM M M M M M

JO

00

OC

M O

o

O tONOO -tl'PO'toO O ONM rOO O POO O ioui

-

LO

•*

POO

N NNN

Tf

-r

3 ^

a

, 1

JO^}[t j-£

POVO"^PON POiOTj-^t^l-iON POPOPO'O-fO *^"PO

IO

N

«£

o

*

c o

O M

^^^M^\\

Sj'N^Sjrl"^^^ PON NPO^ON PO^OPO^ 'N^ON

IO

-f

_o

a

a. a.

, a. , a

j<-)}[ ijj

NNNN^-^^POPOPO^^^PO-O^^N^PO

O

L~ > ~?. — ^

LJ QJ t^ \-/

'O S +3 S 01 II O*

OH^

o

O >0 O O
O f^ >O O

o^itj^'Y^

oo o oo cs ^j t^ I"** o uo *-o >-o w r^- o\ >-o LO ^o ^o ro r^

M«N(NrOOiNC^C4<NC»(NCN(NCSCsWCSC^O|iN

00
N

N O ii '"^ -•* o

M U ^H

1

oo "

jo^j i^j

M«Mn««M««,*H««0««M*C,«

IO

T "
^ v "g ^

^

o

PO

I \ OJ
O O n>

M Q\ ^^

*™

lot^oo O ^ft^r^-Noo POO NOO r^N ^ POt^-N ON
NNNPONNNPOMPOPOPONNPOPOPONPON

IO

°° S^ ^

a. —

„_

&£

1

oo

N

j^5[U

N

~ ^ ~? ~

DO

PO

T "S

10 o rV

M Q\

°^IllA\.

t-OO POOO O>NO O N N 1OONOO "t't'OiOO
N N rOPON N N PON POPOPON POPO^POPON PO

o'

O

O JJ w

4^ SPO

U

c

1 .-'. ' '

rt "*"" s^ M

s

*^ £i ^*

q

3

zr o -j

— •

.5 c u ^

M

c

^

cd

w f- -1— ^

&

to

f-H

o

^~* U P ~* *—"

J

E

1, w ^J ^

( "

fcw

g -w 3 •£ ^

<D

<U

O

'"o

,4_i

>

u

-*-! 'Cj -^

j^j

oi

C

O

U

f_

s

^3

K *J

<H>— "^ *^" *~

M^

!>

^ ^

52** ^3 •-•

<-*-<

H

r- OC

O

3

— 1^

o

>z.

3

t-« ^ '5 o f

<L»

C r"

rj "^ W U

.0

CO

X

4J r- .M

4> ^ W) 7"

p

3

z

o

n

!3

M N PO^'OO t^oo ONO M N PO^LOO r^oo O\O

r-.

£°~~-

E
P
fc

CO

Id

i

w

fc

W

-> d

j <

< 55

u «

Of <

W M

fe to

o o

s «
- ^

H ><

« W
w H
fc X

l§

w

B af

Q s
«

8 S

w
w frl

H CO
§ W

§ W

5 H

OS

U

PH

X

W

h

O

t«

£

co

>H

H
Z
U
£

H

fe

o

H

tn

w

348 J- PAUL VISSCHER AND ROBERT H. LUCE.

determining the percentage of red transmitted and then deducting
that value from the total value on the basis of the effectiveness of
that region in the red.

During the course of this work on cyprids it was also found pos-
sible to test the reactions of the nauplii of four species of barnacles
to these lights as described above. These results were uniform in
that the region of the spectrum between 530 and 545 /A/A was in
each case the most stimulative. This maximum stimulating effect
was the same both when these organisms were negative and when
positive to light.

It is of interest that our results are in substantial agreement with
the work of other investigators using nauplii of barnacles : Loeb &
Maxwell, 1910, Loeb & Westeneys, 1916 and Loeb & Northrup,
1917. Thus it is apparent that the same region of the spectrum
offers the greatest stimulating efficiency in both the nauplius and
in the cyprid which is not surprising in view of the fact that the
nauplius eye is still present in the cyprid.

CONCLUSIONS.

The cyprid larvae of barnacles agree with the larvae of most
other marine organisms in that they react to light. With equal
radiant energies and with the infra-red heat rays eliminated, a dis-
tinct difference in stimulating effectiveness is found between dif-
ferent regions of the spectrum. This difference shows, for the
cyprids of the two species of barnacles tested, a region of maximum
efficiency in the light green, between 530 and 545 /A/A. It is of in-
terest to note that similar results have been found for the larvae of
other marine organism. The larvae of the hydroid, Eudendrium
are most sentitive to blue light (460-480 /A/A) ; the larvae of Are-
nicola, to the green (about 495 /A/A) ; and the larvae of the squid,
to blue green (470-510^).

However from the results obtained by Hess (1910), Mast
(1917) and Hecht (1921), it is apparent that the distribution of
stimulating efficiency in the spectrum is not necessarily the same
for all organisms, which react to light. It is consequently appar-
ent that there is no true color vision in these lower forms — in the
usual interpretation of the term, but rather that the photo-sensitive
materials in the light perceiving organs are more sensitive to the

CYPRID LARV^: OF BARNACLES. 349

effects of certain wave-lengths than to others. Increased intensity
of other wave lengths may however produce similar results.

This investigation has, however, an important bearing on the
problem of the nature and extent of the fouling of ships' bottoms,
an intensive study of which is being completed by the senior author
and of which this work forms a part. It has been found that light
is an important factor governing the attachment of the organisms
which cause fouling and that at the time of attachment many of
these larvae are negative to light, for they are found most abun-
dantly on the shaded portions of the ships' hull.

Since barnacles are the most important factor in fouling it seems
probable that the studies reported above may have some significance
in the solution of this problem, for if a light green paint were used
in place of the red, now universally used, a considerable benefit
should be derived, on theoretical grounds, provided all other fac-
tors were comparable.

BIBLIOGRAPHY.

Blauuw, A. H.

1908 The Intensity of light and the length of Illumination in the Photo-
tropic Curvature in Seedlings of Avena sativa. Kon. Ak. Wet.
Amsterdam, Proc. Meeth., Sept. 26, 1908.
Day, Edward C.

1911 The Effect of Colored Light on Pigment-migration in the Eye of
the Crayfish. Bull. Mus. Comp. Zool., Vol. 53, 1908-1913, pp.
303-344.
Gibson, K. S., Tyndall, E. P. T., McNicholas, H. J.

1920 The Ultra-violet and Visible Transmission of Various Colored
Glasses. Technologic Papers of the Bureau of Standards, No.
148. 27 pp. Washington.
Gross, Alfred O.

1913 The Reactions of Arthropods lo Monochromatic Lights of Equal

Intensities. Jour. Exper. Zool., Vol. 14, pp. 467-513.
Laurens, Henry.

1911 The Reactions of Amphibians to Monochromat:c Lights of Equal

Intensity. Bull. Mus. Comp. Zool., Vol. 53, pp. 251-302.
Loeb, J. and Groom, T. T.

1890 Der Heliotropismus der Nauplien von Balanus perforatus und die
periodischen Tiefenwanderungen pelagischer Tiere. Biol. Centr.,
Bd. X., s. 160-177 und 219-220.
Loeb, J. and Maxwell, S. S.

1910 Further Proof of the Identity of Heliotropism in Animals and
Plants. Univ. Cal. Pub. Physiol., Vol. 3, 1910, pp. 195-197.

350 J. I'AUL VISSCHER AND ROBERT H. LUCE.

Loeb, J. and Westeneys, H.

1917 A Reexamination of the Applicability of the Bunsen-Roscoe Law
to the Phenomena of Animal Heliotropism. Journal Experi-
mental Zoology, Vol. 22, pp. 187-192.
Loeb, J. and Northrup, J. H.

1917 Heliotropic Animals as Photometers on the Basis of the Validity
of the Bunsen-Roscoe Law for Heliotropic Reactions. Proc.
Nat. Acad. Sci., Vol. iii, pp. 530-544.
Mast, S. O.

1911 Light and the Behavior of Organisms. 410 pp. New York.
1917 The Relation between Spectral Color and Stimulation in the Lower
Organisms. Journal Experimental Zoology, Vol. 22, pp. 471-
528.
Merejkowsky, C. de.

1881 Les crustaces inferieurs distinguent-ils coulours. Compt. Rend.

Acad. Sci., Paris, T. 93, pp. 1160-1161.
Pergens, E.

1899 Ueber Vorgange in der Netzhaut bei Farbiger Beleuchtung gleicher
Intensitat. Zeitschr. F. Augenheilk., Bd. 2, Heft 2, s. 125-141.
Visscher, J. Paul.

1928 The Nature and Extent of the Fouling of Ships' Bottoms. In

press. (Bull, of the U. S. Bureau of Fisheries, Vol. 43.)
White, G. M.

1924 Reactions of the Larvae of the Shrimp, Pal&monetes vulgaris, and
the Squid, Loligo pealii, to Monochromatic Light. BIOL. BULL.,
Vol. 47, pp. 265-273.

Vol. LIV.

-1/c/v, 1928

Xo. 5.

BIOLOGICAL BULLETIN

VITALITY OF THE GAMETES OF CUMINGIA

TELLINOIDES.

B. H. GRAVE,

WABASH COLLEGE, SEESSEL FELLOW, YALE UNIVERSITY, 1926-27.
From the Marine Biological Laboratory, Woods Hole, Mass.

Cumingia tdlinoides is a small lamellibranch mollusk found
abundantly in the Woods Hole region. Its breeding season
extends over a period of three or three and a half months, be-
ginning early in June. The production of eggs is continuous
and spawning by each mature individual occurs two or three
times during the course of the breeding season. Eggs may be
had in great abundance during this time. One female may shed
300,000 eggs, although the average number extruded is estimated
to be between 100,000 and 200,000. They are expelled into the
sea water where they are fertilized.

The present paper is a report on experiments to determine the
vitality of the gametes, and more especially the longevity of
the unfertilized eggs. It may be assumed that when eggs are
released from the ovary, they are normal. From the time of
liberation, deterioration begins and if not fertilized they ulti-
mately go to pieces by fragmenting in a characteristic manner.
The extensive work of Goldforb indicates that some deterioration
takes place in sea urchin eggs while stored in the ducts of the
gonad before extrusion. This may also be true of Cumingia,
since two clear cases of physiologically poor lots of eggs have
been found among hundreds of normal ones. Deterioration o
both gametes of Cumingia and of Arbacia is rapid after extrusion
into sea water.

VITALITY OF THE EGGS.

The eggs used in these experiments were obtained by isolating
sexually mature individuals in small stender dishes filled with sea
24 35i

352

B. H. GRAVE.

water into which the eggs were shed. The eggs, so collected in
one dish, may therefore be conveniently referred to in the
description of the experiments as a "batch" or a "lot" of eggs.
The reference in all cases is to the eggs of a single female.

During the summer of 1926 many batches of eggs were kept
without fertilization to test their longevity. During the progress
of the experiments a few eggs were taken from the dishes from
hour to hour and fertilized as a test of their physiological con-
dition and of the rate of deterioration as indicated by decreasing
percentage of fertilization. The experiments reported here deal
principally with the longevity of the unfertilized eggs as indi-
cated by the time of disintegration.

Approximately two per cent, of the eggs of most females are
defective at spawning and fragment within four or five hours.
The other ninety-eight per cent, remain intact and capable of
development for longer periods. The poorest lot of eggs when
tested at a temperature of 20° to 22° centigrade went to pieces
completely in six hours, and it was not uncommon for 98 per cent
of the eggs of some females to fragment in from eight to nine
hours. The great majority of the lots of eggs fragmented in
from nine to fifteen hours, the average being between ten and
twelve hours. When tested at 18.5° to 20° C., on the other
hand, the most vigorous batches of eggs, remained intact from
fifteen to twenty hours, that is, the eggs retained their normal
appearance for that length of time. It was shown that a majority
of the eggs either refuse to develop or develop abnormally, when
fertilized shortly before fragmentation. However, a few normal
embryos come from the oldest eggs, even when the percentage of
fertilization is greatly reduced. When a lot of eggs is fertilized
after fifty per cent, of them have fragmented, it not infrequently
happens that most of those which remain intact fertilize and
cleave. However, most of them die in cleavage or gastrula
stages and few develop into normal embryos. The oldest lots
of eggs to develop normally were ten and twelve hours old and
rarely fifteen hours old. These all showed 95 to 98 per cent,
fertilization and normal development.

The extreme limits of longevity of the unfertilized eggs are
therefore found to be six to twenty-six hours, and the limits of

VITALITY OF THE GAMETES OF CUMINGIA TELLINOIDES. 353

fertilization and normal development in the best lots of eggs are
ten to fifteen hours. These figures are for temperatures normally
experienced by the eggs of the species, or 18.5° to 22° C. They
therefore represent the normal longevity of the eggs.

The statement by Morgan that the eggs of Cumingia will not
stand rough handling is misleading. They are injured by
centrifuging but there is abundant evidence that they develop
normally under ordinary laboratory manipulation.

Tables I., II. and III. deal with various phases of the longevity
and vitality of the eggs and include representative experi-
mental data.

INDIVIDUAL VARIATION.

After any particular lot of eggs begins to fragment, three or
four hours elapse before all or 98 per cent, have fragmented.
The variation in longevity within a single group is, therefore,
approximately four hours. For example, if 10 per cent, of the
eggs have fragmented at twelve hours, the average expectancy
of fragmentation for the remainder would be 30 per cent, in
thirteen hours, 50 per cent, in fourteen, 70 per cent, in fifteen,
and 95 to 98 per cent, in sixteen hours. Two per cent, may
remain intact indefinitely. The rate of fragmentation is shown
in Table I., as is also an actual comparison of the longevity of
five lots of eggs from five females studied under identical
conditions. Tables II. and III. show the same thing in slightly
different form.

When fresh eggs are fertilized by fresh sperm the percentage of
cleavage is usually between 97 and 100 per cent. However, a
few lots of eggs gave from 89 to 95 per cent, cleavage. One lot
of physiologically poor eggs was observed which when fertilized
with physiologically good sperm gave only 20 to 30 per cent,
cleavage. This sperm when used to fertilize other lots of eggs
gave from 97 to 100 per cent, cleavage showing that failure to
cleave on the part of these eggs was not due to defective sperm.
This lot of physiologically poor eggs was slimy from the first
and showed a tendency to cling to the containing dish. This
together with one other lot of defective eggs may be regarded as
confirmation of Goldforb's contention that eggs at the time of
spawning may show all the symptoms characteristic of aging

354

B. H. GRAVK.

w

t— 1

w
H

.

O
O

W

TALITY OF C;

"~ " ^ O
•*J •*"* LJ

= .sl

c rr -

'—• ^- 13

? 2 o

r-1 ® ~*~*

5^ S

"o rt cd
w c3 i-i

'-" x c —

N o O O

1 I r^

•^* ^l — f I

M= gg

,,

O -ti to a;

cd ^

I?

<u.S
Ij

>>•;

•o S

OJ II

I £

5 >

S--S

ai

O 4J

'-^- 3
O CO
^ OJ '

3 C

agmentation began
appearance. Fro

•0 ^

P .—
D. 5

> 01

'-C >

<— 'So— 5

T) S *i +J

J- o JS ni

o 4-> -G _^

,3 S

* ~
O .

o .

P Mi

a rt x cs s ci a

^ rt oj rt

VO

s s s s s s s

s s E E

d

O

o o o o o o o

O O O CO CO O O
O O O Os O\ O\ CO

t-* M M

o o o o

O O N O
in ro

4

E E E E £ E E

E £ E I

0
0

o o o o o o o

c c a c c c c

£§ £? ^x ^x ^§ £x fe§

O O O O O 00 OO

o o o o

s s c c

u"> "1 O O

r-O

I E E I E E I

E I £ E

o'
o

o o o o o o o
c c c c c c c

o o o o o o o
o o o o o o o

o o o o
c c c c

00 to O O
O\ O\ O\ oo

tN

O

o

E S E E

o o o o

C c c c

I— O ""> O

O\ 0

I-H

IH

6
o

100% normal
70% normal
20% normal
o% normal

'. '. '. w ;

FHlNiHlN *^ "^ O M N

ill!

yn

H

88

a —

^ c ^5 .b — M£ .« -5

o^-.

' '

£-3 "2 ^-5 S.s

•= -e «

G ^^ M

s

VITALITY OF THE (iAMETES OF CUMINGIA TI-:i.l,I XOIUES. 355

w
J
m
<
H

=.-

C, •-

r- ^ ~
— r- O

HI

ao
1

—

G

Q

X

U

C

U

K

*T*

t—t

5 °

O <u

2l

J -+->

W g

J o

U|

w S

H

z

B

a

CH

o

H

O

a ^

'Z c

- ,r -
-C C3 >-

-u rt O

J QO

Cj O

C =1

'- x

"S % S

r S

c o

'M c5
rt c.

b' J 3
^ ~ S
c S S

"• v. rt

|^^

r, rt

3T 2<-s u

. •/. S o

o s
73 c

356

B. H. GRAVE.

eggs. The first vertical column of Table II. shows the usual
range of variation in freshly spawned eggs in respect to per-
centage fertilization and cleavage.

TABLE III.

LONGEVITY OF THE EGGS OF Cumingia tellinoides, TEMPERATURE 22° C.

The eggs of the ten females compared in Table II. in respect to percentage
fertilization are here compared as to longevity and time of fragmentation or dis-
integration. It will be noted that the time of reduction in cleavage is almost
synchronous with the time of fragmentation. This relationship does not always
hold because in numerous experiments the percentage of cleavage fell off before
fragmentation. In other words, if the reduction in cleavage due to aging and the
rate of fragmentation were plotted as curves they would not as a rule be super-
imposed or parallel.

Percentage of Fragmentation with Aging.

After 7 Hrs.

After ii Hrs.

After 15 Hrs.

After 16 Hrs.

Female No.
Female No.
Female No.
Female No.
Female No.
Female No.
Female No.
Female No.
Female No.
Female No.

i

o%
o%
75%
o%
73%
o%
o%

0%
0%
100%

o%
45%

100%

2%
100%
0%
0%
0%

7%

100%

50%

100%
100%

72%
100%

40%

99%

12%
100%
100%

100%
100%
100%

90%

100%

50%

100%

60%

100%
100%

2

T

A

*. .

6

7

8

Q

IO

Note. — Many of the eggs of lots 4, 6 and 8 lived beyond sixteen hours. The
other lots of eggs had fragmented completely after sixteen hours.

VITALITY OF THE SPERMATOZOA.

The work of Gemmill ('oo) and F. R. Lillie ('15) has shown
that variations in the concentration of sperm suspensions of
Arbacia make a great difference in the duration of their fertilizing
ability. The writer accordingly used several sperm dilutions in
order to learn whether the same phenomena are exhibited by
the sperm of Cumingia.

Spermatozoa shed by a mature male Cumingia of average size
in 30 cc. of sea water was estimated by actual measurement to
be a 2 to 3 per cent, suspension.1

1 The method of measuring the percentage was to kill the spermatozoa by the
addition of formalin and after settling they were measured en masse. From this
the calculation of the concentration of the original suspension was a simple matter.

VITALITY OF THE GAMETES OF CUMINGIA TELLINOIDES. 357

This concentrated suspension, though at best somewhat vari-
able, was used as a standard suspension from which various
dilutions were made. It was found possible to select suspensions
of approximately the same strength and this is a matter of more
importance than that the exact percentage be known.

Two drops - of standard sperm suspension when added to
25 cc. of sea water was estimated to be a 1/500 to 1/750 per cent,
suspension; two drops in 50 cc. of sea water was considered to
be a i/iooo to 1/1500 per cent, suspension; one drop in 50 cc.
of sea water makes a 1/2000 to 1/3000 per cent, suspension, etc.
The last named suspension when fresh is adequate to fertilize
one hundred per cent, of the eggs whereas greater dilutions
sometimes gave only partial fertilization. What was estimated
to be a 1/4000 to 1/6000 per cent, suspension gave from eighty
to one hundred per cent, fertilization and usually one hundred
per cent. All of these suspensions were used and also the same
percentage suspensions in larger quantities of sea water.

The method of studying the relative longevity of these various
dilute sperm suspensions was to make up, by the proper dilutions,
several dishes of each from freshly shed sperm. To these fresh
eggs were added in turn at hourly intervals until the suspensions
no longer gave fertilizations. It was shown in general that the
weakest suspensions die first. Tables 4 and 5 show that the
longevity of the spermatozoa depends largely upon the degree
of concentration. Even a suspension of 1/500 per cent, shows
some preserving effect in that spermatozoa live for a longer time
in this concentration than in greater dilutions. There is little dif-
ference between a 1/2000 and a 1/6000 per cent, suspension, and
these no doubt represent natural conditions so far as longevity
is concerned.

In work reported at this time the writer had in mind to study
the longevity of gametes under natural conditions. In general
the longevity of the sperm in the most dilute suspensions is
somewhat less than that of the eggs. In a few cases using
suspensions of 1/500 per cent, approximately ninety per cent,
normal embryos developed from sperm and eggs that were

2 The same pipette was used in making all dilutions of an experiment and one
giving approximately one cc. per 20 drops

358

B. H. GRAVE.

twelve hours old and in numerous other cases at nine and ten
hours. As a rule when sperm suspensions of I 2000 to 1/3000
per cent, are used very few fertilizations occur after nine or ten
hours. Numerous experiments show that spermatozoa begin to
die after three and one half or four hours. The indication
therefore is that a majority of the spermatozoa under natural
conditions live from four to nine hours.

TABLE IV.

LONGEVITY OF THE SPERM OF Cumingia IN VARIOUS DILUTION'S AS SHOWN BY

THE PERCENTAGE OF FERTILIZATIONS THAT THEY GIVE

WITH FRESH EGGS (JULY 25).

Age of the Suspen-
sion Tested.

i Drop in
100 Cc. Sea
Water 1/4000
to 1/6000%.

I Drop in
50 Cc. Sea
Water 1/2000
to 1/3000%.

4 Drops in
50 Cc. Sea
Water 1/500
to 1/750%.

8 Drops in
50 Cc. Sea
Water 1/250
to 1/500%.

2 hours

00% +

99% +

100 ' ,

100%

3^ hours

96%

95%

97%

96 %

7 hours. . . .

17%

24%

99%,

97%

10 hours . .

2%

8%,

639?

72%

For this experiment the spermatozoa used were all from the same male. The
eggs were from three females and were not over three hours old when used.
Controls showed them to be practically 100 per cent, normal eggs.

TABLE V.

LONGEVITY OF THE SPERM OF Cumingia IN VARIOUS DILUTIONS

TESTED ON JULY 23.
Sperm all from one male.

Age of
Sperm
Suspen-
sion L'sed.

i Drop in
50 Cc. Sea
Water 1/2000
to 1/3000%.

2 Drops in
50 Cc. Sea
Water i/iooo
to 1/1500%.

4 Drops in
50 Cc. Sea
Water 1/500
to i/75f-'( o-

6 Drops in
50 Cc. Sea
Water 1/300
to 1/500',.

2j hours. . .
4,', hours. . .
75 hours. . .
9 hours. . . .

84% cleaved
35% cleaved
1 5-20% cleaved
o% cleaved

1 00% cleaved
65 % cleaved
30% cleaved
o% cleaved

1 00% cleaved
75% cleaved
50-60% cleaved
2% cleaved

99%+ cleaved
80% cleaved
85-8 7% cleaved
3 5-40% cleaved

This table shows that sperm in 1/2000 to 1/3000 per cent, suspension died in
nine hours. Other experiments have shown that from ten to forty per cent, of
the spermatozoa in these most dilute suspensions often survive for 10 to 12 hours.
Table IV. is more typical in this respect, but even in this case the percentage ot
fertilization is below the average. The point that is shown clearly is that sperma-
tozoa die first in the most dilute suspensions and that a suspension of 1/500 to
T/750 per cent, suspension preserves the life of the sperm beyond their normal
life in the open sea.

VITALITY OF THE GAMETES OF CUMIXCJA TEI.LIXOIDES. 359

When the work of Gemmill and Lillie is taken into consideration
it is apparent that Goldforb's experiments on the aging of
spermatozoa do not represent normal conditions. He apparently
relied upon concentrated dry sperm which are known to age
much more slowly than sperm kept under natural conditions.
The aging of sperm in concentrated suspensions has no particular
significance. His work on the eggs is certainly not open to the
same criticism but I should question his interpretation and
results on the aging of spermatozoa. The work of the other two
authors and this present work show that the life of spermatozoa
in dilute suspensions is brief.

The statement by Lillie that the spermatozoa of Arbacia lose
ability to fertilize eggs in a few minutes in dilute suspensions is
not verified for Cumingia sperm, although the theory that they
gradually lose something to the water which is essential to
fertilization and which ultimately renders them incapable of
bringing about the fertilization reaction may very well be true.

It has been my observation that spermatozoa that are able to
swim actively are capable of fertilizing eggs; and that they do
not lose the ability to initiate the fertilization reaction after a
few minutes as reported by Lillie for Arbacia. Up to the present
time no attempt has been made to learn whether or not the
spermatozoa of Cumingia contain an activating substance,
essential to fertilization. In many ways the spermatozoa of
Cumingia behave as those of Nereis and Arbacia do, but if they
contain a sperm receptor (Lillie, 15) preliminary experiments
indicate that it is dissipated in the sea water more slowly than
in the cases investigated by Lillie. Experiments on this question
are in progress, as well as investigations of the cause of excessive
polyspermy. These experiments are designed for comparison of
the gametes of Cumingia, Chcetopleura and Hydroides with the
work of Goldforb and Gemmill on sea urchin eggs and with
Lillie's work on the spermatozoa of Nereis and Arbacia.

ACTIVATION OF SPERMATOZOA.

Spermatozoa, after having apparently lost their vitality, may
be revived. \Yhen placed in a dish with eggs they become
activated and swim vigorously. The activation of sperm in the

360 B. H. GRAVE.

presence of eggs was observed repeatedly. The stimulus from
the eggs or egg water is evident almost instantly, but not all of
the spermatozoa are so activated. I interpret this to mean that
most of the quiescent spermatozoa are already dead or weakened
beyond the possibility of functioning.

The phenomenon of activation of apparently spent sperm by
the exudations from eggs is interesting and not uncommon.
The physiological value of this activation is evident although
the real cause is obscure. Lillie (13) shows that the spermatozoa
of Nereis and Arbacia are positively chemotropic to weak acids
and to egg secretions and are apparently stimulated by them.

CONCENTRATED SPERM SUSPENSIONS.

It was learned that spermatozoa in concentrated suspensions
retain their vitality for very long periods. Under such conditions
they may swim from twenty-four to thirty-six hours and give
90 to 100 per cent, fertilization and normal development. They
retain sufficient life for four days to show occasional contractions
visible under a compound microscope. It is evident therefore
that there is some protective element in this unnatural con-
centration. Cohn claims that it is carbon dioxide or hydrogen
ion concentration. Dry sperm of fishes has the same extended
longevity although possibly from different causes, not having
had the initial stimulus to swim and use up its limited store

of energy.

DISCUSSION.

Goldforb expresses the belief that deterioration of eggs begins
at the time of maturation while they are still stored in the gonads.
He attributes the great variability of sea urchin eggs to the
differences in time that they remain in storage before spawning
takes place. I find a similar variation in the vitality of eggs of
Cumingia, but it is noteworthy that maturation does not take
place in the eggs of this species until after the entrance of the
spermatozoon which occurs after extrusion into the sea water.
Deterioration in the gonad in this case, therefore, could not be
due to maturation. The fact that eggs vary so much in their
longevity leads to the belief that some deterioration takes place
before extrusion and before maturation. It is hardly likely that

VITALITY OF THK (1AMETES OF CUMINGIA TELLINOIDES. 361

the great difference in vitality that experiment reveals is due
entirely to natural variability. The eggs are stored in the
gonads and their ducts for some time before extrusion and those
that are stored longest may very well show physiological deteri-
oration. Two cases of eggs which showed all the symptoms of
aging at spawning verify Goldforb's contention on this point.
At the present time I am unable to state whether the considerable
variability in the longevity of Cumingia eggs is due principally to
normal variability or is in part due to physiological deterioration
while in storage. Cases of low percentage fertilization in the
eggs of Cumingia are rare. As a rule they give 97 to 100 per cent,
cleavage, although the longevity may be more variable.

SUMMARY.

I. The average longevity of Cumingia eggs when kept at a
temperature of 20° to 22° C. is 10 to 12 hours as judged by
time of fragmentation and ability to give a high percentage of
normal embryos. The average longevity at 18.5° to 20° C. is
12 to 15 hours.

II. Approximately two per cent, of any lot of eggs may be
defective as shown by their early fragmentation which occurs
long before the rest.

III. The outstanding fact is the wide range of variation in the
vitality and longevity of Cumingia eggs. The eggs of a single
individual vary by four hours or approximately 25 per cent.,
while the eggs of different females vary in their longevity from
six to twenty-six hours or over 400 per cent.

IV. Eggs in rare instances show deterioration at spawning,
apparently due to long-time storage in the ducts of the gonads.
Most lots of eggs when freshly spawned give from 97 to 100
per cent, cleavage and normal development.

V. The longevity of the best sperm in suspensions of 1/400
to 1/500 per cent, is 10 to 12 hours as judged by functional
activity and ability to give 90 to 100 per cent, fertilization and
normal development. In sperm suspensions of 1/2000 to 1/3000
per cent, it is 4 to 7 hours, but from 30 to 50 per cent, fertilization
may frequently be expected from suspensions 7 to 12 hours old and
i to 5 per cent, fertilization from suspensions 12 to 20 hours old.

362 B. H. GRAVE.

VI. Spermatozoa in concentrated suspensions of I to 3 per cent,
retain their vigor for many hours. They frequently give 90 to
100 per cent, fertilization after twenty-four to thirty-six hours,
and some individual spermatozoa live for four days, still showing
faint contractions of the tail at intervals. This preservation of
the life of the spermatozoa is attributed to the presence in the
water of COz produced by the activity of the spermatozoa , or to
hydrogen ion concentration.

VII. Spermatozoa after becoming quiescent show activation
in the presence of eggs or egg water.

This paper was read by title before the American Society of Zoologists at the
meeting in Philadelphia, December, 1926. An abstract was printed in the
Anatomical Record.

BIBLIOGRAPHY.
Cohn, E. J.

'16 The Relation between the Hydrogen Ion Concentration of Sperm Sus-
pensions and their Fertilizing Power. Anat. Rec., Vol. II, p. 530.
Goldforb, A. J.

'17 Variability of Eggs and Sperm of Sea-Urchins. Publication No. 251 of

the Carnegie Inst. Washington, D. C., p. 71.
Goldforb, A. J.

'18 Effects of Aging upon Germ Cells and upon Early Development. BIOL.

BULL., Vol. 34, p. 372.
Gemmill, J. F.

'oo On the Vitality of the Ova and Spermatozoa of Certain Animals. Jour, of

Anat. and Physiol., Vol. 34 (N.S., Vol. 14), p. 163.
Lillie, F. R.

'13 The Behavior of the Spermatozoa of Nereis and Arbacia with Special

Reference to Egg-extractives. Jour. Exp. Zool., Vol. 14, p. 515.
Lillie, F. R.

'15 Analysis of Variations in the Fertilizing Power of Sperm Suspensions of

Arbacia. BIOL. BULL., Vol. 28, p. 228.
Lillie, F. R.

'15 Sperm Agglutination and Fertilization. BIOL. BULL., Vol. 28, p. 18.
Nelson, T.

'21 Report of The Biologist, N. J. Agr. Exp. Sta., year ending June 30, 1921,
p. 297.

CONDITIONS DETERMINING THE ORIGIN AND BE-
HAVIOR OF CENTRAL BODIES IN CYT ASTERS
OF ECIIINARACIINIUS EGGS.1

HENRY J. FRY,
WASHINGTON SQUARE COLLEGE, NEW YORK UNIVERSITY.

CONTENTS.
INTRODUCTION

Statement of the Problem Concerning the Origin of Central Bodies 363

Types of Evidence Valid for Proving a Self-perpetuating or a de novo

Origin of a Cell Component 366

EXPERIMENTS

The Effects of Various Fixatives upon Rays and Central Bodies of

Cytasters 3&9

The Effects of Modifications of Environmental Factors upon Rays and

Central Bodies of Cytasters 375

The Effects of Various Intervals in the History of Cytasters upon Rays

and Central Bodies . 379

DISCUSSION

The Origin of Central Bodies in Cytasters of Echinarachnius parma 383

The Method of Study and its Significance for Cytological Research 387

RESUME 392

LITERATURE CITED. ... 395

INTRODUCTION.

Statement of the Problem Concerning the Origin of Central Bodies.
In every cell generation there are differentiated the various
cell components such as chromosomes, plastids, Golgi bodies,
chondriosomes, central bodies, secretory droplets and similar
elements. In the case of certain ones, such as chromosomes, the
synthetic processes producing new units occur only by the growth
and division of preexisting bodies of the same kind. Such a
component has genetic continuity from cell to cell as an indi-
vidualized body, appearing to play some role in its own formation.
Its origin may therefore be termed autogenic or self-perpetuating.
In contrast to this, the synthetic processes forming other com-
ponents, such as secretory droplets, are not localized about pre-
formed structures of the same kind. Such a component has no

1 The experiments were performed at the Mt. Desert Island Biological Station,
Maine, 1923, and at the Marine Biological Laboratory, Woods Hole, Massachusetts,
1926.

363

•

364 HENRY J. FRY.

genetic continuity as an individualized body and its origin is
commonly described as occurring de novo.

Central bodies 2 have special interest with reference to this
problem of protoplasmic differentiation, since in some cases
they arise by the growth and division of preexistent central
bodies, while in other instances they apparently are formed
de novo. In some cells they persist from one cell division to
another, maintaining their identity as individualized structures
during the vegetative period when the mitotic mechanism, of
which they appear to be the formative foci, has disappeared.
At a certain time in the cell cycle they divide and the daughter
cells receive central bodies that have genetic continuity with
that of the mother cell. For example, in spermatogenesis, the
central body of the sperm can frequently be traced as a self-
perpetuating individualized structure to the central body of the
primary spermatocyte.

On the other hand there is also much evidence for a de novo
origin of central bodies. In higher plants where centers are
absent, blepharoplasts which have the characteristics of central
bodies appear de novo in the final divisions of the germ cells. In
animal fertilization it has yet to be proved that the central body
of the sperm aster has genetic continuity as an individualized
structure with the central body of the sperm, except in a few
forms such as Cerebratulus (Yatsu, '09), and in most cases its
formation may be de novo. When eggs are artificially activated
in such a way that cytasters are formed, a study of sectioned
material shows that they have central bodies (Morgan, '96, '99,
'oo ; Wilson, '01 ; Hindle, '10; Herlant, '19; Fry, J2$b;
Tharaldsen, '26; etc.). Since cytasters appear simultaneously
in large numbers and not by division the central bodies are
apparently formed de novo.3

- The term central body is used throughout this paper in its broad non-
committal sense, denoting any differentiated structuie present at the central
region of an aster, exclusive of the inner ends of the rays (and the possible presence
of chromatin). The term involves no implications as to whether or not the central
body is composed of a granule, the centriole, and a surrounding zone, the centro-
some, either of which may be present or absent in various organisms, and both of
which may show many modifications (Wilson, '25, pp. 30 and 673).

3 For a detailed discussion of the literature on this subject see Wilson "The
Cell in Development and Inheritance," 1925, especially Chapter IX.

CENTKAL BODIES IN CYTASTERS. 365

The idea that all central bodies are self-perpetuating structures
received an impetus due to the early theory of fertilization
developed by Boveri and Van Beneden. They assumed that
the chief event in fertilization is the introduction by the sperm
into the egg of a preformed division center and hence that the
central body of the sperm aster has genetic continuity as an
individualized structure with that of the sperm (Boveri, '01).
So generally accepted was this assumption that when artificial
parthenogenesis was discovered, where the aster appears about
the egg's nucleus, it was taken for granted that the egg nucleus
must therefore have such a preformed central body that is
ordinarily quiescent under the normal conditions of fertilization.
To explain cytasters according to this theory it was necessary
either to regard them as "pseudoasters," assuming that if they
do not arise about preformed centers they are not "true asters,"
or to assume that central bodies lie scattered through the cyto-
plasm as a result of hypothetical rapid divisions of the egg's
hypothetical division center (Meves, '02 ; Wassilieff , '02 ; Pe-
trunkewitsch , 04) . These ideas which are without proof are only of
interest to illustrate the strength of the underlying assumptions
which are still generally accepted, i.e., that central bodies have gen-
etic continuity as individualized structures, and that asters ordi-
narily arise about such units on the theory that the central body
is the persistent element of the astral mechanism acting as the
formative focus.

A possible explanation of the seemingly conflicting evidence
that in some cases central bodies have an autogenic origin and
in others a de novo one has been proposed by Wilson. He
suggests that central bodies which seem to arise de novo, may
actually be self-perpetuating structures that are submicroscopic
during most of their history; they grow to a visible size only at
certain times when they give the impression of having a de novo
origin. This suggestion is in harmony with the general assump-
tion that central bodies are individualized structures maintaining
genetic continuity and giving rise to asters. He states: "...
the cytologist is led on to the conclusion that the ultra-
microscopical dispersed particles of the hyaloplasm may be as
highly diversified chemically as are the visible formed bodies;

366 HENRY J. FRY.

that they may be of all orders of magnitude; and that it is they
which constitute the sources, or at least the formative foci, of
those larger formed bodies that we have so often, but erroneously,
assumed to arise de novo. For my part, I am disposed to take
a final step by accepting the probability that many of these
particles (I do not say all) as if they were ultra-microscopical
plastids, may have a persistent identity, perpetuating them-
selves by growth and multiplication without loss of their specific
individual type" ('23, p. 28). "Could we accept such a view
we could more readily meet some puzzling difficulties such, for
example, as the apparent contradiction between the origin of a
centriole de novo, and its origin by division of a preexisting body
of the same kind" ('25, p. 721).

Morgan ('96, '99, 'oo) has made the most careful study of the
relationship between central bodies and the structure of the
surrounding cytasters. As a result of his observations he raised
certain questions, in sharp contrast to the generally accepted
ideas, that have since been largely ignored: "Is this central
deeply staining mass a distinct body, or only the innermost
fused ends of the rays; or is it only a product of the reagent?"
('99, p. 477). The present study is an attempt to resolve these
conflicting hypotheses concerning the origin and behavior of
central bodies in cytasters.

Types of Evidence Valid for Proving a Self -perpetuating or a
de novo Origin of a Cell Component.

The behavior of a cell component in hybrids yields the only
conclusive evidence concerning the nature of its origin. For
example, in hybrids the paternal chromosomes and their con-
stituent genes continue to perpetuate themselves and produce
their effects in the protoplasm of another species. This con-
clusively proves that they are autogenic, playing an important
role in their own synthesis. It demonstrates that under normal
conditions, within the protoplasm of their own species, they are
not passively produced by more active elements, otherwise they
would not continue to grow and divide in an alien protoplasm.
There is similar evidence for the autogenic origin of certain
plastids. When species with different plastids are crossed, the

CENTRAL BODIES IN CYTASTERS. 367

paternal type continues to perpetuate itself in a cytoplasm that
normally contains a different kind of plastid. The autogenic
origin of other cell components has not been subjected to this
conclusive proof of their capacity to perpetuate themselves by
growth and division in the protoplasm of another species. Such
a study of central bodies, in reciprocal crosses between Arbacia
and Echinarachnius, is now in progress, a report of which will
be presented in a later paper.

After reviewing the facts concerning certain plastids, Morgan
('26) states: "It may not appear far fetched to assume that
there may be other bodies in the cytoplasm that grow and divide
and, by extension, it might not seem too extravagant to assume
that protoplasm itself (except for its secretion products) consists
of units that grow and divide and are inherited." He calls
attention, however, to the fact, very significant in this connection,
that when two species are crossed reciprocally they produce
similar hybrids. If some of the cytoplasmic components autono-
mously perpetuate themselves, maintaining their characteristic
form, then in reciprocal hybrids the adults should show cyto-
plasmic differences. In one cross the cytoplasm with any self-
determining units is derived from one species, and in the opposite
cross it comes from the other. Since the cytoplasm of the adults
is similar in both, despite its differing sources in the two crosses,
this shows that its components generally are not autogenic in
such a manner as to maintain their identity. It indicates that
the synthetic processes are under the control of the nucleus
producing similar results in both cases. It is generally to be
expected, therefore, with some exceptions such as certain plastids,
that cytoplasmic components do not perpetuate themselves
autogenically maintaining their characteristic identity.

A component's power of growth and division, therefore, within
the protoplasm of another species is primary evidence for its
autogenic origin ; its power of growth and division in the proto-
plasm of its own species constitutes a second grade of evidence
indicating such an origin. Were there no genetic evidence from
hybrids as to the self-perpetuating nature of chromosomes and
certain plastids and were their behavior studied only in the
cells in which they normally occur, the fact that they produce

25

368 HENRY J. FRY.

new units by the growth and division of similar preexisting ones
indicates an autogenic origin. When centrioles of secondary
spermatocytes are produced by the growth and division of
centrioles of primary spermatocytes, this is evidence that at
least they localize the synthetic processes forming new centriole
substance and therefore are autogenic to that extent. In so far
as there is evidence for the formation of chondriosomes, Golgi
bodies and vacuoles from similar preexisting units, to that
extent is there probability of their autogenic origin.3

Wilson ('oi) describes the division of central bodies in dividing
cytasters of artificially activated eggs of Toxopneustes, noting
the significance of this with reference to the self-perpetuating
nature of such central bodies. Fry ('256) studied the behavior
of cytasters in nucleated and enucleated Echinarachnius eggs in
both living and fixed material, and found division only in those
instances when they secondarily established connection with
chromatin. Tharaldsen ('26) found the same situation in cy-
tasters of Asterias. Although a cytaster may divide, this has no
significance concerning the origin of its central body unless
it is conclusively proved: (i) that the central body is an
individualized structure independent of the rays, and (2) that
its division initiates, and is not just a result of, the division of
the surrounding cytaster. Since both of these points are un-
proved, and since the present study brings strong evidence
against the former, the facts concerning the division of central
bodies in dividing cytasters are in too uncertain a state to use
them as evidence of an autogenic origin.

The foregoing types of evidence concerning a component's self-
perpetuating nature, based upon its power of division, whether
within the protoplasm of another species or within its own,
assumes that it is microscopically visible throughout its history.
In those components where new units do not visibly arise by
division, but apparently are formed de novo, evidence concerning
their origin becomes more difficult to secure. As previously
noted, they actually may be self-perpetuating bodies that are
submicroscopic during part of their history, becoming visible
only at certain times. "Manifestly it is quite illogical to affirm
an origin de novo of any formed body because it first becomes visi-

CENTRAL BODIES IN CYTASTERS. 369

ble at a particular enlargement, even the greatest at our present
command" (Wilson, '23, p. 24). Data can be obtained, however
which may strongly support either an autogenic origin or a
de novo one. If a component is obviously the product of
another component; if its behavior is invariably correlated with
certain changes in surrounding structures, then the presumption
is very strong that it is produced by them and has a de novo
origin. If, on the other hand, it is obviously not the product of
surrounding structures but rather is their formative and con-
trolling factor; if it passes through a certain definite cycle which
is to some degree independent of modifications in surrounding
structures ; such facts, depending upon the details of the evidence,
may indicate its self-perpetuating nature. Data of this type
were secured in the present study.

EXPERIMENTS.

This investigation deals with the origin and behavior of central
bodies in cytasters of artificially activated eggs of Echinarachnius
parma, to ascertain whether they are de novo structures produced
by the cytasters, or whether they are individualized structures
having an identity independent of the surrounding cytaster
which forms about them, and have further characteristics
indicating a self-perpetuating origin. The experiments are
arranged as follows: (i) keeping the environmental factors of
activation constant, to observe the various effects of different
fixatives upon rays and central bodies; (2) keeping the fixative
constant, to observe the various effects of modifications of
environmental factors such as temperature and osmotic pressure
upon rays and central bodies; (3) keeping both fixative and
environmental factors constant, to observe the effects of various
intervals during the astral history upon rays and central bodies.

The Effects of Various Fixatives upon Rays and Central Bodies

of Cytasters (Cliart /.).

The eggs for this group of experiments, as well as the others
to be reported later, were obtained by removing the oral surface
of the animals and securing eggs from those individuals where
they appeared upon the exposed ovaries. The activation techni-

370

HENRY J. FRY.

m
m

m-

M

FIXAT

NflME

[THOSE CONTAINING,
flCETIC PC ID
HRE MflRKED BY
flN HSTERI4K 0

IVE

FORMULfl

PERCE

NO RflYS

NOCENTRai"BODY

[NTftdL

Vfl&UE RAYS
NOCENTR fit BODY

5 OF

DISTINCT -RRYS
NOT EXTENDING
TO CENTER

NO CEKTRRL:&ODY

"CEtfTRRL flREfl STBINEJ
5RME RS "RRY RRtfl]

1 HI

•***'••*. '•• *~ t*/ f' ' -

n

"::->_••:'; :; ':,; -"'

^#14^:^:

i

* FLEMHINa- STRONG

1% Ch.aro*. Atid ^^<>
1% Oimtt acid *fcts
flc«.i«. »c^ 1 V*

i

FLEMMING- STRONG
WITHOUT flCETIC

It, CKrom^. »ClA 15 ^
Z% ObVntt *tvl A t*^

10

3

* FLEMMING-WEflK

St.o^ TWmm«^ ! |<
W»U. 1 f-t

» 2L

n

4

FLEMMING-WERK
WITHOUT flCETIC

sl""».I,'S"5i'* "'""' in

WAt*v J f*

75

5

OSMIC RCID

.015% Osm.t dC»A

IOO

Id

OTROSWESUBLmflTE
COLD - HLKflLINE

SaU.itoA stl.t.on - HlCk -
Cold- M*t «IKJ>-» W.H. N.OH

1 OO

1

COUR05IVE5UBUMRTE
COLD - flClB

S»tu'4l«A boWttO" H^Ck-
CoU - M^ •«••* "•"*'«"

50

50

&

CORROSIVE SUBLIMRTE
HOT - flCID

SaK.iWi Sal-*-<"\ H^CIi
Mt^^A^^^Ot Hi^aClJ Y*4tt*n

100

9

*SUBLIMRTERCETIC11

S«t Solvt.o- H^CIl - 99f<»
Boeho "'i 1 P«

15

Z.5

i

10

* SUBLIMRTE flCETIC 52

Sit StU.t-m H<)Cll- 95 >«»
R<.C*,<L »<..* 5 p»

3&

32>

II

* SUBLIMflTE flCETIC 102

S>t S.W.on H^Ck 90(>>».
Ht^t.c »oi 1° *•>*

15

50

i

12

* POO-flCETC- STRONG

Sit iol P.t.it «oli 1 P"
W.te. 1 f<

«!.<-». t »t'A 5 %,

30

25

13

#PICRO-RCETIC-WEAK

p,t,0-B^t,t-S'«-V It'
W.U, 1 'l

10

1 5

14

* BOUIN

i.t „, P,«,t«,l ^5^

Fo-".o\ L5 ^l

«tlt.t «'i 5P"

\ 0

IO

15

FOUMOL 5%
RLKflLINE

5% To,™.l ~«. ai«.l.-t
w.n MjCOi

IOO

lie.

FORMOL 5&
ACID

5%To,o,.\ ».^».a.«.»v».<>-

14

6b

n

RLCOHOL 11$ 2

100 X, Rll*'*

15

T5

18

flLCOHOl 353

35 Vo Rlcoh6\

&5

15

19

* ZENKER

t'S"""*1"'*™' 95 I"

Mc.ti-.rtt thlo.tlt. 5 qnv

Rtct-t »c-A E. ct

60

20

ZENKEK
WITHOUT RCETIC

5j"lt. »^ »bovt v/-t^0v1

Rce^-c. at i 4

IOO

l\

CAKNOY «* LEBRUN

I oo To RltoVio* 1 ft
Bc£.t->t a«-iA i t-t
Chlorojwm ipt

5

2O

22

CRRNO?"™1 LEBRUN
WITHOUT RCETI

Sa-mt »s. abo.t v-.th«u\
HctV»c. *t»i

Z

45

15

23

*PETRUNKEWITSCH

W.U, -}00 >*b
IOO% HU.ohot ZOO !*!•

c. . 1.. * 1 90 th
N.*"t. at'* i t ^H,

Z2

24

PETRUNKEWITSCH

WTTHOUTflCrtC NNII NITRIC

«nA N.*»it *t»4^.

94

Z

CHART I. THE EFFECTS OF VARIOUS FIXATIVES

Activation: same in all experiments; a slight over exposure of the optimum
butyric-hypertonic technique, cf. footnote 4, p. 372. Fixatives: as listed. Times
of fixation: at frequent intervals after activation, cf. footnote 5, p. 372. Per-
centages of types of cytaslers: averages of detailed counts made at the intervals at
which eggs were fixed, from twenty-five minutes until one hour after activation;
each count based on about 100 cytasters chosen at random from a number of eggs;
examples of such data for Fixatives 3, 10 and 13, shown in Chart III, p. 380.

CENTRAL BODIES IN CYTASTERS.

371

TYPES (

DISTINCT -RRY5
NOT EXTENDING.
TO CENTER

0 CENTRflL BODY
CtKTRBL flREfl STKINO)
C.KTCRTHHN RKV MU_

)F CYTf

IISTINCT "RRYS
EXTENIINa TO
CENTER

NO CENTRRL BODY

ISTEfiS

a \5TINCT RRY5
EJITENDINC, TO
CENTER

CENTRflL BODY
VBGUELY BEMRRKO)
G>RRNULHR

FINE CEI

DISTINCT "RRYS
EJfTENaiNC, TO
CENTER

CENTRRL T>OBY
CLERGY DEMRRKCn
GTtBNULW

MTRAL E

BISTINCT RftYS
EXTENDINC, TO
CENTE"R

CENTRflL BOBY
CLERRLY BENflRKQ
CORPUSCULBR

ODIES

DISTINCT "RHYS
EXTENHWG* TQ
CENTER

CENTRRLBOIY
O.ERRIY BEMHRKED
•RETICULHR

'• ' -Sp

- -irJ!

•-•'v-W

' " ', \^V"?; ' - • " .-

nf^v^s

KSP^

/ -nrrt '

HI

a'S

-"-•^jjjjjfr- .

^'^jpli^l

n

54-

4U

9
u

30

»3

1 1

4 »

1 1

t

5

z.

*

a 5

5

lo

1

&

«9

3O

f\

\

*10

n

1 1

»u

3 1

1

1

*u

10

2 3

z.

113

9

2.0

u

Z9

1

ilf

15

51=

1

U

1

1

|C

ll»

n

Ifc

H9

ZO

20

*Z

1 5

Zl

2.

3U

ll*

5 5

i 3

I

4

UPON RAYS AND CENTRAL BODIES OF CYTASTERS.

Illustrations: show only the central regions, cf. Fig. 2, p. 390. Result: when
eggs containing similar cytasters produced by similar activation are fixed by
various fixatives, central bodies are present only when the fixative coagulates
the cytaster in such a manner that rays are distinct and reach the center. For
further details cf. p. 369. Secondary result: with few exceptions, only fixatives
containing glacial acetic acid (indicated by an asterisk) clearly fix rays. For
further details cf. p. 375.

372 HENRY J. FRY.

que used was that developed by Just (' 19) . An egg sample of each
individual was given a preliminary test of their capacity to form
clear membranes by treating them.with butyric acid solution (2 cc.
i/io TV butyric acid plus 50 cc. sea water). Only those eggs
were used that gave over ninety per cent, membranes.

An artificial activation for Echinarachnius that is optimum for
producing the maximum number of normal embryos yields only
about ten per cent, eggs with cytasters (Fry, '250). The
activation used in the present experiments, therefore, was a
slight over exposure so as to produce about twenty-five per cent,
eggs with cytasters to facilitate their study.4 The eggs show an
average of seven per cent, cytolysis, although there is con-
siderable variation as to this in the different fixatives.

Cytasters of Echinarachnius arise about fifteen minutes after
removal from the activating solutions. None divide except
occasional ones that secondarily establish contact with chro matin,
and if they are numerous, cleavage does not usually occur
(Fry, '256). Thus the majority of cytasters formed soon after
activation remain unchanged for hours except as to modifications
of their detailed structure which are studied by fixing samples of
eggs at various intervals.

Eggs were fixed in twenty-four fixatives being placed in each
at frequent intervals after activation.5 They were sectioned
5 ju thick and stained in Heidenhain's hsemotoxylin. Various
degrees of destaining, carefully studied with water immersion

4 Experiments r to 4 inc. and 7 to 14 inc., Chart I, were activated with a 35
second treatment with butyric acid solution (2 cc. i/io N butyric acid plus 50 cc.
sea water), followed by a 20 minute interval in sea water, followed by an 18 minute
treatment with hypertonic sea water (5 cc. 2.5 N sodium chloride solution plus
50 cc. sea water) at 24° C. Experiments 5 to 6 inc. and 15 to 24 inc. were similarly
activated with butyric acid solution, followed by a similar interval in sea water.
In treating them with hypertonic sea water (same concentration), however, the
eggs were divided into four lots and exposed 15, 20, 25 and 30 minutes respectively.
The experiment was so arranged that all lots came out of the hypertonic solution
at the same time, and then they were mixed. The temperature was 22° C. That
the two groups received an essentially similar activation is shown by the fact that
samples of both contained about the same percentages of various types of develop-
ment at twelve hours after activation.

6 Experiments i to 4 inc. and 7 to 14 inc.. Chart I., were fixed at the following
intervals after activation: 15 min., 17.5, 20, 22.5, 25, 27.5, 30, 35, 40, 45, 50, 55,
I hr., i hr.-is min., 1-30, 1-45, 2 hrs., 3, 4, 5. Experiments 5 to 6 inc. and 15 to
24 inc. were fixed at the following intervals: 10 min., 15, 20, 30, 45, i hr., 2, 3.

CENTRAL BODIES IN CYTASTERS. 373

lenses, show no significant differences in results. Stains other
than hsemotoxylin were not used as it has been generally found
that they yield similar results, and haemotoxylin is the accepted
standard.

The percentages of the various types of cytasters and their
central bodies shown for each fixative, Chart I., are averages
of a number of detailed counts, one made for each interval when
eggs were fixed, from twenty-five minutes to one hour after
activation (cf. footnote 5, p. 372). This is the period during
which cytasters are at their maximum development. Each
count for each interval is based upon the study (at 750 X mag-
nification) of about 100 cytasters and their central bodies selected
at random from a number of eggs. Such detailed counts for
Experiments 3, 10 and 13, Chart I., are shown in Chart III.,
p. 380.

Since sections were cut at 5 ju, and since the majority of
cytasters are 10 to 15 ju in diameter, most of them are found
in two or more serial sections. In the case of each cytaster
only that section was studied that includes its mid-region
(Fig. I, b, p. 384), ignoring the others (Fig. I, a and c). The
drawings show only the central portion of such mid-sections.
The peripheral area is omitted to save space on the charts since
it yields no significant information concerning rays and central
bodies not shown by the central area. An entire mid-section
of a cytaster is illustrated, Fig. 2, p. 390; the dotted line makes
clear how small a portion of the central region is shown in the
chart illustrations. Although the drawings are necessarily dia-
grammatic they are an accurate representation of the various
classes of cytasters and their central bodies. They have the

following classification (cf. Chart I.):

Chart
Rays. Central Body. Symbol.

None None A

Vague None B

Distinct, not extending to center:
Central area stained same as

ray area None Ci

Central area stained lighter

than ray area None

Distinct, extending to center None

Distinct, extending to center Vaguely demarked — granular Ei

Distinct, extending to center Clearly demarked — granular

Distinct, extending to center Clearly demarked — corpuscular £3

Distinct, extending to center Clearly demarked — reticular £4

374 HENRY J. FRY.

When cytasters that are in a similar condition due to similar
activation, are fixed with a variety of fixatives, central bodies are
present only when the fixative coagulates the cytaster in such a
manner that the rays are distinct and extend to the center. The
fact that rays are distinct and reach the center does not guarantee
the presence of central bodies, since many cytasters with such
rays are without them (Type D). In the majority of such
cases the rays are more delicate than in those cytasters having
central bodies, where are usually found rays of maximum coarse-
ness and clarity (Types Ei, £2, £3 and £4). Central bodies are
never present when rays are absent (Type A), or vague (Type B),
or clear without reaching the center (Types Ci and C.2). Differ-
ences in the size of cytasters, which may vary from 5 to 25 (JL,
do not affect this relationship between the presence of central
bodies and distinct central rays, nor does the location in the
egg which may be central or peripheral.

Within any one egg there are found only related types such as,
A and B, or B and D, etc. Unrelated types do not occur to-
gether. There is a similar occurrence of related types among
the different eggs of the same slide, fixed at a given interval
after activation. Since the percentages of the different types
shown for each fixative of Chart I. are averages of a number of
detailed counts made at various intervals after activation, the
variety of types found in the general average is usually greater
than those found at any one interval.

Type A, without any rays, is easily identified as a cytaster
since it is a granular area that stains darker than the surrounding
cytoplasm, of the same size as a typical cytaster, and is frequently
associated with those having vague rays. This rayless type is
often found during the late history of asters when the rays
frequently fade out completely.

In this series of experiments, concerning the effects of fixatives
(Chart I.), the twenty-four lots of eggs which were placed in
the twenty-four killing agents, at various intervals, had received
a similar activation. Hence the living eggs, prior to fixation,
contained about the same array of living cytasters. Therefore
the wide diversity in cytaster structure found on the slides is
the result of differing coagulation products of various fixatives

CENTRAL BODIES IN CYTASTERS. 375

which differently coagulate the same phenomena. It will be
noted, for example, that Type C, with clear rays not extending
to the center, is entirely absent in certain fixatives but present
in others in considerable numbers. Again, some of the reagents
show large percentages of Types D and E having clear central
rays, whereas others show only Types A and B, which are either
without rays or have only vague ones. The significance of the
relation between coarse clearly-fixed central rays and the presence
of central bodies will be discussed later (p. 383).

In those fixatives containing glacial acetic acid (marked on
Chart I. with an asterisk) there is a total of sixty-nine per cent,
cy tasters with distinct rays (Types Ci to £4), and thirty-one
per cent, with rays absent or vague (Types A and B). On the
other hand, in those without glacial acetic acid there is a total
of ten per cent, with distinct rays and ninety per cent, with
no rays or vague ones. The possible significance of acetic
acid with reference to the chemical composition of astral rays
will be discussed in a later paper. Experiments have been
carried out concerning the effects of acetic acid, related acids,
and acetates, at various hydrogen ion concentrations, as well as
similar studies of other fixing agents.

The Effects of Modifications of Environmental Factors upon Rays
and Central Bodies of Cytasters (Chart II.}.

In the preceding experiments the activation was constant and
the fixation was varied; in the present series the fixative is
constant and the environmental factors of activation are varied.
Chart II. presents the various modifications of activation used,
i.e., treatment with both weak and strong hypertonic sea water,
each for forty, sixty and eighty minutes, to study the effects of
differences in osmotic pressure; treatment with butyric acid
solution followed by exposure to hypertonic sea water, for ten,
twenty, and thirty minutes, to study the effects of modifications
of the butyric-hypertonic activation technique; modifications of
temperature, at two degree intervals between 20° C. and 28° C.,
the range within which cytasters occur when using an optimum
activation, to study temperature effects; treatment with one
per cent, and five per cent, ether, each for three, fifteen and

376

HENRY J. FRY.

Ifl

WEflK IB
HYPERTONIC -

fOTUflCATION'-. -TT Q

OF lln

STRONG EB
HYPERTONIC

m«™eNT 1C

EXPERIHENT
NAHE

flND

NUMBER

MODIFICATIONS
OF

TREHTHEKT

1C

MODlFlCRTiONS

or

BUTYRIC IB

HYPERTONIC -

TRERTnENT JHC

ED

IREflTMENT

1C

SIR

5 X ETHER'

TRERTnCNT

1C

ET

fERCEHTKl
OF
ETHER

ADDED TO

SEft
V/HTER

F\C

HER

LENGTH
OF

TREflTMEKI
WITH
ETHER
flNB
SER
WRTER
MIXTJKE

riVATIOt

HUTYRC flCID

LENGTH OF
TREATMENT WITH

ta.lM BUTYSIC

flCID PLUS
Occ SEfl WITTER
FOLLOWED BY
20™™ INTERVftL
N SEfl UftTEH
EFORE HYPtR-
OWC TRtSTHEHT]

^ Dnn

HfftlTDNIC £

NUMBER

or cc OF

1 5 N NaCI

flBIED TO
50cc SEH
UflTER WHO)
MR MUG
HYPERTONC
5Eft
WflTER

rn

Efl WffTER

LENGTH
OF
TRERTMENT
WITH
HYPERTONt
SER
WRTER

TEMP-
RHIIIK
C.

No Et

T» «-*t m

er

No ButjriC
Trtat -man^

So*.

40™.n

n

"

"

1'

H

i, 0 •"»"

••

«

"

K

ti

&0 --"•"

No Et

T»e»tm

•Y

No Tjutiji'ic.

TtettmMt

IZc-c

40 ™n.

n

ti

••

it

II

UO «n

••

H

••

DO TT\1»%

••

No Et

T^«.«tr

\er

«nt

35 sec.

See

10 ™-

n

••

••

••

••

ZOTOn

••

"

"

»

••

30-m.T>

No Et
TVtitn-

««•

e«V

2)5 sec

5c.c

IT -mm

zo

••

• •

•i

-•

»

22

••

•-

»

••

II

Z4

"

• •

•-

••

II

2L,

••

»

••

'•

Z6

1%

2) iriin-

33 sec

5c-t

30 -mi.

18

••

1 5 Tn»n

••

"

"

••

-

35 Tnfn-

II

••

••

«

5%

"i Tniv».

53 Sec.

5c.c

30 -m.0

16

••

15 >r\«n

"

••

••

35 "mm-

-.

••

"

• •

mm
flQE

OF

CYT-

flSTER
ECCS

1 H.. flFJT!

NCTIVRTIOK

»

flC£

or

CYTOL-
1ZED
EGCS

K. flFTER
RCTlVfVTIOtl

5

0

14

5

It

10

1

22

ze>

55

35

52

12

0

ZO

0

1*0

O

0

0

1

0

45

0

.1

33

0

S5

3

0

5

10

la

14-

10

30

.1

4-0

0

&2>

CHART II. THE EFFECTS OF MODIFICATIONS OF ENVIRONMEN-

Activation: modified as listed. Fixative: all experiments fixed with saturated
corrosive sublimate plus 2.5 per cent, glacial acetic acid. Times of fixation: as
listed. Percentages of types of cytasters: each count for each experiment, at each
of the three intervals of fixation, based on about 100 cytasters chosen at random
from a number of eggs.

CENTRAL BODIES IN CYTASTERS.

377

FEKC

1/2 H

HO RRYb

NO CCNT.gCTI

ENTRG

OUR RFTi

VflGUtWVS
NO CEMT KOnv

ES 0

:R flciiv

•StlNCT SIIW
10 CENTER
HCXEMT BODY

rTYP£

[men
flTION

«ftT\«TR«t6

TO cc«rts

UHRMULftR

CCHT. BODY

5 OF

WITH S

1 HOI

NO RAYS

oo-Hiaran

CYTI

uBLiMflTE

)R flFTE

VftGUE *«>S

MCCNT.KODT

pn\

^STEI^

RCETIC

R flCTIVf

Kii*a mis

TO CENTER
NO CENT. Bo&t

'S M

Z.5X)
II ON

DISTINCT RRVS

TO CtMTtR
&RRNULHR
CEW. BODY

0 GEN

Z HOI

NORITii
NO UNT. He IT.

TKRL

ms RFT

HI..I t sn>-

ICCEKTJOIr

BODIE
:R flcii

nsTlNCTHflft
TO CtNTtS
HO CENT.JUIX

:s

/flTION

DISTIKTKDe
TO CENTER

VHM I UK
UNT BOIY

" &•-&

';:£,'.•*.-.
ft"

" J* - "- -

jm

oW*1

#

ms

1

',;'>', - '•

c: .•:•".*

;, > ;;

'. i .''

/Bi'-.i

_^

fi

90

10

10

L,5

Z5

IZ.

11

IU

ao

zo

13

14

13

[N,

> tx(.e

iriment J

31

4-fc

23

50

Z3

Zl

19

50

19

12

[N

o E*|>«

imctA J

UO

31

3

2.5

fc9

t

[N

0 l.f*

imertl j

19

UO

Zl

90

IO

[N

0 L»f-e

iment )

1

13

ZO

at,

14

N

> CvjU

U»b

IZ

3&

50

Z4

UU

!O

N

0 Cijt*'

tt*s

n

39

44

[N

, E.K

vime^t J

1

5b

43

a

11

Zl

[N<

E»^-

r,m«Nt ]

N

0 Cjtd

bttVS,

N

=• cata

3t«.«S

N

. Cyta.

SttfS

34

kU

15

Z5

50

50

45

55

45

Z5

30

61

13

40

UO

11

16

60

ZO

N

0 Cyti

Ste»s

N

> C.jtd

=.tt«s

N

0 Cijti

ste»s

n

l»3

15

5

ZO

LO

ZO

14

ZU

to

95

5

8

n

15

UO

ra

Z2

ii

33

H

45

ZT

Zl

52

Z4

4U

30

10

>5

15

to

3O

IO

15

45

40

zo

UO

ZO

50

zo

30

Z9

Z9

40

Z

N

> Cijtj

&t«»s

N

5 C.jti

^t«.vs

N

c.jU

,tevs

TAL FACTORS UPON RAYS AND CENTRAL BODIES OF CYTASTERS.

Illustrations: show only the central regions, cf. Fig. 2, p. 390. Result: when
a fixative is used capable of clearly fixing rays, central bodies are present only
when the modifications of the environmental factors of activation produce well-
formed cytasters with distinct rays reaching the center. For further details
cf. p. 375-

378 HENRY J. FRY.

thirty-five minutes, preceding a butyric-hypertonic activation,
to study the effect of an anaesthetic. The chart has a column
showing the percentage of eggs in each experiment containing
cy tasters at one hour after activation. Another column shows
the percentage of cytolysis. There are wide variations in both.

The eggs of all these experiments were fixed in sublimate
acetic 2.5 per cent., i.e., saturated solution of corrosive sublimate
97.5 per cent, plus glacial acetic acid 2.5 per cent. They were
fixed at a half hour, at one hour, and at two hours after activation.
Sublimate acetic fixatives (Chart I., Fixatives 9, 10, and n)
produce four types of cytasters, with few exceptions: (i) with
no rays and no central body (Type A) ; (2) with vague rays and
no central body (Type B) ; (3) with distinct central rays and no
central body (Type D) ; (4) with distinct central rays and a
vaguely demarked granular central body (Type El). The
greater variety of central body structure seen in such fixatives
as weak Flemming or Bouin (Chart I., Fixatives 3 and 14) are
not found in fixation with sublimate acetic.

Eggs were sectioned and stained as described above (p. 372).
The percentages of cytaster types shown for each experiment,
at each of the three intervals after activation when eggs were
fixed, are each based on a study of about 100 cytasters and their
central bodies selected at random from a number of eggs. In
certain experiments, however, where cytasters are found in only
one per cent, of the eggs, or less (Chart II., Experiments IV. B,
IV. D and VI. B), the number of cytasters studied was about
twenty-five.

In this group of experiments (Chart II.) the living eggs of the
various ones contained widely differing percentages of various
types of living cytasters, due to the variety of astral phenomena
produced by the wide modifications of activation. These differ-
ences are visible in living cytasters by the use of a high power
water immersion objective. Poorly formed cytasters appear as
homogeneous structureless vesicles. Although well-formed ones
do not show the clean-cut thread-like rays seen in fixed material,
they clearly do show a radiate configuration. These eggs with
their differing cytasters were all placed in the same fixative
(sublimate acetic 2.5 per cent.), in contrast to the preceding

CENTRAL BODIES IN CYTASTERS. 379

experiments (Chart I.) where eggs with similar cy tasters were
placed in different fixatives. The results of both groups of
experiments confirm each other. Using a sublimate acetic 2.5
per cent, fixation, central bodies are present only when the various
modifications of activation produce well-formed cytasters with distinct
rays extending to the center. This relationship holds true whether
eggs having cytasters number o.i per cent. (Chart II., Experi-
ments IV. D, VI. B) or 60 per cent. (III. C) ; whether the
cytasters arise early (III. C, etc.) or late (III. A, III. B) ; whether
there is no cytolysis (IV. A, etc.) or 33 per cent, cytolysis (IV. D,
etc.) ; whether the cytasters are small (5 //) and numerous, or
large (25 n) and few; whether they are located in the egg centrally
or peripherally.

The meaning of the differences in percentages of types of
cytasters at one half hour, one hour, and two hours after activa-
tion, will be discussed in the next section.

•

The Effects of Various Intervals in the History of Cytasters upon
Rays and Central Bodies (Chart III.}.

In the first group of experiments environmental factors of
activation are constant and the fixative is the variable; in the
second group the environmental factors are the variable and the
fixative is constant. In the present group both fixative and
environmental factors are constant and the variable is different
intervals of time during the history of the cytaster.

Chart III. presents the percentages of classes of cytasters
found at frequent intervals from fifteen minutes to two hours
after activation, in the case of three typical fixatives where
central bodies occur, i.e., weak Flemming, sublimate acetic 5
per cent., and weak picro-acetic. The counts for each interval
are based on the study of about 100 cytasters and their central
bodies chosen at random from a number of eggs. This is the
same material from which the data for Chart I. were obtained.
The sectioning and staining technique were previously described

(P- 372).

It will be observed that rays first appear at about fifteen
minutes after activation, remaining vague for some minutes.
Clear rays are not usually found until twenty minutes after

380

HENRY J. FRY.

FIXATIVE

AFTER
flCTIVflTlON

WHEN
COUNTS
WERER/IDE

OFCYTflSTERS
flND
CENTKfllBGMS

PERCEN

NO RRYS
NOCtKTJIffi

TOS

VfliiUE Wf>
NOCENTJIll

OF TYPE5 Of

DISTINCT Riys asnd sws

NOT EXTMLNOT OTEKH
TO CENTER TO CEXTEJ
IWCEIrtgW NOCEVTUt
amtH. »»M CiNTIft. WtR

A3 ROT AREA ! ThHN RRt Mitfl

aw

ibTiNmms
EXTENDINC,
TO CENTER

NOCEUTBW

Si

"ER5/1NB CENTAL BODIES

abTlNCTB«bBbTII<r\'RB6 KlUCTIMfi'KflKT'KjnS

EUENBINCv lUtkiilt EJfTENDINt EXTEN1HNC,
TO CENTER TOCENTEK TO ttNTER TOCENTFR
CUTM.BII CtliraLBTI lEVreSLlDl aNTRflLBW
VflOUELf 1 CLtHRW CLtWU CLcJRLf
DEMftRKEfl 'BEMflRKU DEHflRKra 3EMRRl\a
URftltULKl C^RNUtE CORP«SQ)ll« HETIOIUR

^

. 'i'^ yV ,

y

1

^%

§

FLLM-
MINC

WERK

I 5 mi ** •

IOO

IIS

IOO

10

IS

55

IO

Hi

10

80

15

50

50

215

IOO

I

30

45

IO

35

IO

1

35

10

1 S

35

30

IO

40

1ft

Zft

29

IO

5

45

a

50

n

ft

n

5O

40

30

10

10

55

1 1

41

40

1

1 h .

5

15

35

15

1 O

1 k>. -I5»»>

50

50

1 - -30 •'

34

lie

34"

1 1>

I - -45 "

n

1 U

n

50

I !,«.

Ul»

34

5UBL1-
MMTE
RCETIC
5%

1 5 ™n

34

CE

US

10

30

20

5O

JO

21.5

3?i

34

33

15

2.

IO

t ft

10

11.5

33

34

3^

30

33

34

33

35

90

10

40

45

4-5

5

5

45

50

5O

50

IOO

55

IOO

1 It

[*

Ent.eMmt

>.V3

1 Hr. - 1 5 ">•-

to

40

1 •' -30 -

55

45

1 " • 45 •

70

30

2 hr..

50

50

P1CRO-

ACET1C

WERK

1 5 Trvrt

[No

E.f *<'••*•«

t ]

115

IOO

10

CN«

£>^e>in»

t 1

ZZ-5

15

30

45

15

14

44

Ifc

14

Z1.5

5

95

30

10

2.5

5

5

45

35

15

1 5

1 "i

5

4O

i

40

SO

IL,

34

45

IOO

50

1 Z

31

19

36

*

55

40

30

10

10

1 V>

4O

30

10

\0

1 *r. - 1 5 m.»

40

ID

10

10

1 - -30 -

1 3

15

U

50

u

1 " • 45 "

23

1 0

35

11

1 7.

4

4

il—

50

SO

CHART III. THE EFFECTS OF VARIOUS INTERVALS IN THE HIS-
TORY OF CYTASTERS UPON RAYS AND CENTRAL BODIES.

Activation: same in all experiments; a slight over exposure of the optimum
butyric-hypertonic technique, cf . footnote 4, p. 372; material same as that reported
on Chart I, pp. 370-371. Fixatives: as listed. Times of fixation: as listed.
Percentages of types of cy tasters: each count for each interval of fixation, based on

CENTRAL BODIES IN CYTASTERS. 381

activation. Central bodies do not appear until rays have
become clear, and they are never present unless the rays are
distinct and reach the center. An hour or two after activation
there is a tendency for the asters to become vague. When the
rays become indistinct the central bodies disappear. This con-
firms the foregoing conclusion. Central bodies are present only
when distinct rays extend to the center of the cytaster. Cytasters
pass through a formative period of vague rays when central bodies
never occur; only after the rays have become clear and reach the
center do central bodies appear; if the rays fade in the later history
of the aster the central bodies disappear. Thus clearly correlated
with the appearance and disappearance of distinct central rays
is the appearance and disappearance of central bodies. This
relationship is somewhat modified by the effects of various
fixatives (Chart III.) as to the percentages of types, and as to
the exact time schedule, but the story is essentially the same
in all.

The three fixatives reported in Chart III. are capable of
showing distinctly fixed rays. The vague rays of the early and
later periods of the astral history are not due to poor fixation.
They are caused by the actual condition of the living rays at
the time of fixation, since the same fixatives show distinct rays
at the mid-period of the astral cycle. The central body phe-
nomena shown in these three fixatives are the same as all those
fixatives of Chart I. where rays are clearly fixed. Where rays
are not clearly fixed there are no central bodies.

Similar results are shown in the experiments where activation
is modified, Chart II. The eggs were fixed with sublimate
acetic 2.5 per cent., at one half hour, at one hour, and at two
hours after activation. It will be noted that in most of the
experiments the highest percentage of vague (or absent) rays
usually occurs at a half hour and at two hours after activation,

about 100 cytasters chosen at random from a number of eggs. Illustrations:
show only the central regions, cf. Fig. 2, p. 390. Result: when an activation is
used capable of producing well-formed cytasters, and a fixative is used capable of
clearly fixing rays, if cytasters are studied at close intervals after activation, it
is found that rays are at first vague and there are no central bodies; they appear
only after rays have become clear and reach the center; they disappear when rays
fade in the later history. For further details cf. p. 379.

382

HENRY J. FRY.

'
, -. ..

• « «•

.' . -• r. . •'

/

whereas the peak of clear rays is usually at one hour. Since
central bodies occur only when rays are clear and reach the
center, they too show their optimum numbers at one hour, and
; are less numerous at one half and at two hours after activation.
These relationships are shown in Graph I. The percentages of

.60.

SO.

40.

.30.

20.

10.

LO.

A7/
-&-

\

GRAPH i. The effects of various intervals in the history of cytasters
upon rays and central bodies.

This graph is derived from the data of Chart II., pp. 376-377. It shows the
percentages of the various types of cytasters present at \ hr., i hr., and 2 hrs.,
after activation, when all the detailed experiments concerning the effects of modified
activation are averaged together. A triangle is the symbol for \ hr. after activation,
a circle for i hr., and a square for 2 hrs. Only cytaster types A, B, D and Ei
occur, since the fixative used was sublimate acetic 2.5 per cent. The graph shows
that asters with vague rays (or none) usually occur in maximum numbers at \ hr.
and at 2 hrs. after activation, whereas clear-rayed ones are most numerous at i hr.
Also, the percentages of asters with central bodies follow the same sequence,
indicating that the presence of central bodies is correlated with the presence of
clear rays. For further details cf. p. 381.

CENTRAL BODIES IN CYTASTERS.

383

each of the four types of cytasters (Types A, B, D and Ei), at
one half hour, one hour and two hours after activation, are in
each case the total of all the detailed percentages of the various
experiments for that type for that interval, shown on Chart II.
A few of the individual experiments do not coincide with the
average behavior, but such discrepancies are accounted for by
the considerable differences in astral phenomena produced by
the wide variations of activation. Despite wide modifications
of osmotic pressure, temperature, butyric-hypertonic treatment,
and ether treatment, producing different percentages of cytaster
types, different percentages of eggs containing cytasters, different
percentages of cytolyzed eggs, and different sizes and numbers of
cytasters, the relationship between clear central rays and the
presence of central bodies holds good.

DISCUSSION.

The Origin of Central Bodies in Cytasters of Echinarachnius parma.

The previously described experiments prove that when eggs
have been similarly activated and therefore contain similar
types and percentages of cytasters, if they are fixed with a
variety of killing agents, central bodies are present only when
the fixative coagulates the cytaster in such a manner that rays
are distinctly fixed and extend to the center (Chart I.). When,
on the other hand, a fixative capable of clearly fixing rays is
kept constant, but the various environmental factors used in
activation are modified, the same relationship holds good, and
central bodies are present only when those environmental factors
produce well-formed cytasters with clear central rays (Chart II.).
Finally, when an optimum activation is used capable of producing
well-formed cytasters, and they are fixed by a killing agent
capable of fixing rays distinctly, if the cytasters are studied at
close intervals after activation, it is found that they pass through
an early vague-rayed condition when central bodies are never
present; only after the rays become clear and reach the center
do central bodies appear; and if the rays later fade, central
bodies disappear. This is strong evidence that central bodies
occur only if distinct rays reach the center of the cytaster
(Chart III.).
26

HENRY J. FRY.

Figure I, a, b, and c, shows the appearance of three serial
sections (each 5 M) of a cy taster, Type El, that was about 15 /x
in diameter. Figure I, d, is a schematic reconstruction of the

FIG. i. The structural relations of rays and central bodies shown by a com-
parison of the serial sections of a cytaster with the reconstructed aster.

The serial sections (a, b and c), which were cut at 5 M, are drawn from those of a
cytaster having a vaguely demarked granular central body (Type Ei). A schematic
reconstruction of the entire aster (d) is shown at the left. In the top and bottom
sections (a and c) rays are visible in side view in the peripheral area only, whereas
in the central area they are seen in cross section as dots. In the middle section (b)
many of the rays are visible in side view throughout their entire length, and the
central portion also contains the closely crowded inner ends of the rays that extend
above and below. These converge at the midpoint, and this when coagulated
constitutes a central configuration, /. e., the central body.

entire aster. It is obvious that the top and bottom sections
(Fig. i, a and c) present a side view of rays in only the peripheral
area, and that in their central regions the rays are seen in cross
section as scattered dots. It is equally apparent that the
middle section (Fig. i, b) presents a side view of many of the
rays throughout their entire length from the edge of the cytaster
to its center, and that the center also contains the inner ends of
those rays extending above and below which appear in cross
section as closely crowded dots.

The following hypothesis is therefore proposed as the probable
explanation of the so-called central bodies seen at the center of
fixed cytasters. The rays converge at the mid-point of a cytaster.
When they are coagulated by fixation their closely crowded
inner ends are sufficient to explain a dark configuration at the
center. Since cytasters differ as to number and coarseness of
rays, there is a variety of coagulation products at the mid-

CENTRAL HODIKS IN CVTASTERS.

385

points; they may be granular, corpuscular or reticular. This
study has conclusively proved that central bodies are de novo structures
produced by the cytaster only after it has formed clear rays reaching
the center. There is strong evidence for the further conclusion that
these structures are produced only by coagulated cy tasters and that
they are nothing but the fixed closely crowded inner ends of well
formed rays, and that they have no existence as individualized
structures in the living cytaster.

This explanation of central bodies in cytasters is supported
by the fact that the type of cytaster with clear rays reaching to
the center which nevertheless is without a central body (Type D),
usually has rays that are more delicate than those of cytasters
with central bodies (Types El, £2, £3 and £4). Also if rays
extend to the center but are vague, there are no central bodies
(Type B). Therefore, only coarse well-formed rays form central
bodies.

That central bodies are nothing but the coagulated bases of
rays finds further support from the fact that not only must the
rays be well formed, but they must also extend to the center.
That type of cytaster with coarse distinct rays that do not reach
the center (Type C) is of special significance in this connection.
Were it not for this type it might be legitimately concluded
that central bodies are individualized structures with a chemical
nature similar to that of the rays; hence when rays are poorly
and vaguely fixed, central bodies fail to be coagulated and
appear to be absent; only when rays are well fixed are central
bodies shown. The existence, therefore, of large numbers of
cytasters with well-fixed distinct rays in the peripheral region
that do not reach the center, disproves this possible interpretation
since such cytasters never contain central bodies. If the assump-
tion is correct that fixatives which fix central bodies are ones
that fix rays, then central bodies should be present in this type
of cytaster with well-fixed peripheral rays.

Almost conclusive evidence supporting the hypothesis that
central bodies of cytasters are nothing but the coagulated inner
ends of well-formed rays is found in the history of cytasters
(Chart III.), where the appearance and disappearance of clear
central rays is invariably accompanied by the appearance and
disappearance of central bodies.

. , -. •

386 HENRY J. FRY.

If, in the light of the above evidence, it is still claimed that
central bodies of cytasters have an autonomous independent
existence beyond the limits of microscopic vision, becoming
visible only at certain times, then it must be assumed further
that they always attain visibility simultaneously with the
formation of distinct rays reaching the center, and again become
submicroscopic whenever the rays fade; that they never attain
visibility if clear rays are formed that fail to extend to the center
or if vague rays arise that do reach the center. Were the central
bodies of cytasters independent units having existence apart
from the rays, their presence would not be so invariably de-
pendent upon clear rays that reach the center. The wide
fluctuations in structure and behavior of both central bodies
and the surrounding rays caused by various fixatives, by modi-
fications of environmental factors, and by different periods in
the astral history, would not show the invariable correlation
that they do. There would be at least some faint indications of
central body behavior independent of clear rays reaching the
center.

Although Morgan ('96, '99, 'oo) and Wilson ('01) did not
reach such a conclusion, their numerous detailed observations of
the behavior of central bodies and cytasters in Sphcer echinus,
Arbacia and Toxopneustes are the same as those of the present
study. They note that cytasters are formed by the accumulation
of a substance that only later assumes a radiate structure
(Morgan, '96, p. 356; '99, p. 465, etc.; Wilson, '01, p. 558);
that central bodies are not present at first and make their
appearance only after the aster has passed through its early
history (Morgan, '99, pp. 470 and 477; Wilson, '01, p. 558);
that 'they are most apt to be present in well-formed cytasters
(Morgan, '99, p. 477) whereas poorly developed ones do not
contain central bodies (Wilson, '01, p. 560); that they are not
present in all cytasters and show a considerable variety of
structure (Morgan, '96, '99; Wilson, '01 ; numerous references) ;
and that cytasters frequently fade out in their later history
(Morgan, '99, p. 464; Wilson, '01, p. 554). As a result of these
facts Morgan ('99, p. 477) tentatively proposed as a possibility,
the conclusion established by the present study: "Whether the

CENTRAL BODIES IX CYTASTERS. 387

centers first form and the fibers (rays) arrange themselves
around them, or whether the centers are the result of the focusing
of the first formed rays at a central point is difficult to determine,
for both centers and rays appear at about the same time. All
things considered I am inclined to adopt the latter alternative."
The plates of Morgan ('96, '99) and Wilson ('01), as well as
those of other investigators illustrating echinoderm cytasters
(Herlant, '19; Tharaldsen, '26, etc.), show a diverse array of
cytasters and central bodies, most of which are similar to those
found in the present study of Echinarachnius . Toxopneustes
(Wilson, '01), however, shows certain differences. Professor
Wilson kindly permitted an examination of some of his slides,
and it is clear that Toxopneustes cytasters contain a high
percentage of a central body type somewhat like £2 of the
present study, but it is more compact and very much smaller.
In some cases it is granular; in others it is homogeneous and is
identical in appearance with a typical period-like centriole
(Wilson, '01, plates XV. and XVI.). Such differences, however,
due to different species, are probably of no more significance
that those caused by different fixatives or differing environmental
conditions. Any variations in central body structure of cy-
tasters, no matter what the cause, are meaningless, provided
the conclusion of the present study is true, that such central
bodies are nothing but various coagulation products of the inner
ends of well-formed central rays, having no existence as indi-
vidualized structures in the living cytaster. All previous investi-
gators of echinoderm cytasters, from Morgan ('96) and Wilson
('oi) to Fry ('256) and Tharaldsen ('26), have assumed that the
central bodies seen in fixed cytasters are structures having an
actual existence in the living egg. The present study of
Echinarachnius proves that in this species, at least, they are but
the coagulated focal point of rays. It is probable that this is
true of central bodies in cytasters generally, but proof of this
awaits the application of a similar method of study to cytasters
of other species.

The Method of Study and its Significance for Cytological Research.

If a cell component is studied in a fixed condition it is of course

necessary to check such observations with a study of living

388 HENRY J. FRY.

material if this is possible. The study of living cells stained with
vital dyes may yield information concerning the structure of a
component when alive, to use as a basis of comparison for its
appearance when coagulated. It is further necessary, if possible,
to secure data concerning its chemical composition. Some
components such as central bodies, are so small, however, that
they cannot be seen in the living cell. Observations of such
structures can be made only after the material has been coagu-
lated and sectioned. When such is the case the greatest caution
must be used in interpreting the results, keeping in mind the
fact that a component when coagulated may present a very
different appearance from the same component when alive.

The present study shows that various fixatives cause such a
diversity of structure and behavior of central bodies, that to
draw conclusions from observations made by the use of one or
two killing agents might easily lead to erroneous results. A
description of their structure based on fixation with Bouin's
fluid would not harmonize with the results produced by sublimate
acetic, and would be equally different from conclusions based on
phenomena seen in material fixed with strong Flemming's fluid
(Chart I.). Not only are the types and percentages of cy tasters
and their central bodies modified by the fixative, but the time
schedule of events during the astral history is also changed.
Clear rays and central bodies disappear at forty minutes after
activation when eggs are fixed with sublimate acetic 5 per cent.,
whereas in weak Flemming's solution and in weak picro-acetic
they are present until about an hour and a half after activation
(Chart III.). Since the eggs were similarly activated, these
time differences in the astral history are caused by the fixatives.
The varying effects of different killing agents are also shown by
the different percentages of cytolysis they cause, the data for
which are not shown on Chart I. In the different sets, cytolysis
varies from none to about fifteen per cent, at forty-five minutes
after activation, and shows a much wider variation at later
periods, depending upon the fixative used. Of all the factors of
the environment capable of modifying the structure of a com-
ponent, certainly no one of them can effect more radical changes
than the killing agent. Hence if a component is studied only

CENTRAL BODIES IN CYTASTERS. 389

in a coagulated condition it is essential to know the modifications
produced not only by one or two fixatives, but by a variety of
typical ones. Only when these are known is there any hope
of arriving at a conception of what is a "normal" condition,
and even then it must be accepted with the greatest caution.

Not only is it important to know the effects of different
fixatives, but it is equally important that in the case of each
fixative, accurate counts be made of all the variations of a com-
ponent produced, omitting none. All the coagulation products
must be considered in arriving at a result and the percentages
of each should be known. Had the method of study in the
present investigation been one of random search, selecting
certain types as "normal" and dismissing others as "artifacts,"
and had all the variations in cytaster and central body structure
not been studied quantitatively, it is doubtful if the real relation-
ship between clear central rays and the presence of the central
body would have been apparent. Had that type with clear
peripheral rays not reaching the center and having no central
body (Type C), and the type with clear central rays without
a central body (Type D), as well as vague rayed asters (Type B),
been passed by as due to poor fixation, and had attention been
centered only on those with central body structure, the real
relationship between clear central rays and central bodies would
probably not have been established. It cannot be said that
any of the types are more or less important than the others,
or that any are "normal" or "abnormal," in arriving at the
conclusion.

That cytaster with clear peripheral rays not reaching the
center, in which the central area is homogeneous and stained the
same as the ray area (Type Ci) occasionally contains a minute
deeply staining granule at its mid-point (Fig. 2). This satisfies
all the requirements of one of the a priori conceptions of a
"normal" central body structure, i.e., one having a period-
like centriole, surrounded by a homogeneous zone, the centro-
some, about which is the ray area. In fact, the present investi-
gation began on this assumption, resulting in confusion due to
the inability to explain the other types which were abundant in
well-formed asters. That the central granule of this cytaster

3QO HENRY J. FRY.

(Fig. 2) is not a centriole, and that this aster is without signifi-
cance for an understanding of central bodies in cytasters is
shown by the following facts, (i) These granules are found only

FIG. 2. Cy taster with distinct rays in the peripheral region only (type Ci)
containing a cytoplasmic granule at its midpoint.

This cytaster strikingly simulates a "typical" central body structure, one
having a period-like centriole, surrounded by a structureless zone called the centro-
some, about which is the ray area. Such a cytaster is without significance, how-
ever, concerning central body phenomena in cytasters. It occurs in but 0.5 per
cent, of all cytasters. The granule is identical in appearance with those scattered
through the cytoplasm. Two or three or more granules occur in such asters as
frequently as does a single one.

The area within the dotted lines indicates that small portion of the midsections
of cytasters shown in the chart illustrations, cf. p. 373.

in those cytasters where the rays do not reach the center (Type
Ci) and only in very few of them. In fact this cytaster with a
centriole-like granule occurs in but 0.5 per cent, of all cytasters.
Its relative scarcity was not appreciated until accurate counts
were made of all the types. Until such a procedure is followed
an observer can center attention upon a certain condition as
typical and largely pass by other forms, and just because atten-
tion is centered on it, it may seem to be far more abundant than
it actually is. Whether it occurs with a rarity that is significant
can only be revealed by comparative counts. (2) The presence
of a centriole-like granule in any one cytaster has no correlation
with its presence or absence in other cytasters of the same egg.
It is not a condition of central body structure difficult to fix,
otherwise it would appear more frequently in the wide variety
of fixatives used, and when present in one cytaster of an egg it

CENTRAL BODIES IN CYTASTERS.

would tend to be present also in others. (3) It is important to
note that such granules occur in pairs and groups of three or
more as frequently as they occur singly. Wilson ('or, p. 558)
observed irregular numbers of centriole-like granules, especially
in the early history of cytasters. He also notes (p. 560) that in
well-formed ones they may be double. (4) Most important of
all is the fact that they are identical in size and appearance with
the darkly-stained granules that are abundant in the cytoplasm
generally. Wilson ('01, p. 557) carefully notes that granules,
sometimes occurring at the centers of cytasters, are similar to
those scattered through the cytoplasm. In view of the preceding
points, it is clear that in Echinarachnius cytasters such granules
that simulate centrioles are actually cytoplasmic granules that
happened to be at the center of an area where an aster formed;
they remain there only if the differentiation of rays does not
extend to the center (Type C) ; they are not found at the mid-
point of cytasters with clear rays reaching the center (Type D)
but these contain only the much larger central bodies of a very
different character previously described and illustrated (Types
Ei to £4). These centriole-like granules are crowded out of
cytasters if the differentiation of clear rays extends to the center,
by the same forces that eliminate all relatively large particles
from between well-differentiated rays in asters generally.

Morgan ('99, p. 475) notes the effects of the several fixatives
he used upon the variety of central body structure seen on the
slides. As a result of this he raised doubts that have been
confirmed by the present work: "How large a part does the
reagent play? How much is artificial and how much actually
represents what is actually present in the living egg? As to
nucleus and chromosomes, all series show the same results.
As to cytasters, and especially as to central bodies, the door is
open to skepticism." The present investigation is a further
warning of the necessity of distrusting the coagulation products of
certain cellular components such as central bodies, chondriosomes,
Golgi bodies, etc.; of realizing that the phenomena seen on slides
may be radically different from the living condition. On the other
hand, it shows with equal clearness that results may be dependable,
if the various typical coagulation products are known, if all the

392 HENRY J. FRY.

tvpes produced are studied quantitatively, and if they are all taken
into consideration in arriving at a conclusion.

No profitable comparisons can be made between central bodies
of cytasters and those of nuclear asters in Echinarachnius until
the latter have been studied in a similar manner. The central
bodies of the nuclear asters on the slides of the present experi-
ments with artificially activated eggs, as well as those of cytasters
that secondarily established connection with chromatin, have
not been studied in detail, awaiting a controlled and quantitative
study of central bodies in the nuclear asters of fertilized eggs.
This is now in progress; a report of it will be presented in another
paper together with a study of central bodies in fertilized eggs of
Arbacia and Asterias.

Central bodies of nuclear asters in general show a great
diversity of structure and behavior. In some forms they are
described as autogenic, in others they appear to arise de novo;
in some they seem to give rise to the aster, in others they appear
only after the aster has been formed; in some they persist
throughout the cell history, in others they disappear at certain
phases of the cell cycle. In the light of the present study, this
diversity of behavior, not to mention the variety of structure
and function in the non-mitotic roles of central bodies, makes it
the more necessary to approach their future study in a controlled
and quantitative manner. It is possible that if the present
method is applied to the study of central bodies generally, that
certain of the previous observations may be found to be based
upon a single type arbitrarily selected as "normal" when using
but one fixative, without knowing its quantitative relationships
with respect to other types present on the same slides, and
without knowing the differences caused by other fixatives.
Whether or not such is the case awaits further study.

RESUME.

i. Some cellular components, such as chromosomes, arise by
the growth and division of preexisting bodies of the same kind,
maintaining genetic continuity as individualized structures from
cell to cell. They, therefore, have a self-perpetuating origin.
In contrast to this, other components, such as secretory droplets,

CENTRAL RODIKS IN CYTASTKRS. 393

have a de novo origin, since the synthetic processes forming
new units occur without reference to preformed bodies of the
same kind.

2. Central bodies (a term used in its broad sense without
reference to various detailed modifications) are of special
interest to this problem of protoplasmic differentiation, since
there is evidence that in some cases their origin is de novo, while
in others it is self-perpetuating.

3. Are central bodies of cytasters in artificially activated eggs
of Echinarachnius parma, de novo structures produced by the
aster, or are they self-perpetuating bodies, that are usually sub-
microscopic, becoming visible only at certain times and giving
rise to the aster?

4. One series of experiments proves that when eggs have been
similarly activated and therefore contain similar types and
percentages of cytasters, if they are fixed with a variety (twenty-
four) of killing agents, central bodies are present only when the
fixative coagulates the cytaster in such a manner that rays are
distinctly fixed and extend to the center.

5. A second series confirms the same result. When a single
fixative capable of clearly fixing rays is used, but there are
modifications of the various environmental factors of activa-
tion, such as temperature, osmotic pressure, etc., central bodies
are present only when those environmental factors produce well-
formed cytasters with clear central rays.

6. A third series also supports the preceding conclusion.
When an optimum activation is used capable of producing well-
formed cytasters, and they are fixed by a killing agent capable
of showing distinct rays, if cytasters are studied at close intervals
after activation, it is found that they pass through an early
vague-rayed condition when central bodies are never present;
only after the rays become clear and central do central bodies
appear; and if the rays later fade, which frequently occurs, the
central bodies disappear.

7. The cytaster, therefore, produces the central body only
after it has formed clear rays reaching the center, no matter
what are the modifications caused by fixatives, or environmental
factors, or intervals of time in the astral history. Thus the

394 HENRY J. FRY.

central body has a de novo origin. It is probable, moreover,
that only a coagulated aster produces a central body and that
it has no existence as an individualized body in the living cytaster.
The mid-point of the aster contains the crowded converging
inner ends of the rays. These when coagulated form the so-
called central bodies seen at the centers of cytasters in fixed
material. This hypothesis is supported by the following facts:
central bodies are most abundant when rays are coarse, and
are usually absent when rays are delicate, and always absent
when rays are vague ; there are many cytasters with coarse rays
in the peripheral portion which do not reach the center, and
these never contain central bodies; clearly correlated with the
appearance and disappearance of clear central rays during the
history of cytasters, is the appearance and disappearance of
central bodies.

8. The investigation is another illustration of the need of a
controlled and quantitative method in cytological study. When
a cellular component is too small to be studied in the living
condition, and must be studied only after coagulation, it is
necessary not only to control the environmental factors, and to
study it at frequent intervals during periods of significant
change, but it is also necessary to know its various coagulation
products formed by different typical fixatives. It is equally
essential to know the exact percentages of all these variations,
without exception, that they may all be taken into consideration
quantitatively when arriving at a conclusion. To arbitrarily
select a certain type as "normal" because it happens to coincide
with a certain a priori idea, and to disregard the other types as
"artifacts" produces uncertain results.

9. No profitable comparison of central body phenomena in
cytasters of Echinarachnius can be made with those of its nuclear
asters until the latter have been studied in a similar controlled
and quantitative manner, a report of which will be presented
in another paper.

10. Glacial acetic acid was present in some of the fixatives
used and absent in others. Of all types of cytasters having
clear rays (whether with or without central bodies), eighty-five
per cent, are found in the presence of acetic acid. The possible

CENTRAL BODIES IN CYTASTERS. 395

significance of acetic acid with reference to the chemical com-
position of astral rays is under further investigation.

LITERATURE CITED.
Boveri, T.

'01 Zellenstudien IV; Uber die Natur der Centrosomen. Jen. Zeitschrift,

Bd. 25.
Fry, H. J.

'2sa Asters in Artificial Parthenogenesis. I. The Origin of the Amphiaster in

Eggs of Echinarachnius parma. Jour. Exp. Zool., Vol. 43.

'25b Asters in Artificial Parthenogenesis. II. Asters in Nucleated and Enucle-
ated Eggs of Echinarachnius parma and the Role of the Chromatin.
Ibid., Vol. 43-
Hindle, E.

'10 A Cytological Study of Artificial Parthenogenesis in Slronglocenlrolus

purpuratus. Arch. f. Ent.-mech., Bd. 31.
Herlant. M.

'19 Comment agit la solution hypertonique dans la parthenogenese experi-
mentale (methode de Loeb). II. Le mechanisme de la segmentation.
Arch. Zool. Exp. et Gen., T. 58.
Just, E. E.

'19 The Fertilization Reaction in Echinarachnius parma. III. The Nature of

the Activation of the Egg by Butyric Acid. Biol. Bull., Vol. 36.
Meves, F.

'02 tiber die Frage ob, die Centrosomei Boveri's als allgemeine und dauernde

Zellorgane aufzufassen sind. Verhand. der Anat. Gesellschaft Halle.
Morgan, T. H.

'96 The Production of Artificial Astrospheres. Arch. f. Ent.-mech., Bd. 3.
'99 The Action of Salt Solutions on the Unfertilized and Fertilized Eggs of

Arbacia, etc. Ibid., Bd. 8.
'oo Further Studies on the Action of Salt Solutions and Other Agents on the

Eggs of Arbacia. Ibid., Bd. 10.

'26 Genetics and the Physiology of Development. Amer. Naturalist, Vol. 60.
Petrunkewitsch, A.

'04 Kiinstliche Parthenogenese. Zool. Jahrb., Supplement 7.
Tharaldsen, C. E.

'26 The Origin and Nature of Cleavage Centers in Echinoderm Eggs. An
Experimental and Cytological Study of Astral Phenomena in Starfish
Eggs. Jour. Exp. Zool., Vol. 44.
Wassilieff, A.

'02 Uber kunstliche Parthenogenesis des Seeigeleies. Biol. Centbl., Bd. 22.
Wilson, E. B.

'01 Experimental Studies in Cytology. I. A Cytological Study of Artificial

Parthenogenesis in Sea-urchin Eggs. Arch. f. Ent.-mech., Bd. 12.
'23 The Physical Basis of Life. Yale University Press.
'25 The Cell in Development and Heredity. The Macmillan Company.
Yatsu, N.

'09 Observations on Ookinesis in Cerebratulus lacleus. Jour. Morph., Vol. 20.

ACTION OF SALTS ON FUNDULUS EGG.

I. THE ACTION OF NA, K, AND CA CHLORIDES UPON
THE EGG OF Fundnlus.

JOSEPH HALL BODINE,

ZOOLOGY LABORATORY, UNIVERSITY OF PENNSYLVANIA, AND THE
BIOLOGICAL LABORATORY, COLD SPRING HARBOR.

Interest in the fundamental action of chemicals upon the egg,
egg membrane, embryo, and larva of the marine fish Fundulus
has centered largely around the work of Loeb and his associates
(1-17). The results of this work seem to show that marked
differences exist in reactions of the eggs and hatched embryos
or larvae to different salts. Marked degrees of salt antagonism
are also pointed out. Inasmuch as the conclusions drawn from
this work are based almost entirely upon the egg and hatched
embryo or larva, it was thought highly desirable to recheck
the observations, using eggs and embryos dissected out of the
eggs, as well as hatched embryos or larvae. By means of an
operative technic devised by Doctor J. Nicholas1 of Yale Uni-
versity, the details of which will be published elsewhere, it has
been found possible to remove the outer egg membrane from
the embryo and yolk sac without, in any apparent way, interfering
with the further development of the embryo. As a matter of
fact, embryos removed from the egg develop in sea water in as
normal a way as those enclosed in the egg. It has thus been
possible to compare the action of various salts directly upon the
egg, the embryo dissected free from the egg membrane, and the
larva hatched from the egg.

That Loeb conceived of a similarity in the action of salts
upon the hatched embryo or larva and the embryo contained
in the egg can easily be gathered from the following passages.
In 1911 he says (6) : " . . . if the fish is out of the shell the addition

1 The writer is deeply indebted for the assistance rendered by Doctor J. Nicholas
who was kind enough to carry out the operative technic involved in removing the
egg membrane.

396

ACTION OF SALTS OX FUNDULUS EGG. 397

of CaCU alone is no longer sufficient and the addition of KC1
also becomes necessary." Other remarks in the same article
seem also to indicate that the marked differences in the reactions
to salts between the larva and embryo were not appreciated.
In 1916 he says (8): "This prolongation of life through the
addition of Ca is due not to an action upon the protoplasm but
to a prevention of the diffusion of the NaCl into the egg, since
if we take the embryo out of the egg (or use the newly hatched
embryo) it is killed in 50 cc. 3 M NaCl + I cc. 10/8 M CaCl2
inside of a few minutes." In his last paper on Fiindulus (17)
he says: "In 1905 the writer suggested as an explanation that a
pure NaCl solution, if its concentration exceeded a certain limit,
made the membrane of the egg more permeable, so that NaCl
could diffuse into the egg, killing the embryo, while this increase
in permeability was prevented by the presence of a low con-
centration of Ca." These and similar passages throughout his
works seem to the writer to show that the marked differences
in the reactions to salts between the embryo dissected out or
freed from the egg membrane and the larva were not fully taken
into consideration.

The present paper embodies results obtained from investi-
gations carried out during the summer of 1927 at the Biological
Laboratory, Cold Spring Harbor, on the effects of Na, K, and Ca
chlorides on the eggs, embryos freed from the egg membrane,
and larvae of Fundulus heteroditus.

METHODS.

The eggs were stripped directly from the female fish into
finger bowls containing sea water and then fertilized. They
were kept at room temperature (2O°-25° C.) as well as at lower
temperatures (i5°-i8° C.). Those at the lower temperatures
naturally took longer to reach a given stage in development than
those kept continuously at the higher temperatures. The eggs
were usually washed for varying periods in distilled water before
use in order to free them of adhering salts as suggested by Loeb
(13-17). This procedure, however, seems to the writer necessary
to but a limited degree, since similar reactions are given by eggs
washed for varying periods except when the time factor is

398 JOSEPH HALL BODINE.

greatly lengthened, i.e., to several days as was the case in many
of Loeb's experiments (13-17).

The egg membrane was cut mid-laterally by means of fine
pointed iridectomy scissors and the contained embryo gently
rolled out of the shell by means of a fine probe. Eggs in as
nearly as possible the same stages of development were used
together with embryos dissected from similar eggs. It is of
considerable importance that eggs and embryos be in the same
stages of development, since marked age differences in reactions
to some salts seem to exist. The length of exposure of the egg
to water is also an important factor in changing the general
consistency of the egg membrane. Both eggs and embryos
were placed in covered Syracuse watch-glasses in 10 cc. of the
solution and were constantly observed under a compound
microscope during the course of the experiments. Five to ten
organisms were used in a single watch-glass. It is of much
importance to carry through experiments until the eggs hatch
or the organisms die since in many cases the eggs live in certain
solutions but upon further development and hatching the
embryos quickly succumb. T*he embryos and eggs used ranged
in developmental stages from those in which heart action was
just beginning to those with fully developed circulation and
ready to hatch. Newly hatched larvae were used for comparison
with the dissected-out embryos and eggs. The end-point ob-
served and recorded in all the experiments herein reported was
the time of cessation of the heart beat. Recovery of the heart
beat in sea water was also noted but will be dealt with in a
subsequent communication. The salts used in these experi-
ments were c.p. NaCl, CaCl2, and KC1 made up in distilled water.
Only the effects of normal solutions of these salts will be presented
at this time since the results at this concentration are typical.

RESULTS OF EXPERIMENTS.

Since in general the results of all experiments are qualitatively
alike only typical experiments will be described.

N KCl. — KC1, as repeatedly pointed out by Loeb and his co-
workers (1-17), acts with considerable rapidity on the egg,
dissected-out embryo, and larva. In all, the heart quickly

ACTION OF SALTS ON FUNDULUS EGG.

399

w

27

o

XXX

*> "> 1?

o

-2 5 5

03

1/3 03 W

CJ 4J

)H 1* )H

QJ CJ CJ

'i> oJ

0 ^

CO

_' — ,—'

cd cd cd

OJ O O

4J

n

^

c c: c;

p

T* — •

nj

cd rt cd

n3

_>< "o

1

o f"> _o

ESS

CJ CJ OJ

SEE

d § ~
•S g E

.- O

31

o tf)

CO tO CO

^- C/J

CJ OJ <U
•*-» -*-> -*J

03 03 c^

U t-. ~

is a s

0> C *-;

rn CJ 03

S at!

ii 8

0 CJ CJ

a co o

T<«

U CJ CJ

o !H S

a

CH (^ Ol

£ < CL,

H -3

u CJ

— OS

^

•d cj

d "^^

*^*1 ^~\ x~"*

S3 >

g 1 1

.2 d .S

03 03

CO %

i

I O O
"T CO i-t

N O IO

W CS M

o 5
i^ "»

oJ

"P CO

I_J

_| 1 L

J I L_

S3 CJ

1) >-"

+ + +

+ + +

" 3

+ +

+

CJ

!3 0

^r--

II

•d

'S' E

i

OJ

IH "^

co £3

CJ QJ
"- C

d' d

43 S ^T

— O

oil

'E'S

i i G

S „

*| ,

^ *-» M

t-O I— I

N O CO

O p.

J3 "" SJ

M 10

^^^?

s s

U

+ + 0

+ + +

3 ..

'-3 C
•-) 0

o3 73

o

O *J

s~\

CO CJ

CO

X * f *

rf --i

d a £

S3 S3

d o3

"E E *O

'S E

.2 S

M

10 oo ^~^ u,

O f*O

O. o3

CJ n

W

^-^^-^ "T ^

^ ^— '

CJ II

r^

_1 1— ^JQ O
G

0 + +

o 4.

CJ ^

O

S +

5 -~

M C

o

co •£

? CJ

-

0 03

n •£ — '

tf5.a

^i C J ^^

1 X

CJ rt W

Is

.

fej^fe;

.S c!

•^ *J

1

U+ + +

^ *-*^

M _

lu — u

^ 03 U o3

t^ ^

M ^

V ^/ \7

^-^ r i I-H r i

->i "

UJ U Z

e r] r. ^

^ +

fe;fe;^

«

+

400 JOSEPH HALL BODINE.

stops beating, the relative order of resistance being, larva < em-
bryo < egg. The egg is about 4 times as resistant as the dis-
sected-out embryo, and about 20 times as resistant as the larva.
In the action of KC1 upon the embryo within the egg, about
three fourths of the time necessary to cause cessation of the
heart beat is spent in the passing through the egg membrane
(Table I.). Eggs with embryos in which the heart is just
beginning to beat appear less resistant then older eggs.

N CaClo. — CaCls is also extremely toxic for eggs, embryos,
and larvae, the relative order of resistance being larva < em-
bryo < egg. The egg is about one half as resistant as the
embryo, and about 3 times as resistant as the larva. Relatively
less time is spent by the Ca in going through the egg membrane
than by K, since the embryo outside of the egg is killed in less
than one half of the total time required for cessation of the heart
beat of the embryo within the egg membrane. No marked
age differences in resistance to CaCl2 seem to exist, since the
organisms are killed at all embryonic stages in almost the same
relative time.

TV NaCl. — The action of NaCl is by far the most interesting
of those studied, since the results obtained are quite different
from those reported by Loeb (17). The relative resistance to
NaCl is, larva < embryo < egg. This series, however, must be
modified because with NaCl the question of age enters in a
most amazing way. Freshly fertilized eggs, as pointed out by
Loeb (2), are killed in solutions of NaCl. As the egg grows
older its resistance to NaCl increases up to the time of hatching.
Eggs in which the embryonic heart has just begun to beat are
susceptible to NaCl, while the same embryo removed from the
egg will usually live for days in the same solution. As a matter
of fact, they live almost as long as it takes for the normal embryo
in sea water to reach the time of hatching. The freshly hatched
larvse, however, are killed very quckly when put into the same
NaCl solution (Table I.). A marked change in the resistance
of the animal to NaCl thus takes place when the embryo is
ready to emerge or emerges from the egg.

Equal parts N KCl -\- N CaClz. — In such a mixture the toxicity
for the egg, embryo, and larva is about the same (Table I.).

ACTION OF SALTS ON FUNDULUS EGG. 40!

The time taken to cause cessation of the heart beat, in the
mixture however, is quite different from that required in the
case of solutions of the individual salts, as is shown in Table I.

Equal parts N NaCl + N CaCl2. — In this mixture the larva
is killed in approximately 20 to 25 minutes. The embryo
dissected from the egg membrane can survive from 6 to 24 hours
while the eggs are not killed. As a matter of fact, eggs will
hatch in the solution, and the hatched larva is quickly killed,
showing quite conclusively that while in the egg the embryo is
protected; once out, it quickly succumbs. The embryo dis-
sected from the egg membrane is always found to be much more
resistant than the larva (Table I.).

Equal parts N NaCl + N KCl. — Such a mixture is quite toxic
for the egg, embryo, and larva, the relative resistance being:
larva < embryo < egg. The egg is approximately 2 times as
resistant as the embryo and about 5 times as resistant as the

larva (Table I.).

DISCUSSION.

The above results seem to the author to be of interest inasmuch
as they definitely point out and show that any assumption as to
the interior condition of the egg cannot be relied upon until
satisfactorily tested and proven. Loeb (1-17), in most of his
work on Fundulus, seems to have assumed that the resistance of
the hatched embryo or larva was comparable to that of the
embryo within the egg membrane. In the case of NaCl in
particular this view is shown to be erroneous, and it is quite
possible that further investigation will yield equally interesting
results. That the egg membrane changes in consistency with
age is quite apparent in the ease with which the membrane can
be cut. Young fertilized eggs have membranes quite tough and
turgid; eggs exposed to distilled water and at low temperatures
for long periods of time have much softer and more pliable
membranes. Associated with these structural changes are doubt-
less the marked physiological ones observed. Further proof of
marked physiological changes in the egg membrane have come
out of unpublished investigations recently conducted by Miss
E.'Yagle of this Laboratory, on the exosmosis of H2O from the
Fundulus egg of different ages. Loeb (6), also points out in

4O2 JOSEPH HALL BODIXE.

several instances that such changes occur — e.g., "... since the
newly fertilized egg is killed more rapidly by a m/2 solution of
NaCl than it is killed by the same solution one or two days after
fertilization." Young eggs, therefore, are very susceptible to
NaCl while older ones are quite resistant. Young embryos, on
the other hand, are much less susceptible than are very old ones.
The hatched embryo or larva, however, is extremely susceptible
to NaCl. The question as to the fundamental action of NaCl
upon the younger eggs with hearts just beginning to beat ap-
parently is not one of a purely chemical nature but rather of
an action on the membrane in which the embryo is doubtless
killed by some secondary effects, possibly osmotic. In eggs
thus killed the contained embryo always appears much shrunken
in size.

By means of the dissection technic in liberating the embryos
from the egg it is possible to study quantitatively the relative
effects of the salt upon the membrane and contained embryo
and also upon the embryo free from the egg membrane. KC1,
for example, seems to spend about 75 per cent, of its total time
effect upon the egg membrane and approximately 15-50 per cent,
on the embryo. CaCla on the other hand, spends about 60
per cent, on the membrane and 30-40 per cent, on the embryo.
NaCl must exert most of its effect on the membrane in those
cases in which it is toxic since for embryos removed from the
egg it is relatively non-toxic.

The site of action of the three salts Na, K, and Ca chlorides
has always been of considerable physiological interest. Loeb
(17), in the case of Fundulus, seems to have attributed the
fundamental action of these salts to the membrane — the process
being purely a diffusion phenomenon. In young eggs NaCl
perhaps does not penetrate the membrane while in older eggs
the membrane seems freely permeable, since the contained
embryo is not killed in the same solutions when removed from
the egg. KC1 and CaCl seem to penetrate the egg membrane,
the Ca entering in a relatively slightly shorter time than the K.
The Na + Ca antagonism as suggested by Loeb (2) must be a
membrane phenomenon since the embryo removed from the egg
or the hatched larva is quickly killed in such a mixture. Combi-

ACTION OF SALTS ON FUNDULUS EGG. 403

nations of Na + K, however, act more like KC1 alone and little
if any antagonism seems to exist (Loeb, 1,2). The combination
K + Ca, on the other hand, seems to exert additive effects and
to kill eggs and embryos in almost the same time.

Several additional facts noted in these experiments are of
especial interest since they seem intimately concerned in any
fundamental explanation of the salt effect upon the egg and
embryo. The space between the egg membrane and embryo is
at first very small, due to the large size of the yolk sac. As the
embryo develops rather a large amount of fluid accumulates
between the egg membrane and embryo. This increase in fluid
perhaps has much to do in modifying the rates at which the
embryo is killed while in the egg. Around the yolk sac and
embryo there is also a delicate vitelline membrane, the properties
of which seem of much importance in respect to the resistance
of the embryo to various salts. If with a very fine pointed
needle a minute hole is made in this membrane immediately
ventral to the eye and the embryo, still normal, transferred to
a solution of NaCl or left in a solution of NaCl which is not
ordinarily toxic, it is quickly killed. It seems that this membrane
is an important factor in determining the resistance of the
embryo to NaCl. Further investigations are to be carried out

on these points.

SUMMARY.

(1) By means of an operative technic it is possible to remove
the egg membrane from the egg of Fundulus heteroditiis and to
compare experimentally the action of salts on the egg, the
embryo freed from the membrane and the newly hatched larva.

(2) The effects of normal solutions of K, Na, and Ca chlorides
upon the above are reported.

(3) The embryo, freed from the egg membrane, is quite
resistant to NaCl solutions while the hatched larva is quickly
killed in the same solution.

(4) The resistance of the eggs to NaCl increases with age.

(5) K and Ca chlorides kill the dissected-out embryo much
more quickly than the egg, while the recently hatched larva is
much more sensitive to the two salts than is the embryo.

(6) Combinations of these salts show antagonistic action.

404 JOSEPH HALL BODINE.

Na + Ca mixtures are not toxic for eggs but are markedly so
for the embryo freed from the egg membrane and for the newly

hatched larva.

LITERATURE CITED.

1. Loeb, J. Archiv. f. d. ges. Physiol., 1894, 4, 530.

2. Loeb, J. Archiv. f. d. ges. Physiol., 1901-02, 38, 68.

3. Loeb, J. Archiv. f. d. ges. Physiol., 1905, 107, 252.

4. Loeb, J. Amer. Jour. Physiol., 1899-1900, 3, 327.

5. Loeb, J. Amer. Jour. Physiol., 1901-02, 6, 411.

6. Loeb, J. Science, 1911, 34, 653.

7. Loeb, J. Science, 1912, 36, 637.

8. Loeb, J. Science, 1916, 44, 574.

9. Loeb, J., and Wasteneys, H. Biochem. Zeit., 1911, 31, 450.

10. Loeb, J., and Wasteneys, H. Biochem. Zeit., 1911, 32, 155.

11. Loeb, J., and Wasteneys, H. Biochem. Zeit., 1911, 33, 489.

12. Loeb, J., and Wasteneys, H. Biochem. Zeit., 1912, 39, 167.

13. Loeb, J. J. Biol. Chem., 1915, 23, 139.

14. Loeb, J. J. Biol. Chem., 1916, 27, 339, 353, 363.

15. Loeb, J. J. Biol. Chem., 1917, 32, 147.

16. Loeb, J., and Cattell, McK. J. Bio. Chem., 1915, 23, 41.

17. Loeb, J. J. Gen. Physiol., 1922-23, 5, 231.

ON THE ABILITY OF CERTAIN MARINE INVERTE-
BRATES TO LIVE IN DILUTED SEA WATER.

A. S. PEARSE,
DUKE UNIVERSITY.

There are many reasons for believing that animal life originated
in the ocean and has gradually spread through the ages into
freshwater and land habitats (15, 21). In the past annelid
worms invaded the soil, probably by a rather indirect route
which led them first into freshwater and gradually out into the
land (23). At the present time many animals in various parts
of the earth show varying degrees of ability to live in diluted
sea water (i, 3, 10, u, 12, 21). Marine invertebrates which
have been studied have an osmotic pressure in their blood which
is approximately equal to that in sea water, but the mineral
salts are somewhat less and the pressure is maintained by other
substances, largely organic, which are present (i, 8, 9). The
skins of various animals differ greatly in ability to control the
exchange of chemicals between the body fluids and the surround-
ing medium. Adolph (i) studied the exchanges of substances
through the skins of annelids and found that there was little
resistance to them.

The writer felt that it would be of interest to determine the
ability of representative marine annelids to live in diluted sea
water, and during August 1927 made some observations in the
Marine Biological Laboratory at Woods Hole. Thanks are due
to Drs. J. A. Dawson and R. Bennitt who made suggestions and
helped in the identification of several species. Mr. A. M.
Hilton and his staff of collectors also made special efforts to
secure materials.

Animals were brought in fresh from the field and placed as
soon as possible in clean glass finger bowls containing 250 cc.
of water. Tubicolous worms were removed from their tubes,
except in the case of Hydroides, which was studied both in and
out of tubes, and of Cystenides, which was left in its own tubes.

405

406 A. S. PEARSE.

The water in all bowls was changed each morning, and oftener
when it showed indications of becoming stagnant or when a
worm bled. Dilutions were made with fresh water from the
taps in the Marine Biological Laboratory. Sea water came
from the same source. Page (13) has made a careful analysis
of this' water for mineral constituents.

The results of the observations are given in Table I. Nereis
virens, Laonice viridis, and Limulus polyphemus showed a con-
siderable degree of toleration for sea water diluted to one fourth
its normal salinity. Arabella opalina, Glycera dibranchiata, and
Lepidonotus squamatus lived for many days in one half sea water
and one half fresh. Most of the worms tested lived several
days in three fourths sea water. Nereis in higher concentrations
of sea water climbed out of the dishes at intervals and was
found in varying degrees of desiccation, hence some individuals
probably died sooner than some of those in more dilute solutions.
Every Laonice studied lived throughout the period of observation
and was active at the end. The Hydroides in tubes did not
live as long as those which were removed. This was probably
due to the fouling of the water by small organisms in and on
the mollusc shells to which the tubes of these worms were
attached. The Limuli used were small individuals, less than
10 cm. long. All that were tested in solutions as low as one
fourth sea water survived to the end of the observations. One
individual lived 26 hours in a solution of one eighth sea water,
and another lived two hours in fresh water.

Nereids have been observed in various localities to be note-
worthy for their ability to endure considerable dilution of sea
water (3, 7, n, 13). In India there is a species of Limulus which
lives in brackish water (3). Vaughn (24) found that several
species of corals survived a reduction of twenty per cent, in the
salinity of the sea water and he interpreted this as indicating
that the ocean has been in the past less salty than now. The
observations described in this paper show that many worms have
similar toleration. The writer cannot see that general features
of bodily structure and habitat are especially correlated with
ability to survive in diluted sea water. Apparently delicate
branchiate worms, like Ch&topterus, Diopatra, and Laonice,

MARINE INVERTEBRATES IX DILUTED SEA WATER.

407

endure fresh water about as well as apparently tougher worms,
such as Phascolosoma and Lumbricillus. The only oligochaete
observed, Lumbricillus agilis, was not as hardy as many
polychaetes when placed in diluted .sea water. In low salt con-
centrations all the animals studied showed a tendency to swell
and became turgid and extended. Laonice perhaps showed this
least; Lepidonotus and Lumbricillus perhaps most; Amphitrite
frequently bled and died soon. The individual Nereis which
lived and was active for twenty-one days in a medium containing
only one fourth sea water became very active and soon began to
shrink to its normal size when replaced in undiluted sea water.

TABLE I.

IIME IN HOURS WHICH ANIMALS LIVED IN SEA WATER A'ND
VARIOUS DILUTIONS OF IT.

A indicates that an animal was apparently in good condition when the obser-
vations were discontinued.

Name of Animal.

No.
Ob-
served.

Sea Water.

i

3

4

i

i

4

1
8

Amphitrite ornata (Leidy)

12

5

2
IO

4
13
3
i?
9
7

12

8
13
13

5
i

2
13

4
5

65

403/1

475^4
487.4

28oA
498.4
228.4
26o.4
230^
474-4
igiA
8

356
236 +
252.4
303-4

H5
302.4

403-4

475-4
487.4

35
478-4
498-4
228.4
26q.4
230.4
474-4
191.4

100

"5
259.4
298

7
302.4

4
8

38
256.4
30
498-4

4
105
230.4
379-4
192.4
130
8

500/1
-3
35

2

• 7

2

I

3
•3

4

.2

• 7
•5
•5
230.4

-3
192.4

•3
500,4

.2
26

•4

40

3

Arabella opalina (Verrill)

Arenicola marina (L.)

ChcEtoptenis pergamentaceus Cuvier. . .
Cirratulus grandis Verrill

P Clymanella torquala (Leidy)

'Cystenides gouldii Verrill . . .

Diopatra citprea (Bosc)

Glycera dibvanchiata Ehlers

Harmothoe imbricala (L.)

Hydroides hexagonus Bosc

Laonice viridis (Verrill)

Lepidonotus squamatus (L.)

«. Limulus polyphemus (L.)

Lumbricillus agilis Moore

Maldane urceolata (Leidy)

Nephthys bucera Ehlers.

Nereis Tirens Sars . .

Phascoloscoma gouldii (Pourtales) ....
Pisla palmate. (Verrill)

SUMMARY.

The ability of Limulus, Phascolosoma, and eighteen marine
annelids to survive in various dilutions of sea water was studied.
Most of the animals lived for a week or two in a mixture of three

408 A. S. PEARSE.

fourths sea water and one fourth fresh; several species lived in
one half sea water. Limulus, Laonice, and Nereis lived and
were active for periods of two to three weeks in one fourth sea
water, but died in weaker solutions.

BIBLIOGRAPHY.

1. Adolph, E. F.

'25 Some Physiological Distinctions between Freshwater and Marine Or-
ganisms. BIOL. BULL., 48: 327-335.

2. Andrews, F. B.

'25 The Resistance of Marine Animals of Different Ages. Publ. Puget
Sound. Biol. Station, 3: 361-363.

3. Annadale, N.

'22 The Marine Element in the Fauna of the Ganges. Bijdr. Dierk. Amster-
dam (Feest-num., M. Weber): 143—154.

4. Benham, W. B.

'01 Polychaete Worms. Cambr. Nat. Hist., 3: 241-346.

5. Blum, H. F.

'22 On the Effect of Low Salinity on Teredo navalis. U. Cal. Publ. Zool.,
22: 349-368.

6. Dambovinceanu, A., and Raplsine, L.

'25 Sur le pH interieur de ceitains elements du liquide cavitaire chez
Sipunculus nudus. C. R. Soc. Biol. Paris, 93: 1364.

7. Dehorne, A.

'25 Observations sur la biologic, de Nereis diversicolor. C. R. Acad. Sci.

Paris, 180: 1441—1443.
8. Dekhuysen, C.

'21 Sur la semipermeabilite biologique de parois exterieures des Sipunculides.

C. R. Acad. Sci. Paris, 172: 238-241.
9. Duval, M.

'24 Recherches sur le milieu interieur des Invertebres marins. Comparison
entre le teneur en cholorure de sodium de ce milieu dt celle de 1'eau
de merexterieure. Bull. Sta. biol. Arcachion, 21: 33-39.

10. Duval, M.

'25 (Same title.) Ann. Inst. Oceanographique Paris, N.S., 2: 233-407.

11. Ferronniere, G.

'01 Etudes biologique sur les zones supralittorales de la Loire-Inferieure.
Bull. Soc. Sci. Nat. Quest Nantes, //: 1-45.

12. Flattery, F. W., and Walton, C. L.

'22 The Biology of the Seashore. N. Y., xvi + 336.

13. Florentin, R.

'99 Etudes sur la faune des mares salees de Lorraine. Ann. des Sci. Natur.
(8), 10: 209-346.

14. Fredericq, L.

'22 Action du milieu marin sur les animaux Invertebres. Bull. Acad. Roy.
Bruxelles (5), 8: 423-426.

15. Hesse, R.

'20 Die Anpassung der Meerestiere an das Leben in Susswasser. Bonn, 1-6.

MARINE INVERTEBRATES IN DILUTED SEA WATER. 409

16. Labbe, A.

'22 La distribution des animaux des marais salants dans ses rapports avec
la concentration en ions hydrogens. C. R. Acad. Sci., Paris, 175:

9I3-9I5-

17. McClendon, J. F.

'18 On Changes in the Sea and their Relations to Organisms. Pap. Dept.
Mar. Biol., Carnegie Inst. Wash., 12: 213-258.

18. Packard, E. L.

'18 A Quantitative Analysis of the Molluscan Fauna of San Francisco Bay.
U. of Calif. Publ. Zool., 18: 299-336.

19. Page, I. H.

'27 The Composition of Woods Hole Sea Water. BIOL. BULL., 52: 147-160.

20. Pearse, A. S.

'26 Animal Ecology. N. Y., ix + 417.

21. Pearse, A. S.

'27 The Migration of Animals from the Ocean into Freshwater and Land
Habitats. Amer Nat., 61: 466-476.

22. Redeke, H. C.

'22 Zur Biologic der Niederlandischen Brackvvasser typen. Bijdr. Dierk.
Amsterdam (Feest-Num, M. Weber): 329-335.

23. Thomson, J. A.

'22 The Haunts of Life. N. Y., xv + 272.

24. Vaughn, T. W.

'19 Corals and the Formation of Coral Reefs. Smithsonian R. (1917):
189-276.

THE EFFECT OF THE QUANTITY OF CULTURE

MEDIUM ON THE DIVISION RATE

OF OXYTRICHA.

H. B. YOCOM,
DEPARTMENT OF ZOOLOGY, UNIVERSITY OF OREGON.

The hypothesis advanced by Robertson (1923) that unicellular
organisms produce an autocatalytic substance which aids in
bringing about their own division, has been subjected to investi-
gation by a number of workers who find little if any evidence
supporting Robertson's view. In a recent paper Greenleaf
(1926) presents the results of a series of experiments concerned
with the effect of the amount of culture medium and cell
proximity on the division rates of various Protozoa. From these
results he is unable to find any evidence in favor of the auto-
catalytic effect. On the other hand he finds evidence corrobo-
rating the findings of Woodruff (1911), namely, that in small
quantities of culture medium the division rate of isolated indi-
viduals is retarded. This is due, according to the author, to
the accumulation of excretory products and their greater con-
centration in small amounts of medium. This retardation effect
is lessened as the quantity of medium is increased.

These differences in findings leave the problem still open for
investigation. It is the purpose of this brief article to set forth
the results of some experiments made in an attempt to determine
the relation of the amount of culture medium to the division rate
of certain protozoa. No attempt has been made at this time to
go into all the phases of the whole subject. The experiments
were carried on with a species of Oxytricha. One of these
organisms was isolated from a laboratory aquarium into ten
drops of sterile .05 of one per cent, beef extract solution, on a
depression slide. This was kept in a moist chamber made of a
covered glass disk 7x4x4 inches. A pure line was established
by reisolating every day for several days. When it was desired
to start the experiments one or more slides were allowed to stand

410

DIVISION RATE OF OXYTRICHA. 4! I

48 hours or until there were sufficient organisms in one slide to
start the desired number of cultures. All cultures of a given
experiment were therefore descendants of a single individual
isolated 48-72 hours before.

Oxytricha was chosen after a number of preliminary trials with
various ciliates, because of its rapid division rate. At laboratory
temperature there are usually two or more divisions in 24 hours
and in rare cases as many as six generations have been produced
in a day. The large number of individuals produced seemed to
make the organism a more t favorable subject for these experi-
ments than other more slowly dividing ciliates.

The experiments consisted in comparing the division rate of
sister cells in four and ten drops of .05 of one per cent, of sterile
beef extract.

In order to keep the organisms in as near the same atmospheric
environment as possible slides with two depressions were used.
One depression contained the organism in four drops, the other
the culture of ten drops. Four slides were used in the earlier
experiments, but the number was later increased to ten. The
pipettes for measuring the quantity of culture medium were
drawn out so that one drop was about one one-hundredth cubic
centimeter.

The first series of experiments was carried on at room tempera-
ture but as this fluctuated greatly it was decided that controlled
temperature conditions were essential for reliable result. For
temperatures above ordinary room temperature a Freas electric
incubator was used and for temperatures below room tempera-
tures a Cenco water bath was used, in which the moist chambers
were sunk almost to their tops in the water and a constant
temperature of 18° C. was maintained.

In all cases counts were made at the end of the 24-hour period.
One or more slides were set aside to develop "seed" for the next
day's experiments. Division rate was measured by the total
number of organisms produced in 24 hours from an isolated
individual in a given amount of culture medium.

In making the isolations care was taken to take individuals of
as near the same size as was possible. In a culture divisions do
not occur simultaneously and there are therefore small individuals

4I2

H. B. YOCOM.

—the result of very recent division, large individuals almost
ready to divide, and a majority of an .intermediate size, probably
those full grown. This latter group was selected for the experi-
ments. All organisms used in one experiment were of the same
line of protoplasm and as near the same age with respect to the
last division as could be determined.

SERIES I.

Series I. was carried on at laboratory temperature which
fluctuated very much from day to night, varying probably from
1 6° C. to 22° C. The results, however, are fairly constant.

Experi-
ment.

4 Drops.

Experi-
ment.

10 Drops.

No. of
Cases.

Total No.
Individuals.

No. of
Cases.

Total No.
Individuals.

i

4
4
4
4
4
4
4
4
4
4

28
10
13

22
24
20
27

45

15

IS

i

4
4
4

4
4
4
4
4
4
4

27
8
13
15
14
16
36
20

14
16

2

2

1

•3

A. .

4

S- -

c .

6

6

7. .

8

8

o .

o

10 .

IO

40

219

40

179

Average, 5.48

Average, 4.47

This indicates that at room temperature the individuals in
four drops have a much higher division rate. The average
division rate is 24.6 per cent, higher in the former than in the
latter.

SERIES II.

Series II. was similar to Series I., except for the fact that the
temperature was maintained at about 24° C. The division rate
was therefore higher than in Series I.

Here again the four drops show a higher division rate than ten
drops. The average for four drops is 12.3 per cent, higher than
for ten drops.

DIVISION RATE OF OXYTRICHA.

413

4 Drops.

10 Drops.

Experi-
ment.

No. of
Cases.

Total No.
Individuals.

Experi-
ment.

No. of
Cases.

Total No.
Individuals.

1 1

4
4
4
4
4
4
4
4

2

50

21

26

37
38
3i

10

76
59

ii

4
4

4
4
4
4
4
4
4

28
17
27
35
48
28
15
47
82

12

12

T7

13

1A

14

I e

15

16

16

17

17

18

18

IQ

19

34

348

36

327

Average., 10.2

Average, 9.08

SERIES III.

Since the difference in division rate in the two given amounts
was larger in the lower (room) temperature (Series I.) it seemed
advisable to compare the differences in rates at different con-
trolled temperatures. Therefore, a series of experiments was
run with individuals grown in four and in ten drops at 18° C.
and at 23° C. The organisms used in any one day were from
the same isolated individual. They were presumably as near
the same age as it was possible to get them and all had been
under identical environmental conditions.

18° C.

23° C.

4 Drops.

10 Drops.

4 Drops.

10 Drops.

Exp.

No.
Cases.

In-
divid.

Exp.

No.
Cases.

In-
divid.

Exp.

No.
Cases.

In-
divid.

Exp.

No.
Cases.

In-
divid.

B

10

19

B

9

15

A

7

4i

A

10

45

D

9

38

D

10

32

C

10

59

C

IO

77

F

10

32

F

10

3i

E

10

83

E

10

80

H

10

46

H

10

36

G

10

103

G

10

82

L

10

44

L

10

34

K

IO

138

K

10

142

N

10

26

N

10

24

M

IO

87

N

9

60

P

9

36

P

10

29

O

IO

97

O

IO

76

68

241

69

2OI

67

608

69

562

Average, 3.54

Average, 2.91

Average, 9.07

Average, 8.1

H. B. YOCOM.

As in Series I. and II. the division rate in Series III. is higher
in four drops than in ten drops. At the lower temperature the
difference in the division rate in the two groups is greater than
it is at the higher temperature. In the former the organisms in
four drops have a higher rate by 21.64 Per cent, while in the
latter the four drops have a higher rate by 10.19 Per cent.

In all of these experiments the division rate in four drops
exceeds that in ten drops by more than ten per cent. If we
consider only the amounts of culture medium and disregard
temperature the results are as follows :

4 Drops.

10 Drops.

Series.

No. of Cases.

Individ.

Series.

No. of Cases.

Individ.

I

40

34
68
67

219
348
241
608

I

40
36
69
69

179
327

2OI
562

II

II

Ill 18°
23°-...

Ill 18°. ...
23°..-.

209

1,416

214

1,269

Average, 6.77

Average, 5.93

The division rate in four drops is considerably higher than
that in ten drops — a difference of 14.17 per cent. This result
is perhaps more significant than the others since there are over
200 cases considered.

The results of these experiments have a direct bearing on two
important problems. The first of these is the question of the
autocatalase. Robertson (1923) quite definitely showed that
with the infusorian Enchelys, isolated in a limited amount of
culture medium, there was an increasing reproductive rate during
the second twenty-four hours of a culture over that of the first
twenty-four hours. He also found that, within certain limits,
division rate of Enchelys was higher in the first twenty-four
hours in small quantities of culture medium than in larger
quantities and that individuals isolated into one cubic centimeter
or more of medium rarely survive.

In his interpretation of these and other experiments Robertson
was forced to the conclusion that "the only possible inference

DIVISION RATE OF OXYTRICHA. 415

that can be drawn from this phenomenon is that infusoria dis-
charge into the culture medium some substance which accelerates
their own multiplication."

As further test of this theory he showed that reproduction
would occur in sterile culture medium or in distilled water and
cites the experiments of Peters (1921) with Colpidium to show
that bacteria as food are not essential for reproduction of infusoria
if the proper ingredients necessary for growth are put into the
medium.

The second question for consideration is that concerning the
effect of the excretory products on the division rate of protozoa.
Woodruff (1911) has shown that the rate of division is very
markedly lowered by the accumulation of excretory products.
Greenleaf (1926) has taken up the problem and using several
different infusorians arrived at the same conclusions as those
reached by Woodruff fifteen years before. These results are in
direct contradiction to those of Robertson and Greenleaf con-
cludes that as far as the organisms with which he worked are
concerned there is no evidence of the autocatalytic effect. He
further concludes that infusoria tend to multiply faster in larger
amounts of culture medium and that one individual in a given
amount of medium multiplies faster than each does when two
individuals are in the same amount. This difference is explained
on the basis that two organisms would raise the concentration
of the excretory products to a degree injurious to themselves
more rapidly than a single individual would do.

The results of the experiments upon which this paper is based
seem to favor the idea advanced by Robertson. In all series in
which four and ten drops of culture medium were used the
average division rate in four drops was higher than in ten drops.
Since the culture medium was sterile when the organisms were
introduced, the question of food does not seem to be one of
prime importance in our consideration. The most plausible
explanation therefore is that Oxytricha produces a substance
which is liberated into the surrounding medium and which
reacts on the infusorians to bring about cell division, and that
the concentration of the substance necessary for bringing about
cell division is reached in the smaller amount of culture medium
before it is in the larger amounts.
28

416 H. B. YOCOM.

The conflicting conclusions reached by various workers leave
the question still unsolved. It is possible that the various
results have been due to the fact that organisms with a high
division rate show the effect of the autocatylitic substance more
readily than do organisms which divide only once or possibly
twice in twenty-four hours. Robertson used Enchelys which
divides many times a day. Woodruff and Greenleaf used larger
infusorians which have a much slower rate of division. Between
these extremes Oxytricha has an intermediate division rate and
in the experiments cited undoubtedly shows a higher average
rate of division in the smaller amounts of culture medium.
There seems to be no explanation of this so satisfactory as that
suggested by Robertson for Enchelys, namely, that Oxytricha
produces a substance which reacts on the organism to bring
about division, and that the concentration of this substance is
reached in sufficient quantities to bring about cell division in
small quantities of culture medium before it is in greater
quantities.

BIBLIOGRAPHY.
Calkins, G. N.

'26 The Biology of the Protozoa. Philadelphia. Lea and Febiger.
Greenleaf, W. E.

'26 The Influence of Volume of Culture Medium and Cell Proximity on the

Rate of Reproduction of Infusoria. Jour. Exp. Zool., 46.
Peters, R. A.

'21 The Substance Needed for the Growth of a Pure Culture of Colpidium

colpoda. Journ. Physiol., 40.
Robertson, T. B.

'23 The Chemical Basis of Growth and Senescence. Philadelphia, J. B

Lippincott.
Woodruff, L. L.

'n The Effect of Excretion Products of Paramcecium on the Rate of Reproduc-
tion. Jour. Exp. Zool., 10.

CONJUGATION, DIVISION, AND ENCYSTMENT IX

PLE UROTRICIIA LA NCEOLA TA .l

REGINALD D. MAXWELL, M.A.,
SCHOOL OF HYGIENE AND PUBLIC HEALTH, JOHNS HOPKINS UNIVERSITY.

INTRODUCTION.

Evidence has been steadily accumulating of recent years that
many closely related species differ from each other in number of
chromosomes, the numbers often being multiples, or multiples
of some common factor. Because of this fact and because the
processes of conjugation have been thoroughly studied in only
a very few hypotrichs, the present investigation was undertaken.
Oxytricha fallax and Pleurotricha lanceolata, although placed in
different genera, are morphologically quite similar, and it was
thought that a comparison of the two species with respect to
chromosome number and details of conjugation would be of
interest. It has been shown before in at least one case (Chilodon
uncinatus, MacDougall, 1925) that new species may arise de novo
from old ones among the protozoa, the chief differences being in
chromosome number, and it is not unlikely therefore that this
often happens in nature. A study of division and encystment
was undertaken as a natural corollary.

Pleurotricha lanceolata AND SIMILAR SPECIES.

Pleurotricha lanceolata was first described accurately by Stein
in 1859, although two other species which he regards as the
same had been described previously — Stylonychia lanceolata and
Keratium calvitium, the former by Ehrenberg in 1832 and the
latter by Miiller. But Ehrenberg's organism had 16 to 18 cirri
on the dorsal surface as well as the full complement on the

1 From the Department of Protozoology, School of Hygiene and Public Health,
Johns Hopkins University and the Department of 'Biology, Amherst College.

I wish to acknowledge the kindly interest and assistance of Dr. Harold Plough
of Amherst College, Dr. Robert Hegner of Johns Hopkins University, and Dr.
Mary Stuart MacDougall of Agnes Scott College. Without their aid and sug-
gestions the work would have been much more difficult, if not impossible.

417

41 8 REGINALD D. MANWELL.

ventral surface, and Miiller failed to see the abdominal cirri, so
that priority of description and name belongs to Stein.

Pleurotricha lanceolata was also described and figured by Kent
(1880 to 1882), but his description is slightly different from that
of Stein. In addition to the double row of marginal cirri on
the left-hand side, Stein shows a very short third row in the
middle region of the body which Kent omits, as well as the
anterior third of the inner marginal row.

The race which I have used (Fig. i) fits Kent's description very
closely, but is only about half as large, and the inner marginal
row of cirri is found only in the middle third of the body. Stein
gives the size as varying between 83 and 143 ju, and Kent makes
it just about twice as great. I have found the average size to
be about 140 /*, the extremes being 100 to 165 M, if exconjugants
which are always exceptionally small are omitted.

Oxytricha bifaria (Stokes, 1887) also resembles Pleurotricha
lanceolata, differing principally in having only a single marginal
row of cirri extending around the body, and in having eight
instead of six cirri on the anterior quarter.

There is also a considerable resemblance between Oxytricha
fallax and the chief species under discussion. The former has
five anal cirri arranged in an oblique row, the two posterior
extending only slightly beyond the body margin.

MATERIAL AND METHODS.

The cultures were started from a single individual obtained
from a mass culture of leaves and swamp water, which in turn
was secured from Orient Springs, near Amherst, Massachusetts,
November 5, 1925. These cultures were continued until May
1927, the medium being hay infusion, to which in some cases a
small amount of peptone was added.

All preparations were fixed with Calkins' modification of
Schaudinn's fluid, and stained with hemotoxylin, both the long
and short methods being used.

DIVISION.

The first indication of division is seen in the macronuclei, in
each of which a nuclear cleft, or kernspalt, appears at the extreme

CONJUGATION", DIVISION, AND ENCYSTMENT. 419

outer ends, and this gradually moves across the nuclei to finally
disappear at the inner ends. Simultaneously, or very shortly
afterward, a new adoral zone begins to form in the midregion
of the body. This stage of division is shown in Fig. 2.

Very soon the micronuclei begin to enlarge and a clear space,
or halo, appears about each. At about the same time a division
center appears and divides, the two products remaining con-
nected by a centrodesmose (Figs. 3 and 4). I have never been
able to detect this endosome in the resting micronucleus however,
and it is difficult to observe even in division. As the dividing
nucleus enlarges it elongates, and the chromatin appears to be
arranged in very fine strands, some of which seem to be in
reality rows of granules (Fig. 4 A). At this time it stains very
faintly. The chromatin next appears to become more or less
concentrated at the poles (Fig. 4 B). As all this goes on the
macronuclei change their shape, becoming almost round and the
kernspalt finally passes off the inner ends (Figs. 4, 5, 6). The
new adoral zone meanwhile increases in size, and new cirri may
be seen developing beside the old ones of the old adoral zone
(Fig. 6). In Fig. 5 the new ventral cirri may also be seen close
to the adoral zones, where they first arise. As the time ap-
proaches for final separation of the daughter individuals these
cirri gradually move to more nearly their normal positions
(Fig. n).

In the meantime the chromatin of the micronuclei appears to
condense into discrete masses, which take the form of threads
of varying length (Fig. 5 B). These gradually combine, until
the nucleus appears to be composed of closely packed and more
or less tangled threads, as in Figs. 5 A and 6 B, and finally a
definite spindle appears (Fig. 6 A}. The poles of the latter
move apart and a very characteristic figure is produced which
is shown in Fig. 7 C. Here the chromosomes appear as heavily
staining Vs, the limbs of which are connected by heavy and
deeply staining strands. Apparently division is really longi-
tudinal, the two daughter chromosomes taking on a V shape and
appearing connected at the ends for a while thereafter. Their
final separation is shown in Fig. 7 A. Fig. 8 A shows a some-
what later stage of the metaphase.

42O

REGINALD D. MAXWELL.

At about this time the macronuclei begin to elongate (Fig. 8)
and the remaining stages in division are passed through very
rapidly, the time required being about fifteen minutes at ordinary
temperatures, as compared to more than an hour for the preceding
steps. The micronuclei enter the anaphase, which is quickly
completed, requiring only a very few minutes, and which may
be seen in Figs. 9, 9 A, 9 B, 9 C. The telophase is shown in
Figs. 10, n, ii A, II B, and 12. The macronuclei divide at this
time (Figs, n and 12), and the parent animal very shortly
separates into two daughter individuals (Fig. 13).

DISCUSSION.

The process of division in ciliates is so well-known that few
comments are necessary, since different species differ only in
minor details, and fission in Pleurotricha lanceolata follows the
general plan. There are, however, a few points in which this
species differs from others, and in other respects there are
interesting resemblances.

The presence of a kernspalt is said to be common to all
hypotrichous ciliates, even during the major part of the resting
stage (Calkins, 1919), but I have never been able to see it in
Pleurotricha except when the animal was about to divide, as
shown by the presence of a rudimentary adoral zone. This
happens even before there is any visible change in the micro-
nuclei.

Apparently the early appearance of a new cytostome with its
accompanying peristomial apparatus is characteristic of most
dividing infusoria, e.g. Uroleptus mobilis (Calkins, 1919), Chilodon
uncinatiis (MacDougall, 1925), although there are other cases in
which its development is more delayed as in Paramcecium
trichium (Wenrich, 1926).

It seems to be generally accepted that in division there is a
complete reorganization of the entire cell, extending even to the
disappearance and subsequent reformation of all the organelles,
including trichites in organisms which have them, undulating
membranes, cirri, and even cilia. According to Wallengren
(1900) the cirri of the old adoral zone in hypotrichous ciliates are
gradually absorbed as new ones take their places, and I have

CONJUGATION, DIVISION, AND ENCYSTMENT. 42 1

been able to confirm this observation for Pleurotricha. That this
reorganization does not always extend to the old oral apparatus,
however, is indicated by Wenrich (1926) who was unable to
find evidence that the old cytostome disappeared and was
reformed in Paramoscium trichium.

It is a remarkable fact that all the cirri, with the exception of
the marginal ones, arise in the immediate neighborhood of the
adoral zones, and indeed the first indications of their appearance
are visible very shortly after the adoral zone itself appears.
That this was the case in the Oxytrichidae was noticed by Sterki
(1878), and has been noted for other groups by other investi-
gators. MacDougall (1925) found that new cilia in a dividing
Chilodon uncinatus first appear near the new pharyngeal baskets.

The presence of a centrosome which divides as an initial step
in division has until recently not been noted in ciliates. It was
reported by Calkins in the first maturation division of Uroleptus
mobilis however (Calkins, 1919), although he evidently did not
see it in ordinary division, and MacDougall found it in Chilodon
uncinatus. To this list can now be added Paramoscium trichium
(Wenrich, 1926), and Pleurotricha lanceolata.

ENCYSTMENT.

Encystment is a common occurrence in the life cycle of
Pleurotricha bifaria, although what the conditions are which
determine its occurrence I have been unable to discover. It
may occur in old cultures, and is perhaps most frequent under
those circumstances, but it also occurs in pedigreed drop cultures,
which are changed every day, and in which the division rate is
rapid. Of four individuals which have arisen by division from
a single ancestor during the preceding twenty-four hours it is
not uncommon to find one or two encysted, while the others
continue to divide actively.

The account of nuclear phenomena which follows is admittedly
inadequate, but it is given since no other description has been
made of this stage in the life cycle of this particular species, and
because the history of the cytological changes which occur in
ciliates during encystment is notoriously incomplete.

The first change appearing in an individual about to encyst is

422

REGINALD D. MANWELL.

like that occurring in other infusoria under similar circumstances.
The shape becomes spherical, the contractile vacuole gradually
ceases to pulsate, the food vacuoles and all the organelles dis-
appear (Fig. 14). The next step is the secretion of the cyst wall,
shown in Fig. 15. It is thick, quite pliable but very tough, and
without visible openings. The outside is covered with irregular,
short spines. Such a cyst is shown by Stein (1859) for P.leuro-
tricha lanceolata, but Cienkowsky (1855), who gives three figures
of the cyst formation of Stylonychia lanceolata, does not show the
spines. Nevertheless these two species are regarded by Stein as
the same, as already stated.

Soon (probably under normal conditions in about twenty-four
hours) the old macronuclei, and apparently one of the micro-
nuclei, are bodily extruded through the cyst wall, although the
opening through which they pass is never visible under any other
circumstances (Fig. 16). This leaves the encysted animal in
the condition shown in Fig. 17. The remaining nucleus then
proceeds to divide but I am unable to say much as to the details
of this division. Apparently some of the micronuclei shown in
Figs. 18 and 19 are in process of division, and the process appears
to be a typical mitosis. As to how a cyst in the condition of
that in Fig. 19, just referred to, reaches that shown in Fig. 20
I am unable to say. Cysts are exceedingly difficult to stain and
destain at this stage.

Figs. 20 and 21 show the final stages in the reorganization.
As may be seen there has been a great decrease in the quantity
of cytoplasm, and a corresponding decrease in chromatin.
Animals at this stage may be seen swimming about within the
cyst, but I have never been able to get them to leave the cyst,
despite washing, prolonged observation in cultures in which the
media was changed daily, and attempts to rupture the cyst wall
artificially. The latter always failed because of the tough
pliable nature of the wall. Both distilled water, tap water and
Ringer's solution were used for washing.

Moore (1924) found that Spathidium spathula behaved in a
similar fashion, though she was more successful in liberating
encysted individuals by puncturing the cyst wall. Under ordi-
nary circumstances few encysted animals would excyst, and

CONJUGATION, DIVISION, AND ENCYSTMENT. 423

drying, which Calkins (1919) found necessary to induce the
excystment of Uroleptus, and the addition of fresh medium,
found efficacious by Moody (1912) in securing the excystation of
Spathidium, both proved unsuccessful.

But the cytological changes which occur during the encyst-
ment of Spathidium spathula appear to be markedly different
from those which take place in Pleurotricha. There is no ex-
trusion of nuclear or cytoplasmic material from the cyst in Spath-
idium, but the macronuclei degenerate and finally disappear, the
fragments being absorbed by the cytoplasm. The micronuclei
remain practically unchanged. Shortly before redifferentiation
occurs, preparatory to leaving the cyst, new macronuclear anlagen
appear which strongly resemble those formed in the late stages
of conjugation. It is possible that more study would show that
the new macronuclei arise in a similar way in Pleurotricha, but
I have seen no indication of it in any of my preparations.

That an extensive nuclear reorganization takes place in many
infusoria during encystment has been known for a long time,
but I have found no mention of the extrusion of nuclear material
in any other species, although Prowazek (1899) believed that
such a phenomenon occurred during conjugation. Fermor (1913)
found that in Stylonychia there was a fusion of the two micro-
nuclei previous to the formation of a new nuclear apparatus,
and it is possible that the same thing occurs in encysted indi-
viduals of Pleurotricha. But in one instance I have actually
observed one of the two micronuclei being extruded along with
the macronuclei.

CONJUGATION.

During the eighteen months in which the cultures were main-
tained conjugation occurred but rarely, and epidemics of it were
never observed. Even in mass cultures containing thousands of
individuals it was usually necessary to look for an hour or more
to obtain a few pairs. Because of the small number of conjugants
no sections were made and all the results herein described were
obtained from whole mounts, fixed with Calkins' modification of
Schaudinn's fluid, stained with hematoxylin, and mounted indi-
vidually. Various methods of inducing conjugation have been

424

REGINALD D. MAXWELL.

suggested by investigators who have found them more or less
useful with certain species, and a number of these methods were
tried in this case, but with practically no success.

Conjugating individuals fuse by the adoral surfaces (Fig. 22),
the entire peristome of one member of the pair disappearing
completely. There seems to be no reason why it cannot function
in the other conjugant but there is very little evidence that it
does so, for food vacuoles become fewer in number as conjugation
progresses, and exconjugants are always small. The time from
fusion to final separation varies from eighteen hours to five days,
but the usual duration is twenty-four hours.

BEHAVIOR OF THE MICRONUCLEI.

The micronuclei normally go through three maturation di-
visions, only two taking part in each. The others degenerate
more or less rapidly, although some may persist even after the
interchange. The pronucleus usually undergoes two cleavage
divisions. Of the four products one enlarges and eventually
gives rise to the macronuclei of the reorganized exconjugant, one
degenerates, and the remaining two form micronuclei.

THE FIRST MATURATION DIVISION.

This division requires more time than any of the others — at
least eight hours — and is also strikingly different in type. The
micronuclei at first show no change but soon increase in size,
become surrounded by a clear space, or "halo," and stain more
faintly than usual. The chromatin takes on a finely-granular
appearance. Shortly an endosome, or division center, appears,
and in favorable preparations two may be seen (Fig. 28 A, C),
but I have never been able to detect an intradesmose connecting
them, although in vegetative division and in the other divisions
of conjugation it can often be seen. This division center in-
creases in size until it becomes hemispherical and stains very
heavily (Fig. 28 C, D). In the meantime spindle fibers begin
to appear, and the spindle takes on the typical parachute appear-
ance (Fig. 28 C, D, E), The chromatin in the expanded top of
the parachute, which is at first composed of very small, dimly
staining granules, condenses into larger granules which stain

CONJUGATION, DIVISION, AND ENCYSTMENT. 425

more heavily and these pair to form heavily staining chromo-
somes of dumbbell shape. These granules are altogether too
numerous to count, and in most preparations this is also true of
the chromosomes, but I have made counts in a few cases which
will be discussed at greater length below. The chromosomes
now move to the center of the spindle and divide longitudinally
(Fig. 28 F, G). Stages of the anaphase and telophase are shown
in Figs. 28 H to M. The most characteristic feature of this
division, aside from the dumbbell shape of the chromosomes,
is the way in which they lag and the peculiar curve at the poles
of the spindle.

THE SECOND MATURATION DIVISION.

This follows rapidly on the first, and is over in much less
time, with the incidental result that material showing it is
difficult to obtain. It is also of very different character. The
micronuclei which are to divide enter the prophase almost
before the telophase of the preceding division is complete (Fig.
32 A). At this stage they stain faintly, and rows of granules
appear which are arranged in a more or less "whorled" fashion
(Fig. 32 B). Fig. 32 C represents a stage which is seldom seen,
but it is presumably earlier than those shown in the two preceding
figures. The nucleus next enlarges and the rows of granules
become threads (Fig. 32 D). This may well be in reality a
leptotene stage, since reduction occurs during this division.
There is soon evidence of a definite spindle which is at first of a
peculiar oval shape (Fig. 32 E, F). I have been unable to get
preparations showing the metaphase and later stages of the
anaphase, but the early anaphase is shown in Fig. 32 F. I have
made a number of counts of the number of chromosomes con-
cerned in this division, the average result being forty. Two
stages of the telophase are shown in Fig. 32 G and //.

THE THIRD MATURATION DIVISION.

The third division differs in type from both the preceding, and
requires more time than the one just described. The micro-
nuclei which are to divide enlarge, and become very finely
granular. I am uncertain as to whether there is an intradesmose

426 REGINALD D. MANWELL.

connecting the products of the division of the intranuclear
endosome, but I believe that there is. What is apparently one
end of it can be seen in Fig. 36 A. These granules now condense
to form rows, and the latter in turn become much coiled chromo-
somes which often appear double (Fig. 36 B, C). The meta-
phase is shown in Fig. 36 D, and is considerably like that of
ordinary vegetative division. The chromosomes in both the
second and third maturation divisions are rod-shaped, in con-
trast to the dumbbell shape which they have in the first, and the
V and irregular rod shape of the cleavage divisions. The long-
pointed anaphases are characteristic of the third division, and
long drawn-out telophase (Figs. 36 E to /).

The interchange is a rapid process, and the wandering nuclei
do not appear to differ in any respect from the stationary ones.
New adoral zones appear at this time or very early during the
first cleavage division (Fig. 37).

THE FIRST CLEAVAGE DIVISION.

The amphinucleus is shown in Fig. 40, and is apparently
divided into two parts, doubtless representing the two pronuclei.
The stages in the first cleavage division are shown in Figs.
41 A to H. The nucleus, which increases very much in size
shortly after fusion, becomes smaller, and forms a peculiar sort
of spindle (Fig. 41 A), in which the chromatin appears to be
condensed into a heavy ribbon, twisted more or less upon itself,
and quite definitely double. Chromosomes soon appear which
are arranged in bouquet fashion, about a small endosome, and
which at this stage seem to be, in some cases at least, in pairs
(Fig. 41 -B). They then straighten out and form a characteristic
spindle which has at first two definite parts, possibly representing
the two pronuclei (Fig. 41 Q. At this time the chromosomes
are definitely double, and apparently twisted about each other.
The metaphase is figured in 41 D, and the anaphase soon follows.
Apparently the chromosomes in the later stages of this division
are shaped like very acute Vs. Figs. 41 E to H show the anaphase
and telophase. The former is rather characteristic in its early
stage because the chromosomes are so widely separated within
the receding plates.

CONJUGATION, DIVISION, AND ENCVST.MENT. 427

THE SECOND CLEAVAGE DIVISION.

The first cleavage division which has just been described is a
rapid process, but the second one is much slower. A division
center appears, as in two of the three maturation divisions, and
divides, the products remaining connected by an intradesmose
(Fig. 46 ^4). The nucleus then becomes elliptical, and the
chromatin becomes arranged in long deeply-staining strands
which are at first quite regular in arrangement (Fig. 46 B and C).
These elongate and become twisted (Fig. 46 D, £), some of them
appearing double. The whole spindle resembles nothing so
much as a tangled skein of yarn at this stage, and possibly the
stage represented in Fig. 46 D is in reality a spireme. The
strands in the following figure are definitely polarized and
probably are really chromosomes. The metaphase is shown in
Fig. 46 H. It does not differ very much from that in the first
cleavage division. The anaphase is very peculiar, as can be
seen from Figs. 46 J to M. At the poles of the spindle a "cap"
of very deeply-staining chromatin is formed and the chromosomes
lose the definite shape which they had in the metaphase, ap-
parently coalescing into more or less irregular masses which
take a very heavy stain. These are very characteristic and
make the later stages of the second cleavage division differ from
all the others. The strands persist for some time in the anaphase
(Fig. 46 L, M).

THE OCCASIONAL THIRD CLEAVAGE DIVISION.

Following the second division there may in some cases be
another, but when it occurs only two of the four nuclei take
part as a rule, so that exconjugants in which this third division
has taken place have six nuclei, rarely eight. All the stages of
this division which I have observed are like corresponding ones
of the second division.

THE BEHAVIOR OF THE MACRONUCLEI.

The anterior of the two macronuclei elongates during the
maturation divisions, and may occasionally divide by a process
of mass division, but only in very rare cases. At the same time
it undergoes a slow fragmentation, so that the cytoplasm usually

428 REGINALD D. MAXWELL.

has a number of heavily staining masses of chromatin at this
time, which often simulate degenerating micronuclei derived
from previous divisions. The posterior macronucleus may elon-
gate slightly, but I have never observed it to divide and it
undergoes much less fragmentation than the other.

The chromatin fragments and degenerating micronuclei usually
disappear about the time of the interchange, or soon afterward,
but may in some cases persist until after the cleavage divisions.

REORGANIZATION OF THE EXCONJUGANT.

Reorganization begins before separation, and usually requires
several days. One of the four nuclei resulting from the second
cleavage division becomes very large and coarsely granular, and
at the same time loses much of its capacity to stain with hemo-
toxylin. It is destined to form the macronuclei of the re-
organized exconjugant and is shown in Figs. 47, 48, 49, 50,
52 and 53. This nucleus divides at the first division of the
exconjugant, and then divides again without corresponding
division of its possessor, thus restoring the normal macronuclear
condition (Fig. 55).

But it is apparently possible for reorganization to become
complete without prior division of the exconjugant, and an
individual in which this has happened is shown in Fig. 54. It
may be that this can happen in individuals in which there has
been a third cleavage division.

The micronuclei of the reorganized conjugant are derived
directly from two of the three which remain from the second
cleavage division. The third persists for a time but eventually
degenerates. As to what happens when there have been three
cleavage divisions I am unable to say, since I have been unable
to observe enough cases.

The remains of the old macronuclei persist for a day or two as
more or less circular, deeply staining and vesicular masses of
chromatin (Figs. 49 to 52), but they eventually disappear
completely, leaving no trace (Fig. 53).

THE NUMBER OF CHROMOSOMES.

In order to determine the number of chromosomes a large
number of counts were made in several different stages, par-

CONJUGATION, DIVISION, AND KNCY.STMKKT. 429

ticularly from an especially favorable preparation showing the
early anaphase of the first maturation division, and from several
preparations of the anaphase of the third maturation division.

Since the total number of chromosomes is so large in the
anaphase of the first maturation division accurate counts are
difficult, but ten were made as a check for the other counts
described below. The mean of these ten counts was eighty-six,
which would indicate that the diploid number is forty-three.
The standard deviation was found to be 15.4, the coefficient of
variation 17.9, and the probable error ± 3.46. There is no
doubt therefore that the counts are very significant.

A few counts were made of the chromosomes in the early
anaphase of the second maturation division, and the average
was close to forty, again indicating therefore that this figure is
close to the diploid number.

Since the most favorable preparations for chromosome counting
were those of the third maturation division these were given the
most study. To insure results as free from error as possible
seventy counts were made from several preparations showing
the anaphase of this division. A curve was constructed of these
counts and the mode found to be 19. The mean was 19.62, and
the probable error ± .295. The coefficient of variation was
18.6 per cent., and the standard deviation 3.66. It is therefore
certain that the counts are highly significant, and there is no
doubt that the haploid number of chromosomes is close to
twenty, with a high degree of probability that it is exactly that
number, making the diploid number forty.

Since the number of chromosomes in Oxytrichafallax is twenty-
four (Gregory, 1923), there is therefore no obvious relation as
far as number of chromosomes is concerned between this species
and Pleurotricha lanceolata.

DISCUSSION.

Although only a relatively small number of ciliates have been
studied with reference to the phenomena of conjugation the
process appears to be essentially the same in all. From his own
observations Maupas (1888) divided it into eight phases — A, the
period of preparation preceding the first meiotic division; B,

43O REGINALD D. MANWELL.

the first division; C, the second division; D, the third division;
E, the interchange and fusion of the pronuclei; F, the first
cleavage division; G, the second cleavage division; and H, the
period of reorganization preceding the first fission of the ex-
con jugant.

These phases have been shown to hold for all infusoria so far
studied, including Pleurotricha lanceolata which is the subject of
this paper, although in a few forms such as the Vorticellidse,
Ophryoscolecida?, and Euplotes patella, there are one or more
preliminary divisions before the meiotic divisions begin. (In
the case of the Vorticellidee this happens only in the case of the
microgamete).

Ciliates as far as at present known fall into two classes
according to the behavior of the micronuclei in the first phase—
those which undergo a prophase like Paramcecium, the micro-
nucleus being drawn out into a crescent, and those in which a
parachute or candelabra-like figure is formed, to use Calkins'
term. To the last group belong Onychodromus grandis (Maupas,
1888), Burs aria truncatella (Prowazek, 1899), Didinium nasutum
(Prandtl, 1906), Anoplophrya branchiarum (Collin, 1909), Uro-
leptus mobilis (Calkins, 1919), Oxytricha fall ax (Gregory, 1923)
and Chilodon uncinatus (MacDougall, 1925). The character of
the first division, already described, makes it necessary to add
Pleurotricha lanceolata to this list.

As might perhaps be expected from the close morphological
resemblance of Oxytricha fallax and Pleurotricha lanceolata the
details of the conjugation process are much alike, and while
they are also similar to those which occur in Uroleptus mobilis
as described by Calkins, the resemblance is much less striking.

The formation of the parachute preliminary to the first
maturation division more nearly resembles the corresponding
stages in Uroleptus than in Oxytricha, but differs from both.
In the latter, although the parachute fibers are focused on a
single granule derived by division from the endosome just as
they are in Pleurotricha, there is also a centrodesmose connecting
the two halves of the endosome which I have never been able to
observe in Pleurotricha. In Oxytricha there appears to be no
endosome, and the place of the basal granule is taken by a row
of granules.

CONJUGATION, DIVISION, AND ENCYSTMENT. 43!

The chromosomes in this division appear to be formed by the
fusion of the granules into which the chromatin is divided early
in the prophase, but whether the number of these granules bears
any constant relation to the number of chromosomes, as Gregory
believes it does in the case of Oxytricha, I am unable to determine.
These chromosomes which at first make up the top of the para-
chute appear to move down the fibers until they reach the
center of the spindle when they form an equatorial plate. The
other pole of the spindle appears to be formed by one of the two
halves of the endosome which remains in the top of the parachute,
the other as already stated, forming the granule at its base. This
is apparently the same as in Uroleptus.

The number of nuclei taking part in any of the three maturation
divisions is apparently never greater than two in the case of
Pleurotricha, although forms in which there are three micro-
nuclei occur. Of the four products of the first division all but
two degenerate, and the same thing happens after each of the
other two maturation divisions. The number of nuclei taking
part in each maturation division in Oxytricha appears to be
variable, although there are always at least two. In other
multinucleate ciliates which have been studied there is con-
siderable variation in this respect. In Bursaria all the sixteen
or more nuclei may take part in the first meiotic division
(Prowazek, 1899), and in Uroleptus there may be two, three or
four primary spindles (Calkins, 1919), but these are exceptions.

The second maturation division is apparently of different
character from the first in most ciliates, and lacks the elaborate
preliminaries of the latter, with the result that it requires much
less time. The micronuclei in most cases do not return to the
resting stage between the first and second division, although
there are exceptions (e.g., Chilodon, MacDougall, 1925). In all
these respects therefore Pleurotricha lanceolata is typical.

In nearly all cases in which a careful study has been made
reduction has been found to take place in the second meiotic
division, although the first division was thought to be the reducing
division in Paramcecium caudatum (Calkins and Cull, 1907), and
in Oxytricha fallax (Gregory, 1923). But Dehorne (1920) regards

the third division as the one in which reduction occurs in the
29

432 REGINALD D. MANWELL.

former form. In the case of Pleurotricha there is no doubt but
that it occurs in the second, for the following reasons: The total
number of chromosomes taking part in the first division in the
beginning anaphase has been determined to be in the neighbor-
hood of eighty, as already stated, whereas there are only half as
many concerned in the second division, and the average number
moving toward each pole of the third maturation spindle has
been determined, as stated above, to be about twenty.

The only protozoa in which reduction is known to occur at
any other time (with the two exceptions already mentioned),
are Diplocystis (Jamieson, 1920), and Aggregata eberthi (Dobell,
1915). It is suggested by Dobell and Jamieson that this may
be a universal occurrence among the Telosporidia.

The third maturation division and the interchange take place
exactly as in Oxytricha, and are very much alike, even to details
in the form of the spindle. Nothing is said as to the existence
of a division center or intradesmose however. The dumbbell
character of the chromosomes is not lost until the third division
in the case of Oxytricha, and it is stated that the chromatin
streams toward the poles in the form of granules. I find that
the chromosomes are definitely rod-shaped in Pleurotricha, in
both the second and third divisions, although the rods are
heavier and shorter in the former, and may give place to rows of
granules in the late anaphase stages of the latter. The dumbbell
character is never regained after the first of the three maturation
divisions.

The two cleavage divisions follow the usual course, again
resembling closely those in Oxytricha, although not many details
are given in Gregory's paper. The products of the second
cleavage division are equal as in Oxytricha, but differ in this
respect from the state of things in Uroleptus mobilis, in which
one of the two nuclei derived from the first division gives rise
to the macronuclei and the other to the micronuclei of the
reorganized exconjugant (Calkins, 1919). I have found no other
case in which the anaphase of the second cleavage division
resembled that in Pleurotricha however, and it differs strikingly
from that in Oxytricha.

The occasional third cleavage division is also a striking charac-

CONJUGATION, DIVISION, AND KXC YSTMKXT. 433

teristic, and as far as I can discover it occurs in no other hypotrich
so far described, although it is admitted that the number studied
is thus far small. In other forms such as Paramoecium candatnm
(Calkins and Cull, 1907), and several of the Vorticellida-, a
third division of the amphinucleus occurs, but as a regular thing.
In Bursaria (Prowazek, 1899), there are four divisions before
differentiation occurs.

The behavior of the macronuclei in Pleurotricha follows the
usual rule — degeneration eventually occurs, although it is not
completed until after separation. But division of the macro-
nuclei, although apparently the rule in Oxytricha, is very rare in
Pleurotricha.

The time required for reorganization and the details of the
process differ considerably in different cases, but the process as
it occurs in Pleurotricha is not significantly different from the
essential features of reorganization in other species. Prowazek
believed that the remnants of the old macronuclei must be
extruded bodily from the cell in Stylonychia pnstnlata and
Bursaria truncateUa (1899), but I have seen no evidence of this
in the present instance, nor has it ever been observed by anyone
else. He based his hypothesis on the fact that nucleins are
indigestible.

THE PHYSIOLOGICAL EFFECT OF CONJUGATION.

The significance of conjugation has been much disputed, one
school holding that it is the necessary result of senescence in
any given race, and that by it the death of the race could be
averted and rejuvenation, expressed chiefly in an acceleration of
the division rate, be secured. Maupas held this view and since
his time Calkins has been its leading exponent. Another group
has regarded conjugation as a process useful chiefly as a means of
producing variations, which are much greater and much more
frequent in exconjugants than in ordinary vegetatively repro-
ducing individuals. Jennings has been one of the chief ex-
pounders of this view.

Many experiments have been tried to settle this question and
since the results have differed considerably for different species
I have endeavored to find out the effect of conjugation in Pleura-

434 REGINALD D. MANWELL.

triclia lanceolata, as measured by the division rate of exconjugants
when compared to closely related non-conjugating lines.

The experiment consisted in isolating one hundred excon-
jugants, and following them daily as long as possible. Of this
number 92 per cent, died without division, usually about four
days after separation, and of the remaining eight only one gave
any evidence whatever of rejuvenation. The others divided
two or three times and then all died. This single individual
gave rise to a line which divided between two and three times
a day for six days, and from then until the death of the line four
weeks later, the average daily division rate gradually declined
until during the last two weeks it was much less than one per day.

The controls during this period averaged from one to two
divisions a day, and showed -no evidence of senescence for the
six months for which pedigreed cultures were maintained. Both
controls and exconjugants were in the same media.

It is realized that one hundred cases is scarcely enough to
draw final conclusions from, and yet the percentage of deaths is
so high that it does not seem likely that a greater number of
cases would have given significantly different results.

Jennings (1913) in a long and comprehensive series of experi-
ments demonstrated that with Paramcecium "Conjugation de-
creases the rate of fission, causes a great increase in variation in
fission rate, brings about many abnormalities, and greatly
increases the death rate," but Mast (1917), using Didinium
nasutum, was unable to secure evidence of any such effect. He
found that there was no appreciable effect on either death rate,
fission rate, or increase in variation of fission rate, thus proving
however that there was also no rejuvenation.

Calkins, in a long series of experiments with Uroleptus mobilis,
has shown that in this species at least conjugation appears to
have a genuine rejuvenating effect (Calkins, 1919 and 1926),
and Woodruff and Spencer (1924) conclude that conjugation has
a similar effect in the case of Spathidium spathula, although they
prefer the term "high survival value" to that of "rejuvenation."

It has been shown by many investigators working with various
species that as culture methods are improved the longevity -of
cultures without conjugation can often be indefinitely increased,

CONJUGATION, DIVISION, AND ENCYSTMENT. 435

so that it can probably be safely concluded that conjugation is
at least not a necessary process if environmental conditions are
sufficiently favorable.

Since cultures of Pleurotricha have been maintained for
eighteen months with very little conjugation (none in the
pedigreed lines which were carried for six months), and since in
the cases mentioned above none of the exconjugants gave rise
to lines which continued for any length of time, although kept
under identically the same conditions as controls which main-
tained a uniform and fairly high division rate, it can be concluded
I think that conjugation is not an indispensable part of the life
cycle. This conclusion is supported by Baitsell (1914) who
carried on experiments of this nature with both Pleurotricha
lanceolata and Oxytricha fallax.

It is nevertheless apparent that a process which is so universal
among infusoria as conjugation must serve some useful purpose,
and that in nature exconjugants do not always die. Why then
do they die in cultures? It is impossible to give a definite
answer, but it has been suggested that media which is suitable
for active vegetative multiplication may not always be suitable
for conjugation and exconjugants, and this may well be the

explanation.

SUMMARY.

1. Pleurotricha lanceolata is a hypotrichous ciliate belonging to
the family Pleurotrichidae. The species has as its chief charac-
teristic six anal cirri, divided into two groups. The anterior
of the two groups consists of four cirri — three very large ones
arranged in an oblique row, and a smaller one, a little forward
of the posterior end of the row. The second group includes two
cirri, both of them very long and projecting well beyond the
posterior end of the animal. The usual size of vegetative
individuals is 140 /JL.

2. The process of division does not differ particularly from
that in other infusoria, except that in the very early stages an
endosome appears and divides, the products remaining connected
by an intradesmose. This has been described in Paramcecium
trichium but in few if any other species. The process of division
is initiated by the appearance of a rudimentary adoral zone and

436 REGINALD D. MAXWELL.

a kernspalt in each of the macronuclei. The micronuclei divide
by typical mesomitosis. When division of the latter is virtually
complete the macronuclei follow suit, dividing amitotically, and
then the cell itself divides. All organelles appear to be re-
generated, the old ones being absorbed.

3. Encystment may occur at any time, and appears to bear no
relation to periods of depression, to division or to conjugation.
The old macronuclei are extruded bodily from the cell, and a
single micronucleus remains. It is uncertain whether the other
is always extruded with the macronuclei or fuses with the first
micronucleus. An uncertain number of micronuclei are formed
from the single remaining micronucleus, and from these the
normal nuclear complex is rebuilt, the process being complete
at the time the animal is ready to leave the cyst.

4. The nuclear changes which occur during conjugation are
essentially the same as those described for other ciliates. There
are three maturation divisions, an interchange of pronuclei, and
two or rarely three cleavage divisions. The four products of
the second division are at first alike, but one soon enlarges and
eventually gives rise to the new macronuclei of the reorganized
exconjugant. One of the others degenerates and the other two
form the new micronuclei. The old macronuclei degenerate after
separation of the exconjugants. There may or may not be one
division of the latter before reorganization is completed.

5. Reduction occurs in the second maturation division. The
diploid number of chromosomes is forty, as nearly as can be
determined, and the haploid number twenty. The chromosomes
are dumbbell-shaped in the first maturation division, and rod-
shaped in the second and third. They are also rod-shaped in
the cleavage divisions, but the shape then is unlike that in the
maturation divisions or in vegetative division.

6. Each of the divisions in conjugation differs from all the
rest and from vegetative division.

7. Conjugation occurs but rarely in the race of Pleurotricha
used, and under the conditions of culture virtually always
results in the death of the conjugants a few days after separation.
In only one instance out of a hundred did it result in anything
like rejuvenation, and even in this case the daughter race died
within a month.

CONJUGATION, DIVISION, AND KNCYSTMKNT.

437

8. Although the details of conjugation are much alike in
Oxytricha fallax and Pleurotricha lanceolata there is no obvious
relation between the number of chromosomes of these two very
similar species.

CONCLUSION.

The cytology of conjugation and division in Pleurotricha
lanceolata has been described in detail, together with some of
the cytological phenomena of encystment. The processes of
division do not differ particularly from those which are known
to occur among ciliates in general, with the exception of
the dividing endosome and connecting intradesmose. Certain
features in the division of the micronucleus are strikingly like
some which occur in the mitosis of metazoan tissue cells.

A remarkable feature of encystment is the extrusion of the
old macronuclei, and perhaps one of the micronuclei, through
the cyst wall. After this has happened the normal nuclear
complex is rebuilt from the single micronucleus remaining.
Whether the presence of the latter is regularly due to the extru-
sion of the other micronucleus from the cell, or to the fusion ot
the two original micronuclei is not altogether certain.

The cytological changes of conjugation are in general similar
to those previously described in other ciliates, and especially
resemble those which occur in Oxytricha fallax, but there are
important differences in detail. Reduction is shown to occur in
the second maturation division. The diploid number of chromo-
somes is too great to determine exactly, but is probably forty.
There may be a third cleavage division in addition to the two
which regularly occur, and in this respect Pleurotricha lanceolata
differs from any other hypotrich so far described.

Conjugation appears to be not only an unnecessary part of
the life cycle of this species, at least as long as environmental
conditions remain favorable, but is a very dangerous event, for
92 per cent, of one hundred exconjugants died without further
division, and only I per cent, showed any indication of an
accelerated fission rate.

438 REGINALD D. MANWELL.

BIBLIOGRAPHY.

Baitsell, G. A.

'14 Experiments on the Reproduction of the Hypotrichous Infusoria. II. A
Study of the So-called Life Cycle of Oxytricha fallax and Plenrotricha
lanceolata. Jour. Exp. Zool., Vol. 13, pp. 211-234.

Calkins, G. N.

Uroleptus mobilis, Engl. I. — History of the Nuclei during Division and
Conjugation. Jour. Exp. Zool., Vol. 27, pp. 293-357.
Uroleptus mobilis, Engl. II. — Renewal of Vitality through Conjugation.
Jour. Exp. Zool., Vol. 29, pp. 121-156.
'26 Biology of the Protozoa. Philadelphia. 623 pp.
Calkins, G. N., and Cull, S. W.

'07 Conjugation of Paramecium aurelia (caudatum). Arch. f. Protistk., Vol. 10,

pp. 375-415.
Cienkowsky, L.

'55 Uber cystenbildung bei infusorien. Zeitschr. f. wiss. Zool., Vol. 6, pp.

301-306.
Collin, B.

'09 La conjugaison d'Anoplophrya branchiarum St. Arch. d. Zool. Exp. et

Gen. — 5th Series — T. I, pp. 345~388.
Dehorne, A.

'20 Contribution a 1'etude comparee de 1'appareil nucleaire des Infusoires
cilies, des Euglenes et des Cyanophysees. Arch. Zool. Exp. et Gen.,
Vol. 60, pp. 47-176.
Ehrenberg.

'32 Die infusionsthierchens. S. 373, Taf. XLII.
Fermor, X.

'13 Die Bedeutung der Encystierung bei Stylonychia pustulata Ehr. Zool.

Anz., Bd. 42, pp. 380-384.
Goodey, T.

'23 The Excystation of Colpoda cucullus from its Resting Cysts and the Nature
and Properties of its Cyst Membranes. Proc. Roy. Soc., Vol. 86, pp.

427-439-
Gregory, Louise H.

'23 The Conjugation of Oxytricha fallax. Jour. Exp. Morph., Vol. 37, No. 3,

PP- 555-581.
Jamieson, A. Pringle.

'20 The Chromosome Cycle in Gregarines, with Special Reference to Diplo-

cystis schneideri. Quart. Jour. Mic. Sci., 64: 2, pp. 207-266.
Jennings, H. S.

'13 The Effect of Conjugation in Paramecium. Jour. Exp. Zool., 14- 279-391.
'20 Life, Death, Heredity and Evolution in the Protozoa. Boston. 233 pp.
Kent.

'8o-'82 Manual of the Infusoria. London. 2 vols. 913 pp. 51 plates.
MacDougall, M. S.

'25 Cytological Observations on Gymnostomatous Ciliates, with a Description
of the Maturation Phenomena in Diploid and Tetraploid Forms of
Chilodon uncinatus. Quart. Jour. Mic. Sci., 60: Pt. 3, pp. 361-384.

CONJUGATION, DIVISION, AND ENCYSTMKNT. 439

Mast, S. O.

'17 Conjugation and Encystment in Didinium nasutum with Especial Reference

to their Significance. Jour. Exp. Zool., Vol. 23, No. 2, pp. 335-359-
Maupas, £.

'88 Recherches experimentales sur la multiplication des Infusoires cilies.

Arch. d. Zool. Exp. et Gent.. Vol. 6, pp. 165-277.
'89 Le rajeunissement karyogamique chez les cilies. Arch. d. Zool. Exp. et

Gen., 2d Series, T. 7, pp. 149-517.
Moody, J. E.

'12 Observations on the Life History of Two Rare Ciliates, Spalhidium spathula

and Actinobolns radians. Jour. Morph., Vol. 23, pp. 349~39O-
Moore, E. Lucile.

'24 Endomixis and Encystment in Spathidium spalhula. Jour. Exp. Zool.,

Vol. 39, No. 2, pp. 3I7-337-
Prandtl, H.

'06 Die konjugation von Didinium nasutum O.F.M. Arch. f. Protist., Bd. 7,

pp. 229-258.
Prowazek, S.

'99 Protozoenstudien (Bursaria truncalella und Stylonychia pustidata). Arb.

a. d. Zool. Inst. Univers. Wien., Bd. n, pp. 195-268.
Stein, F. R.

'59 Der organismus der infusionsthiere. Leipzig- Abdruck i, 206 pp.
Sterki, V.

'78 Beitrag zur Morphologic der Oxytrichen. Zeit. f. wiss. Zool., Bd. 31.

pp. 29-58.
Stokes, A. C.

'87 Ann. and Mag. Nat. Hist., Aug. 1887, pp. iio-m.
Wallengren, H.

'oo Zur Kenntnis der vergleichende morphologic der hypotrichen infusorien

Bih. T, k. Svenska vet. Akad. Handl., Vol. 26, 31 pp.
Wenrich, D. H.

'26 The Structure and Division of Paramecium trichium Stokes. Jour. Morph.

and Phys., Vol. 43, No. I, pp. 81-102.
Woodruff, L. L., and Spencer, H.

'24 Studies on Spathidium spathula. II. The Significance of Conjugation.
Jour. Exp. Zool., Vol. 39, No. 2, pp. 133-196-

4/jO REGINALD D. MAXWELL.

EXPLANATION OF THE FIGURES.
Magnification (unless otherwise stated) X 550; spindles X 1700.

PLATE I.

Early Stages of Division.

FIG. i. A typical vegetative individual. Occasionally individuals are found
with only one micronucleus, or with three.

FIG. 2. The first stage in division. A kernspalt appears at the outer end of
each macronucleus, and the beginning of a new adoral zone can be made out (a. z.).

FIG. 3. The micronuclei show an endosome, or division center, in process of
division, the halves being connected by an intradesmose. The kernspalt has
moved nearer the center of each macronucleus

FIG. 4. The micronuclei have become somewhat larger, and the macronuclei
have become almost round. The kernspalt is almost ready to pass off the latter
entirely, and the new adoral zone is more conspicuous.

FIG. 4 A. An elongated, faintly staining micronucleus; seen in very early
division.

FIG. 4 B. Same as above, but a little later.

FIG. 5. The chromatin in the micronuclei has condensed into definite threads.
The macronuclei have become almost completely spherical and the new adoral
zone is well developed. New cirri are making their appearance in the neighborhood
of both adoral zones. Enlarged figures of the micronuclei are given under 5 A
and B.

FIG. 6. The anterior micronucleus, shown enlarged in Fig. 6 A, has formed a
definite spindle.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE I.

SB

(spindles -I7oo)

XS50

REGINALD D. MANWELL, M.A.

442 REGINALD D. MANWELL.

PLATE II.

Later Stages of Division.

FIG. 7. Both micronuclei have formed spindles, which are shown enlarged in
7 A and 7 B. The cirri are larger, and the macronuclei still almost round.

FIG. 7 C. A spindle, magnified about 3500 times, showing the shape of the
chromosomes, and the way in which they divide. The heavy strands connecting
them are to be noted particularly.

FIG. 8. An individual with but one micronucleus, which is in the metaphase.
The macronucleus have elongated, and the new cirri are conspicuous and large.

FIG. 9. An individual in which the micronuclei are in the anaphase.

FIG. 9 C. A later stage in the anaphase.

FIG. 10. A still later anaphase.

BIOLOGICAL BULLETIN, VOL, LIV.

PLATE II.

7A

Spindles- X17OO

organisms-*

REGINALD D. MANWELL. M,A.

30

444 REGINALD D. MANWELL.

PLATE III.

Final Stages of Division; Cysts.

FIG. ii. The macronuclei are beginning to divide, and some of the old cirri
have apparently disappeared. The new ones, originally formed near the adoral
zones, are approaching their normal positions.

FIG. ii A. A late stage in the telophase, enlarged from Fig. n.

FIG. 1 1 B. An earlier telophase.

FIG. 12. The final stage of division.

FIG. 13. An individual which has just divided. The shape differs somewhat
from that of a typical vegetative individual, and the anal cirri are not yet quite
in their normal positions.

FIG. 14. An early stage of encystment, in which the individual has rounded up,
but some of the cirri still remain, and the contractile vacuole is still functional.
X 750.

FIG. 15. An encysted animal, in which the cyst wall has just been secreted.
X 750.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE III.

11 —

X550

REGINALD D. MANWELL, M.A.

446 REGINALD D. MANWELL.

PLATE IV.
Cysts in Various Stages.

FIG. 16. Showing the extrusion of the macronuclei and a small portion of the
cytoplasm.

FIG. 17. An encysted animal in which only one micronucleus remains. X 75°.

FIG. 1 8. A cyst containing three nuclei and some darkly staining material
which is however not chromatin. X 75O.

FIG. 19. A cyst containing seven nuclei, several of which appear to be in
mitosis. X 750.

FIG. 20. A cyst in which reorganization is almost complete. X 75°.

FIG. 21. Cyst containing a completely reorganized animal. X 75°.

BIOLOGICAL BULLETIN, VOL. LIV

PLATE IV.

X
20

REGINALD D. MANWELL, M.A

448 REGINALD D. MAXWELL.

PLATE V.

First Maturation Division.

FIG. 22. Initial stage in conjugation, showing the manner of fusion.

FIG. 23. A pair in which the micronuclei have begun to enlarge, preparatory
to the first maturation division.

FIG. 24. Here the division centers have made their appearance, and in one
nucleus the spindle fibers are becoming visible. Enlarged in 28 B and 28 C.

FIG. 25. Parachute stage, just prior to the formation of chromosomes. Shown
enlarged in Fig. 28 E.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE V.

REGINALD O. MANWELL, M.A.

450 REGINALD D. MAXWELL.

PLATE VI.
First Maturation Division.

FIG. 26. Here the micronuclei have entered the anaphase. Two of the
spindles are shown enlarged in Figs. 28 H and 28 K.

FIG. 27. A pair in which a telophase may be seen in each member.

FIG. 28 A-M. Various stages of the first maturation division arranged con-
secutively.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE VI.

REGINALD O. MANWELL, M.A.

31

452 REGIXALD D. MANWELL.

PLATE VII.
Second Maturation Division.

FIG. 29. The left-hand member of the pair shows the final stage in the telophase
of the second maturation division. On the right there is a prophase and an early
anaphase.

FIG. 30. Each individual of this pair shows a telophase of the second matu-
ration division, and a prophase of the third division is also to be seen in the left-
hand member.

FIG. 31. Two prophases of the second maturation division are visible in the
left-hand member and a prophase of the third division may be seen on the right.

FIG. 32 A-H. Various stages in the second maturation division. "A "shows
both products of the first maturation division, one of which is already in the
prophase of the following division. "C" is also a prophase. and should probably
precede "A."

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE VII.

31

REGINALD D. MANWELL, M.A.

454 REGINALD D. MAXWELL.

PLATE VIII.
Third Maturation Division.

FIG. 33. A pair undergoing the third maturation division. Two anaphases
appear on the left and a prophase and anaphase on the right.

FIG. 34. Each member of the pair shows a metaphase, and in addition there
is an anaphase in the left-hand member and a prophase in the one on the right.

FIG. 35. A metaphase and anaphase appear on the left, and a metaphase and
telophase on the right.

FIG. 36 A-I. Stages of the third division, arranged consecutively.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE VIII.

36/Y

367

REGINALD D. MANWELL, M.A.

456 REGINALD D. MAXWELL.

PLATE IX.
Interchange and First Cleavage.

FIG. 37. The interchange. The wandering nuclei have already migrated but
have not yet fused. Shown as large and relatively faintly staining bodies. New
adoral zones are beginning to form.

FIG. 38. The left-hand member of this pair shows the fusion nucleus.

FIG. 39. In the left-hand individual the pronucleus has not yet entered the
prophase of the first cleavage division, but division in the right-hand member is
well advanced.

FIG. 40. The pronucleus of Fig. 38, enlarged X 1700.

FIG. 41 A-H. Stages of the first cleavage division, arranged consecutively.

BIOLOGICAL BULLETIN, VOL. LIV.

REGINALD D. MANWELL, M.A.

458 REGINALD D. MANWELL.

PLATE X.

The Second Cleavage Division.

FIG. 42. Shows several prophases.

FIG. 43. In this pair a prophase and anaphase are visible in the left-hand
individual, and a prophase and telophase in the right.

FIG. 44. A prophase and telophase may be seen in each of the members of this
pair, and also several degenerating micronuclei which still persist from the matu-
ration divisions.

FIG. 45. The cleavage divisions are complete in the left-hand individual but
a prophase of the second division is still to be seen in the other member.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE X.

':''~'v:'i.:":><:''' ''\'^'": ••'f ^£r ' :'. ':: ':*. •<*•«- ^'--A

REGINALD D. MANWELL, M.A.

32

460 REGINALD D. MANWELL.

PLATE XI.
The Second Cleavage Division and Early Stages of Reorganization.

FIG. 46 A-M. Stages of the second cleavage division, arranged consecutively.

FIG. 47. Shortly after the conclusion of the last cleavage division. Each
individual contains several micronuclei, one of which is increasing in size and will
give rise to the macronuclei of the reorganized exconjugant.

FIG. 48. The stage is similar to that in the preceding figure, except that the
old macronuclei are becoming vesicular and are beginning to stain much more
heavily than normally.

FIG. 49. A pair more completely fused than usual, and with the distribution
of micronuclei irregular. The macronuclei are noticeably degenerating.

BIOLOGICAL BULLETIN, VOL. LIV.

REGINALD D. MANWELL, M.A.

462 REGINALD D. MAN WELL.

PLATE XII.
Reorganization.

FIG. 50. A pair the members of which are about to separate. The fourth of
the nuclei arising from the last cleavage has apparently degenerated in each
individual, and the macronuclei have taken on the typical circular and vacuolated
appearance which persists until their final degeneration. ,

FIG. 51. Exconjugant shortly after separation. Typical of those individuals
in which a third cleavage division has occurred.

FIG. 52. Exconjugant somewhat longer after separation than the one shown
in Fig. 51.

FIG. 53. Exconjugant in somewhat more advanced stage than the preceding
The remnants of the old macronuclei have finally disappeared.

FIG. 54. Reorganization in this individual has apparently been completed
without preliminary division.

FIG. 55. One of the daughter individuals produced by the first division of
an exconjugant. The macronucleus is in process of division by which the normal
nuclear constitution will be finally restored.

BIOLOGICAL BULLETIN, VOL. LIV.

PLATE XII.

REGINALD D. MANWELL, M.A.

Vol. LIV June, 1928 No. 6

BIOLOGICAL BULLETIN

NODES AND CHI ASM AS IN THE BIVALENTS OF

L/L/r.17 WITH REGARD TO SEGMENTAL

INTERCHANGE.

JOHN BELLING,

•

CARNEGIE INSTITUTION OF WASHINGTON, DEPARTMENT OF GENETICS. COLU

SPRING HARBOR, N. Y.

The bivalents of several species of Lilinin, during the interval
between the double-thread stage and the metaphase, have already
been examined by a number of good observers. Hence for fur-
ther fruitful work three things seem more or less necessary: a
narrowly limited aim ; perfect fixation, which is probably best
attained by fixing the pollen-mother-cells outside the anther; and
accurate microscopy, which will doubtless include the use of a
small circular diaphragm on the source of light, a corrected im-
mersion condenser, and appropriate light-filters.

To attain uniform terminology, the writer ventures to use in
this paper the terms, diaphase, instead of diakinesis; diplophase,
for diplotene stage ; etc.

Bulbs of Lilinin longiflbrum (clonal variety, Easter lily) were
potted late in the year, in a cool greenhouse. Flower-buds some-
what less than two centimeters long usually showed the required
stages of the pollen-mother-cells; namely, late diplnphase, diaphase
(early, middle and late), and metaphase. One anther at a time
could be removed, so that six stages at intervals of an hour or two
could be prepared from each bud. The anther was cut up. and
the short segments squeezed out in a drop of iron-acetocarmine.
(Spreading out the pollen-mother-cells on a slide, fixing with
Flemming's solution, and staining with iron-haematoxylin or iron-
brazilin, gave preparations less suited for the purpose.) The thin
layer of liquid was allowed to evaporate under a large coverglass
for several hours, until the cells were flattened by the capillary

30 465

466 JOHN BELLING.

pressure. Then the edges were sealed (by a thick solution of
dammar in xylol, or better) with melted paraffin wax. For oil-
immersion objectives the tube-length required to be slightly in-
creased, according to how many microns the chromosomes ex-
tended below the coverglass. The apochromatic objective for
water immersion, of 1.25 aperture, and the binocular microscope,
were mainly used.

Flower buds of L. regale, from the greenhouse, gave nearly as
clear figures as those of L. longifloruin. Buds from L. candidum,
grown in the open, in June, gave excellent metaphases. But in
late July and August, buds of L. tigrinum, L. speciosum and L.
a n rat uni had granular pollen-mother-cells ; and though these
cleared up in time in acetocarmine, yet they were not so clear at
diplophase and diaphase as corresponding stages of L. longiflorum.
At anaphase also, the chromosomes were shorter and thicker in
the plants grown in the summer than in those species raised in the
greenhouse in February.

An examination of the chromosomes of L. hngiflonun, at the
anaphase of the first division, shows that there are two large J
chromosomes ; one with the constriction perhaps one-third from
one end, and the other with the constriction near one-quarter from
the end. The remaining ten chromosomes have the primary con-
striction more or less near to, or at, the end ; so that the separated
segment is small or invisible at this stage. (One of these chromo-
somes shows a secondary constriction.)

The leptophase and zygophase closely resemble Newton's figures
for Tulipu (Newton, 1927). In the late diplophase, before the
bivalents are separate enough to be countable, nodes may be ob-
served in the double thread. At the earliest diaphase, as soon as
separate bivalents can be distinguished, many nodes are visible.
It is difficult to find a cell, however, in which all twelve bivalents
are free enough from overlap to permit the counting of all the
nodes. However, Fig. i is a camera drawing of a cell showing
the 12 bivalents at the early diaphase. There is here a total of 39
nodes; made up by one bivalent with 5 nodes, four with 4 nodes,
four with 3 nodes, and three with 2 nodes. Other cells at the
same stage seemed to have about the same numbers of nodes,
though all 12 bivalents could not be counted. In L. regale, also,

NODES AND CHIASMAS IN L1L1UM. 467

this stage seems to have a larger number of nodes than the fol-
lowing stages ; as has, indeed, been noted by previous observers in
both plants and animals. Tn Tulipa, which is allied to Liliitin.
Newton's drawings allow the estimation of about 34 nodes in 1 1
of the 12 bivalents at the latest diplophase or earliest diaphase.
Here, there seem to have been one bivalent with 5 nodes, one with
4, seven with 3, and two with 2 nodes.

Between the late diplophase and the late diaphase, many of these
nodes disappear. At middle diaphase, so far as incomplete counts
have been made, the numbers are less than in early diaphase and
more than in late diaphase. At late diaphase, the numbers of
nodes in five cells of Lilinin longi/lornin, where all 12 bivalents
were widely spaced and the nodes readily counted, were 24, 24, 23,
21, and 21, respectively. The average of these is 22.6. Fig. 2
is a camera drawing of one of these cells. In Tulipa, Newton'^
drawing shows about 2\ nodes at late diaphase. Thus in Lilium
we have a diminution in the number of nodes, from the earliest
diaphase to late diaphase, of 43 per cent. ; while in Tulipa the
estimated loss is also slightly over 40 per cent. It is possible that
if the counts could have been made earlier in the diplophase the
loss would have been shown to be somewhat larger. Hence we
may say that, in Liliitin and Tulipa, at least, somewhat less than
half of the original nodes disappear before the metaphase.

At the metaphase, the numbers of nodes were about the same as
in late diaphase; namely, 24 in one metaphase of L. longifloruin
(Fig. 3), and 21 and 20 in two metaphases of L. regale; averaging
21.7, as compared with 22.6 for the late diaphase. In these cases
the metaphase chromosomes were squeezed from the cells, and
cases were readily procured in which the nodes of all 12 bivalents
could be accurately counted.

The nodes at late diaphase and at metaphase are doubtless
chiasmas (Janssens, 1924), as they are also in Uvularia, Hyacin-
ilins, and All in in (Chodat, 1925). A chiasma can be distin-
guished microscopically by the following points. ( I ) When the
bivalent is normally flat, or is flattened by pressure, two chro-
matids may be seen to pass on, and two to form an X. Thus Fig.
4 shows a metaphase bivalent of L. candlduui, slightly com-
pressed, at three different levels, the focus of the microscope de-

468 JOHN BELLING.

scencling from left to right. There are two chiasmas visible, of
which the left one is the clearer. Fig. 5 shows another flattened
bivalent from the same metaphase plate as Fig. 4, drawn as usual
with shifting focus. Here the two chiasmas are clearly alternat-
ing with regard to the chromatids. (2) In a chiasma at the
diplophase or cliaphase, the cross junction can often be seen to
be thinner than the continuous threads. In a half turn of a
spiral, the cross junction would of course be of double thickness.
(3) When sufficient pressure is applied to the diaphase bivalents,
after being some days in acetocarmine, they seem to break across
only at the nodes. The separation of the strands in a chiasma
would doubtless make it a point of weakness. (4) At late dia-
phase and metaphase, the portions of the bivalents on each side of
a chiasma are in planes more or less at right angles to each other
(Fig. 3). (5) At the first anaphase, the constituent chromatids
do not separate in the halves of vertical V's and rings, while they
do in horizontal V's or rings. Fig. 6 shows this separation in
L. longifloniin.

The nodes (chiasmas) have been counted in 96 bivalents of
LUium, from eight different cells, at late diaphase and metaphase.
Of these, 31 bivalents had one node, 49 had two nodes, and 16
showed three nodes. Calculating the points of segmental inter-
change, or points of genetic crossing over, in the resulting chromo-
somes (four from each bivalent), we find: 119 chromosomes with
no point of interchange; 184, with one point of interchange; 73,
with two points of interchange ; and 8 with three points of inter-
change. In percentages these are: 31, with none; 48, with one;
19, with two; and 2, with three points of interchange. These
figures agree fairly with the figures for the four large bivalents of
Hyacinthus (Belling, 1927), and perhaps also with the genetical
results in species of Drosophila.

The nodes which disappear between diplophase and late dia-
phase do not seem to be all or mainly twists. Clearly recognizable
twists (half turns of a spiral) are apparently rare in Liliitin longi-
flontin and L. regale. Nor do these vanishing nodes seem to be
chiasmas which open out; for if so, this process should have been
visible, as it is at early anaphase. Their nature awaits further
investigation. However, the drawings of early observers, which

NODES AND CHIASMAS IN LILIUM. 469

show bivalents in Liliitin. or pirn its allied to Liliitin, forming, at
the late diplophase or early diaphase. regular right-handed or left-
handed spirals of severrd turns, do not correspond to what is to he
observed in Liliitin hnujiflorutn. with perfect fixation, yellow-green
light, a nine-tenths condenser cone, and apochromatic oil-immer-
sion objectives of 1.3 and 1.4 aperture.

Summary. — (i) The bivalents of six species of Liliitin were
studied, from the double-thread stage to the early first anaphase,
in pollen-mother-cells fixed in iron-acetocarminc. Thirty-nine
nodes were found at late diplophase or early diaphase.

(2) In 5 late diaphases, the average number of nodes was nearly
23, and in 3 metaphases the average number was nearly 22. Thus
there was a loss of 43 per cent, of the nodes between late diplo-
phase and late diaphase.

(3) Out of 96 bivalents, there were 31 with one node, 49 with
two nodes, and 16 with three nodes.' This would result in chromo-
somes having 31 per cent, with no point of segmental interchange,
48 per cent, with one point, 19 per cent, with two points, and 2
per cent, with three points of segmental interchange.

(4) Dixon's term, " strepsitene," seems to have been a mis-
nomer in the case of Liliitiu. Fifty-seven per cent, of the nodes
were clemonstrably chiasmas.

LITERATURE CITED.

Belling, J.

'25 Homologous and Similar Chromosomes in Diploid and Triploid

Hyacinths. Genetics. 10: 59-71.
'26 Single and Double Rings at the Reduction Division in Uvularia.

BIOL. BULL., 50: 355-363.

'27 Configurations of Bivalents nf Ilyacinthus with regard to Seg-
mental Interchange. BIOL. BULL., 52: 480-487.
Chodat, R.

'25 La chiasmatypie et la cinese de maturation dans 1'Allium

ursinum. Bull. Soc. Bot. Geneve, 1925.
Janssens, F. A.

'24 La chiasmatypie dans les insectes. Cellule, 34.
Newton, W. C. F.
'27 Chromosome Studies in Tulipa and some related Genera. Linn.

Soc. Journ. Bot, 47: 339-354.
References to earlier literature may be found in these papers.

47O JOHN BELLING.

EXPLANATION OF PLATE I.

FIG. I. Early diaphase in a pollen-mother-cell of Lilium longiflonini.
Camera drawing from an iron-acetocarmine preparation (as were also the
other five figures). Cytoplasm and chromosomes squeezed from the cell-
wall, and flattened on the coverglass. Viewed with the apochromatic oil-
immersion objective 60, of 1.3 aperture, used with the binocular attachment.
Tube-length increased to correct for watery layer. Wratten film, No. 66,
and water-immersion condenser giving aperture of 1.2, used. There were
39 nodes counted. Some of them were seen to be chiasmas, but the detail
was too fine to be shown in this drawing. Only one was proved to be a
half twist.

FIG. 2. Late diaphase of Lilium longiflonini. Preparation and optical
apparatus as in Figure I, except that lower eyepieces were used. There
were 24 nodes. Many could be seen to be chiasmas. Probably all were
chiasmas. Stages like this could be had in abundance, and the nodes could
be easily counted.

FIG. 3. Metaphase bivalents of Lilium longiflonini. Magnification, etc.,
as in Figure 2. The nodes can be counted with accuracy at this stage, and
can be shown to be chiasmas. There were 23 nodes. This stage is common.

FIG. 4. Metaphase bivalent of Lilium candiduin which has been squeezed
flat by due pressure, drawn in three focal planes. Viewed with apochro-
matic water-immersion objective 70. Description in text.

FIG. 5. Another bivalent from the same cell as that drawn in Fig. 4.
Camera drawing with shifting focus of microscope. Described in the text.

FIG. 6. Separation of the homologues in the metaphase and early anaphase
of Lilium longiflonnn. The first four above show stages in the separation
of the two larger bivalents. The two below, and the one on the right are
.shorter bivalents. Chiasmas are evident.

BIOLOGICAL BULLETIN, VOL. LIV

PLATE I

JOHN BELLING.

THE NEUROMOTOR APPARATUS OF
CHLAMYDODON SP.

MARY STUART MACDOUGALL.

From the Department of Biology, Agnes Scott College, and the Marine
Biological Laboratory of Woods Hole, Mass.

INTRODUCTION.

"While working on the life cycle of Chhniiydodon, a system of
fibers around the mouth was found, and, on account of the rela-
tion of these fibers to the pharyngeal basket and other organelles,
it seemed worth while to make a further study.

As most of the literature on the subject of conductile fibers,
especially in the Protozoa, has been reviewed by Taylor (1920),
Rees (1922), and Calkins (1926), it seems unnecessary to take
up the subject here.

Kofoid applied the term Neuromotor Apparatus to the complex
fibrillar system associated with blepharoplasts, parabasal bodies,
etc., found in Giardia. Other systems of this kind have been de-
scribed for the flagellates by other workers.

In the ciliates Sharpe (1915) was the first to give the term
Neuromotor Apparatus to a complex system of fibrils, having a
center or motorium, which he found in Diplodiniwm ecaudatum.
Since that time, similar systems have been described for seven
other ciliates as follows: Yocom (1913) in liuplotcs patella; Mac-
Donald (1922) in Balantidium coli and suis; Rees (1922), Para-
•mcchim caudatuni; Visscher (1925), Dilcptus giyas; Campbell
(1926 and 1927), Tintinnopsis nucnla and Favclla; and Picard
(1927) Bovcria tcrcdinldi. Euplotes represents a highly special-
ized type of Neuromotor Apparatus, Paramecium caiidatuin a
generalized type, while that of Chlamydodon represents a com-
bination of the two types.

ACKNOWLEDGMENTS.

The work was begun at the Marine Biological Laboratory and
completed at Agnes Scott College. I wish to thank Dr. Gary N.

31 47i

472

MARY STUART MACDOUGALL.

Calkins, of Columbia University and the Marine Biological Labora-
tory, for advice and helpful criticism during the progress of the
work, and Miss Ruth B. Rowland for help that made the work of
microdissection possible.

MATERIAL AND METHODS.

The animals used in this investigation were collected from
Knowlton's ditch at Woods Hole, Mass. A small portion of
algae, chiefly Oscillatoria, from the same ditch, placed in a Syra-
cuse dish in Knop's solution one part, distilled water ten parts,
furnished an excellent culture medium. By making sub-cultures,
the animals were kept in the laboratory at Agnes Scott College
for two winters.

Various killing and fixing solutions were used, the best results
being obtained with Schaudinn's Fluid, Bouin and strong Flem-
ming. If the material was to be stained with Mallory's connective
tissue stain, it was fixed with Zenker's or Picro-mercuric fluid,
using the time schedules recommended by Sharp and Yocom.
Some of the best preparations were obtained by the use of .3 per
cent, hsemotoxylin, long method.

Two methods of embedding were used: (i) by the aid of a
LeFevre embedding dish, and (2), by killing a whole culture in a
Syracuse dish, and embedding small bits of algae to which the
animals were attached. This was found to be the most satis-
factory method, as there was little danger of losing the animals,
and the material could be quickly handled ; also it gave an abun-
dance of material. Sections were cut from 2-4/1 thick. On ac-
count of the thickness of CJilainydodon, it is impossible to work
out the system of fibrils in whole mounts.

All drawings were made with the camera lucida, except Fig. 15,
and with the use of 1/7 Leitz oil immersion lens and 12 X oculars.
Details were worked out with 12 X oculars and a Zeiss 1.5 mm.
apochromatic lens. A 25o-watt light in a Zeiss microscope lamp
was used for illumination.

THE GENUS Chlainidodon.

The genus Chlamidodon was named and described by Ehrenberg
in the Proceedings of the Berlin Academy in 1835. He states

THE NK.run. MOTOR APPAKATl'S OF CIILAMV o X SI'. 473

that he discovered Chlaniidodon inncuiosync in the waters of the
Baltic Sea near \Vismar, Aug. 26, 1834. In 1838, in his Infu-
sionsthierchen als volkomene Organisnu-n, he mentions this form
again, and. in addition to the " teeth apparatus " mentioned in his
first paper, he descrihes, a " colorless oval shield," projecting on
all sides beyond the body, and covering its body. Stein (1859)
differs with Ehrenberg in some details, e.g., he says that he found
only eight instead of sixteen trichites in the basket, and he thinks
the "oval shield' is just part of the body. Entz (1884) de-
scribes Chlamydodon cyclops Entz as having fifteen and sixteen
trichites in the pharyngeal basket, while Erlanger (1890) figures
Chiamvdodon iniicinosyiic with sixteen. Stein changed the spell-
ing from Clavnidodon to Chlamydodon.

Ehrenberg placed the genus in the family ' Euplota " ; Stein
placed it in the family " Chlamydodonta."

*n - - .*- • •-

../•'• ' -Iff-t^SC^- •-& .1

Zone of origin of
rows of cili

oneme around the mpjfth
Pharyngeal baske
iacronucleus
Mioronuoleus (?)
-"Railroad track."

I

.

Fie. I A. Clilainyiloiion sp. (Prob-
ably a variation of mncmosyne.)
Ventral view.

FIG. i B. Side view.

The form most common at \Yoocls Hole is probably a variation
of Chlamydodon mncmosyiie, though it differs very much from
Erlanger's description of that species. There is another species of
Chlamydodon, found at Woods Hole, which the writer has seen
only twice, and only isolated specimens. It has a very short

474

MARY STUART MACDOUGALL.

pharyngeal basket situated at the extreme anterior end. It dif-
fers from all described species in the character of its cilia.

The species which was used as a basis for the present study,
Fig. I, measures from 60-70^ in length, and 40-45^ in breadth.
Like all of the species, it is ciliated on the ventral side only, the
anterior cilia being much longer than those on the other parts of

FIG. 2 A. Arrangement of cilia and fibers traversing the circular myoneme.

FIG. 2 B. Fan of fibers from cilia just before they join motorium. Delicate

cross fibers from basal bodies of cilia.

the body. The cilia are very fine, and arranged in rows set close
together. To get the true arrangement of cilia, the ventral sur-
face has to be removed and flattened out. This is possible only
when the " railroad track " has been cut, as this structure bounds
the rows of cilia, and, at the anterior end pulls a part of the ventral
surface over on the dorsal side, Fig. i B and Fig. 4. The anterior
cilia take their origin from a zone on the right-hand side (ventral
side up), extend around the anterior end in a half circle, thence to
the posterior boundary of the " railroad track '; in a somewhat
curved line, Fig. 2. The cilia below the zone on the right-hand

THE NErROMOTOK Al'I'AKATfS OF C 1 1 1. A M V! )ODOX SP. 475

side, take their origin as said zone, and extend in curved lines to
the posterior boundary of the railroad track. The three central
rows begin at the posterior end of the myoneme surrounding the
mouth, and extend to the posterior end of the " railroad track " in
the same manner as the other rows, Fig. 2. Fig. I B shows the
manner in which the " railroad track " bounds the rows of cilia.

FIG. 3. Sections showing various changes in shape of the " railroad track."
FIG. 4. Sections of entire animal showing changes in the shape of the " rail-
road track," and changes in the shape of the animal.

The peciliar organelle known as the " railroad track " is a band
of trichites encircling the animal in the manner shown in Fig. i B.
In some of the older papers, the exact location of this band is a
matter of discussion. We were able to pierce the dorsal wall of
the animal with a micro pipette, and blow out the cell contents, in-
cluding the pharyngeal basket ; nothing was left except the body
wall, almost entire, and the " railroad track." Its position was un-
changed. It is covered with a thin pellicle, and fastened tightly
to it. With the micro dissection needles, the track was cut out
and pulled apart. While it was easier to tear apart where it was
thin, we could not duplicate the effect shown by Erlanger. The
figure in Erlanger's paper seems to show the parts of the " railroad
track " easily separable into round masses of protoplasm, each hav-
ing a trichite in the center.

If the animal is viewed from the ventral side, the structure
looks flat, and the trichites arranged after the manner of crossties,
hence the name " railroad track." A side view of whole mounts,
easily obtainable, and sections, shows that the ordinary shape of
this structure is half a circle. The organelle has the power of
changing its shape. This is readily shown in sections ; it may be
almost completely closed, or wide open, Figs. 3 and 4. After this

4/6

MARY STUART MACDOUGALL.

observation was made, a careful study of the living animal showed
that when the lip was bent back very far the structure was closed,
forming a complete ring. The anterior end was closed during
ingestion of food.

The ends of the trichites seem to be heavier than the middle
portion, and between each two trichites there is a thin portion with
a place in the center easily pulled apart. When the organelle is
cut and straightened out, it is seen that the trichite is thickened
and the thin portion reduced, the end view reminding one of an
accordion, Fig. 17.

FIG. 5. The pharyngeal basket of
Chlamydodon >]>., with circular myo-
neme which covers it removed.

Fir,. 6. Side view, showing relation
of basket to mvoneme.

The band or " railroad track " is not interrupted, as shown in
Erlanger's figures of Chlamydodon inncinosyiic. Just before di-
vision, a break near the center may be observed.

The pharyngeal basket is very large in proportion to the size of
the animal, it is very different from the basket described by
Erlanger, and figured in Doflein, p. 54. There are ten heavy
trichites, the anterior ends being expanded, and the trichites show-
ing distinctly. About half way down, the trichites seem to fuse,
their identity becoming lost as the basket narrow's, resembling
Chilodon nncinatiis in this respect. The posterior end has a small

THE NEUROMOTOR APPARATUS OF CIILAM YDODON SP.

477

filament wound to the right (ventral view). At the anterior em!,
each trichite has a sort of cap fitted to it somewhat after the man-
ner of a hinge, Figs. 5 and 6, and extending to the mouth opening.
These caps are in the shape of a triangle, and form a lid. This
lid is, in turn, covered by a circular myoneme, like a sphincter
muscle, in the center of which is the mouth opening. As Figs.

10

FIG. 7. Circular myoneme surrounding the month, with traversing fibers.
FIG. 8. Tangential section through mouth region, showing relation of basket

to myoneme and fibers.

FIG. 9. Fibers in caps of trichites and top of trichitcs.
FIG. 10. Caps removed from the trichites of the basket. Note fan of fibers.

6 and 8 show, the edges of the mouth opening in this myoneme fit
under the points of the triangles. It is suggested that the triangles
are pulled back by the myoneme, and so opens the mouth.

The macronucleus is divided into two parts, the anterior part
taking the stain more lightly than the posterior half. In the rest-

478

MARY STUART MACDOUGALL.

ing stage, the granules in the anterior half form a complete horse-
shoe. These granules change in size and position as cell division
proceeds. There are, in addition, many small granules. The
posterior half has a large division center, and many other smaller
granules. Between the two halves there is a split, or querspalt.
Often a granule is found in this split, and it has been observed to
divide. The two granules migrate into the cytoplasm. It has not

FIGS, ii, 12, 13, 14. The Motorium, and its relation to other structures.

been possible to follow their further history. In division, the
split disappears, the division center moves up the center of the
nucleus, pulls out and divides. All of the stages of division have
not yet been worked out.

The position of the micronucleus is uncertain. Entz pictures it
on the side of the nucleus in Clilainydodon cyclops, and Erlanger

THE NEUROMOTOR APPARATUS OF CIILAMYDODOX SP. 479

shows it imbedded in the anterior end of the macronucleus. It is
quite small, and some of my preparations show it at the posterior
end of the macronucleus, and some at the side. It cannot be seen
in all preparations.

THE NEUROMOTOR APPARATUS

The first part of the neuromotor apparatus to be observed was
a system of fibers around the mouth or oral opening, Figs. 2, 6, 7.
As described above, the pharyngeal basket in Clihnnydodon is very
large and heavy, and the circular myoneme around the mouth is
seen to be traversed by many fine fibers, with a small granule for

FIG. 15. Diagrammatic representation of the relation of fibers to the " rail-
road track."

i

each one on the edge of the myoneme, Fig. /. In only one or
two favorable preparations could it be seen that these fibers run
out into the cytoplasm in the general direction of the " railroad
track," Fig. n. With the aid of micro dissection needles, it is
possible to lift off the ventral surface of the animal. If this is

480

MARY STUART MACDOUGALL.

then properly stained, it is seen that the inyoneme around the
mouth is continuous with the body wall, but much thicker. The
fibers from the myoneme extend under the rows of cilia. They
are very fine, and it is impossible to trace them far.

Fibers run from one trichite to the next, and there are cross
fibers on each of the caps, Fig. 9. These fibers all come together
in two heavy strands, joining other fibers around the mouth, and
connecting the motorium at the lower right-hand corner, Figs. 12
and 13.

The basket seems to be lined with a very thin membrane, and
when the caps are removed from the trichites, fibers can be seen
in this lining, Fig. 10.

The motorium is just below the anterior end of the basket, and
is placed a little slantingly, Figs, u, 12, and 13. It is a bilobed
mass, which stains with haemotoxylin. It, and the other fibers in
the system, stain bright red with Mallory's connective tissue stain.

16

17

FIG. 16. Portion of the "railroad track" greatly enlarged.

Fie. 17. Portion of the " railroad track " dissected out and flattened.

Shows peripheral fibers.

FIG. 18. Section showing dorsal and ventral fibers from the " railroad

track."

THE NEUROMOTOR APPARATUS OF CHLAMYDODON SP. 481

It is impossible to see the motorium in the whole mounts. On ac-
count of its proximity to the pharyngeal basket, it \vas overlooked
for a long time, for the basket is heavy, and when destaining is
carried on long enough to differentiate it, the motorium is de-
stained. In sections, however, the structure is easy to see. With
the aid of microdissection needles, the basket can be dragged out
whole, and the motorium is sometimes pulled out with it. It is
then seen to be a refringent body, while in stained sections its gen-
eral structure appears granular. Fig. 12 shows the fans of fibers
joining the motorium.

The peculiar organelle known sometimes as the " striped band "
or " railroad track " mentioned above, has a very complex system
of fibers connected with it. Sections show the fibers very well,
but their paths are hard to trace on the ventral side on account of
the presence of the cilia.

At each end of each trichite, there is a plate or mass, Fig. 16.
In sections this mass appears single, Figs. 13 and 18; when viewed
from the ventral side of the animal, it appears bilobed or double,
Fig. 1 6. There is a possibility that when the granules were ob-
served from the ventral surface, they were in the process of di-
vision, but, when they were seen at all from this angle, they ap-
peared double. The presence of fine rows of cilia on the ventral
surface obscures everything; on the ' dorsal surface the body is
much curved in the region of the railroad track, making it im-
possible to see the granules or masses so close to the organelle.
The fibers that go through the trichites are connected with these
masses. Fig. 15, in a somewhat diagramatic way, shows the re-
lation of one set of ventral fibers with the masses and the
trichites. There are two sets of fibers, one set that enters the
trichites, and one set runs through the thinner portion of the
"railroad track," Fig. 15. These fibers are from two levels.
When the mouth is open, during the ingestion of food, the rail-
road track changes its shape from a half circle to a closed circle,
Fig. 4. From the evidence of the sections, and the observations
upon the living material, it is reasonable to suppose that one set
of these fibers is connected with the myoneme around the mouth.
The presence of ciliary lines and fibers make it almost impossible
to trace these fine lines.

MARY STUART MACDOUGALL.

The dorsal fibers extend only a short distance underneath the
pellicle, then they turn and go in the direction of the motorium.

Only two preparations showed with any clearness the fans of
fibers joining the motorium. The basal bodies of the cilia are
connected both by longitudinal and cross fibers, Fig. 2 B. The
longitudinal fibers of the cilia turn in at the zone of origin of the
cilia, and connect with the motorium at the anterior end, Fig. 12.
This end of the motorium also receives the fibers from the ventral
surface of posterior end of the animal. Fig. 15, and some fibers
from the mouth region. Figs. 13 and 14. The dorsal fibers join
the motorium at the posterior end.

As described above, between each two trichites, there is a thin-
ner portion, and in the center of this is a series of small granules.
The granules are connected with fibers which pass over the trich-
ites to the next set of granules, Fig. 17. In some sections, the ac-
cordion like arrangement is more pronounced than others, sug-
gesting that there is power of movement, a sort of folding of the
trichites. No observations on the living material settled this point,
but further evidence of the possibility of the movement suggested
lies in the fact that the trichites are sometimes closer together than
at other times.

MlCRODISSECTION.

After the location of the motorium, twenty-five animals were
successfully cut, freehand, with the aid of micro-dissection needles
given to me by Dr. Robt. W. Chambers of Cornell Medical Col-
lege. Later, Miss Rowland, of New York University, operated
upon several with the Chambers micro-dissection apparatus.

The cilia of Chlamydodon are fine, the anterior ones being longer
and easier to observe than the posterior ones. The motorium
cannot be seen in unstained specimens but its position with relation
to the basket is known, so that it is quite simple to destroy it. If
the motorium be destroyed, there is a marked disturbance in the
action of the cilia, in no way comparable to the disturbance of the
cilia if other parts of the body are injured. The cilia still have
their wave like motion, however, and this is to be expected when
one takes into consideration the relation of connecting fibers of
the cilia, Fig. 2 A and B. However, the cilia do not reverse after

THE NEUROMOTOR A I'l'A RATUS OF CI1I.A M YDODOX SP. 483

the destruction of the motoritim as is usual in intact animals.
Isolated pieces hehave the same way, as has been observed by
Jennings and Jamieson (1902), and by Rees (1922). Animals
without conspicuous motile organs are not favorable material for
the study of the coordination by means of microdissection.

SUMMARY.

A description of Chlamydodon, probably a variation of the
species mnemosyne, found in brackish water at Woods Hole, is
presented.

There is a complex neuromotor apparatus, including a coordinat-
ing center, and systems of fibers connected with cilia, the mouth
opening, the phryngeal basket, and the " railroad track."

The connection of the fibers with the organelles and the central
mass, or motorium, the behavior of the animal after the destruc-
tion of the motorium, seem to suggest a coordinating function for
the system.

LITERATURE CITED.

Calkins, Gary N.
1926 Biology of the Protozoa. Lea & Febiger, Philadelphia and

N. Y.
Campbell, Arthur Shakleton.

1926 The Cytology of Tintinnopsis nucnla. University of Cal. Pub.

in Zoc5L, Vol. 29, No. 9, 179-236.

1927 Studies on the Marine Ciliate Fuvella. Ibid., Vol. 29, Xo. 16,

429-452.
Doflein, F.

1916 Lehrbuch der Protozcenkunde. Fischer, Jena.
Entz, Geza.

1884 Ueber Infusorien des Golfes von Xeapel. Mitt, der Zool. Sta-
tion zu Neapel, Band 5, 340-345.
Erlanger, R. v.

1890 Zur Kenntnis einiger Infusorien. Zeit. f. Wiss. Zool., Band

49, 645-659.
Ehrenburg, C. G.

1835 Abhandlungen Akademic der Wissenschaften zu Berlin.
1838 Atlas iiber Infusionsthierchcn.
Klein, M. Bruno.

1927 Die Silberlinienensysteme der Ciliaten. Arch. f. Protist.,

Band 58, Heft 1, 54-140.

1926 Ueber ein neue Eigentumlichkeit der Pellicula von Chilod<>n
Hiicitialits. Zool. Anz., Band 67, Heft 5/6.

484 MARY STUART MACDOUGALL.

Kofoid, C. A., and Sweezy, O.

1915 Mitosis and Multiple Fission in Trichomonad Flagellates.

Proc. Amer. Acad. Arts and Sci., Boston, 51, 289-378.
MacDonald, J. D.

1922 On Balantidium coli (Malmsten) and Balantidium suis (sp. nov.),
with an Account of their Neuromotor Apparatus. Univ. Cal.
Pub. in Zool., Vol. 20, No. 10, 243-300.
Maier, H. N.

1903 Ueber den feineren Bau der Wimperapparate der Infusorien.

Arch. f. Prot., Vol. 2, 73-179.
Pickard, Edith A.

1927 The Neuromotor Apparatus of Boveria teredinidi. Uni. Cal.

Pub. in Zo61, Vol., 29, No. 15, 405-428.
Rees, C. W.

1922 The Neuromotor Apparatus of Paramcchim. Univ. Cal. Pub.

in Zool., Vol. 20, No. 14, 333-364.
Sharp, R. G.

1914 Diplodinium ecaudatum, With an Account of its Neuromotor

Apparatus. Univ. of Cal. Pub. Zool., Vol. 13, 42-122.
Taylor, C. V.

1920 Demonstration of the Function of the Neuromotor Apparatus
in Euplotes by the Method of Microdissection. Univ. of
Calif. Pub. Zool., Vol. 19, 403-470.
ten Kate, C. G. B.

1927 Uber das Fibrillensystem der Ciliaten. Arch. Prot., Band 57,

Heft 3, 362-426.
Visscher, Paul.

1925 The Neuromotor Apparatus of Dileptits. MSS.
Yocom, H. B.

1918 The Neuromotor Apparatus in Euplotes patella. Univ. Calif. Pub.
Zool., Vol. 18, 337-396.

OBSERVATIONS OX Till-; FOOD REACTIONS
OF ACTINOPHRYS SOL.

JAMES BURDINE LOOPER.i
INTRODUCTION.

Observations on the ability of the protozoa to choose their food
date back to 1838 at least. At that time Ehrenberg observed that
various organisms ingest carmine which, as food, is nothing more
than inert matter. From 1838 until 1902 most of the workers in
this field held that protozoa do not possess the power of choosing
their food. Even at the present time, as stated by Visscher ('23)
in the introductory to his work on Dilcpius f/igns, it is evident
that the selection of food has thus far been positively demonstrated
in a very few forms.

Kepner and Whitlock ('21), in their work on Amoeba, give proot
of qualitative efforts on the part of Ama-ba to meet certain con-
tingencies. Up to the present time proof of such qualitative effort
in other rhizopods has not been recorded so far as I have been
able to determine.

This work was done under the direction of Professor William
A. Kepner, to whom I am deeply indebted for helpful suggestions
regarding these experiments and the preparation of this paper.
I am also indebted to Professor Bruce D. Reynolds for valuable
criticisms and suggestions.

CULTURE METHOD.

The best cultures were obtained in an infusion made by boiling
three grains of wheat for five minutes in 100 cc. of filtered spring
water. The infusion, along with the three grains of boiled wheat,
was placed in a small, clean, glass aquarium and allowed to stand
at least twenty- four hours before the animal was introduced.
After about a month had passed Jctinoplirys could be secured

1 From the Laboratory of Biology of the University of Virginia, and
the Laboratory of Histology, University of Mississippi.

485

486 JAMES BURDINE LOOPER.

from such a culture usually in abundance. The cultures were kept
covered with thin sheets of plate glass and were kept in the darkest
portion of a well lighted room. To keep the cultures going it was
found necessary to add to each culture 5 to 10 cc. of boiled spring
water per week. The Heliozoa subsisted chiefly upon Pcrancma
and Colpldlo, the progenitors of w'hich were introduced when the
cultures were inoculated.

METHODS OF OBSERVATIONS.

All observations were made under the compound microscope
—the 4 mm. objective and number 10 eyepiece being used in
practically every observation made. A number 10 eyepiece con-
taining a corrected micrometer, and having a camera lucida at-
tached was kept at hand in case sketches or measurements were
to be made.

Cultures in Petri dishes were examined under the 16 mm.
objective. In this way observations were obtained on movement
and general behavior in a habitat to which the animals were
thoroughly accustomed.

A few observations were made on specimens in a drop of water
on an ordinary micro-slide under a cover glass, but most obser-
vations were made on specimens mounted in a hanging drop. The
former method is inconvenient because water has to be supplied
to compensate for evaporation. This often disturbed the Actl-
noplirys in such a way as to cause it to lie quietly for some time,
hence, much valuable time may be lost. The hanging drop method
alleviated this trouble once the mount was properly made. To
prepare the mounts, first a glass ring about i cm. in height and
1 1/2 cm- m diameter, was placed on a clean slide and attached
thereon by applying vaseline around the proximal end of the ring.
The ring was then supplied with well-oxygenated water until the
meniscus came up to half the height of the ring. Clear water in
which Elodca or Chara was actively growing was found to be best.
The supply of oxygen in the water in the ring tends to replenish
the supply in the hanging drop as it is used up by the mounted
Actinophrys sols. The desired number of the above mentioned
specimens was then isolated by means of a capillary pipette and
placed on a clean number one coverglass. Next, with another

FOOD REACTIONS OF ACTINOPHRYS SOL. 487

capillary pipette, the desired objects of prey were applied to the
drop on the coverglass containing Actinophrys sol. The upper
end of the ring, which had previously been prepared, was smeared
with a small amount of vaseline. The coverglass was then in-
verted and gently pressed down on the ring. Mounts, prepared
as described above, often lasted two weeks for observation. In ob-
serving these mounts under the 4 mm. objective, better lighting
was obtained by removing the sub-stage condenser and employing
the concave mirror.

SPECIAL PSEUDOPODS.

The special pseudopods for taking in prey are always composed
of hyaline ectoplasm. There are three general types of these
special pseudopods. First, if the object to be ingested is very
small and relatively motionless, a small, straight pseudopod is ex-
tended, and upon coming in contact with the object spreads out into
a cup which encircles and closes in closely on the object. These
pseudopods often resemble a dipper if the protruding end happens
to pass beyond the food object before contact is made. Second,
if the motionless object be large, a wide, hyaline outgrowth (Fig.
i, i a) of protoplasm advances towards the prey. When it gets
quite near to, or comes in contact with the prey, the tip of this
special pseudopod expands in all directions laterally. These lateral
expansions close about the large motionless food object usually in
close contact with it. In the few instances when close contact
may not be made about the large motionless food object, the food
vacuole is always decidedly smaller than is the case with animate
objects mentioned in the next or third class. Third, large sack-
like pseudopods (Fig. 2, i a) are sent out to encircle active ani-
mals. The victim is often cut off from a means of escape by a
palisade of rays on one side, and by the approaching special pseu-
dopod on the other.

REACTIONS TO INANIMATE OBJECTS.

In experiments on reaction to inanimate objects, only such ma-
terials were used that were thought to be non-toxic to Actinophrys.
In addition, the materials chosen were of such nature as to be
easily distinguishable in the vacuoles. The following were tried :
32

... .••{•'.•• x-'-''--

A- . : .' -,<

>• /

1
* >

1 ;- \ \ \ \
i 'i *

• '.
i

i

i

. ••. / I /. ,' ,/ A^

'

FIG. i. Sketches illustrating the ingestion of an inanimate object — granule

of wheat starch.

1. Actinophrys sol with the granule partially enclosed by the food-
getting pseudopod, a; b, ray; c.t'., contractile vacuole.

2. Same specimen one half hour later in which a has grown to o'. r.r-.',
Protoplasmic projections representing broken wall of emptied contractile
vacuole. (X 825.)

FOOD KKACTIOXS OF ACTINOPHRYS SOL.

489

carmine, alizarine, powdered glass, plain wheat flour, corn starch,
and powdered graphite. Ten experiments were tried with each of
these materials, but evidence of ingestion was seen only in the
experiments where carmine, wheat flour, and corn starch were
used. In the experiments with each material care was taken to
try individuals which were in different physiological conditions;
e.g., small, underfed specimens, well fed specimens, and speci-
mens which were accustomed to different types of food. Only
experiments with materials which .-Ictinoplirys ingested will he
recorded in this paper. The following tables contain the results:

TABLE I.
REACTIONS TO CARMINE.

Experiment.

Xo. of Indi-
viduals in
Hanging
Drop.

Xo. which
Did Xot
Take Any
Particles
within 2
Hrs.

Xo. Taking
Only One
Particle
within 2
Hrs.

Xo. Taking
Two Parti-
cles within
2 Hrs.

Xo. Taking
Three
Particles
within 2
Hrs.

Xo. Taking
More than
Three
Particles
within 2
Hrs.

I

8

c

2

I

o

Q

2 ...

e

I

o

O

Q

3

4.

1 1

9

c

I
I

I
o

o

I

O

Q

f

A

7

I

Q

o

O

6

e

•3

2

Q

Q

7

6

r

Q

I

o

8

-2

2

I

o

o

o

9 ....

1 c

12

2

I

o

o

10

c

i

I

o

o

o

Total. .

60

^

12

A

I

o

Per cent

7c 4

I 7.4

5.8

I I

o

In each experiment observations were made as soon as the
hanging drop mount was made, then every fifteen minutes there-
after for a period of two hours. In no case did it seem probable
that ingestion followed by egestion took place within the time be-
tween observations. This is because Actinoplirys reacts slowly.

Table I. shows that fifty-two individuals (75.4 per cent.) did
not take carmine at all. Twelve individuals (17.4 per cent.) took
only one particle of carmine. The number taking two particles
was four, only one individual took as many as three particles,
while none took more than three. Hence, it is evident that car-

490

JAMES BURDINE LOOPER.

mine was ingested by only about one fifth of the total number of
individuals, and these took carmine in relatively small quantities.
I was fortunate in seeing eight of the specimens ingest the
particles. In each case small, straight, food-getting pseudopod.^
were extended and the particles were closely embraced. This
phenomenon will be referred to again later.

TABLE II.

REACTIONS TO WHEAT FLOUR.

Experiment.

No. of Indi-
viduals in
Hanging
Drop.

No. which
Did Not
Take Any
Particles
within 2
Hrs.

No. Taking
Only One
Particle
within 2
Hrs.

No. Taking
Two Parti-
cles within
2 Hrs.

No. Taking
Three
Particles
within 2
Hrs.

No. Taking
More than
Three
Particles
within 2
Hrs.

I

5

2

I

2

O

O

2

5

3

2

O

0

O

7

7

4

2

O

I

O

4

4

2

I

0

0

I

9

3

I

0

I

6

7

6

o

I

o

0

7

IO

7

2

0

I

o

8

3

2

I

0

o

o

0 . .

6

3

2

I

0

o

IO ...

8

5

2

I

0

o

Total

64

39

16

6

2

2

Per cent

61

25

9-3

2.3

2-3

In Table II. are found the results of ten experiments with plain
wheat flour. Thirty-nine individuals (61 per cent.) did not in-
gest particles of the flour. Twenty-five per cent, ingested only
one granule each, and three tenths per cent, ingested two granules,
while two and three tenths per cent, ingested three and four
granules, respectively. None were observed to take in more than
four granules.

The experiments with flour were carried out under practically
the same conditions as the experiments with carmine. However,
the results of the former tend to show that particles of wheat
flour are taken in by a larger number of individuals, and in rela-
tively larger amounts, than carmine was by individuals under
similar conditions. This can probably be explained by the fact
that particles of wheat flour may be coated with a thin film of
protein material which Actinophrys is able to digest and assimilate,

FOOD REACTIONS OF ACTIXOPHRYS SOL.

491

though I was not able to detect any change in granules which lay
in the food vacuoles for as long as five days.

The type of pseudopod employed in engulfing particles of
wheat flour belongs to the second class previously described for
the taking in of relatively large motionless objects. As with the
carmine particles, the pseudopod closely embraces the particles of
flour, though the process is very much slower. The slowness of
ingestion is probably due to the larger size of the particles.

Table III. contains the results of ten experiments with corn
starch. Fifty-eight individuals (85.3 per cent.) did not take in
granules at all. Eleven and seven tenths per cent, engulfed only
one particle each. Three per cent, took in two particles each,

TABLE III.
REACTIONS TO CORN STARCH.

Experiment.

No. of Indi-
viduals in
Hanging
Drop.

No. which
Did Not
Take Any
Particles
within 2
Hrs.

No. Taking
Only One
Particle
within 2
Hrs.

No. Taking
Two Parti-
cles within
2 Hrs.

No. Taking
Three
Particles
within 2
Hrs.

No. Taking
More than
Three
Particles

within 2
Hrs.

I . .

12

IO

2

o

o

0

2

8

8

o

o

o

o

•3

c

A

o

I

o

Q

4

o

7

I

I

o

o

S • .

9

8

I

o

o

o

6

3

3

o

o

o

o

7
8

6

s

5

T.

I

2

o
o

o
o

o
o

0 . .

4

I

o

o

o

IO. . .

6

6

O

o

o

o

Total

68

58

8

2

o

o

Per cent . . .

8=5.3

II. 7

3

o

o

while none took in more than that number. By referring to
Tables I., II., and III., it may be easily seen that these results
show that corn starch was taken in by fewer individuals than
either carmine or wheat flour.

The method of ingesting corn starch was observed to be practi-
cally identical with the method referred to under experiments
with wheat flour, viz., inanimate particles are closely embraced by
the food-getting pseudopods.

Since it is more convenient to merely state the results of the

492 JAMES BURDINE LOOPER.

following experiments, tables will not be given. In all experi-
ments the procedure as outlined under the reactions to inanimate
objects was followed and the data were obtained in a correspond-
ing manner.

REACTIONS TO MOTIONLESS ANIMATE OBJECTS.

Under this heading reactions to living yeast, motionless algae,
and Eiiglcna cysts are recorded.

A. Desmid s.

a. Sccncdcsuius. — In experiments with Scenedesmus forty-
eight Actinophrys were involved. Sixty per cent, did not take the
algre. Those engulfing only one mass constitute thirty and three
tenths per cent. The remaining nine and six tenths per cent, of
the individuals engulfed two masses each. These desmids were
taken in by the same type of pseudopod as were the inanimate
objects.

b. A Larger, Desmid. — Fifty animals involved in ten experi-
ments were tried with this desmid. Only four per cent, attempted
to engulf the larger alga?, and in one case only was the attempt
successful!. In this case one Actinophrys sent out a pseudopod
and partially surrounded the plant but was not large enough to
completely surround it. This animal, while attempting to sur-
round the desmid, was joined by another animal. This union ap-
peared to be a complete fusion of the cytoplasm of one with that
of the other, the nuclei remaining separate. After this fusion
the pseudopod of the first animal was enlarged to completely sur-
round the desmid. The desmid remained in the food vacuole of
the multiple individual for a period of five hours and was then
slowly forced out through the outer membrane. Nothing re-
mained of the plant except the cellulose wall and a few particles
which remained on the inside of it. A similar phenomenon in
the Rhizopods was referred to by Delage and Herouard ('96), as
a " Socictc dc Consultation" Distaso ('08) called attention to
the grouping of A sol for the capture of large food objects.

B. Yeast.

In experiments with yeast seventy-two animals were used.
Twenty per cent, of these took in one or more of the plants.

FOOD REACTIONS OF ACTINOPIIRYS SOL. 493

Since yeast plants are so small it was impossible to keep check on
the number taken in. When taking in objects of the above kind
the pseudopod was of the small straight type referred to pre-
viously in this paper. The plants in all cases observed were
closely embraced by the pseuclopod until digestive fluid began to
collect around them, filling a space between the outside of the
plants and the wall of the vacuole.

C. Cysts of Alga.

Enc/leiut cysts and cysts of Chlamydomonas were taken in in
practically the same proportion. In each case one cyst each was
taken in by fifteen individuals (20 per cent.). Four per cent.
took in two cysts each, while only one per cent, ingested three
each. There were not any taking in more than three cysts. The
close embrace was employed in all ingestions observed.

The experiments with motionless animate objects show that
Sccncclcsiints is taken more often than either living yeast, Eiujlcnn
cysts, or cysts of Chlamydomonas ; also, that a larger desmid was
rarely chosen as an object of food. The ingestion of the larger
algae required the cooperation of two or more individuals.

REACTIONS TO MOVING ANIMATE OBJECTS.

In reacting to moving animate objects Actinophrys presents an
entirely different type of food-getting pseudopod. Previously,
this was referred to as a relatively large, cup-like pseudopod. In-
stead of the close embrace of the object by the pseudopod, which
was referred to under reactions to inanimate objects, we have
what may be referred to as a " subtle embrace." A description
of the capture and ingestion of a Colpiiliitin will serve to make
clear what is meant by a " subtle embrace."

^A'hile watching the movements of an . Iclinoplirys a Colpicliiun
was observed to encounter the tips of some of the rays of the
former. The Col '/> id him became motionless almost instantly, as
if completely paralyzed by the encountered rays. While watching
to see the ultimate fate of the ciliate a relatively large mass of
cytoplasm was observed to be protruding from the surface of the
Actinophrys. This mass, which proved to be a food-getting pseu-
dopod, came forward as a single mass until it was four micra

FIG. 2. Sketches illustrating the ingestion of a moving animate object—

Chilomonas.

1. Actinophrys sol with a pseudopod, the contour of which is becoming
adapted to that of the prey before coming into contact with the latter, a,
food-getting pseudopod.

2. Same specimen fifteen minutes later, a has grown to become a, which
shows wide detour about the motile prey. (X7SO.)

FOOD REACTIONS OF ACTINOPHRYS SOL. 495

a\vay from the quiet Colpidium; then, without approaching farther.
it gave off two branches (Fig. 2, i a). These branches continued
to protrude, encircling the victim, and fusing on the opposite side
of it. Thus far, the Colpidimn was not approached closer than
two micra by the branches of the engulfing pseudopod. Hence,
the pseudopods were thrown subtly, but not intimately, around the
prey. Immediately after the fusion of the pseudopods beyond
the prey they spread in such a way as to form a hollow sphere
or vacuole containing the Colpidimn on the inside (Fig. 2, 2).
The vacuole then contracted around the prey, which in the mean-
time became stimulated to exhibit frantic efforts, apparently for
the purpose of escape. These efforts were kept up for a period
of at least five minutes before the imprisoned animal finally be-
came quiet again. The vacuole containing the prey was eventually
incorporated into the body and grew smaller as digestion pro-
ceeded, and disappeared completely when the indigestible particles
were finally ejected through the cell membrane of Actinophrys.

A. Ciliatcs.

Colpidium. — Ten experiments with Colpidia reveal the fact that
forty individuals (50 per cent) took in one or more Colpidia.
Thirty-seven and seven tenths per cent, engulfed one Colpidium
each. Twelve per cent, took in two specimens each ; while only
three tenths per cent, engulfed three each. Apparently, the
reason why there was such a small per cent, taking in as high as
three each lies in the fact that Colpidium is relatively too large
to be engulfed in larger numbers by Actinophrys within two hours.

Loxoccphalus gramilosiis. — Loxoccphalus is taken in very
readily and in relatively large numbers by Actinophrys. In the
experiments with the above mentioned ciliate only thirty-six (40
per cent) of the Heliozoans failed to ingest one or more. Thirty
per cent, of the individuals engulfed one ciliate each. Two each
were taken in by twenty per cent., while ten per cent, were ob-
served to have three vacuoles containing one Loxocephalus each.
The type of pseudopod used in taking in this organism was ob-
served to be practically identical with the type described under
reactions to Colpidium.

Stylonychia mytilus. — This ciliate is much larger and stronger

496 JAMES BURDINE LOOPER.

than Actinoplii'vs. Some of the cilia have been modified to form
liristle-like cirri which may be used as organs of offence, or de-
fence. Hence, it is hardly possible for Actinophrys to capture
such an organism single-handed, and so far as known no one has
ever observed such. I have seen four different cases where in-
dividual StvlonvcJiia were engulfed by a group of Actinophrys.
These groups in two cases were assembled after one Actinophrys
sol had tried to capture and ingest the larger and stronger animal.
In the other two cases the groups had been previously formed.
In no case was a Stylonychia taken in by a group of less than four
individuals. The group action here, if it may be called group
action, differs in at least one respect from the association of in-
dividuals described by Kepner ('25). In the case he 'reported,
three individuals united to take in a motionless desmid, and the
pseudopods of the former embraced the latter closely. In the
case of a group taking in a StylonycJiia the pseudopods presented
the phenomenon of the loose, or subtle embrace described under
reactions to Colpidinin.

OtJicr Ciliatcs. — Ten experiments were carried out with Fron-
tonia lucas, Nassnla oniata, and Spirostomwn ainbignnin respec-
tively. Some of the experiments in each case were kept under
observation for five days. However, in no case were any of the
above-mentioned ciliates engulfed by Actinophrys; while Colpidia
and Chilonionads, which happened to be among these ciliates, were
all engulfed sooner or later. This certainly indicated a definite
discrimination in regard to the type of ciliate taken in. On what
is this discrimination based? Whatever the basis, it is hard to
believe that either Frontonia, Nassula or Spyrostomum would
be more difficult, from a mere physical standpoint, to capture and
engulf than Stylonychia.

B. Reactions to Flagellates.

Cltilonionas parainecinin. — Reactions to this flagellate were prac-
tically the same as described under reaction to the ciliate, Col-
pidinin. Forty-one individuals (51 per cent.) of Actinophrys sol
did not engulf CJiiloinonas. Of the remaining forty-nine per cent,
forty per cent, engulfed one each. Six per cent, took in four each.

Pcnmcnia trichophorum. — Pcrancnia proved to be engulfed

1001) REACTIONS OF ACTIXOPHRYS SOL. 497

very readily by ActinopJirys. Only twenty-five individuals (42
per cent.) failed to take in this flagellate. Thirty-five per cent,
engulfed one per individual. Nineteen and one tenth per cent,
took in two each, and three per cent, three each. In no case were
as many as four \Pcranenia engulfed by an individual. The com-
plete process of capture and ingestion was observed in eleven
cases. In each the subtle embrace was employed.

Enylcna -ciridis. — Enylena riridis was taken more readily than
any other food material. Thirty-five individuals (58.4 per cent.)
of the Heliozoan ingested one each. Ten per cent, took in two
each, and one per cent, three each.

Enylcna, as a rule, was embraced more closely than either of
the other moving animate objects. I have not been able to ob-
serve definitely why this was the case. However, this deviation
from the usual procedure in ingestion of moving animate objects
probably has some connection with the fact that I'.nylcna was often
brought in contact through a reaction with the body proper of
Actinophrys before a food-getting pseudopod was sent out. This
contact with the body proper was brought about by a sudden jump
of Enylcna in the direction of the body proper of Actinophrys
while reacting to a contact with the rays. The exact reason for
this sudden jump on the part of Enylcna has not been determined,
though it has been suggested that the long flagellum of l:.u<ilcna
became attached to the body of Actinophrys, probably, by a
mucilaginous property of the latter. Then in an effort to free
itself, the Eiiglcna pulled itself into contact with the body of Ac-
tinofihrys to which the end of its flagella was fixed. Thus the body
of the flagellate became fastened by the same property. If this
be true, the food-getting pseudopod which arises from the outer
region of the Heliozoan's body, would quite probably adhere
closely to the flagellate's body as it encircled it. Should, however,
the Euylena begin to squirm, the cup of the pseudopod would
open up to form a larger cavity around the victim. Thus in case
the Enylcna began to struggle, what was started as a close em-
brace was forthwith transformed into a subtle embrace.

D. Reactions to Rotifers.

Only individuals belonging to one or two of the small species
of Rotifers were taken in by Actinophrys. Rotifers are the only

498 JAMES BURDINE LOOPER.

multicellular animals which were observed to be captured and
ingested by the Heliozoan. Here, as with Stylonichia, a group of
less than four was not strong enough to capture a Rotifer. I was
fortunate to see the complete struggle involved in the capture and
ingestion of five specimens by five different groups of Actinophrys.
Here, as in the case of the capture and taking in of Stylonychia,
some of the groups were formed after a single individual had
become attached to the prey (Fig. 3). In the formation of two
of these groups individuals which were entirely out of the low
power field of the microscope joined the fray, fusing with their
comrades which were already involved. In one of the above men-
tioned cases four individuals were not in the field when the Rotifer
was encountered by an Actinophrys which was in the field. In
the other case, likewise, three were outside of the field of vision.
In these cases the rays of the individuals which came into the field
of the microscope to join in the feast were not adjacent with the
rays of those already involved. This seemed to indicate that there
must be some way of sending out a call for assistance when the
object was too large and pugnacious to be conquered and swal-
lowed by one individual. The reason that it cannot be said defi-
nitely that this is the case is because the outsiders might be drawn
in by currents set up in the medium by the floundering of the at-
tacked prey. The fact that a number of Actinophrys sol fused to
form an aggregate when brought together by agitating the sur-
rounding medium with a glass filament, strengthens the above
mentioned idea. Whatever may be the cause of the formation of
temporary colonies, the phenomenon is, at least, of interest. By
virtue of this fact, it becomes easier to conceive of how multi-
cellular animals may have arisen by the association and cooperation
of individual cells. Be this as it may, the formation of multiple
individuals by Actinophrys sol makes possible the capture and
ingestion of larger objects of prey. This is essentially in agree-
ment with Distaso's ('08) observations on A. sol. Apparently this
temporary colony formation also serves a protective role. It is
probably of interest to note that an animal so low in the scale of
organic life exhibits the phenomenon of cooperation in the funda-
mental activities of its life.

By comparing the results of the experiments carried out with

FOOD REACTIONS OF ACTINOPHRYS SOL.

499

FIG. 3. A. Five Actinophrys sols later to be involved in the capture and
ingestion of a rotifer. Specimens 4 and 5 already show advanced co-
operation, a. Specimens 3 and 4 show the inception of cooperation through
the fusion of two of their slender rays, b. Specimens 2 and 3 are also in-
volved in the cooperation through the fusion of two rays at c. Specimen
I is yet independent.

B. Same group showing progress made within two hours. Complete in-
gestion has been accomplished and some digestion has ensued, i', 2', 3', 4',
and 5' shows the position taken by i, 2, 3, 4, and 5, respectively, of Fig. 3 A.
(X26o.)

500

JAMES BURDINE LOOPER.

A

V

A

i V '/ 'lA-T-^

FIG. 4. Sketches illustrating the fusion, also, the cooperation of two
Actiiwpltrys sols in the capture and ingestion of a Colpidium.

1. Pseudopodial fusion in the union of the two individuals.

2. Cooperation of the same two individuals in ingestion of Colpidium by
one individual, A, sending out a and individual B sending out b to form in
common one large food-engulfing pseudopod.

3. Same specimens one and one half hour later. Prey engulfed by fusion
of pseudopods a and b about it. (X 325.)

FOOD REACTIONS OF ACTIXoiMIRYS SOL. 50!

•

the various types of objects it may easily be seen that . Ictinophrys
sol feeds chiefly on the smaller ciliates and on the flagellates. The
experiments with moving animate objects show that certain of
these objects are eaten more readily and in greater numbers than
are others. Difference in the size and vigor of the different food
objects, no doubt, plays an important role in causing this di —
crimination. However, it is not probable that difference in size
and vigor alone is the only factor causing the discrimination noted.
The experiments with moving animate objects also show that
such objects are not embraced intimately by the food-getting
pseudopods of Actinophrys sol; but on the contrary, the moving
animate objects, as a rule, are loosely or subtly embraced by the
food-getting pseudopods.

SUMMARY.

1. Actinophrys sol is essentially omnivorous.

2. Certain inanimate objects are ingested more often than
others. Some inanimate objects apparently are not ingested at all.

3. Motionless animate objects are accepted, but not in appre-
ciably greater numbers than inanimate objects.

4. Many free-swimming ciliates and flagellates are readily eaten.
but certain ciliates apparently are not taken.

5. A degree of selection or discrimination was exercised be-
tween different food objects in every class of materials used.

6. Large objects of prey may be taken in bv temporary colonies
of AclinopJirvs sol which are formed for the purpose of taking in
such objects.

7. As a rule, motionless animate objects and inanimate objects
are closely embraced by the ingesting pseudopod; objects which
struggle while being ingested are subtly embraced.

BIBLIOGRAPHY.

Belar, K.

'21 Untersuclumj^vn iiher rk'ii F<>rm\vechsel von Aclinofhrys sol.

(Vorl. Mitt.!. I'.icl. Xrntralblatt., 1UI. 41.

'23 Untersuchun.nvii an .•Iclinof'lirys sol Ehrb. I. Arch. f. Protis-

tenk., I'.il., 40.

'24 Untersuclnnix-fii an .-Iclinoplirys sol Ehrby. II. Beitra.yv zur

Phv-.ii >]< >;,rie des Forimvechsels. Arch. f. Protistcnk., Bd. 48.

JAMES BURDINE LOOPER.

Bronns, H. G. and Butschli, O.

'8Q-'89 KlassL-n uiul Ordungen des Thierrichs, Bd., 1. Protozoa,

pp. 261-331.
Calkins, G. N.

'01 The Protozoa, pp. 33-71.
'09 Protozoology, pp. 69-100.
Delage et Herouard

'96 La Cellule et Les Protozoaires, pp. 156-162.
Distaso, A.
'08 Sui process! vegetativi e sull' incistidamento di Actinophrys sol.

Arch. f. Protistenk., Bd. 12.
Kepner, W. A.

'25 Animals Looking- Into the Future, pp. 107-147.
Kepner, W. A., and Whitlock.

'21 Food Reactions of Amoeba protcus. Jour. Exp. Zool., Vol. ^9-.
Leidy, Joseph
'79 Fresh-water Rhizopods of North America. Report U. S. Geol.

Survey of the Territories, Vol. XII, pp. 235-241.
Sondheim, M.

'16 tiber Actinophrys ocitlata Stein. Arch. f. Protistenk., Bd. 36.
Schaeffer, A. A.
'10 Selection of Food in Stcntor cccntlciis. Jour. Exp. Zool., Vol.

8, pp. 75-132.
Visscher, J. P.

'23 Feeding Reactions in the Ciliate Dileptus gigas. BIOL. BULL., Vol.

45, pp. 113-143.
Weston, I.

'56 On the Actinophrys sol. Quart. Jour. Micro. Sc., Vol. 4.
Wetzel, A.

'25 Zur Morphologic und Biologic von Raphidocystis infestans N. Sp.
einen temporar auf Ciliaten parasitierenden Heliozoon. Arch f.
Prtistenk, bd. 53.

"

THE HONEY-GATHERING HABITS OF PO LIST US

WASPS.

PHIL RAU,
KIRKWOOD, Mo.

Wheeler, in his valuable paper on " Vestigial Instincts in In-
sects and Other Animals," * gives a resume of what is known as
the honey-gathering propensities of wasps, principally those of
the tropics. He also records his own discovery on November 3
of a nest of Palis! cs nictrica in New York City, hanging from the
eaves of a boathouse. This nest of a few dozen cells had four
females, inactive from the cold, clinging to the comb, and six cells
'contained half-grown, much contracted but still living pupae.
There were small drops of colorless liquid in many of the other
cells, and one of them was half full of this substance, which
tasted like and proved to be honey of an agreeable flavor. These
drops, he tells us, hung suspended in the angles of the cells, but
were without any definite arrangement, and varied much in size.
" This honey must have been collected some weeks previously
from the autumn flowers and stored, now that nectar and insect
food were no longer to be had, for the purpose of bringing the
few remaining larvae to maturity. The belated brood undoubtedly
accounted for the presence of the female insects at so late a date.''

Wheeler assembles the details of the observations on honey-
storing by tropical wasps made by Lepeletier, Rouget. Marchal
and others, and then goes on to show that in the northern species
this habit of honey-storing is a vestigial instinct, and is called into
play only when the insect is facing unusual conditions. To use
his exact words: '' The great quantities of honey collected by the
tropical wasps are, of course, stores of provisions for the winter,
for, as the von Iherings have shown, many of the species, unlike
the northern Polistcs, do not abandon their nests on the approach
of the unfavorable season and start new ones in the spring, but
continue to add to their combs and keep on raising their brood

* Am. Jouni. ljs\clwl., 19: 1-13. 1908.
33 503

504

PHIL RAU.

throughout the year. The naturalists are unquestionably right in
deriving the conditions seen in our northern Polistes from those of
the tropical species. There can be no doubt that Polistes has ex-
tended its range into North America and Europe since the close
of the glacial epoch. The storing of honey for the winter has
been discontinued, and the life of the species has been saved by a
new set of adaptations involving the abandonment of the nest, the
temporary suspension of the breeding instincts, and the hiberna-
tion of a small number of fertilized females. The drops of honey
occasionally stored in the nests are all that remains to point to a
once very important means of tiding over the flowerless season and
preserving the life of the individual colony. Rouget, Brongniart
and myself have observed this vestigial instinct only in the autumn,
and this would seem to be the most likely time for the feeble dis-
play of the old habit.

' Not only has the honey-storing instinct of our northern
Polistes been reduced to a feeble and useless vestige by the adapta-
tion of this insect to life in a temperate zone, but the nest-building
instincts, when compared with those of the allied tropical wasps,
show unmistakable signs of similar degeneration."

This interpretation was to me highly fascinating, for I too had
occasionally seen similar drops of this clear, honey-like substance
attached to the walls of the cells of Polistes paUipcs. Other things
crowded out further investigation of this point until early in 1920,
when the above-mentioned paper fell into my hands, and I decided
to look fully into the matter, with the following results.

FIG. i. A cross section of a nest of Polistes pallipcs, showing globules of
honey in the cells. (Exact size.)

Investigation No. i. — On June 16, 1920, twelve nests of
Polistes paUipcs were found in the old buildings at Wickes, Mo.,
and one nest was far in among the branches of a closely foliated
cedar tree. None of these cells had workers, but each nest was

HOXKY-C.ATIIKRIXr. IIAII1TS ()]•• !•()[. ISTKS WASPS. 505

presided over 1>y the queen. A close examination was made of
each nest, and the twelve in the buildings showed no evidence nf
having drops of this jelloid substance in their cells. The one nest
out in the cedar tree, however, gave positive evidence. This was
a nest of fourteen cells, with the four central ones filled with half-
grown larvae; these cells contained none of the jelloid material,
but the ten surrounding cells each contained an egg and from one
to three drops of this material glued to the angle of the cell. Some
of the globules were quite large. An idea of the relative size
when the terms large and small are used may be gained from Fig.
i, which illustrates the globules, exact size. The substance hail a
decidedly sweetish taste. This was an absolutely newly made
nest; there was consequently no possibility that this material could
have been left over from the preceding year in an old nest. The
twelve nests which gave negative results also contained eggs, so
of the thirteen mothers which worked in this neighborhood, only
one provided this material for her young.

Investigation No. 2. — The one nest containing the globules in
the previous study was taken; the other twelve were undisturbed,
with the idea of studying them later. By mid-summer (July 21),
the workers had matured and covered the nests, so that observa-
tions from day to day were impossible on account of their inac-
cessibility, the ferocity of the occupants and the poor light. \Yith
so important a problem, I could do nothing but knock the nests
down with a stick, and after the infuriated wasps had flown back
to the place where the nest had been, harvest my booty.

In this study, thirteen nests were taken and carefully examined ;
ten of these were nests of Pollstcs paUipcs, one was P. variatns
and two were P. annnhiris. Every nest except one of this lot gave
evidence more or less abundant to show that all of these three
species store drops of this " honey " in their cells containing eggs
or very tiny larvae. Throughout this study I have never found
these drops in cells containing larvae more than a few days old.
The data herewith give the details for the thirteen nests, all taken
on July 21, 1920.

In the foregoing- table, the first ten nests were P. pullipcs, " k "
was P. variatus, and "I" and "m" were P. annitlaris. These
few nests show that all three species follow this habit of supplying

506

PHIL RAU.

Cells

Cells

Cells

Cells

Cells

Total

Nest.

Only

Cont.

Cont.

Cont.

Cont.

Cells in

Remarks.

Begun.

EKKS.*

Honey.

Larvae.

Pupae.

Nest.

a

12

I

=

12

b

— .

30

O

—

7

37

c

=

15

14

6

10

31

From i to 3 beads of various

size. Note: much honey

but no larvae.

d

6

30

25

—

—

36

e

—

12

3

5

5

22

/

• —

24

24

3

9

36

One to 3 drops of various size.

g

—

26

14

5

7

38

h

49

13

13

IOI

Drops varied in size, number

and consistency.

i

—

36

32

6

18

60

7

61

40

7

10

78

From i to 4 beads, many of

them very large.

A;

—

20

20

10

M

44

From i to 3 globules of var-

ious size.

/

— •

22

2

13

4

39

in ....

—

38

I

23

4i

IO2

1.3

6

401

234

9i

138

636

* This signifies eggs or newly hatched larvae less than a day old.

their young with this honey, but of course from such limited data
we can have no idea whether any one species is more constant in
this habit or furnishes more profuse supplies. The totals show at
a glance how common is the occurrence of this food in the egg-
cells of wasps even at the middle of the summer season. Out of
the 407 cells containing, or about to contain an egg, 234 of them or
almost 58 per cent had globules of this honey-like material. The
actual occurrence of the honey was probably even more frequent
than the numbers indicate, for the reason that the first-day larvae
were included with the eggs, and in some cases they had doubt-
less already devoured their rations before I counted them.

Investigation No. 3. — In the foregoing experiment, I removed
and carried home the thirteen nests of Polistcs, leaving the adults
behind. Three days later, on July 24, I returned to the scene to
see if they had rebuilt the nest, and especially to find if, without
the stimulus or the aid of the larvae present, they would supply
these drops of sweet food in the new cells. The details of their
behavior in rebuilding are given elsewhere, but here I record that
of the nine colonies which did rebuild (eight of P. pallipes and one
of P. variatns) each nest had from fourteen to twenty-two cells,
all very shallow at this stage, but each and every cell contained its
globule of jelly along with its egg.

HONEY-GATHERIXC, IIAIMTS dK I'OI.ISTKS WASPS.

507

Investigation No. 4. — Records of two nests taken at Wesco,
Mo., on August 3, gave the following data. The first was a nest
of P. variatus, which contained thirty-five cells with eggs, sixteen
cells with larva? and forty-four with pupa?, and had ahout a dozen
adults on the nest. This large nest gave absolutely no evidence of
the presence of the honey. The other was a nest of P. pallipes,
which was in all probability the second one made by this colony,
since at this late date it had only small, newly-made cells which
would hardly hold a half -grown larva, and there were no cells large
enough to have harbored the five adults which were then on the
nest. These shallow cells, fifty-two in number, gave no evidence
of having a supply of honey, although about one half of them con-
tained eggs.

Investigation No. 5. — Since I had removed all of the accessible
nests at Wickes, the next scene of operations was at Meramec
Highlands, west of St. Louis and about thirty miles distant from
Wickes. These nests were gathered about two weeks later, on
August 6. 1920. In this collection, fourteen nests (all P. pallipes}
were knocked from the walls and eaves with long poles and carried

Nest.

Cells

Just
Begun.

Cells
Cont.
Eggs.

Cells
Cont.
Honey.

Cells
Cont.
Larvae.

Cells
Cont.
Pupae.

Total
Cells.

Remarks.

aa. . . .

10

35

35

—

3

48

All ten new (egglcss) cells, and

25 of the 35 cells with eggs

contained honey.

bb

—

23

8

32

14

69

cc . . . .

—

IO

6

25

28

63

dd....

— .

ii

6

21

22

54

ee. . . .

IO

20

25

32

45

107

All 10 new (cggless) cells, and

15 of the 20 cells with eggs

contained globules.

ff....

6

23

25

IS

7

54

Again, all <> empty cells con-

tained drops of honey.

gg

—

14

IO

38

20

72

hh....

—

54

35

2 **
•"• i

2O

101

Some of the globules im-

mensely large.

ii. . . .

—

12

IO

14

3

29

jj • • • •

—

IO

10

4

3

17

kk....

28

II

39

8

21

68

All new empty cells and all

cells with eggs supplied, i

to 4 drops per cell, some

very large.

II

—

57

56

1 6

25

<,8

mm.. .

—

36

15

36

65

137

nn . . .

3

4

3

—

7

NYw iu'~t with no workers.

14..

57

320

283

271

2/6

92 1

508

PHIL RAU.

home for examination. The data from these will also be more
convenient if presented in tabular form.

In this table we find that out of 377 cells containing or ready for
an egg, 283, or about 75 per cent, of them, contained drops of this
honey-like substance.

The next spring, the following additional experiments were

made.

Investigation No. 6. — On this date, April 28, 1921, I found
forty-two small nests of P. pallipes which had only recently been
begun by their queens. In only one cell of all these nests did I find
one drop of this honey suspended in the angle of the cell. This
one drop is sufficient to show that very early in the nesting season
the queens sometimes store this substance. At this season the
queens were not so bellicose, so instead of knocking down the nests,
at intervals for examination I climbed up to the nests and, with
the aid of a flash-light, examined the nests and contents in situ.

Investigation No.?. On May 13, twenty-nine nests, with the
number of cells varying from ten to eighteen, were minutely ex-
amined, and fourteen were found to contain drops of honey. Nest
No. i, with thirteen cells each containing an egg, had a drop in
each of two cells. One cannot say that this had been placed there
for the young about to hatch, for these two cells were adjacent
and were at the margin of the nest, whereas the first eggs to hatch
are as a rule in the center of the comb ; in this case the eggs in the
two center cells were even then much inflated and ready to give
forth their life. Nest No. 56 had ten cells all containing eggs, the
two central ones large and ready to become larvae. It was in pre-
cisely these cells, that the drops of honey were found. Nest No.
15 comprised fourteen cells having eleven eggs and three larvae
in the center cells. Two other cells adjacent to these had each a
drop of honey. The larvae were several days old, and no evidence
remained in their cells to tell whether or not they had had this food
in their infancy. The eggs provided with honey were inflated and
ready to hatch next. Nest No. 22 had ten cells each containing an
egg; the largest of these, already swollen with its embryo, was
provided with a drop of honey in its cell.

Thus we see four nests out of twenty-nine at this early date con-
taining honey-drops ; I suspect that the larva, upon emerging, gets

HONEY-GATHERING HABITS OF POLISTES WASPS. 509

its first meal from this food, and in three of four of the cases ob-
served on this date this material had been placed in the cells con-
taining first-laid eggs, the larvae from which would he first in need
of food. However, since only about one-eighth of the nests ex-
amined contained honey in even a. part of their cells, it seems very
likely that some of the young lack the advantage of this food, or
at least do not have it waiting for them before their birth. The
most interesting bit of evidence that this day gave forth, however,
was the one cell, which on April 28 had contained a drop of honey,
on this date, May 13, contained a small larva but no honey! The
period of incubation in this species has elsewhere been found to
be a little more than two weeks, so it is not surprising if the mother
cannot estimate exactly the date of the need of food for each of
her young.

Investigation No. 8. — The accompanying table summarizes the
details of the observations on twenty-seven nests on May 27 and
28, 1921. At this time, when the nests were larger and contained
more life, I found less of the honey; it occurred in only three of
the twenty-seven nests, and as usual only in the cells containing
eggs; out of 142 cells only 13 had drops of honey. It seems at
first puzzling that a mother would supply this extra provision to
some cells containing eggs, and not supply it to all, and that if one
mother is so provident, not all are likewise. All were /v////v.s-.

I previously formed the hypothesis that all young, when they
first hatch from the egg, must be fed this material probably pre-
digested by the mother; that frequently she is on hand at the time
and administers the close personally, and that sometimes she gets
an over supply or gathers it before the young are hatched, in which
event she relieves herself of it by placing it in the angles of the
cells containing the eggs which will soon hatch. The following
notes will to a degree bear out the first part of this hypothesis.

While restlessly waiting for a belated train at a rural flag-
station at 10 A.M., I fell to watching a queen which had charge of
a nest built on the wall of the station shelter. I noted that re-
peatedly she inserted her head deep into two cells and kept it there
far out of sight for periods of three to seven minutes. I knew that
trophallaxis occurs between the larvae and the workers, but this
queen had crowded her body so deep into the cells as to completely

PHIL RAU.

Nest.

Cells
Just
Begun

Cells
Cont.
Eggs.

Cells
Cont.
Honey

Cells
Cont.
Larvae

Cells
Cont.
Pupae.

Total
Cells.

Remarks.

5 1 ....

3

—

13

—

16

69--..

—

II

— •

5

—

16

gA..

—

5

—

6

— •

II

11...

—

2

—

1 1

—

13

12. . . .

—

7

7

7

—

14

Honey with all the eggs; none

with the larvae.

I . . .

—

4

— •

IO

—

14

2. . . .

—

IO

—

2

—

12

3- - - •

—

—

—

16

2

18

13...

—

8

—

IO

18

8... .

—

2

—

7

9

35---

—

7

—

8

IS

34- •

—

5

—

IO

IS

31...

—

6

—

12

18

54- •

—

18

6

14

32

Honey all in egg cells; this

nest had 3 queens.

14...

—

2

—

IO

12

I5--

—

—

10

2

12

21. ...

—

4

—

4

8

17.. .

—

6

—

6

2

14

58....

—

i

—

10

4

IS

59....

—

2

—

IO

I

13

2O. ...

—

2

—

IO

— •

12

23...

—

3

—

ii

—

14

21. ...

—

—

—

IO

—

IO

12...

— •

4

3

—

—

4

73-.

—

12

—

—

—

12

74-.

—

12

— •

—

—

12

22. ...

—

6

—

8

—

14

27 •

—

142

13

220

1 1

373

cover the thorax ; this indicated at once that the cell must contain
only a very small organism, and thus the present process could
not he trophallaxis. Presently she removed her head from the
cell from time to time and went about examining three other cells
which contained eggs ; six other cells nearby contained good-sized
larvae, but these she ignored. Her attention was always centered
about the two cells whose contents I could not see ; whenever she
left one or the other of these two cells, she would poke her head
momentarily into each of the three egg-cells and then go back to
the others. When visiting the egg-cells, she would always vibrate
the antennae, bring them close together and point them in front of
her head as she entered, insert the head for an instant and with-
draw, repeat the process exactly for the other two, and then return
and creep deep into one of the mysterious cells for several minutes.
When she withdrew from either of these, she moved her mandibles

HONEY-GATHERING HABITS OF I'OI.ISTKS XVASI'S. :; I I

v '

in a significant manner, hut very different from the " licking-her-
chops " manner of the workers after they take the liquid from the
larva?, for in the latter case not only were the mandibles in motion,
but also the labium, palpi, etc., were actively registering satisfac-
tion or function.

The approaching train brought the observations to a sudden
close. Hastily I pulled out the mother wasp and threw the gleam
of the flashlight deep into the cells, and lo ! in the far corner of
each lay a tiny larva, hardly larger than an egg ! I could solve the
mystery of her long stops in these two cells only by concluding that
she was feeding them with the material from her own mouth, since
I was sure she brought nothing to the cells in her jaws. I
pondered on the number of times she had poked her head inquir-
ingly into the three cells containing eggs ; then I realized that her
frequent visits here were probably for the purpose of ascertaining
if they had yet hatched or were ready for her attention. The
larvae she did not need to mind just now, but the new-born twin-
and the eggs had all her attention.

My most recent observations have led me to believe that the
queen does not deposit this honey with the intention of making a
self-feeder for the larva, so that when the egg hatches the larva
has only to thrust its head into it and imbibe, but rather that she
stores it here temporarily where it will soon be needed and where
she can easily administer it to the waspling as soon as it is ready.
The ground for this conclusion is the fact that these drops are al-
most always placed far below the egg in the cell, or on the opposite
wall, so that it is quite impossible for the new-born larva? to reach
them. I have seen dozens of young larva? in early spring nests
which had been brought into the laboratory, die with large drops of
honey before their very eyes but out of their reach; since the tail
of the larva is glued to the wall, its progression to the drop is im-
possible. In contrast to this, I have seen a few larva? survive on
this substance when it was artificially placed near their mouths. I
have also fed them on that taken from other cells ; they ate it
readily when served to them on a pin-head. 1 have also assisted
larvae to get this food by pressing the opposite wall of the cell
toward them, bringing the drops nearer to their mouths ; and in the
past two years I have actually seen two cells in which the drops of

512

PHIL RAU.

honey were placed very near to the larvse, and they were stretching
down and actually imbibing the substance.

The above instances indicate that the newly-hatched larva gets
its first meal from the mouth of its mother. This of course can
be independent of the storing of the honey as we 'find it in some
cells, but it might point very definitely to the origin of this habit.
How logical it seems for all the mothers to feed their new-born
young upon predigested food from their own crops, and how pos-
sible it could be for the crops to be filled too full or filled too soon.
In that event the next logical step would be for the mother to dis-
pose of this surplus by placing it in the cells where she would
normally get rid of it at an early date. While one does not wish
to speculate too far, one can also see the possibility of another
pretty adaptation, in that mothers with only a few offspring at
that stage can feed the delicate food to them personally, but when
the young are too numerous and the mother's attention must be
divided, food may be supplied to them in advance — a habit well
known among the solitary wasps. To go further, how possible it
is for a social species to show vestiges of a habit of the solitary
species,1 i.e., to supply the egg with food in sufficient amount to
carry it to maturity.

Investigation No. o. — On June 21, 1921, I examined the con-
tents of 20 nests of P. pallipcs. The egg-laying was over for the
time being. It seems that the queen of this species spends her time
at this season in caring for the young, and not in adding new cells
and depositing more eggs. It was then thought that perhaps the
building activities would be resumed after the workers emerged.
Only six of 'the twenty nests contained a very few eggs; hence
one would expect proportionately fewer drops of honey in the
cells. That is precisely the condition that I did find ; in all of the
twenty nests examined, only one was found containing one drop
of this material ; this was in an empty cell of a nest which con-
tained five cells with larvae, one with an egg, and five empty cells.

Investigation No. w. — This paragraph relates merely the acci-
dental discovery of honey in a nest of P. variatits. This nest was
in an inverted, rusty teakettle in a city dump-lot. My attention
was attracted to it by seeing the mother enter through a crack near

1 Bouvier (" Psychic Life of Insects," p. 335) says that the social insects
have developed from solitary ancestors.

HONEY-GATHERING HABITS OF POLISTES WASPS. 513

the bottom, where she was obliged to alight and creep under the
vessel. By carefully lifting the whole outfit I could observe the
contents. The nest was a strong one of about twenty cells with
eggs and larvae. Here P. variatiis had again demonstrated her
fondness for low situations by building her nest in this unusual
site near the earth.

On the morning of May 17, I surprised the mother by lifting
up the old teakettle to the level of my eye for examination. She
was for a time too much bewildered to move — and the sight that
met my eyes affected me likewise, so it was some time before I
recovered myself to set it down again. Each one of the eight cells
containing eggs had also a large globule of this shining, trans-
parent liquid, so large that it half filled the cell. The next morn-
ing, upon similar examination, I found the globules much reduced
in size and more viscous in nature. The mother was then on the
nest, where I had evidently caught her in the act of bringing in
more of it, for in her jaws was a large, shining globule, surpris-
ingly clear at this stage. Three days later these large, watery
globules were still present. The mother was again on the nest,
but this time she was chewing a yellow ball of caterpillar meat.
This was undoubtedly her own food, since I could see no larvae in
the cells to which she could feed it. Thus the evidence indicates
that from the very first, P. variatiis stores a sugary liquid, which
with evaporation condenses into a jelly-like substance; that she
adds to the store from time to time, and eventually the residue is
in the form of sticky balls or drops. It is interesting also that for
her own food the queen sometimes catches caterpillars and brings
them home to her nest to devour them, as a variation from her
nectar diet.

Two weeks later these cells were again seen to contain large
globules of liquid, like huge, yellow dew-drops. The eggs had all
hatched ; the larvae were still tiny. This set me to wondering
whether, in a large, dry field where there was no possibility of
getting water, the mother did not gather drops of dew each morn-
ing and store her supply of drink for the day. The next day I
was greatly disturbed to find that the ubiquitous small boys, who
had been spying on my strange behavior in the dump-lot, had dis-
covered the quaint nest and destroyed it.

514

PHIL RAU.

Investigation Xo. n. — At the end of May, 1921, only a few
small drops of honey were found in a large number of nests, to
be exact, only three in twenty-seven nests. In May, 1922, ex-
tensive examinations showed that this year the honey content was
so abundant as to be very conspicuous in a large proportion of the
nests ; twenty-five of the thirty nests each contained the drops in
many cells, and larger than usual in size. These were only in the
egg cells and the new cells ; none were in the cells with larvae.
Just in what way this year afforded conditions which produced this
excess, I can only surmise. It does seem probable that the abun-
dance of flowers or the climatic conditions would be factors in-
fluencing the supply. On cold, rainy days the queen seldom leaves
the nest.

Glancing at the data as a whole one sees at once that this honey-
storing habit is much more common in these three species of
Pollstcs than was generally thought to be the case. Furthermore,
it is probably not a local condition, since the collections from two
places thirty miles apart showed this characteristic about equally
well established.

Early in the course of this investigation I thought that this sub-
stance was the product of the large larvae. I had observed the
workers gathering the saliva-like secretion in large amounts from
the mouths of the larvae, and knew that the adult wasps seemed
to regard this as a delicacy and enjoyed it as food. So, since the
color was similar, I suspected that the worker, after the material
had perhaps undergone some chemical change in their gullets,
would place it in the. cells, so that the new-born waspling would
find a supply of the most delicate food for its first meal.

Soon the evidence from the observations began to militate
against this theory. The results of the third group of observa-
tions left my belief well-nigh untenable, since the groups of adults
whose nests were taken away built clusters of new cells at once,
and within three days had the majority of these tiny cells well
supplied with this substance. This was done after all their own
large larvae had been carried away. Of course the possibility re-
mains that the adult workers, of which there were plenty at that
time, might already have had their crops or gullets full of this,
gathered from the nestlings when the catastrophe came, and

HONEY-GATIIERIXC HABITS OF POLISTES WASPS. 515

that they retained this until they could deposit it in the new cells,
but this supposition seems far-fetched. Also in Investigation No.
2, nest " (/," we find a nest entirely without larvae, yet twenty-five
of the thirty cells of the egg stage contained honey. Surely the
workers would not venture to go to other nests and rob the larva;
of their salivary secretion ; moreover the queens had been stor
ing it at a time when there were no larvae from which they could
gather it. In consideration of these points I felt compelled to
doubt or even abandon the idea, and turn to the hypothesis that
these drops are vegetable matter, gathered from the flowers and
fed to the tiny larva or stored in the cells awaiting its use, like the
honey stores of other Hymenoptera. It is unfortunate that we
must thus abstractly speculate — that no one has yet actually ob-
served the wasps collecting, making or storing this material, so
that we may know with certainty just how it is clone and may fig-
ure out the relation of the habit to the life or even to the history
of the insect. Here, however, I can only continue to give the
evidence derived from the foregoing data, in the hope that it may
pave the way to conclusive observations in the near future.

In several of the nests observed, especially the new nests de-
scribed in the third lot, we have in all probability a very definite
clue to the nature of the material in these drops. In all nine nests
in the third lot, which we knew to be very new, and in other nests
where there was evidence that the cells had been made and stored
very recently, the globules were exceptionally large and of very
watery consistency. This gave rise to the idea that the material
is probably similar to that gathered by bees and that it must like-
wise be subjected to evaporation, as is the nectar in the bee-hive,
after which it retains the form of little, firm beads. In the large
nests the comparative age or the chronological order of the cells
could not be definitely determined, but one small, new nest lent its
evidence to this theory. The three central cells each contained a
drop, one of them condensed and bead-like, the two others in an
intermediate condition. The last cell, a new and incomplete one.
contained a huge drop of very watery substance. I placed the
entire nest in a tiny tin box to take it home for study, but the
large drop was scattered or absorbed by the paper walls t-n route.
This little nest seemed thus to show the natural evolution of a

516 PHIL RAU.

drop of this gelatinous material, from big, watery sweetish drops
in the newest cell to smaller, waxy beads in the older cells.

Thus the material from distinct localities shows that this honey
is used much more generally than investigators have heretofore
suspected. Furthermore, it is present not only at the end of the
season, provided to tide over some tardy larvae, but at all times
throughout the season, less frequently during the spring when
the queen alone tends the nest, and most frequently and in greater
abundance toward the end of the summer. Whether this greater
abundance in August is due to a larger supply of flowering plants,
or possibly to the greater leisure and number of the workers, or
whether they actually gather this material to feed certain larvae in
order to develop queens, remains for further investigation to de-
termine. The frequent lack of this food in the earliest cells of
the season might point toward this latter suggestion.

In concluding one would say of Wheeler's theory, that tht
honey-gathering habit in Polistcs is a vestigial instinct which is
called into play to carry over tardy larvae, can hardly apply to P.
pallipcs, P. annularis and P. variatus, since in these species the
drops of honey are to be found at times throughout the season.
The evidence seems to show that this instinct is not vestigial but
functional, and is probably necessary for the survival of the young.

The watery consistency of the drops might give a clue to a
solution of the evolution of the habit. We know Polistcs are
heavy drinkers, and probably they give water to the young, as do
honey-bees. It might happen that for convenience this water was
disgorged on the walls of the cells ; by and by a portion of nectar
from their throats was accidentally added, and then with evapora-
tion there remained but a small, transparent globule.

Sharp says,1 in speaking of the habits of social wasps ; " . . .
the eggs soon hatch and produce larvae that grow rapidly; the
labors of the queen wasp are chiefly directed to feeding the young.
At first she supplies them zvith saccharine matter, which she pro-
cures from flowers or fruits, but soon gives them a stronger diet
of insect meat. . . . The hornet is particularly fond of the honey-
bee." Unfortunately Sharp does not give us the specific name of
the wasp upon which he bases this statement, but that this is based

1 " Insects," Pt. II., p. 84. [Italics are mine.]

MONEY-GATHERING HABITS OF POLISTES WASPS. 517

careful observations on some species and not upon mere theory
is corroborated by my notes on Polistcs.

To theorize upon the phylogenetic position of this habit is a small
task, and evidently can be substantiated by direct observation.
It seems probable that the flower-loving insects evolved from non-
flower-frequenting species about the time of the appearance of
flower-bearing vegetation. The flower-visiting Aculeates, the bees,
were pollen- and nectar-gatherers. If we assume that the bees and
wasps had, 'way back in prehistoric time, the same ancestry, we
can gather that, phylogenetically speaking, honey-gathering is an
older trait in Aculeates than insect-gathering.

Here we have wasps that gather nectar for their own food and
honey for their young, and super-imposed upon these traits we
have the newer habit of insect gathering for the young. This
honey-gathering is not a recent addition to their feeding instincts,
but a vestige which proves their ancestry from the bees.1

A step further was made when probably certain bee-like, honey-
feeding Aculeates back in geologic time, from scarcity of flower
food or drought, or through pure " change of mind " or by acci-
dent, began to prey upon other insects that sought the same
flowers, for their own food or for food for the young in their
nests. Sharp tells us that the hornet is particularly fond of honey-
bees, which suggests the thought to my mind that the change in
habit from the use of flower products to meat may not have been
so wide and abrupt as the lines would indicate. It seems so easily
possible for certain ones to get a little additional food by licking-
one another's faces and mouths, just as we see Polistcs nibigiiwsus
doing today ; - in time of scarcity of food this habit might easily be
exaggerated even to the point of robbery or squeezing out the
sweet liquid from the honey-crops, as Fabre describes in the
modern behavior of Philanthus. From this stage it would indeed
be a poor Aculeate that would not eventually learn to abbreviate

1 I am aware that the opposite view is generally held, for Wheeler
(Scientific Monthly, 15: 68-69, 1922) says: "The structure and behavior of
the Sphecoids and Vespoids show that they must have arisen from what
have been called Parasitic Hymenoptera, and the structure of ants and bees
shows that they in turn must have arisen from the primitive Sphecoids or
Vespoids."

2 Paper soon to be published.

518 PHIL RAU.

the work hy making a perforation in the body wall to reach the
nectar crop directly (just as Robertson describes for Odyncnts
t'onuninatns that bites an opening into the base of the flowers of
Pentstcmon Iffvigatus}- for the same purpose, or to carry home
an entire insect, making it serve as a vessel to hold the honey, with-
out any further trouble. This puncturing of the living honey-pot
would lead to a taste for animal juices which we know now exists
in certain Pompilids. Even if the habit of carrying this prey
home had not by this time been acquired, it would be but a step to
learn to carry it to their young after they had learned to enjoy it
themselves.

An additional bit of evidence to indicate that the nectivorous
habit probably antedates the carnivorous habit in these insects is
that, as Sharp tells us and as we have seen in the observations just
recorded, the mother feeds to the new-born young first a nectar
compound and later, animal food. If we accept the principle that
ontogeny recapitulates phylogeny, it is easy to find significance in
the fact that the immature organism passes through the nectar-
feeding stage before it arrives at the carnivorous stage. Among
the surviving descendants of the primitive flower-loving Aculeates,
some have digressed to the extent of finally evolving into carnivor-
ous creatures with certain feeding traits that are vegetative.

In my opinion, the honey as I found it was merely stored honey
that had been carried in the crop of the queen or the worker for
the purpose of direct feeding into the mouths of the young, and
this over supply was deposited for convenience on the cell wall.
Since these drops are not present at all times in all cells with eggs,
in the nests of the three species of Polistcs, the evidence lends some
merit to the theory that these drops are stored by law of supply
and demand : a queen having many young larvae to care for must
needs fill the crops of the young directly from her own gullet and
has no surplus to store in drops ; on the other hand, when the eggs
hatch at intervals, the mother has sufficient time during the period
of watchful anticipation to store up a surplus. Another factor in
regulating the number and size of the drops is the abundance of

2 In order to get the nectar which she could not ordinarily reach, this
wasp cut a hole in one side of the base of the tube with her sharp jaws and
inserted her tongue ; then she cut a hole in the other side and again inserted
her tongue. (Trans. Acad. Sci., St. Louis, 5: 569, 1891.)

HONEY-GATHERING HABITS OF POLISTES WASPS. 519

flowers, and still another reason for the presence of so much in
the latter part of the season may he that then the workers are also
on hand to help bring it in.

I often wonder if the honey-storing habit in the honey-bee was
not evolved in the same fashion. Is it not possible that originally
nectar was gathered and the regurgitated material fed direct to the
larvae; that finally, when workers became abundant in the colony
and more was brought in than could be consumed, they resorted
to dumping it into cells without larvae, and that this may have been
the forerunner of the honey-storing habit? A contrasting trend
of development may be found in the honey-gathering ants, Myr-
inccocystus horti-deoruwi, which do not eject this substance, and
their distended abdomens as they hang to the roof of the honey-
chamber are the actual store-houses of the honey.

The reader must bear in mind that the foregoing explanations
of this phenomenon are only speculative, but I have tried to use
as much logic as I could instil into the ideas. In this I take refuge
under the justification of C. H. Eigenmann, when he says: " The
imagination is in Biology as elsewhere the guiding spirit. The
trouble is our imaginations are sometimes too heavily loaded with
statistics, and at other times they fly without the balancing kite's
tail of facts. The Palaeontologists have contributed to speculative
Zoology because their imaginations have been kept alive by bridg-
ing their numerous gaps, and because they have not been hampered
by too great a wealth of material.''

34

ENDODERMAL FLAGELLA OF HYDRA O LI G ACT IS

PALLAS.

PAUL R. BURCH.
UNIVERSITY OF VIRGINIA.

The subject of flagella upon the endodermal cells of hydra was
first opened prior to 1871 by Hatchek, who figured a single taper-
ing flagellum upon each cell, all of which he considered similar.
F. E. Schulze ('71). figured a single flagellum to each cell of an
optical section of a tentacle. Kleinenberg ('72) in transverse
sections of the living animal observed one or more flagella in con-
nection with several cells and claimed that the flagella were not
fixed structures but that they could be protruded and retracted
again, while at the same time the cells sent out pseudopodial
processes. T. J. Parker ('80) saw one, two, or three cilia on cell
after ceil. Because of this observation he drew the inference that
the endoderm was ciliated throughout. K. C. Schneider ('90)
distinguished at least six types of endodermal cells ; but he stated
that only three of these types bore flagella. The digestive or
epithelio-muscular cells usually bear two flagella, which project
into the enteron. The glandular cells bear twro or three flagella.
The sensory cells bear one, two or no flagella. Schneider gave
us, therefore, the first proper description of the flagellated condi-
tion of hydra's endoderm. Haclzi ('09) did not carry the knowl-
edge of the flagellated condition of the endoderm beyond
Schneider. I have been able to corroborate the work of Schneider
and to carry his observations a step further by the use of the fol-
lowing method.

Hydra oligactis was macerated by Mundie's ('26) method. In
this case a hydra was placed upon a slide and the water drawn off
until but a film remained covering the polyp. The slide was then
placed over the mouth of a bottle containing Looper's fixing fluid
(made up of equal parts of 95 per cent, alcohol, glacial acetic acid
and 40 per cent, formalin1). At the end of eight or ten minutes

1 This method was developed in this laboratory by Dr. J. B. Looper.

520

KXDODERMAL FLAGELLA OF HYDRA. 52 1

the polyp was rinsed in one or two drops of water, as much of this
water as possible drawn off and a drop of Gramm's iodine solu-
tion added. The polyp was teased into fragments with needles,
a drop of 40 per cent, glycerine added and a coverglass applied.
The fragments were then examined under the oil immerson ob-
jective. Being in glycerine, the cells could be preserved for a
relatively long time, a month or more. During this time the
flagella persisted and the cells did not deteriorate. The tissues
may even be further treated. For example, if the- iodine fades,
more iodine may be drawn beneath the coverglass. I have also
carried one per cent eosin-licht gruen (95 per cent, alcoholic)
solution beneath the coverglass and have, in this manner, stained
nuclei, food vacuoles, flagella and other details well. In passing
it may be stated that the use of the licht-gruen solution brings out
conspicuously the pseudopodia of the epithelio-muscular cells of
the endoderm.

The epithelio-muscular cells are columnar. A myoncme runs
at right angles to the polyp's axis through their broad bases. The
distal end bears one or more flagella. There are many food
vacuoles in epithelio-muscular cells from well-fed specimens. The
secreting cells of the general endoderm are club shaped with the
smaller end directed toward the mesoglea. The distal end bears
one or two flagella. The distal half of the cell is much vacuolated
and in well-fed specimens these vacuoles contain darkly staining
material called by Schneider, " Sekretballen." The basal end is
darkly granular and bears no myoneme. The sensory cells are of
the tall columnar type — almost filamentous. Each contains an
ellipsoidal nucleus. These cells bear at their bases slender
nodulated processes similar to those figured by 1 lad/i, in Table
II., Fig. 7 and 8. In this type of cell we encountered a flagellum
as did Schneider (Fig. 3).

I now come to the point at which my work goes beyond the
work of Schneider upon the flagella of hydra. Some of the in-
vestigators describe the flagella as being tapering protoplasmic
processes. Schneider shows them to be slender and of uniform
caliber from base to tip. He does not, however, show a structure
that is typically found associated with a flagellum, vis., a blepharo-
plast. All of my preparations show the flagella of epithelio-mus-

522

PAUL R. BURCH.

- - -

\

"V;

'

•,. ^^^l ;:::Jr;@;4
i ^&55^-v-:;"

tv.

I
I

I

I

nuc*

my.

|,

1 •'• • ;' !J:;M

. • •-; • •;. .',

^::,'r\

}>:^m

. > i

^B

f

FIG. I. Epithelio-muscular cell; Wr/>/*.. blcpharoplast ; //., flagellum ;
f ood-vacuole ; >ny., myoneme ; nnc., nucleus. X 1000.

FIG. 2. Secreting cell.
FIG. 3. Sensory cell; up., nodulated process.

EXDODERMAL FLAGELLA OF IIYDR \. 523

cular cells, glandular cells, and sensory cells to be associated with
a blepharoplast. My Fig. i is of an epithelio-muscular cell that
shows two flagella each extending into the cell's cytoplasm and
ending upon a blepharoplast (Fig. I, blcph). Fig. 2 shows, like-
wise, that in a secreting cell the flagellum enters the cytoplasm and
ends upon a blepharoplast. Finally Fig. 3 indicates that the flagel-
lum of a sensory cell enters the cytoplasm and terminates in a
blepharoplast.

SUMMARY.

The cells of hydra's endoderm — epithelio-muscular, secreting,
and sensory — are flagellated. The flagella are typical in that they
are non-tapering lash-like processes which terminate in blepharo-
plasts.

LITERATURE.
Hadzi, Jovan.

'09 Uber den Nervensystem von Hydra. Arb. a.d. Zool. Inst. d.

Wien, Vol. 17.
Kleinenberg, N.
'72 Hydra, eine anatomisch-entwicklungsgeschichtliche Unter-

suchung. Leipsig.
Schneider, K. C.

'90 Histologie von Hydra fusca mit besonderer Beruchtigung des
Nervensysten der Hydropolypen. Archiv f. mikr. Anat., Bd.
35.

'02 Lehrbuch der vergleichenden Histologie der Tiere.
Schulze, F. E.

'71 Ueber den Bau und die Entwicklung von Cordylophora lacusti-is.

(Allman), etc., Leipsig.
Mundie, J. R.

'26 Maceration of Green Hydra, Science, Vol. LXIV., Xo. 1667,

Dec. 10, 1926.
Parker, T. J.

'80 Histology of Hydra. Proc. Roy. Soc., London.

A NEW HISTOLOGICAL REGION IN HYDRA
OLIGACTIS PALLAS.

WM. A. KEPNER AND LULU MILLER,
UNIVERSITY OF VIRGINIA.

Three anatomical regions are to be recognized in Hydra: (i)
the oral end bearing tentacles, peristome and mouth; (2) the
middle region; and (3) the basal region. In this oral region
Kepner and Hopkins ('24) described, in detail, the peristomal
endodermal glands that lie about the mouth. They also demon-
strated the presence of sphincters at the bases of the tentacles.
These sphincters operate against pressure from the enteron to-
wards the tentacles but not in the reverse direction. The histology
of the middle region has recently been intensively studied by Hadzi
('09). In this region he finds three types of cells presented by
the endoderm : ( i ) epitheliomuscular cells, that are capable of in-
gesting small food-particles; (2) gland-cells, which lay down
enzymes upon larger masses of food within the enteron; (3)
(3) sensory cells. Hadzi, in this respect followed closely
Schneider ('90), ('02) who had seen that each of these types of
cells bore one or two flagella. Burch ('28) has carried this one
step further in demonstrating that the flagella of the three types of
endodermal cells arise from blepharoplasts. Not sufficient atten-
tion has been given, as yet, to the peculiar cells that lie laterally
disposed in the basal region of Hydra. These cells are highly
vacuolated and larger in calibre than are the endodermal cells of
the middle region. In addition to these features, the endoderm of
this region is peculiar because of the absence of gland cells. It is
not yet clear to us that these cells bear flagella.

The lateral ectoderm of the basal region of hydra does not dif-
fer from that of the middle region. The ectoderm of the basal disc,
however, has long been known to be peculiar. Korotneff ('80)
wrote " Die Epithel-Muskelzellen des Fusses unterschieden sich
von den ubrigen Ectodermzellen. Sie besitzen eine cylindrische
Form, enthalten eine stark lichtbrechende Fibrille und zeigen in
ihrem oberen Drittel eine gleichfalls stark lichtbrechende mucose
Ausscheidung, clurch welche die Anheftung des Thieres an fremde

*b>

524

A XKW IIISTOLOGICAL REGION IN HYDRA. 525

Korper bedingt \vird " (s. 165). We have not gotten beyond this
early interpretation of the histology of the ectoderm of the basal
disc.

With reference to the endoderm of this basal <li>r. however, \ve
have something to add.

It had been noticed in this laboratory for sometime that hydras
are sometimes found at low levels in aquaria with small gas-
bubbles attached to the basal discs. Mr. George Dare made a
careful series of observations upon isolated specimens and found
that this gas was developed at the basal disc of hydra independent
of the presence of bacteria or other organisms and of temperature
changes. We then undertook the histological study of gas-
elaborating specimens. We were able to make longitudinal sec-
tions that involved the complete or unbroken wall of the gas-
bubble. This was done by gently transferring a hydra, that had a
small bubble at its end in a drop of water to a slide. The slide
ivas then placed upon a pinch of salt, that lay upon a block of ice,
in such manner that the drop containing the hydra lay immediately
over the salt. This resulted in rapid lowering of the temperature
of the hydra. When ice crystals began to form about the margin
of the water in which the hydra lay, it was flooded with Bouin's
fluid. Thus the specimen was fixed and sectioned. The sections
show the wall of the gas-bubble to be fixed to the marginal cells of
the basal disc (Fig. i, g}, and the general or central cells to be in
contact with the lumen of the gas-bubble, there being neither
secretion material, bacteria, nor other substance lying over their
free ends (Fig. i, //). It was thus seen that the wall of the gas-
bubble is composed of the mucus-like secretion of the basal disc.
The basal ectoderm is, therefore, concerned with the elaboration of
both the mucus-like material with which the hydra fixes itself to
some submerged surface, and with the elaboration of a gas.

Our chief point of interest, however, lies not in the ectoderm of
the basal disc but in the endoderm ; for here we find a feature that
has not been described. The endodermal cells of the basal region
are larger than those of the middle and oral region, and bear great
vacuoles. They do not ingest food particles (Fig i, c). In this
respect they again differ from the middle and oral regions' endo-
dermal cells. Until recently these cells were supposed to overlie
the basal disc's ectoderm Indeed Curtis and (Juthrie (/_'/) have

526 WM. A. KEPNER AND LULU MILLER.

illustrated these peculiar cells as lining the fundus of the enteron.
In our gas-secreting specimens we find that the fundus of the
enteron is lined with endodermal cells that bear numerous food-
vacuoles. They are tall, columnar cells having myonemes in their
bases. They resemble the epitheliomuscular cells of the middle
region of the body. We have not been able to isolate them for
maceration methods and cannot, therefore, say that they bear
flagella. But they so closely resemble the epitheliomuscular cells
of the middle region that our inference is that they do bear flagella.
The presence of these cells, therefore, in the fundus of the enteron
gives us an epithelial region that is like that of the oral and middle
regions of the body except that there are no gland-cells in this
basal endoderm.

The new histological region that we have thus discovered lies in
the wall of the basal disc. The ectoderm of this region we have
found to conform to the descriptions of earlier investigations ; but
the endoderm is peculiar in that it is an epithelial disc composed
of columnar epitheliomuscular cells, that contain food-vacuoles,
and which presents no gland-cells.

The presence of this endodermal disc, that is active in food-
appropriation, is of interest when we 'keep in mind the dual
function of the basal disc. The elaboration of gas takes place
relatively rapidly. The rapid elaboration of gas would, therefore,
involve rapid metabolism. The basal disc, then, is a region of
relatively high metabolism. The presence of this food-getting
endoderm in this region falls in line with the observation of
Tannreuther ('09) when he says that " Those endodermal cells of
the region of growth, ' the formation of buds and sexual organs ',
are the most active in ingesting partly digested food from the
enteron and preparing it for diffusion into the ectodermal cells "
(p. 211).

The basal disc has, therefore, a peculiar endoderm correlated
with the dual function of mucus- and gas-elaboration.

The elaboration of gas is done in order that the specimen may
be lifted in the water. The gas is discharged into the sac of mucus
until the bubble formed lifts the animal to the surface of the water.
As the polyp rises it has its basal end directed up and oral end
pending. When the bubble encounters the surface film of water,
its wall ruptures and forms a somewhat circular disc of mucus,

A NEW HISTOLOGICAL REGION IN HYDRA. 527

that is closely applied to the surface film of the water to form a
raft by which the polyp hangs (Fig. 2, g). Wave-action will
destroy this raft and then the specimen will sink. Or the hydra
may ahstrict itself from the raft of mucus and sink.

SUMMARY.

We have observed that the basal disc's ectoderm in hydra not
only secretes adhesive material, with which the polyp fixes itself
to some submerged surface, but that it also under certain con-
ditions elaborates a gas. This gas is caught within the mucus-
like secretion of the basal disc and retained therein. The bubble-,
thus retained, increases in size until the hydra is lifted to the sur-
face by it. At the surface of the water the retaining vesicle of
mucus breaks and spreads as a circular raft from which the hydra
hangs beneath the water's surface.

The ectoderm of the basal disc is thus seen to have a double
function. Associated with this region of the ectoderm of hydra
there has been found a pecular region of endoderm. The en-
doderm of this region is characterized by its component cells
having the appearance of the epithelio-muscular cells of the oral
two thirds of the body as over against the highly vacuolated
epitheliomuscular cells of the aboral third of the body.

LITERATURE.

Burch, Paul M.

'28 Endodermal flagella of Hydra. BIOL. P-ru.. Present Ynl.
Curtis, W. C. and Mary J. Guthrie

'27 Textbook of General Zoology. New York.
Hadzi, Jovan

'09 Uber das Nervensystem von Hydra. Arb. a. d. Zool. Inst. YVien..

i".
Kepner, Wm. A. and D. L. Hopkins

'24 Reactions of Hydra to chloretone. J. Exp. Zool. Vol. 38.
Korotoneff, A.

'80 Anatomischer biologischen mid embryologisch.cn Beobachtungcn an

Hydra. Zool. Anz., Bd. 3.
Schneider, C.

'90 Histologie von Hydra fnsca mit besondercr Berucksichtigung des
Nervensystems der Hydropolypen. Arch. f. micr. Anat. Bd. 35.

'02 Lehrbuch der vergleichended Histologie der Ticre. Jena.
Tannreuther, Geo. W.

'09 Budding in Hydra. BIOL. BULL., Vol. 16.

528 WM. A. KEPNER AND LULU MILLER.

EXPLANATION OF PLATE.

FIG. I. Longitudinal, axial section of hydra, a, peristomal gland-cells
of endoderm ; b, gland-cell of middle region's endoderm ; c, epitheliomuscular
cell of endoderm, bearing two flagella and two f ood-vacuoles ; d, epithelio-
muscular cell of endoderm ingesting a food particle ; c, epitheliomuscular
cells of lateral endoderm of basal region ; /, endoderm of basal disc ; g,
mucus-wall of the gas bubble ; h, basal disc's ectoderm that is exposed to
lumen of gas bubble ; i, ovary ; k, testes.

FIG. 2. Basal disc of polyp at surface film of water, s, surface film of
water ; g'g', circular raft formed by ruptured wall of gas bubble now ap-
plied to surface-film of water; h', ectoderm of basal disc that is now in con-
tact with air.

BIOLOGICAL BULLETIN. VOL. LIV.

WM. A KEPNER AND LULU MILLER.

HISTOLOGICAL FEATURES CORRELATED WITH GAS
SECRETION IN HYDRA OLIGACTIS PALLAS.

\V.\I. A. KEPNER AND W. L. THOMAS, JR.,
UNIVERSITY OF VIRGINIA.

The ectoderm of the basal disc of hydra ordinarily elaborates
mucus by means of which the polyp adheres to some submerged
surface. It has recently been observed that this same ectoderm
may, under certain circumstances, elaborate gas. This gas is held
within a sack of mucus (Kepner and Miller, '28). This fact has
suggested to us that there may be some interesting histological
change correlated with the elaboration of gas. Kepner and Miller
have found, to begin with, that there is a peculiar region in hydra's
endoderm in the basal disc. This endoderm supplies the food
demanded by the metabolism that yields the two products of the
basal disc's ectoderm — mucus and eas.

O

Their observation concerning this local accumulation of food
was interesting; for it carried the point, arising out of Tann-
reuther's ('09) observation, a step further. Tannreuther's ob-
servation was that in the region of a developing gonad or of an
incipient bud there were local and pronounced accumulations of
food within the endoderm. Yoder ('26) found that there was a
marked accumulation of glycogen in these regions of most active
metabolism. Finally Kepner and Hopkins ('24) observed that
there was not a wide diffusion of material absorbed by the en-
doderm. For example, they recorded that " there is no extensive
diffusion of absorbed chloretone through the tissues of the body.
A diploblastic animal, therefore, cannot possess anything com-
parable to a circulatory medium" (p. 448). Our observations
further strengthen the hypothesis that local 'needs must be met
locally.

The detailed histology of the basal third of a hydra, that had
not been secreting gas, shows the lateral endodermal cells to be
stout and highly vacuolated (Fig. i, <?)• The extent of distribu-
tion of these lateral cells varies greatly in different specimens and

529

530

\YM. A. KEPNER AND \V. L. THOMAS, JR.

perhaps also in the same specimen at different phases of the
polyp's activity. The cells of the endoderm of the basal disc are
more slender and present a denser cytoplasm than do the cells of
the above lateral endoderm. They also carry within their cyto-
plasm food vacuoles (Fig. i, /). Food-vacuoles are not present
in the lateral endoderm of the basal part of hydra. The lateral
ectoderm of the basal third does not in any manner differ from
that of the general ectoderm of the body. The ectoderm of the
basal disc, however, presents features that are peculiar to it. In
the first place, there are no nematocysts in this region, of the outer
epithelium. These cells, moreover, when actively discharging
mucus have conspicuous inclusions within their cytoplasmic bodies.
These may be designated secretion-granules (Fig. I, m g). These

FIG. i. Partly diagrammatic, axial section of base of hydra that had
been elaborating mucus in order to fix itself to substratum, c, lateral
endoderm of basal third of body ; /, peculiar endoderm that lines the f undus
of the enteron and contains f ood-vacuoles ; m, mesoglea ; ing, mucus-granules
arising within the mucus-secreting cells of ectoderm. X 650.

cells extend over the entire basal disc. All attached hydras will
not show the secretion-granules in the ectodermal cells through-
out the extent of the basal disc. A specimen that has long been
attached to the substratum may not yield, when fixed and stained,
sections that display secretion-granules. Only specimens caught
at the time they are fixing themselves to a substratum will present

GAS SECRETION IX IIVDKA OU< iACTIS I'AI.r.AS. 531

these secretion-granules in sections (Fig. i, in g) . The most
emphasis must here be placed upon the condition of the mesoglea.
In a mucus-secreting specimen, the mesoglea of the basal disc
differs in no manner from that of the body proper. It presents a
uniform, unbroken contour (Fig. I, m). All this in no manner
presents anything new concerning the histology of hydra.

In the gas-secreting hydra some conspicuous histological fea-
tures appear that have not been recorded. In the first place, the
endodermal cells of the basal disc appear to be larger and more
active in the gas-secreting specimens than in the non-secreting
ones. Next we find that the ectodermal cells of a wide peripheral
region of the disc elaborate mucus. These, in other words, do
not have their usual function changed. The axial cells in the
basal disc's ectoderm, however, do have their function altered.
They no longer present secretion-granules and therefore stain (in
haematoxylin) less than do the peripheral cells of the disc. We
have now an epithelium the periphery of which elaborates a re-
taining wall of mucus, while the axial region secretes gas into the
mucus to form a buoy or lifting float for the polyp (Fig. 2, me and
gc}. The most conspicuous feature of the gas-secreting polyp lies
in the mesoglea's modification. The mesoglea, in this disc at this
time, becomes greatly swollen and highly vacuolated as though
the endoderm had flooded it with a deposit of metabolic substance.
Within this broken region of the mesoglea there appears, in fixed
material, a substance that suggests a coagulated plasma (Fig. 2,
•in'). The presence of this plasma within the mesoglea may signify
one of two things: (i) It may be that food is being deposited
there by the endoderm in order to meet the demands of an intense
metabolism taking place during gas-elaboration, or (2) It may be
that metabolic substances are being dammed back from the rela-
tively active axial ectoderm while it is elaborating gas. If the
first alternative be correct, a plasma, as it were, is thrown down
locally into the axial mesoglea of the basal disc in order that the
gas-secreting cells may be abundantly supplied with food during
the peculiar metabolism involved. If the second alternative be
correct, it means that the axial ectoderm is throwing metabolic
wastes into the mesoglea. Ordinarily the ectoderm discharges its
metabolic wastes externally through a moist mucus when the

532

\VM. A. KEPNER AND W. L. THOMAS, JR.

polyp is fixed, or directly into the water when the polyp lies free.
\Yhen, however, a layer of gas is deposited upon the outside of
this basal, axial ectoderm the osmotic drainage is blocked. The
presence of the gas no longer lets the metabolic wastes drain by
means of an osmotic exchange through the free ends of the axial,
ectodermal cells. Hence metabolic wastes, instead of metabolic
food, back into the mesoglea to flood it and form the vesicle that
we have observed. As the elaboration of gas advances the vesicle
of the mesoglea, together with its included plasma, decreases until,
at the time of the gas-buoy's attaining its maximum size, the
mesoglea has returned to the condition characteristic of the gen-
eral mesoglea.

FIG. 2. Axial section of base of hydra that had been elaborating a gas-
bubble, c, lateral endoderm of basal third of body; /, peculiar endoderm
that lines fundus of enteron and contains f ood-vacuoles ; gc, gas-secreting
region of ectoderm; iny, mucus-granules; g, mucus wall of gas-bubble;
me, mucus-secreting region of ectoderm ; m unbroken, but deeply staining
region of mesoglea ; m ', plasma-like material within distended region of
basal mesoglea. X 650.

It matters not which of the above alternatives be the correct
interpretation, the interesting point may be made that, in the basal
disc of hydra secreting gas, we have a situation arising that makes
a peculiar demand upon the passive mesoglea. As a result, the
mesoglea is sometimes flooded either with metabolic food or with
metabolic wastes. Thus we have the mesoglea foreshadowing the

GAS SECRETION IN HYDRA OLIGACTIS PALLAS.

function of a true circulatory medium such as is found in the
plasma of the Turbellaria or even in that of the hlood and lymph
of the higher triploblastic animals.

SUMMARY.

The basal disc of hydra has a two-fold secretory function: (i)
It secretes a mucus by means of which the polyp is anchored.
(2) It secretes gas that is retained within a wall of mucus by
means of which the polyp is lifted to the surface of the water.
This dual function of the basal disc places a peculiar metabolic
demand upon the disc's endoderm, ectoderm and mesoglea. As
a result of this peculiar metabolism and the conditions arising out
of gas-secretion, the mesoglea becomes flooded with a plasma and
thus handles either metabolic food or metabolic wastes in a man-
ner that foreshadows the plasma of the circulatory media of
triploblastic animals.

LITERATURE.

Kepner, Wm. A., and D. L. Hopkins.

'24 Reactions of Hydra to Chloretone. J. Exp. Zool., Vol. 38.
Kepner, Wm. A., and Lulu Miller.

'28 A New Histological Region in Hydra. BIOI.. lU'i.i.. Present Vol.
Tannreuther, Geo. W.

'09 Budding in Hydra. BIOI.. BTLL., Vol. 16.
Yoder, M. C.

'26 The Occurrence, Storage, and Distribution of Glycogen in Hydra
z'iridis and Hydra fusca. J. Exp. Zool., Vol. 44.

35

THE LIGHT-RECEPTIVE ORGANS OF CERTAIN

BARNACLES.1

DORIS EDNA FALES.

From Western Reserve University, Cleveland, Ohio and the Marine
Biological Laboratory, Woods Hole, Mass.

CONTENTS.

Introduction 534

Materials and Methods 535

The Nauplian Stage 535

The Cyprid Stage 53$

The Adult Stage 54<>

Summary 545

Literature Cited 546

INTRODUCTION.

Although brief references to the light-perceptive organs of
barnacles are found throughout the literature, no really compre-
hensive study has been made of them since that by Pouchet and
Jobert in 1876. The relationships between the " eyes " in the
larval stages and those in the adult have never been fully worked
out. Accordingly the object of this investigation has been a study
of the various light-perceptive organs found in the larval and adult
barnacles and the determination of their relationships.

Investigations relating to the light-receptors of barnacles are
usually found combined either with studies of other structures
of the barnacle or with studies of the eyes of Crustacea in gen-
eral. Among the former type Darwin's " Monograph of the Cir-
ripedia " (1851 and 1854) is outstanding. More recently Gruvel
(1905) has published a detailed account of the morphology and
taxonomy of this group. Grenacher (1879), Claus (1891),
Brooks and Herrick (1892), and Demoll (1917) have studied the
eyes of Crustacea. General problems in this connection have been
taken up by Parson (1831), Patten (1886 and 1887), and Mark
(1887).

1 The author is indebted to Dr. J. P. Visscher of Western Reserve Uni-
versity for the suggestion of the problem and for assistance during the
course of the investigation.

534

TIIK LICHT-RECEPTIVK OKOAXS OF KAKXACLKS.

MATERIALS AXD METHODS.

The ivory barnacle, Hahtints churncns, and the rock barnacle,
Balanus balanoides, were used in this study. Collections were
made at \Yoods Hole, Mass., during the spring and summer of
1926. Living forms were observed at this time and material was
preserved for subsequent study.

Adults of Balanus cbnnicits brought to the laboratory often con-
tained Nauplei nearly ready for hatching. Unhatched Nauplei
were obtained by removing the ovigerous lamellae from these
adults, and free swimming forms were collected after hatching.
A few metanauplei and Cyprids were obtained from towings madi
in the vicinity.

In order to test the actual light-perceptive function of the adult,
the shell was broken so that the eyes could be exposed. Speci-
mens with the shell thus broken were tested to make sure that
the light reaction was unimpaired. The eyes were then removed
by a hot needle. In all cases included in the data, the eye ad-
hered to the needle and was removed without apparent injury to
the surrounding tissue.

Material was fixed in picro-acetic formol (Bouin's) or in 10
per cent, formalin, and was either stained in Ehrlich's hema-
toxylin or mounted unstained. Most of the study was done by
means of sections which were prepared by the ordinary paraffin
method and cut from §/* to IO/JL in thickness. Grenadier's tech-
nique for the removal of pigment (Lee, 1890) was used on sec-
tions of the Cyprids. In addition a modification of Cajal's silver
nitrate technique as developed by Hess (1925) was used in the
study of the adult eye.

THE NAUPLIAN STAGE.

The formation of the median eye in Cirripedes has never bivn
worked out. In I>alanns churncns the eye first appears in the
unhatched Nauplius as an elongated area of reddish pigment.
Certain authors, for example Darwin (1851), have believed that
the nauplian eye might arise from the union of two anlagen but.
at least in Bolanns clutrncus, this does not appear to be the case.
A large number of specimens in which the eye was just forming
were studied but in all cases it appeared as a single area with no

536 DORIS EDNA FALES.

evidence of a double origin. In succeeding stages the pigment
becomes darker and the eye appears as a bilobed structure.

The morphology of the nauplian eye of Cirripedes has usually
been treated incidentally in connection with studies of the com-
plete animal and consequently its detailed structure is known in
only a few forms, chiefly among the Lepadse. The pigmented
area in the anterior region of the Nauplius was noticed by the
earliest authors but their descriptions mention only the general
shape of the pigment cup. A fairly complete study of the light
receptors in the Cirripedia was made by Pouchet and Jobert in
18/6. These authors described the nauplian eye as formed of two
parts each of which represented a simple eye and was composed
of a pigmented body with finely granular, rose-colored pigment
and small oval bodies which stained black with osmic acid and

P.O.

o.c

FIG. I. Diagrammatic drawing of the median eye in the cyprid stage
(longitudinal vertical section X 500). c.c.. cortical cells; c.t., connective
tissue; ;;., nerve; [>.c., pigmented cells; i'.c. visual cells.

which they thought might be designated as lenses, although the
analogy was somewhat questionable. The median eye of Crus-
tacea in general was described by Demoll (1917) as an inverted
eye composed of two pigment cells forming a cup in which were •
the visual cells Which extended proximally in a nerve fibre. These
visual cells were in the nature of " Stifchensaume " and the whole
eye cup was lined with fine, reflecting scales which formed the
" tapetum." A cellular lens was said to be present in some forms.
\Yhen sections of fully developed Nauplei are studied the
median eye (Fig. T) can be seen lying between the so-called
"cerebral ganglia." The eye at this period measures about I5/A
in diameter and, due to this small size, an interpretation of its
structure is very difficult. The central region is composed of

T1IK UCHT-RECEPTIVE ORGANS OF BARNACLES. 537

heavily pigmented cells which form a bi-concavr pi^nu-im-d cup.
In each concavity there is a non-pigmented area coated hy cells
with large nuclei, similar to ones which form the cortical area of
the cerebral ganglia. The non-pigmented area appears to be made
up of two visual cells surrounded by a small amount of connective
tissue. Each area sends off a nerve which is probably formed by
an extension of these visual cells as in the adult. These nerve--
leave the posterior region of the eye and by very short connec-
tives enter that part of the ganglia which is slightly posterior to
the eye.

A lens as described by Claparede (1863) for Lcpas is not
present in Bakinns. The non-pigmented area composed of visual
cells and connective tissue was probably mistaken for a lens by
Pouchet and Jobert (1876). This description is not in accord
with that given for the median eye of other Crustacea by such
authors as Hesse (1901), and Demoll (1917). The eyes of the
two Cirripedes studied show ganglion-like visual cells instead of
the " Stabchen " or " Stiftchensaume " described bv these authors.

j

While it is difficult to determine accurately the morphology of the
median eye at this period, the structure, as described, agrees fully
with that observed later in development. Hanstrom (1927) in a
recent article has described a somewhat similar structure in the
larval eye of NynipJion s/ronii. Whether the lack of agreement
concerning the structure is due to error on the part of the fore-
going authors or whether there is a real difference in the median
eye of these Cirripedia is a point which has not yet been deter-
mined.

In addition to the median eye characteristic of the Xauplius.
the Metanauplius of Balaints cbuniciis has a pair of compound
eyes which are generally known as the cyprid eyes since they are
the most noticeable light-perceptive organs of that stage. The
cyprid eyes in the Metanauplius appear on either side of th^
median eye near the base of each of the first pair of nauplian ap-
pendages. Their structure at this period is superficially the same
as in the Cyprid but material was not available for histological
verification of the appearance. The nauplian eye retains its char-
acteristic position and appearance during the metanauplian stage.

In common with the Nauplei of other barnacles those of Hnlann*

538

DORIS EDNA FALES.

cbiinicus show definite reactions toward light. In a dish contain-
ing Nauplei these were ordinarily found congregated in the area
having the greatest illumination. Occasionally a few individuals
were observed which collected at a point just opposite this bright-
est spot. The reason for this variation has not been determined
but it seems to be constant in such specimens and therefore is not
the same type of reaction as that described by Groom and Loeb
(1891) in which the same individuals responded differently under
different conditions.

THE CYPRID STAGE.

The early authors had not observed the Metanauplius stage of
Cirripedia and so did not always clearly recognize the distinction
between the median and compound eyes. Burmeister (1834)
described the median eye as a rounded, black spot which became
divided and modified to form the < paired, compound eyes of the
Cyprid. This error persisted through several of the later works
and Darwin (1854) reported Burmeister's observation although
he considered it " scarcely possible that the eye of the larva of the
first stage can be changed into the double eyes of the second stage."
It was not until the two types of eyes were observed present at
the same time that this misapprehension was fully removed.

The fate of the compound eyes at metamorphosis into the adult
has been variously described. Darwin (1851) and Hesse (1874)
noted that the eyes fell from their capsules during metamorphosis
while Von Willemoes-Suhm (1876) stated that before meta-
morphosis the eyes lost their original position and might be seen
only as black pigment spots which were later absorbed. A similar
observation has been made by Coar 1 in an unpublished study of
Balanus balanoides and by Hanstrom (1927). The author has
observed the extruded eyes in slides of Balanus amphitrite col-
lected by Dr. J. P. Visscher at Beaufort, North Carolina. How-
ever, the condition in Balanus cburncus seems to be similar to that
described by Von Willemoes-Suhm and Coar. The two late
cyprid stages which were observed at metamorphosis showed no
compound eyes, the region of these being occupied by pigment
masses which may have been the degenerating eyes.

The compound eye, as seen in total mounts, has been described

1 Personal correspondence, 1926.

Till-: LIGHT-RECEPTIVE ORGANS OF BARNACLES. 539

by all authors studying the Cyprid. In such preparations the eye
appeared as consisting of a black pigment body and eight to ten
globules or lenses surrounded by a large capsule, and is described
as such by Darwin (1851). Hesse (1874), and Von Willemoes-
Suhm (1876). When studied in section the compound eyes of
I-iahnius balanoldcs (Fig. 2) are found to be situated in pockets
near the bases of the antennules. In contrast to the rather unique
structure of the median eye, their appearance is very similar to
that of the compound eyes of other Crustacea. The eight to ten
lenses described for similar forms by early authors are present
and represent the same number of visual elements or oinmatidia.
Each ommatidium contains a cuticular lens surrounded by " cor-
neagen cells," which are reported by Patten (1886 and 1887) to

c.c-

FIG. 2. Diagrammatic drawing of the compound eye in the cyprid stage
(vertical section X 300). c., cornea; c.c. corneagen cells; /., lens; ;/.,
nerve.

secrete this body, and other cells known as the cells of the crystal-
line body. The lens or crystalline body is made up of three units
forming an egg-shaped structure. The proximal portion of the
ommatidium contains the rhabdomes surrounded by heavily pig-
mented retinular cells. This region is much shorter in the com-
pound eyes of Balanus than in most Crustacea. It was not pos-
sible to study the rhabdome in detail but it appears to be like that
of other Crustacea, of reticular nature and penetrated by nerve
fibrils. The retinular cells pass into the optic nerve of the com-

540

DORIS EDNA FALES.

pound eye. The ommatidia are surrounded by a common corneal
sheath. The structure of the compound eye of Balanus cburncus
appears to be similar, as far as could be determined from a study
of total mounts.

In 1851, Bate noted that the pigmented area of the Nauplius was
represented by a similar region in the Cyprid although he did not
believe that the area was a light-receptor in either period. Claus
(1869), in a drawing of an unknown Cyprid, pictured the per-
sistence of the unpaired eye up to the moment of metamorphosis,
although the eye is not labelled and it is probable that he did not
realize the significance of his observation. Pouchet and Jobert
(18/6) recognized and discussed this persistence of the median
eye through the cyprid stage.

The median, or nauplian, eye (Fig. i) persists throughout the
cyprid period. It has enlarged to five times its former diameter
(i^/jL-yg/j.) but its structure is the same. The cells with large
nuclei still coat the non-pigmented areas and because of the
similarity of these nuclei to those of the regular visual cells, it is
extremely difficult to determine the exact number of the latter.
The bi-concave pigmented area is composed of many cells in con-
trast to the condition reported for Lcpas, where only two cells are
found. In each pigmented area there are two non-pigmented
zones each containing two visual cells which form a nerve con-
nection with the cerebral ganglia. These nerve connections have
elongated while the amount of connective tissue surrounding the
visual cells has also become greater and the reticular nature of
part of it is evident in most sections.

It is difficult to say just what part the compound and median
eyes play in the light perception of the Cyprid. Probably both
are functional. Since the median eye is functional in the stages
preceeding and following this period, as well as after the degen-
eration of the compound eyes, it is unlikely that it would com-
pletely lose its function at this time.

THE ADULT STAGE.

Early observers denied the presence of an eye in adult barn-
acles and indeed the relatively degenerate structure and enclosing
shell of the adult tended to support this view, as well as the fact

Till', I.IC.IIT-RECEPTIVE ()]<(,. \NS OF !:.\KXA( I.I -. 541

that all eye structures were thought to he lost at metamorphosis,
with the disappearance of the cyprid eyes. The first report on
the existence of eyes in the adult barnacle was made hy Leidy
(1848) on f-!alaiuis niyosus (sp?). In 1854. Darwin substan-
tiated Leidy 's report by finding eyes present also in the adult of
Jntlaints tintinnabulum. A summary of previous studies of the
barnacle eye was made by Gerstaecker (1866). These were con-
cerned chiefly with the fact of occurrence or with external ap-
pearance of the eyes. The most complete study of the barnacle
eye in the adult was that made by Pouchet and Jobert in 18/6.

There seems to be no published report of the origin of the eye-
in the adult. Darwin (1854) pointed out that they were not de-
veloped from the eyes of the Cyprid, since the new eyes were
formed at some distance from the compound, but thought they
might have been formed from the nauplian eye since they occupied
a similar position. Coar l reported that in llalanns halanoitlcs the
adult eyes were formed by a division of the median or nauplian
eye, and the author has found that the same situation occurs in
Balanus cburncus. The eye of the Nauplius, which has persisted
throughout the cyprid stage, divides into two parts during the
metamorphosis of the Cyprid into the adult. These, together with
a part of the mantle which becomes modified around them, form
the simple, paired eyes of the adult. In animals which have just
completed metamorphosis the two eyes may be seen completely
separate although still very close together (Fig. 3 A). They
move apart during the succeeding period and, at about the fifth
day, are in the position which they occupy in the adult (Fig. 3 B).
The eyes were found to lie in the mantle between the scutum and
the juncture of the rostrum (rostrum coalesced with rostro-lateral-
Darwin 1851) and the lateral plates of the shell. Immediately
around the eye and optic nerve the mantle lacks its usual pig-
mentation and for this reason the eye appears prominent. The
part of the eye toward the body of the barnacle is heavily pig-
mented while the region toward the shell is without pigment.

Pouchet and Jobert (1876) described the eye as a rounded
structure partly covered by pigment and adherent to the surround-
ing tissue " en arriere." They believed that this pigment func-

1 Personal correspondence, 1926.

542

DORIS EDNA FALES.

tioned as a choroid coat while the non-pigmented area might pos-
sibly be called a cornea. When the eye was macerated in Miiller's
fluid they noticed the presence of a cell which owing to its volume,
granular nature, and distinct nucleolus, they felt was undoubtedly
a nerve cell. The existence of a double optic nerve suggested to
them the possibility that there were two such cells. However,
they never observed more than one and were inclined to consider
the situation analogous to that in the Lepadse where they had
found a double nerve but always a single nerve cell. The light
was described as reaching the eye by traversing the tissue which
united the valves and which contained no pigment in the vicinity
of the eye. They found that barnacles were sensitive to light,

A ^- — -^ B

FIG. 3. Diagrammatic drawing showing adult of Balamis cburneus. A,
immediately after metamorphosis; B, five days after metamorphosis.

but did not believe that they possessed object vision. No men-
tion is made of the origin of the adult eye although development
from the median eye is suggested by the nature of the remarks.

When sections of the eye of the adult are studied it is seen to
be composed of two main divisions ; an outer covering which is a
modification of the mantle and an inner part which is developed
from the divided nauplian eye (Fig. 4).

The outer covering is composed of irregularly shaped cells
similar in appearance to those of the mantle. Pigmented cells
make up about half the area of this coat and are continuous with
like cells in the mantle, while non-pigmented cells form the rest
of the covering and are a continuation of similar mantle cells.

Between this covering and the inner portion of the eye is a
region of loose collagenous connective tissue fibers which serve

THE Lir,iiT-KK( i:n IVE ORC.ANS OF BARNACLES.

543

to hold the inner region in position as well as to support the outer
covering. In the living specimen the interstices of these fihers are
filled with fluid which helps to maintain the contour of the eye.

The inner part of the eye is a sphere containing pigmented cells,
collagenous fibers, reticular fibers, and visual cells with their
nerves. The pigmented region forms a cap over about half of
this portion of the eye and lies directly beneath the pigmented
region of the outer coat. Beneath this, is an area of rather loose
connective tissue fibers and cells similar to those between the outer
and inner regions. The rest of the inner eye is composed of
a reticulum of connective tissue fibres surrounding two large
ganglion or visual cells. A slight division is observable in this

o.np.c

r.f-r-

o.p.c.

o.*.

i.p.o.

c.t.

v.o

FIG. 4. Diagrammatic drawing of the simple eye of the adult (vertical
section X2o). c.t., connective tissue fibres; i.p.c., inner pigmented cells;
;;.. nerve; o.np.c., outer non-pigmented cells; o.p.c., outer pigmented cells;
r./., reticular fibers; 7-.c. visual cells.

reticulum, indicating the two units of nervous elements. These
two visual cells send off branches which ramify throughout the
reticular fibers. Each cell also gives off a nerve fiber which is
surrounded by a sheath. These fibers, with their individual
sheaths, become united as a double nerve enclosed within a com-
mon connective tissue coat before passing through the outer cov-
ering and continuing to the supraoesophageal ganglion as the optic
nerve.

The enlargement of the nerve just prior to its entrance into

544

DORIS EDNA FALES.

the pigment mass, as described by Gruvel (1905), is not apparent
in the sections and it seems probable that he has included the non-
pigmented region of the outer coat as a part of the nerve. Since
his work did not include studies of sections of the eye, this error
is very natural.

Papers on the median eye of Crustacea often describe the pres-
ence of a tapetum next to the two (or more) cells which form the
eye cups. By homology the inner pigment region of the adult
barnacle eye corresponds to these cells and the tapetum might
therefore be either the loose connective tissue or the reticular
regions of this eye. However so little uniformity is found in the
use of this term by investigators that it is considered inadvisable
to apply the name to any particular region of the barnacle eye.

Barnacles in their natural environment will retract their cirri
when stimulated by light, and this fact is mentioned by several of
the early authors. Pickering (1848) brought it forward in con-
firmation of Leidy's report as to the existence of eyes in the adult.
Darwin (1854) observed the reaction to light in Balanus bala-
noidcs, Balanus crcnatus, and Chthaniahis stcllatus and found that
they were all sensitive to a shadow produced by passing his hand
between them and the light. Gerstaecher (1866) reported ex-
periments by Fr. Muller who found that Balaiuis tintinnabuluin
would react to a shadow when the body was removed and the eyes
and certain muscles were left in the shell. (" dass Balanus tin-
tinnabuluin auf eine Beschattung mit der hand auch claim reagire,
wenn er mit Zuriicklassung seiner Augen an clem Manteldeckel,
von diesem abgelost werde. Ein in dieser Weise entblosstes, mit
halb entrollten Ranken im Wasser liegendes Exemplar zog
dieselben jedesmal schnell ein, wenn es beschattet wurde.")

In our study it was found that the adults of both Balanus bala-
noidcs and Balanus cburneus close the opercular valves if there
is a sudden change in light intensity although very gradual changes
do not excite the reaction. Balanus cburneus is more sensitive to
such changes than is Balanus balanoidcs. A slight shadow may
sometimes be cast upon the latter without affecting them but it
was never possible to do this with Balanus cburneus.

Since it is shown that the adults possess some light-perceptive
mechanism, it becomes necessary to find what part the so-called

THE LIGHT-RECEPTIVE ORGANS OK P.ARX ACI.ES.

" eye '' ])lays in causing this reaction. Accordingly, the light
reactions' were studied in twenty-five adults of BaSainis cbunicus
which had been deprived of these organs and it was observed that
such animals showed no reaction to light changes, however sudden
or intense, although when the eyes were intact they had all with-
drawn their cirri and closed the valves under similar conditions.
Therefore it is concluded that these organs are the sole light-
perceptors present in the adult barnacle.

The function of certain parts of the eye is problematical. Since
the ganglion or visual cells send branches throughout the reticulum
of connective tissue fibers immediately surrounding them, it is
evident that this part serves in the transmission of the impulse.
The looser collagenous fibers outside this area do not have any
self-evident function. As they are not pigmented they would not
prevent the passage of light and it seems probable that their func-
tion is merely that of support. The inner pigmented cells may
be either protective or reflective in nature. The structure of the
eye and its inverted nature lend some support to the latter possi-
bility. The light enters the eye through the non-pigmented area
of the outer coat while its entrance through the other cells is pre-
vented by the pigment. The fluid which is found in the eye of
living specimens must act as a refractive as well as a supporting
medium.

SUMMARY.

1. The development, structure, and function of the light-per-
ceptive organs are described in the nauplian, cyprid, and adult
stages of Balanns cbnrncns and I'alainis balanoides.

2. The light-perceptive organs present in the various stages are:
(a) nauplian — a median eye, (/>) metanauplian — a median eye and
two compound eyes, (c) cyprid — a median eye and two compound
eyes, (</) adult — two simple eyes.

3. The median eye in Jntlmnis cbnniciis originates as a single,
pigmented mass in the unhatched Nauplius and persists with no
change, except in si/.e. until the metamorphosis of the Cyprid into
the adult.

4. The compound eyes first appear in the metanauplian st
and remain functional throughout the cyprid stage.

546

DORIS EDNA FALES.

5. These compound eyes are resorbed at the time of the meta-
morphosis of the Cyprid into the adult.

6. At the metamorphosis into the adult the median eye divides
into two parts which form the simple paired eyes of the adult.

/. Each of the paired eyes in the adult is the morphological
equivalent of half of the median eye plus an outer covering.

8. The simple, paired eyes are the sole light-perceptive organs
of the adult.

LITERATURE CITED.

Bate, C. S.

1851 On the Development of the Cirripedia. Annals and Mag. of Nat.

Hist, Second Series, Vol. 8, pp. 325-332.
Brooks, W., and F. Herrick

1892 The Embryology and Metamorphosis of the Macroura. National

Acad. of Sciences, Vol. 5.
Burmeister, H.

1834 Beitrage zur Naturgeschichte der Rankenfiisser. Berlin.
Claparede, A.

1863 Beobachtungen iiber Anatomic und Entwickelungsgeschichte wir-

belloser Thiere. Leipzig.
Claus, C.

1869 Die Cypris Aehnliche Larvae der Cirripedien und Ihre Verwand-
lung in das Festsitzende Thier. Schriften der Gesellschaft zur
Beforderung der gesammten Naturwissenschaften zu Marburg.,
Bd. 9, pp. 1-17.

1891 Das Medianauge der Crustaceen. Arbeiten aus dem Zoologischen
Institute der Universitat Wien und der Zoologischen Station in
Triest, Wien.
Darwin, C.

1851 A Monograph on the Sub-class Cirripedia with Figures of All

Species. I. Lepadidae. London.

1854 A Monograph on the Sub-class Cirripedia, II. Balanidse. London.
Demoll, R.

1917 Die Sinnesorgane der Arthropoden, Ihre Ban und Ihre Funktion.

Braunschweig.
Gerstaecker, A.

1866 Bronn's Thierreich, Arthropoda, Bd. 5, pp. 406-564. Leipzig und

Heidelberg.
Grenadier, G. H.

1879 Untersuchungen tiber das Sehorgan der Arthropoden, insbesondere

der Spinnen, Insecten und Crustaceen. Gottingen.
Groom, T., and J. Loeb

1891 Der Heliotropismus der Nauplien von Balanus perforates und die
periodischen Tiefenwanderungen pelagischer Tiere. Bio. Cen-
tral., Bd. 10, pp. 160-177.

THE LIGHT-RECEPTIVE ORGANS OF BARNACLES. 547

Gruvel, A.

1905 Monographic cles Cirrhipedes ou Thecostraces. Paris.
Hanstrom, V. B.

1927 Neue Beobachtungen iiber Augen uncl Sehzentren von Ent<>-
mostraci'ii, Scbizopodcn und Pantopodcn. Zool. Anzeiger, Bd.
70, pp. 236-251.
Hess, W.

1925 Photorcccptors of Lninhricits tcrrcstris. Jour. Morph. and

Physiol., Vol. 41, pp. 63-93.
Hesse, C. E.

1874 Description dr la Serie complete des Metamorphoses qui sul>i>-
sent, durant la periode embryonnaire, les Anatifes designes sous
le nom de Scalpel oblique ou de Scalpel vulgaire. Revue des
Sciences Nat. Montpellier, T. 3. pp. 1-14. 206-212.
Hesse, R.

1901 Untersuchungen iiber die Organe der Lichtempfindung bei niederen

Thieren. Zeit. fiir Wissen. Zool., Bd. 70, pp. 347-473.
Lee, A.

1890 The Microtomist's Vade-mecum. Philadelphia.
Leidy, J.

1848 Proceedings of the Academy of Nat. Sciences, Philadelphia, Vol.

4, p. I.
Mark, E. L.

1887 Simple Eyes in Arthropods. Bull. Mus. Comp. Zool. Harvard

College, Vol. 13, pp. 4-105.
Parson, G.

1831 An Account of the Discoveries of Miiller and Others in the Organs
of Vision of Insects and Crustacea. The Mag. of Nat. Hist.
Edinburgh, Vol. 4, pp. 124-134, 220-234. 363-372.
Patten, W.

1886 Eyes of Molluscs and Arthropods. Mitteil. Zool. Stat. Neapel.,

Vol. 6, pp. 542-756.

1887 Studies on the Eyes of Arthropods. Jour. Morph., Vol. i, pp. 193-

227.
Pickering, C.

1848 Proceed of the Acad. of Nat. Sciences, Philadelphia, Vol. i, p. 2.
Pouchet, C., and C. Jobert.

1876 Contribution a 1'Histoire de la Vision chez les Cirrhipedes. Jour.

de 1'Anat. et de la Physiol., T. 12, pp. 575~5"4-
Von Willemoes-Suhm, R.

1876 On the Development of Lcpus fuscicitlaris and the " Archizoea " of
Cirripedia. Phil. Trans. Royal Soc. of London, London. Vol.
166, pp. I3I-I54-

BIOLOGICAL BULLETIN

OF THE

flDartne Biological Xaboratorp

WOODS HOLE, MASS.

VOL. LIV JANUARY, 1928 No. i

CONTENTS

CHILD, C. M. I'.\j>i'rimen1a1 Transformation^ oj l-'i polar

l-'ornis in Corymorpha pal ma ........... i

CHILD, C. M. Physiological Polarity and Dominance in thr

Holdfast System of Corymorpha ......... 15

HALL, RICHARD P., AND

POWELL, WILLIAM N. Morphology and Binary Fission of Peranenni

trichophoruni (Ehrbg.) Stein ............ V

HYMAN, LIBBIE H. Miscellaneous Observations on Hydra, icitli

Special Reference to Reproduction ....... 65

PUBLISHED MONTHLY BY TIIK

MARINE BIOLOGICAL LABORATORY

PRINTED AND ISSUED BY

LANCASTER PRESS, INC.
LANCASTER, 'PA.

AC.KNT FOR GREAT BRITAIN
\YHKLDON & WESLEY, LIMITED

2, ,j and 4 Arthur .S'/'v</, N<iv (\\fnrd Stn-,'1, l.nadnn, IT'. C. 2

Single Numbers, S1.OO. Per Volume (6 numbers). S4.5O
Entered October 10, ipoa. at Lancaster, Pa., at second-class matter under Act of Congress of July 16, 1894

EOttortal Statt

GARY N. CALKINS — Columbia University.

E. G. CONKLIN — Princeton University.

M, II. JACOBS — University of Pennsylvania.

FRANK R. Ln.ME— University of Chicago.

GKOROE 1'. MOORE— The Missouri Bolamc Garden.

T. H. MORGAN — Columbia University.

W. M. WHEELER— Harvard University.

E. B. WILSON — Columbia University.

Managing EMtor

C. R. MOORE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. I5t.h to June I5th, or
to Biological Bulletin, Woods Hole, Mass., June 15th to Sept. I5th,
Subscriptions and other matter should be addressed to the Biological
Bulletin, Prince and Lemon Streets, Lancaster, Pa.

Notice to contributors. Every paper to appear in the Biological
Bulletin should be accompanied by an author's abstract presenting
the chief results of the investigation. The abstract should not exceed
225 words in length.

For indexing purposes there should be, in addition to the title,
one or more subject headings indicating in a word or two the divi-
sions of the subject discussed in the paper. The entire name of the
author (of each author if a joint paper) and the year of birth is also
desired.

BIOLOGICAL BULLETIN

OF THE

Marine Biological Xaborator?

WOODS HOLE, MASS.

VOL. LIV

FEBRUARY, 1928

No. 2

BECKER, ELERY R.

CONTENTS

Streaming and Polarity in Mnstigina
hylce (Frenzel) 109

TUCKER, MARY ELIZABETH. Studies on the Life Cycles of Tico Species

of Fresh-Water Mussds Belonging to
the Genus Anodonta . 117

UHLENHUTH, E., AND
KARNS, HILDA.

QUIGLEY, J. P.

SKOWRON, STANISLAW.
PATTERSON, J. T.

The Morphology and Physiology of the
Salamander Thyroid Gland, III 128

Reactions of tin Elasmobranch (Squalus
sucklii) to Variations in the Salinity
of the Surrounding Medium 165

The Luminous Material of Microscolex
phosphoreus Dug 191

Functionl'-ss Males in Tico Species of
Neuroterus 196

PUBLISHED MONTHLY BY THE

MARINE BIOLOGICAL LABORATORY

PRINTED AND ISSUED BY

LANCASTER PRESS, INC.
LANCASTER, PA.

AGENT FOR GREAT BRITAIN
WHELDON & WESLEY, LIMITED

2, j and 4 Arthur Street, New Oxford Street, London, W. C. 2

Single Numbers, $1.00. Per Volume 16 numbers), S4.5O

Entered October 10, igoa. at Lancaster, Pa., as secoud-class matter under Act of Congress of July 16, 1894

Statf

GARY N. CALKINS — Columbia University.

E. G. CONKLIN — Princeton University.

M. H. JACOBS — University of Pennsylvania.

FRANK R. LILLIE — University of Chicago.

GEORGE T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN — Columbia University.

W. M. WHEELER — Harvard University.

E. B. WILSON — Columbia University.

Editor

C. R. MOORE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. I5th to June I5th, or
to Biological Bulletin, Woods Hole, Mass., June I5th to Sept. I5th.
Subscriptions and other matter should be addressed to the Biological
Bulletin, Prince and Lemon Streets, Lancaster, Pa.

Notice to contributors. Every paper to appear in the Biological
Bulletin should be accompanied by an author's abstract presenting
the chief results of the investigation. The abstract should not exceed
225 words in length.

For indexing purposes there should be, in addition to the title,
one or more subject headings indicating in a word or two the divi-
sions of the subject discussed in the paper. The entire name of the
author (of each author if a joint paper) and the year of birth is also
desired.

BIOLOGICAL BULLETIN

OF THE

flDarine Biological Xaborator^

WOODS HOLE, MASS.

VOL. LIV MARCH, 1928 No. 3

CONTENTS

PATTERSON, J. T. Sexes in the Cynipidce and Male-Pro-

ducing and Female-Producing Lines 201

SAYLE, MARY HONORA. Factors Influencing the Rate of Alctab-

_olism of JEshna umbrosa lymphs. . 212

CLEVELAND, L. R. Further Observations and Experiments

on the Symbiosis between Termites
and Their Intestinal Protozoa. ... 231

BISSONNETTE, THOMAS HUME. Notes on a 32 Millimeter Freemartin . 238

BABKIN, B. P., AND

BOWIE, D. J. The Digestive System and its Function

in Fundulus hcteroclitus 254

PUBLISHED MONTHLY BY THE

MARINE BIOLOGICAL LABORATORY

PRINTED AND ISSUED BY

LANCASTER PRESS, INC.
LANCASTER, PA.

AGENT FOR GREAT BRITAIN

WHELDON & WESLEY, LIMITED

2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2

Single Numbers, $1.00. Per Volume »6 number*), $4.50

Entered October 10. 1902. at Lancaster. Pa., as gecond-dass matter under Act of Congreos of July 16. 1804

EOttorial Statf

GARY N. CALKINS — Columbia University.

E. G. CONKLIN — Princeton University.

M. H. JACOBS — University of Pennsylvania.

FRANK R. LILLIE — University of Chicago.

GEORGE T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN — Columbia University.

W. M. WHEELER — Harvard University.

E. B. WILSON — Columbia University.

C. R. MOORE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. I5th to June isth, or
to Biological Bulletin, Woods Hole, Mass., June I5th to Sept. I5th.
Subscriptions and other matter should be addressed to the Biological
Bulletin, Prince and Lemon Streets, Lancaster, Pa.

Notice to contributors. Every paper to appear in the Biological
Bulletin should be accompanied by an author's abstract presenting
the chief results of the investigation. The abstract should not exceed
225 words in length.

For indexing purposes there should be, in addition to the title,
one or more subject headings indicating in a word or two the divi-
sions of the subject discussed in the paper. The entire name of the
author (of each author if a joint paper) and the year of birth is also
desired.

BIOLOGICAL BULLETIN

OF THE

noarine Biological laboratory

WOODS HOLE, MASS.

VOL. LIV

APRIL, 1928

No. 4

ROWLAND, RUTH B.

WHITING, P. \Y.
HUMPHREY, R. R.

HEATH, HAROLD.
VISSCHER, J. PAUL.

CONTENTS

Grafting and Reincorporation in A ctinos p>h ce-
rium eichhornii Ehr 279

Mosaicism and Mutation in Habrobracon . . 289

Ovulation in the Four-toed Salamander,
Hemidactylium scittatnni, and the External
Features of Cleavage and Gastrnlation . . . 307

Fertile Termite Soldiers 324

Reactions of the Cyprid Larva of Barnacles
at the Time of Attachment 327

VISSCHER, J. PAUL, AND

LUCE, ROBERT H. Reactions of the Cyprid Larvce of Barnacles

to Light with Special Reference to Spectral
Colors 336

PUBLISHED MONTHLY BY THE

MARINE BIOLOGICAL LABORATORY

PRINTED AND ISSUED BY

LANCASTER PRESS, INC.
LANCASTER, PA.

AGENT FOR GREAT BRITAIN

WHELDON & WESLEY, LIMITED

2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2

Single Numbers, $1.00. Per Volume >6 numbers). $4.50

EntereJ October 10, looa. at Lancaster. Pa., at second-class matter under Act of Congress of July 16. 1804

EDttorfal Statf

GARY N. CALKINS — Columbia University.

E. G. CONKLIN — Princeton University.

M. H. JACOBS — University of Pennsylvania.

FRANK R. LILLIE — University of Chicago.

GEORGE T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN — Columbia University.

W. M. WHEELER — Harvard University.

E. B. WILSON — Columbia University.

/Ibanaging EMtor

C. R. MOORE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. I5th to June I5th, or
to Biological Bulletin, Woods Hole, Mass., June I5th to Sept. I5th.
Subscriptions and other matter should be addressed to the Biological
Bulletin, Prince and Lemon Streets, Lancaster, Pa.

Notice to contributors. Every paper to appear in the Biological
Bulletin should be accompanied by an author's abstract presenting
the chief results of the investigation. The abstract should not exceed
225 words in length.

For indexing purposes there should be, in addition to the title,
one or more subject headings indicating in a word or two the divi-
sions of the subject discussed in the paper. The entire name of the
author (of each author if a joint paper) and the year of birth is also
desired.

BIOLOGICAL BULLETIN

OF THE

firmrine Biological laboratory

WOODS HOLE, MASS.

VOL. LIV

MAY, 1928

GRAVE, B. H.
FRY, HENRY.

CONTENTS

I 'itality of the Gametes of Cumin gin tcllinoides. 35 i

ConditionsDetcriiiining the Origin and Behavior
of Central Bodies in Cytaste.rs of Echinn-
rachniiis Eggs 363

No. 5

BODINE, JOSEPH HALL. Action of Salts on Fundnlus Egg. 1 396

PEARSE, A. S.
YOCOM, H. B.

On the Ability of Certain Marine 1 nrc/l>l>nitcs
to Live in Diluted Sea Water 405

The Effect of the Quantity of Culture Medium
on the Division Rate of Oxytricha 410

MANWELL, REGINALD D. Conjugation, Division, and Encyst men t in

Pleurotricha lanceolate 417

PUBLISHED MONTHLY BY THE

MARINE BIOLOGICAL LABORATORY

PRINTED AND ISSUED BY

LANCASTER PRESS, INC.
LANCASTER, PA.

AGENT FOR GREAT BRITAIN

WHELDON & WESLEY, LIMITED

2, 3 and 4 Arthur Street, New Oxford Strcrt, London, W. C. 2

Single Numbers, $1.OO. Per Volume <6 numbers). $4.50

Entered October 10,1902. at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894

Staff

GARY N. CALKINS — Columbia University.

E. G. CONKLIN — Princeton University.

M. H. JACOBS — University of Pennsylvania.

FRANK R. LILLIE — University of Chicago.

GEORGE T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN — Columbia University.

W. M. WHEELER — Harvard University.

E. B. WILSON — Columbia University.

Managing Boitor

C. R. MOORE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. I5th to June I5th, or
to Biological Bulletin, Woods Hole, Mass., June I5th to Sept. 15th.
Subscriptions and other matter should be addressed to the Biological
Bulletin, Prince and Lemon Streets, Lancaster, Pa.

Notice to contributors. Every paper to appear in the Biological
Bulletin should be accompanied by an author's abstract presenting
the chief results of the investigation. The abstract should not exceed
225 words in length.

For indexing purposes there should be, in addition to the title,
one or more subject headings indicating in a word or two the divi-
sions of the subject discussed in the paper. The entire name of the
author (of each author if a joint paper) and the year of birth is also
desired.

BIOLOGICAL BULLETIN

OF THE

flDarine Biological laboratory

WOODS HOLE, MASS.

VOL. LIV JUNE, 1928 No. 6

CONTENTS

BELLING, JOHN. .\odes and Chiasmus in the Bhnlnits

of Liliuni with Regard to Scgniental
Interchange 465

MACDOUGALL, MARY STUART. The Neuromotor Apparatus of Chi a my -

dodon sp 471

LOOPER, JAMES BURDINE. Observations on the Food Reactions of

Actinophrys Sol 4x5

RAU, PHIL. The Honey-Gathering Habits of Politics

Wasps 5u3

BURCH, PAUL R. Endodermal Flagella of Hydra oliyntix

Pallas 520

KEPNER, WM. A., AND

MILLER, LULU. A New Histological Region in Hydra

oligactis Pallas 524

KEPNER, WM. A., AND

THOMAS, W. L., JR. Histological Features Correlated with

Gas Secretion in Hydra oligactis
Pallas 529

FALES, DORIS EDNA. The Light- Receptive Organs of Certain

Barnacles 534

PUBLISHED MONTHLY BY THE
MARINE BIOLOGICAL LABORATORY

PRINTED AND ISSUED BY

LANCASTER PRESS, INC.
LANCASTER, PA.

AGENT FOR GREAT BRITAIN

WHELDON & WESLEY, LIMITED

2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2

Single Numbers, $1.00. Per Volume '6 numbers), S4.5O
Entered October 10,1902. at Lancaster. Pa., as second-class matter under Act of Congress of July 16, 1894

Eottorial Staff

GARY N. CALKINS — Columbia University.

E. G. CONKLIN — Princeton University.

M. H. JACOBS — University of Pennsylvania.

FRANK R. LILLIE — University of Chicago.

GEORGE T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN — Columbia University.

W. M. WHEELER — Harvard University.

E. B. WILSON — Columbia University.

EMtor

C. R. MOORE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. isth to June 15th, or
to Biological Bulletin, Woods Hole, Mass., June I5th to Sept. I5th.
Subscriptions and other matter should be addressed to the Biological
Bulletin, Prince and Lemon Streets, Lancaster, Pa.

Notice to contributors. Every paper to appear in the Biological
Bulletin should be accompanied by an author's abstract presenting
the chief results of the investigation. The abstract should not exceed
225 words in length.

For indexing purposes there should be, in addition to the title,
one or more subject headings indicating in a word or two the divi-
sions of the subject discussed in the paper. The entire name of the
author (of each author if a joint paper) and the year of birth is also
desired.

MBL WH01 I.IHKAKY