BIOLOGICAL BULLETIN

OF THE

fiDarine Biological Xaboraton?

WOODS HOLE, MASS.

EMtorial Staff

E. G. CONKLIN — Princeton University.

GEORGE T. MOORE — The Missoiiri Botanic Garden.

T. H. MORGAN — Columbia University.

W. M. WHEELER — Harvard University.

E. B. WILSON — Columbia University.

/IDanacjincj

FRANK R. LILLIE — The University of Chicago.

VOLUME XXXVIII.

WOODS HOLE, MASS.
JANUARY TO JUNE, 1920

PRESS OF

THE NEW ERA PRINTING COMPANY
i_ANC«STER, PA.

CONTENTS OF VOLUME XXXVIII.

No. i. JANUARY, 1920.

REIGHARD, JACOB. The Breeding Behavior of the Suckers

and Minnows I

STRONG, LEONELL C. Roughoid, A Mutant Located to the
Left of Sepia in the Chromosome of Drosophila melano-
gaster 33

No. 2. FEBRUARY, 1920.

HESS, WALTER N. Notes on the Biology of Some Common

LampyridcB 39

DAWSON, A. B. An Exception to Bateson's Rule of Secondary

Symmetry 77

SHAFFER, E. L. A Comparative Study of the Chromosomes of

Lachnosterna (Coleoptera) 83

No. 3. MARCH, 1920.

ROBBINS, H. L. AND CHILD, C. M. Carbon Dioxide Produc-
tion in Relation to Regeneration in Planaria dorotocephala 103

BALDWIN, FRANCIS M. Susceptible and Resistant Phases of
the Dividing Sea-Urchin Egg when subjected to Various
Concentrations of Lipoid- Soluble Substances, especially
the Higher Alcohols 123

BERRY, S. STILLMAN. Light Production in Cephalopods, I . . 141

No. 4. APRIL, 1920.

BERRY, S. STILLMAN. Light Production in Cephalopods, II. 171
GOODRICH, H. B. Rapidity of Activation in the Fertilization

of Nereis 1 96

DANCHAKOFF, VERA. Immunity and the Power of Digestion. 202
HARMAN, MARY T. Chromosome Studies in Tettigidae. II.
Chromosomes of Paratettix BB and CC and their Hybrid
BC 213

iii

IV CONTENTS

BRIDGES, CALVIN B. White-Ocelli — An Example of a

"Slight" Mutant Character with Normal Viability 231

ISHII, O. Observations on the Sexual Cycle of the Guinea Pig. 237

No. 5. MAY, 1920.

JONES, D. F. Selective Fertilization in Pollen Mixtures . ... 251

WODSEDALEK, J. E. Studies on the Cells of Cattle with Special
Reference to S per mato genesis, Oogonia, and Sex-Deter-
mination 290

HEILBRUNN, L. V. Studies in Artificial Parthenogenesis.
III. Cortical Change and the Initiation of Maturation
in the Egg of Cumingia 317

HAUSMAN, LEON AUGUSTUS. A Contribution to the Life

History of Amoeba Proteus Leidy 340

No. 6. JUNE, 1920.

HYMAN, LIBBIE H. The Axial Gradients in Hydrozoa, III.

Experiments on the Gradient of Tubidaria 353

SHAFFER, ELMER L. The Germ-Cells of Cicada (Tibicen}

septemdecim (Homoptera) 404

Vol. XXXVIII. January, 1920. No. i.

BIOLOGICAL BULLETIN

THE BREEDING BEHAVIOR OF THE SUCKERS AND

MINNOWS.

I. THE SUCKERS.1

•

JACOB REIGHARD.

CONTENTS.

I. Introduction to the Series i

II. Breeding Behavior of the Suckers 3

A. The White Sucker 3

1. The Breeding Grounds and Breeding Season : . 3

2. General Activities of Breeding Fish . 4

3. Coloration and Color Changes 6

4. Sexual Differences 6

Color 6

Pearl Organs 7

Fin Length 8

5. Breeding Activities 9

B. The Common Red-horse 15

1. Breeding Grounds and Breeding Season 15

2. Sexual Differences 16

3. Coloration 18

4. Breeding Activities 19

C. The Hogsucker 20

1. General Activities 20

2. Sexual Differences 21

3. Breeding Activities 22

III. Summary of Observations.

A . The White Sucker .... 24

B. The Red-horse 26

C. The Hogsucker 28

IV. Conclusions 29

V. Literature 31

i. INTRODUCTION TO THE SERIES.

This paper is the first of a series on the breeding habits of
nine species of suckers and minnows. These fishes form a well-
defined group by some systematists united into a single family,

1 Contributions from the Zoological Laboratory of the University of Michigan.

I

2 JACOB REIGHARD.

by others separated into the Catostomidse and Cyprinidae. To
accumulate the field notes and sketches on which the papers are
based has been the work of many seasons. If there were no bad
weather, no university duties, and no human interference with
breeding environment or breeding fish, such work might be
carried on with as little interruption as that of the laboratory;
but it would not progress as rapidly, for observations need to be
many times repeated. The behavior is often so complicated or
so rapid that it is only by analyzing it into its elements and
observing -each of these repeatedly that a degree of certainty is
possible. Such attitudes as these shown in Fig. 3 of this paper
may be correctly represented only after many observations on
the position of each fin and on every other detail. The observa-
tions upon which my descriptions are based have been very many
for each element of the behavior, often in the neighborhood of a
hundred, sometimes probably several hundred, or even thousand.
Although the work has been spread over a period of years no
one should suppose that the record is complete. I have studied
only the breeding behavior and that in but few fishes; I have not
the life history or the full natural history of any one. Of the
breeding behavior I shall try to give for each species a composite
picture taken from many fish through many seasons. This I
give in such detail as I have, because that seems necessary to
clearness. It further affords a better basis for the discussion of
theories to be considered in the final papers of the series. The
behavior that I describe may be easily seen in suitable places
and at the proper season. Yet few are likely to take the trouble.
This has seemed to me an added reason for fullness of treatment.
The suckers and minnows that I have studied, with the ex-
ception of the blunt-nosed minnow, Pimephales notatus (Rafin-
esque) breed .in running water. During the breeding season the
males of all the species studied have pearl organs, hard, tough,
white, usually conical elevations of the skin which consist of
cornified epidermal cells. They occur in many situations.
Some of them are so small as to be visible only under a lens;
others may be seen with the naked eye at a distance of ten or
twelve feet. They commonly make the surface of the breeding
males distinctly rough to the touch and are in that case referred
to as "effective" pearl organs.

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 3

The drawings that show fish in action have been made from
field notes and sketches with the help of specimens and in some
cases with the aid of models or photographs. I am indebted
to Mr. Charles R. Knight for valuable suggestions in connection
with the drawings but he is in no way responsible for their
obvious shortcomings. To my former student, Professor
Norman H. Stewart, of Bucknell University, I am indebted for
data from his unpublished manuscript on the distribution of the
pearl organs. To the United States Commissioner of Fisheries,
Doctor Hugh M. Smith, I am indebted for permission to publish.

2. BREEDING BEHAVIOR OF THE SUCKERS.
A. The While Sucker (Catostomus commersonii LeSueur).

i. The Breeding Grounds and Breeding Season. — A little way
from the State Bass Hatchery, near Grand Rapids, Michigan,
Mill Creek is spanned by a bridge of the Pere Marquette Railway.
Just below the bridge, where the stream is three or four rods
wide, a line of four-inch waterpipe is laid across its bed. An
eighth of a mile above the bridge the stream is dammed to furnish
water for the hatchery. Between the dam and the pipe-line it
is made up of a series of alternating rapids and pools. For a little
distance above the bridge and beneath it the water runs six to
twelve inches deep, swift and smooth. Below the pipe-line it
breaks into ripples. Under the smoother water the bottom is
sandy gravel, but under the broken water this gives place to
larger stones. Over the gravel in the smoother water a thin
brown mantle of silt and algal growth usually stretches without
interruption from shore to shore and obscures the bottom.

Rapids of this sort are typical of a drift-covered country and
afford the characteristic breeding ground of the white sucker.1
In the spring patches of the gravel bottom in the upper water
of the rapids often look as though they had been scoured by a
broom. The silt mantle is absent from these patches and the
bright colors of the clean gravel and sand throw them into

1 The usual name in Southern Michigan; known also as common sucker, fine-
scaled sucker, brook sucker.

I am indebted to Professor W. M. Smallwood for permission to use his un-
published notes on this species. He records seeing it in the spawning attitude on
stony bottom in Lake Clear in the Adirondacks in late June (cf. Reighard, 1915)-

4 JACOB REIGHARD.

sharp contrast with the surrounding silt-covered bottom. The
patches show where the suckers have been breeding.

The water of a rapid is rarely so smooth that one can see
readily into it. Not often is such a rapid near a bridge or
other perch; but under this Mill Creek bridge the surface of the
stream is but little broken and on a sunny, cloudless day, when
the water is clear and there is no wind to ruffle iits surface, it is
possible to see in detail what happens on the rapid. With field
glasses the fish may be studied almost as readily as though they
wrere in air. From this vantage point I watched the white
suckers at intervals.1

My work was done by day. It is well known that in the spring
suckers ascend small streams in great numbers at night and it is
possible that their breeding activities are continued at night.
They are often interrupted by colder weather or roily water.

2. General Activities of Breeding Fish. — About 2 o'clock on
April 23, 1903, I cautiously took my place on the Mill Creek
railroad bridge. Numerous white suckers were on the rapids.
Although I walked with extreme slowness and made no sudden
movements of any part of my body the fish were at once aware
of my coming and scurried to the shelter of the banks and nearer
pools. I sat quiet and in the course of fifteen minutes they began
to reappear in the shallow, swift water. Thereafter, for an
hour, any quick movement on my part resulted in the fish
starting swiftly up stream, but if the movement was not repeated
they dropped slowly down-stream to where they had been.
To get the field glasses to the eyes or the hand to the notebook
without startling the fish needed a movement so slow that it
must have been scarcely perceptible at the distance of twenty-
five or thirty feet at which the fish were. It was probably about
two inches per second. As time went on the fish became grad-
ually used to my presence and after an hour were no longer dis-
turbed by slow movements. By three o'clock twenty suckers
from eight to twelve inches long were on the rapid and were
moving slowly up stream in small groups. The fish stopped

1 From April 23 to May 6, 1903. In Honey Creek near Ann Arbor, I saw them
breeding April 27 to May 2, and in Mallet Creek, Ann Arbor, on May 10, 1909.
The two creeks last named are only about a third the width of Mill Creek (Grand
Rapids) at the point at which the suckers were seen.

BREEDING BEHAVIOR OF SUCKERS AXU MINNOWS. 5

here an'd there in the rapid on the patches of cleaned gravel
and were seen to take gravel into the mouth and spit it out.
They were presumably in quest of eggs that had been laid in
the disturbed gravel areas and their occurrence in small groups
is perhaps in part the result of the distribution of such areas.
In their search they were accompanied by numerous small
minnows doubtless on the same quest. Other suckers were
seen crossing the pipe-line. At four o'clock no suckers were to
be seen from the bridge, but a dozen were found just below the
dam an eighth of a mile further up stream. The fish had appar-
ently covered this distance in an hour. At nine o'clock the
same evening the search light showed suckers still crossing the
pipe line bound up stream.

In my experience the white sucker is one of the most difficult
of our native fishes to approach in the open. Ordinarily it
becomes accustomed to the observer with extreme slowness and
at no time permits him any great freedom of movement. Con-
fined with other fish in an aquarium it is among the last to
become accustomed to the observer or to take food.1

1 I have noted but two exceptions to this general fact, (i) On May 6, in the
morning, I placed in an outdoor aquarium at the Mill Creek Hatchery four males
and two females that had been captured in a seine on the previous evening. At
three o'clock on the same day the fish were moving about and feeding and by four
o'clock they were spawning. They did not react to an observer within two or
three feet of them. (2) At Douglas Lake in Cheboygan County, Michigan, suckers
are found in rather deep water. At night they come into shallower water to feed
and are occasionally seen there at dawn. At such times they flee to the deeper
water at the first glimpse of a moving object. lu late June the log perch (Percina
caprodes) are laying their eggs in the sand in very shallow water. At that time
suckers enter shallow water in the day time and feed on the eggs of the log perch.
Each sucker is accompanied by a group of log perch which appear to be feeding on
eggs uncovered by him and perhaps on other crumbs from his table. At this time
the suckers may be approached with little trouble and I have come close enough
to photograph them as they lay at my feet. I have thought this absence of the
suckers' usual wariness due to the presence of the log perch. These are breeding
and are then unafraid. In deeper water the sucker has probably found freedom
from disturbance where they were present. Safety and log perch have been closely
linked in his experience. So now, so long as the log perch are on the shallows, he
is not easily startled and feeds there undisturbed by sights that would otherwise
send him hurrying to shelter.

To these observations may be added one of Smallwood (unpublished notes).
At the end of June he found Catostomus commersonii and two other species of
sucker on stonv bottom in the shallow water of Lake Clear in the Adirondacks.

6 JACOB REIGHARD.

3. Coloration and Color Changes. — The mature white sucker,
when seen in its native waters or in captivity, is ordinarily uni-
formly olivaceous on the back and sides and white below. There
is no color pattern nor are there color differences between the
sexes. The suckers seen from the bridge (Fig. 4) were so different
in coloration from all that I had seen before, that I was at first
doubtful as to their identity. Each had a broad yellow-white
stripe which crossed the occiput and extended thence down the
sides. When some of these fish were seined they were found to
have the usual uniformly olivaceous color. They were placed
in an aquarium, males and females together, and four hours
later the males had begun to move about and to feed. Shortly
afterward the light stripe appeared across the occiput and down
the sides. Beneath the light stripe was a broad dark stripe
(Fig. 4) and in one of the males this had a rosy tinge. During
the actual pairing described below, the rosy tinge gave place to a
brilliant crimson. Later I often saw the light stripe appear in a
few seconds on uniformly colored males that were on the rapids.
This happens regularly in the breeding season when the sexes are
together.

4. Sexual Differences. — It is at first difficult to discriminate
between males and females. As seen from the bridge the paired
fins are transparent white in both and in both the yellowish
white stripe crosses the occiput and extends down the sides.
But the males are on the average smaller than the females and
slenderer. It is soon apparent that the occipito-lateral stripe is
whiter in them and that their backs may be flecked with white
especially between the dorsal fin and the occiput (Fig. 4, male
at right of female). In the region of the occiput the white
flecks may form a distinct patch which, seen from a distance,
looks like fungus. The white flecks are perhaps not always
present in males, but I have never seen them in females. In
pairing males, seen in the aquarium, the dorsal half of the eye is
lighter colored than in females, but I do not know that this is
not the case at other times. The differences so far noted afford

C. commersonii and one of the other species, probably C. catosiomus, were seen in
the spawning attitude. The fish were not disturbed when a boat, in which were

•

two children and a barking dog, was poled about above them so that they were
not more than five feet from it.

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 7

excellent field characters by which the sexes of the breeding
fish may be distinguished in their native waters. Less striking
differential characters are the greater length of the caudal and
lower fins of the male, the difference in length of his caudal lobes
and his possession of pearl organs.

Effective pearl organs (vide introduction) occur on the male
in the following situations (Fig. i): (i) Large, sharp-pointed

FIG. i. Lateral view of a part of a male and female of Catostomus commersomi
drawn to the same scale. The black dots represent pearl organs. The anal fin
and lower lobe of the caudal are longer in the male and bear large pearl organs .
All the scales of his sides bear small pearls, somewhat larger behind the caudal
The female has no pearl organs.

organs are found on the anal fin and on the lower part of the
caudal. They may be visible to the unaided eye at a distance
of three or four yards. (2) The caudal margins of the scales
on the sides bear small hemispherical organs, which are effective
behind the dorsal. (3) The upper surfaces of the pectorals and
both surfaces of the pelvic fins bear small organs. (4) The rays
of the dorsal fin bear small organs. Pearl organs do not occur
in the female (Fig. i).

8

JACOB REIGHARD.

The difference between the sexes in fin length are shown in

Table I.

TABLE I.

SHOWING IN MILLIMETERS THE AVERAGE LENGTH OF FINS IN MALES AND FEMALES

OF Catostomus commersonii OF EQUAL LENGTH, THE DIFFERENCE IN

AVERAGE LENGTH OF FINS AND THE PERCENTAGE DIFFERENCE.

Length, Tip

Cauda!

of Snout to

Lower

Anal.

Dorsal,

Pelvic,

Pec-

Base of

Caudal.

Lobe,
Length.

Length,

Length.

Length,

toral,
Length.

Males, M

^12

62

6s. S

47.8

42

1:7.1;

Females, F

"?I2

S4.6

40-8

4S-7

-27.7

ci

Difference, D == M - F

OOO

+ 7-4

+ 15-7

+ 2.1

+ 4-3

+ 6.5

Percentage of Difference,

% D = (M — F)-F.

ooo

+ 1^.4

+ 71. c;

+ c o

H-II d

+ 17.8

(+31- 1)

By fin length is meant the greatest distance from base of fin
to its margin, approximately the length of the longest fin ray.
The table is based on nine fish of which six were males. The
males ranged in length from 260 to 405 mm. with an average
of 312 mm. All had wrell developed pearls. The females aver-
aged 263 mm. in length (210, 215, 365). In order to compare
fish of equal length the female average length has been made
equal to that of the male and the average fin lengths of the
females as obtained from measurements have been corrected in
proportion. The figures in the lower horizontal line therefore
show in percentages the sex difference in length of fins in fish
of the same length. Since the caudals of two of the females
were broken there is added for the caudal a corrected percentage
(31.1) obtained by comparison of a single male of 340 mm. length
with a female of 365 mm.

From the table it appears that in fish of equal length the anal
and caudal (lower lobe) of the male are about 31 per cent, longer
than those of the female; the pectoral about 18 per cent, longer;
the pelvic about n per cent, and the dorsal only 5 per cent.
In the female the caudal lobes are of about equal length whereas
in the male the lower lobe in two perfect specimens averaged
about 10 per cent, longer than the upper. The fins appear to
differ in robustness in about the same proportion as in length.
This is shown by the width at mid-length of the longest anal ray
in a male and female of equal length; in the male 4.5 mm. in the

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 9

female 2.6 mm. In smaller individuals there is no great differ-
ence between the length of fins in individuals of opposite sexes,
but these differences appear with increasing size. A male of 135
mm. when compared to a female of 137 mm., had slightly shorter
dorsal, caudal and pectorals, but somewhat longer anals and
pelvics.

It is to be noted that in adult specimens those fins of the
male (lower caudal lobe, anal) that bear the largest pearl organs
also exceed the corresponding fins of the female by the largest
percentage. (Cf. Table I. and Fig. I.) Indeed the fins of the
male may be divided into three groups on the basis of the per-
centage by which they exceed the corresponding fins of the
female; the anal and caudal with 31 per cent, excess; the pec-
toral and pelvic with 10 to 20 per cent, and the dorsal with
5 per cent. The size of the pearls borne on the fins of these
three groups is roughly proportional to these percentages.

When the fish are seen close at hand, or under favorable
conditions with field glasses, the pearl organs of the male and
the greater size of his anal fin suffice to distinguish him from the
female (Fig. i). But were the sexes identical in form, size and
color, the behavior differences described in another place would
differentiate them.

5. Breeding Activities. — During the. breeding season males are
at all times much more numerous on the rapids than females
and during the greater part of the time none but males are
present. It does not follow that males are actually more num-
erous than females. On the contrary data that I have collected
at other localities and at times when the fish were not breeding
indicate that the males and females are equally numerous.
(Reighard, 1915.) In the breeding season the females do not
mingle with the males on the rapids until ready to lay their eggs.
But from time to time a female comes from her retreat in the
deeper water above or below the rapid or from beneath the
bank and takes her place on the rapid. If no males happen to
be near she may lie quiet in one place for a considerable time.
But if males are near they at once approach her, sometimes one
or two, sometimes as many as ten. Pairing is best seen when
but two males are involved and will be first described under these

10

JACOB REIGHARD.

conditions. As the males approach the resting female she
hurries forward as though to escape, but presently stops with
her belly. on the bottom. As the males again approach she hur-
ries forward a second time, but soon stops as before. .Thus she
appears to be driven here and there over the spawning ground,
too "coy" to allow the males near her. After a varying number
of apparent efforts to escape her "coyness" vanishes and she
rests quietly on the bottom and permits the males to come near.
In what follows I describe the spawning behavior as I saw it very
many times from the bridge or from the banks of Mill Creek.
Details were observed several times in the aquarium (Fig. 4).

When a male comes within a few inches of the waiting female
he is often seen to stop, spread his pectorals, erect his dorsal and
protrude his jaws (Fig. 4, second fish from bottom). Then, for
perhaps a second, his head trembles with a slight, rapid vibration
from side to side. The movement is not unlike the tremor of
a palsied hand. It is like the tremor that one may produce in-
his own head by strong continuous contraction of the muscles of
the neck. This tremor may be seen not only when a male
approaches a female, but often when he approaches another
male on the spawning ground. I have never seen it in a female.

FIG. 2. A female of Catostomus commersonii engaged in pairing, with a male
on each side of her. The body and tail of one male are shaded. The pearl organs
are shown on the anal and caudal of the nearer male, but not elsewhere. The
figure represents the pairing act near its end with the head of the female well above
the bottom, which is represented by the irregular horizontal line. Drawn with the
help of a photograph (see Fig. 4).

It is probably nothing more than a beginning of the tremor of
the whole body which accompanies spawning. It necessarily
produces a vibration in the water which may be of such a rate
as to stimulate other fish through the skin, ear or lateral line

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS.

II

organs (Parker, 1917). In many fishes the spreading of the
dorsal at the breeding season results in the display of a con-
spicuous marking or color pattern (Reeves, 1907 and citations).
This is true of some of the minnows to be described in later num-
bers of this series. //• is noteworthy that in the suckers, although
the dorsals are unmarked, the display movement (Fig. 2) does not
differ from that of those fishes that have a conspicuous dorsal.

As the male approaches the female another change becomes
evident. The dark stripe on his side (Fig. 4), which may have
become rosy a little while before, now suddenly turns to a vivid
scarlet and remains so during the pairing. At the same time
the eye becomes red and continues so while the female is present.

The males with fully extended pectorals and erected dorsals
now press close against the female one on either side (Figs. 2,3).

FIG. 3. Diagram showing the pairing of Catostomus commersonii as seen from
the dorsal side. Compare Fig. 2.

The smallest males that I have seen pairing were about six
inches long. The two are usually of the same size and commonly
are a good deal smaller than the female. But they are not always
of the same size and one or both may be as large as the female.
When the males are in position (Figs. 2, 3) their adjacent
pectorals are spread beneath the female. Their caudal fins press
on either side against that of the female, but may extend for
some distance behind and below it, so that their distal parts
press against each other. Their anals are spread and extended

12 JACOB REIGHARD.

downward so that they press against the sides of the tail and
sometimes against the anal of the female. The backs of the
males are arched and their dorsal fin-rays, spread like the ribs
of an extended fan, stretch the membrane between them (Figs.

2,3)-

The lateral surfaces of the anal and caudal and the sides behind
the dorsal, all of which are roughened by pearl organs (Fig. i)
are thus pressed vise-like against the female so that she is firmly
held (Figs. 2, 3). But the roughened caudals and anals of the
two males also press against each other where no part of the
female separates them. Thus the pearl organs aid males to keep
their positions with reference to the female as well as to each
other.

When the fish have come into position there is a rapid vibration
of the whole bodies of all three together. This is wide and
vigorous behind the dorsal fin, while in the region of the head
it is a little more than a tremor. At this time the fish are often
in water so shallow that their backs are exposed. The powerful
movement of the tails of the three fish stirs up the gravel and a
cloud of sand is released and washed downstream. At this
time, in the aquarium, one may see milt spurt from the genital
openings of both males and cloud the water. No doubt the
eggs are extruded at the same instant and buried in the gravel,
but the water is made so turbid by sand and milt that I have not
seen the eggs laid. I have estimated the length of the spawning
act at a second and a half but have not actually timed it. It is
often repeated especially by the larger males. Large males are
often taken in which the front edge of the anal fin is raw and
\vorn by rubbing against the gravel. I have seen nothing of the
sort in females, but their anal fins are smaller than those of
males and are protected by them during the spawning act, while
the number of pairings of the average female can be but half
that of the average male.

When a pairing act is completed the female moves on, usually
upstream, and presently pairs with other males on another part
of the breeding ground. Her eggs are thus scattered in small
lots over a considerable area, very likely over more than one rapid
and are commonly fertilized by many pairs of males. The two

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 13

males separate, the red stripe on their sides and the red in their
eyes fades, but the white occipito-lateral stripe remains for
some time. Each male now moves about and feeds as before.
And so they continue until another female appears when one or
both may succeed in pairing with her and this may happen
on any part of the spawning ground. Thus the eggs fertilized
by one male may lie anywhere in the gravel of a rapid or in that
of several rapids. The breeding activities are in no way centered
about individual males, for the cooperation of two males in
pairing makes it impossible to know what eggs are fertilized by
an individual male.

I have never collected from the bottom the eggs laid at a
single pairing. But the smaller fish lurking in the neighborhood
tell one plainly enough where they are. The black-nosed dace
(Rhinichthys atronasus) and the rainbow darter (Etheostoma
coeruleum) gather at once in great numbers over the spot where
the pairing suckers were. They come in a straight line from
down stream attracted, no doubt, by the trail of milt, eggs or
bottom materials swept down by the current. They gather in
an area six or eight inches across and each burrows in the bottom
with its snout as though seeking eggs. The whole little area is
soon concealed by their wriggling tails, close-set like threads in
the pile of velvet. Some of the eggs may have been swept down
stream, but many of them must be buried where the small fish
are rooting.

When more than two males follow a female (Fig. 4) it may be
difficult to see what happens. When she finally stops the two
males nearest or most vigorous in the pursuit attempt to pair
with her. But the others at once crowd about and try to force
their way between her and her mates. They try either to squeeze
in at the sides of the female from above so as to force her mates
outward, or to wedge themselves beneath the pairing males
from the side so as to force them up and take their places. But
once the two males have the female firmly held between them
it is difficult to dispossess either and I have never seen this
happen. As many as ten males have been seen with a single
female during the spawning act, and the act was nevertheless
completed; but often, when many crowd about, she interrupts

JACOB REIGHARD.

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 15

the pairing and moves upstream as though seeking an oppor-
tunity to pair unmolested. This she may secure and the pairing
then takes place in the manner already described. Although
supernumerary males may be present and may attempt to inter-
fere they take no part in the normal pairing. While they crowd
about the female and attempt to gain a place at her side, there
are no combats between them. When not at the side of the
female they seem to pay no attention to one another.

When the breeding season is over the male suckers lose their
pearl organs. They were beginning to shed them in southern
Michigan on May 13, 1913. Both sexes become uniformly
olivaceous on sides and back. They are no longer commonly
seen on their breeding grounds, and in Michigan most of them
seek deeper waters.1

B. The Common Red-Horse (Moxostoma aureolum'Le Sueur).

i. Breeding Grounds and Breeding Season. — In southern
Michigan the red-horse breeds in May. My two dates are
May 17, 1904, and May 4, 1905. On these dates the breeding
season of the white sucker was nearing its end. While the breed-
ing grounds of the two species are the quieter, upper parts of
rapids in shallow water with gravel bottom and, while the two
species often breed in the same rapids, I have not found the
red-horse in streams as small as those sometimes frequented by
the breeding white sucker. The sucker may spawn in brooks
so small that one may step across them but I have found the

1 In Walnut Lake Hankinson (1908) reports the species in water 15-40 ft. in
depth in summer and a few were found in a restricted part of the lake in water 80 ft-
in depth in April and May. I have found it in Douglas Lake, Cheboygan County,
Mich., from July to September in water up to 45 ft. in depth (Reighard, 1915).
Smallwood (unpublished notes) reports the return of suckers to Lake Clear after
breeding in Sucker Brook. On the other hand Forbes and Richardson (1908) say:
" It is with us essentially a species ot creeks and smaller rivers, nearly four times
as common, according to our data in the former as in the latter. . . . Our col-
lections show that it is much more likely to be abundant on bottoms with more
or less rock and sand than on a completely muddy bottom and that it has also a
decided preference for clear, swift water." Without a knowledge of the dates at
which Forbes and Richardson's collections were made or of the size of the fish
taken it is not possible to say to what extent adult suckers collected by them occur
on the rapids at other than the breeding season. Certain it is that they are abun-
dant in the deeper water of some inland lakes of Michigan and in the Great Lakes
when not breeding.

i6

JACOB REIGHARD.

breeding red-horse in streams not less than thirty or forty feet
in width.

2. Sexual Differences. — There are no known color characters
by which the sexes of the red-horse may be distinguished with
certainty at any season but there are structural differences.
According to Forbes and Richardson (1908) the lower fins are
"longest in the male." Table II. has been made in the same
way as Table I. (See p. 8.) The averages were obtained from
five males and two females, all breeding fish. It shows that
all the fins of the male are longer than those of a female of the
same length with the possible exception of the caudal. The
caudal appears to be 5 per cent, shorter in the male. But since
some of the caudals are imperfect a comparison was made of a
single perfect male of 205 mm. with a perfect female of 280 mm.
on the basis of equal length. This shows that the upper lobe
of the caudal is 12 per cent, longer in the male and the lower lobe
33 per cent, longer. The latter value is included in parentheses
in the table. The anal and pectoral are longer in the male by
about 15 per cent. The dorsals and pelvics are longer by about
10 per cent. The lower lobe of the caudal of the male is not
only longer than that of the female, but about 14 per cent,
longer than the upper lobe. In the female the two lobes are
of equal length.

TABLE II.

SHOWING ix MILLIMETERS THE AVERAGE LENGTH OF FINS IN MALES AND FEMALES

OF Moxostoma aureolum OF EQUAL LENGTH, THE DIFFERENCE IN

AVERAGE LENGTH OF FINS AND THE PERCENTAGE DIFFERENCE.

Length, Tip

Caudal

of Snout to

Lower

Anal,

Dorsal.

Pelvic,

Pect-

Base of Cau-
dal. Mm.

Lobe.
Length.

Length.

Length.

Length.

oral,
Length.

Average for s males, M

2T.Q

^6.8

SO. 4

47

T.6.2

4Q.2

Average for 2 females, F

2?O

60

44

4.3

T. 3

•21. C

Difference, D = M — F ....

OOO

— 3.2

+ 6.4

+ 4

+ 3-2

+ 7-7

Percentage of Difference,

% D = (M - F) : F. .

OOO

q.7

+ 14 c

4-Q7

+ 04

+ IS-8

(+33)

The breeding males are further distinguished by the possession
of conspicious pearl organs. (Figs. 5, 6.) On the end of the
snout and sides of the head as far as the caudal margin of the
preoperculum are numerous large sharp-pointed organs, more

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS.

f^'

M?

QI en c<

S4 & '—

b "« rt

.2 u M

•s x ^

en

e 8

0 'a

» a

rt y

bJO —

^ O

u rt
-~ Q>

^ =

en '^3

c
o

« H

o^o

£

CC

as..

. ~ - .2f

bii -3 d,

S c "S w

rt i;

a 3
a —

•'• rt

4-) G,

§ 1

£ O

T? e u
o g ^:

* S

5 -a

^ ..3

O «

o «

1) _x

-— - £J

H •=

c

o °

in ^

CJ nj

"rt g

S d

" TJ

o rt

^ rt

•o •—

^ o

? t

^ S

^ n

Q X

•g -a

s ~

d rt

«i <->

rt o

E t3

o

C -*-> £U

C '-*— ' --—

O 4J

en

c /• "^

M rs 3

s r: -- 1

rt '

fc S ~ i
SI'S &

1 8 JACOB REIGHARD.

than two hundred in a well-developed specimen. On the sides
of the anal and caudal fins and on the tail for a little distance in '
front of the caudal fin are large disc-shaped organs (not pointed).
Smaller, pointed organs occur on all the scales of the sides and
back, on the dorsal, on the upper surface of the pectoral and on
both surfaces of the pelvic. All these organs are effective in
proportion to their size. Here again the caudal and anal fins
of the male, which exceed those of the female by the largest
percentage, bear the largest and most effective pearl organs. It
is probable that further data would make possible a more precise
statement of the relation between size of pearls and relative size •
of fins.

This species is one of the few in which pearl organs have been
noted in the female. In a single specimen I have found minute
organs on the top of the head and on the first few scales of the
back behind the head. They were especially numerous about
the upper end of the opercular opening. They were too small to
be effective and care was needed to see them at all.

3. Coloration. — At 2:30 P.M. on May 17, 1904, I found ten
to twrelve red-horse lying quiet in shallow water near the bank
of Mill Creek, near Ann Arbor. The water was smooth and I
was able to come within ten feet of them and watch them with
field glasses from an elevated position. The fish were from
twelve to fourteen inches long from tip of snout to tip of tail.

The red-horse, seen at other seasons, whether in its native
waters or in aquaria, has the sides and back in both sexes uni-
formly olivaceous, but somewhat darker above. The belly is
smoky white. The sides show tinges of salmon in front of the
dorsals and the lower fins have some orange near the base.

The red-horse before me were of such exceptional coloration
that they were at first not recognized. Pectorals, ventrals and
anals were bright salmon. Along the sides and running forward
above the eye was a white stripe similar to that of the white
sucker. It was more prominent in the darker colored individuals
but in none of them was a red stripe visible beneath it as in the
white sucker. In a few individuals infrequent, elongated white
spots were seen above the lateral stripe and running lengthwise
of the back. With field glasses pearl organs were visible on

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. IQ

anals and caudal. With good lighting these could be seen with
the naked eye at a distance of ten or twelve feet. Subsequent
observation showed that all these fish were males.

4. Breeding Activities. — The fish were quiet most of the time,
but now and then one dropped downstream a few feet and then
slowly returned to his original position. Two of them were seen
to pick stones from the bottom as though feeding. After an hour
and a half a female joined the group of males. She was longer
and relatively thicker bodied than the males, not spotted with
white on the back and without visible pearl organs. She was at
once approached by five males, two of which took position one on
either side of her while the other three crowded down from above.
One of the upper males was seen to vibrate his tail for a moment
but no actual spawning took place at .the time. The fish re-
mained grouped for but an instant and then separated. The
female went upstream a little way and then dropped back
among the males. The group reformed, but immediately broke
up again. In these aborted attempts at spawning it was noted
that after a single male had placed himself by the side of a female
a second male, upon approaching on the same side, turned at
once to the unencumbered side. He behaved as though he
discriminated between the sexes of the two fish, but by what
means this was accomplished I could not tell.

The female again went upstream but this time to a greater
distance and followed by two males only. When she had come
to rest on the bottom one of the males approached and placed
himself by her side. After half a second the other male took his
place on her opposite side and spawning occurred very much
as in the white sucker (Figs. 5, 6). The backs of the two males
were strongly arched so that their dorsal fins were carried well
above that of the female and fully spread. Their caudal and
anal fins were close pressed against those of the female and against
each other as in the wThite sucker. Their snouts were turned
inward and pressed close against the sides of the head of the
female below her eyes (Figs. 5, 6).

While the males were in this position with backs bowed and
heads straining inward and upward the female was held firmly
by the functional pearl organs of their snouts, caudals, anals

20

JACOB REIGHARD.

and tail. The spawning vibrations lasted two or three seconds.
The fish then separated and the female went up stream. Repeat-
edly after this she dropped down among the males, but when
approached by them moved away and did not again spawn. The
spawning was not accompanied by any change in color (red
stripe) such as was noted in the white sucker.

Except in the attitude of the pairing males the breeding be-

FIG. 6. Dorsal view of a female of Moxostoma aiireolum pairing with two
males. Effective pearl organs are shown. Compare Fig. 5.

havior of the red-horse does not differ essentially from that of
the white sucker. The eggs are presumably buried in the bottom
as in the white sucker. Those of a single female are scattered
in small lots over a considerable bottom area. Those fertilized
by a single male are also widely scattered.

C. The Hogsucker (Calostomus nigricans Le Sueur).

i. General Activities. — Catostomus (Hypentelium) nigricans is
known locally in Michigan as the black sucker or pugamoo. Over
a wider territory it is known as hogsucker, hogmolly or stoneroller.
When not breeding it may often be seen feeding on the rapids of
our brooks, creeks and smaller rivers. In feeding, the fish puts
its snout under a stone and roots it up or thrusts it sidewise.
It then sucks up the slime between the stones and with it obtains
immature insects which form its chief diet (Forbes and Richard-
son, 1908). 1 have once seen the fish when thus engaged each

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 21

accompanied by ten or twelve small shiners (Notropis) and by
an occasional Campostoma. These formed a little school at
his sides and below him, and seemed to be waiting for fragments
from his feeding.

2. Sexual Differences. — In color the sexes appear to be alike
at all times. The back and sides are olivaceous, darker above,
with, dark irregular cross blotches. The belly is satiny white;
the lower fins are dull red. Thus the colors of the fish blend
with those of the stony bottom on which it is commonly seen.
I have determined the following percentage differences in fin
lengths (percentage D - = (M • - F) : F) for males and females of
the same size based on a single female and the average of two
males: caudal - 4.5, anal 17.4, dorsal - 0.2, pelvic 4.2, pec-
toral - 2.5. These differences are probably within the limits
of individual variation except in the case of the anal. The anal
of the male thus shows the greatest excess in length over the
corresponding fin of the female. The upper and lower caudal
lobes are of about equal length in both males and females.

This is one of the few species in which pearl organs are de-
veloped in both sexes. In the male (Fig. 7, lower male) they
occur on both surfaces of all the fins, on the upper surface of
the head, on the opercle and on every scale of the body and tail
except those of the ventral surface. The largest organs are
those on the anal fin, on the ventral half of the caudal and
on the sides of the tail, especially near the caudal fin. Those
of the anal reach a diameter of 0.8 mm. and a length of 0.28 mm.
in a fish 12 cm. in length, while on the caudal of the same speci-
men the organs are about 0.5 mm. wide, high and sharp pointed.
On the remainder of the fish the organs are small, 0.08 mm. to
0.25 mm. in diameter. All these organs make the surface rough
to the touch and are effective in proportion to their size. It is
again noteworthy that in the male the anal bears the largest
pearls for it is of all the fins the one that shows the greatest
percentage of excess length over- that of the female. In the
female the organs are smaller than those of the male but have a
similar distribution. They are absent from the dorsal fin and sides
of the body and from the ventral surfaces of pectoral and pelvic.
On the anal they are nearly as large as in the male, while on the

22 JACOB REIGHARD.

caudal they are somewhat smaller, still smaller on the sides of
the tail. Those on the anal are distinctly perceptible to touch,
those on the caudal and sides of the tail are barely perceptible
while the rest are quite imperceptible. Probably only those on
the anal are in any degree effective and they are not sharp. It is
evident that neither the coloration, the length of the fins, nor
the pearl organs afford means of discriminating the sexes in the
field. For this purpose one is compelled to rely on the difference
in average size and in behavior.

3. Breeding Activities.- — I have several times seen a single large
hogsucker moving upstream in rapids and accompanied or fol-
lowed by three or four smaller. In one case the large fish was
some twelve inches long and the four following her half as long.
Occasionally she stopped and one of the smaller fish placed
himself by her side. But nothing further occurred and the fish
presently moved on.

My only opportunity to observe the actual spawning of this
species was on May 4, 1904, in Mill Creek, near Ann Arbor, at
the point at which the spawning of the red-horse had been already
seen. I was watching the rapids about 4:30 P.M. when a large
hogsucker came upstream followed at a short distance by half a
dozen others of two thirds her length. Size and behavior indi-
cated the larger fish in this and other cases to be a female.
Presently she stopped and remained quiet on the bottom while
the males pressed against her three on either side, so close as to
hide every part of her except the head and tip of the caudal fin
(Fig. 7). The seven fish remained together for several seconds
and during this time the female several times made rapidly re-
peated movements of protrusion and retraction of the mouth.
She was not seen to make any other movement nor was any
seen in the males. After remaining thus grouped for a fraction
of a minute the fish moved on, the female leading. Two some-
what larger males now approached and when the female again
stopped these added themselves to the other six, so that the
eight of them formed a complete mantle over her back and
sides from which only her head and caudal projected. Again the
female was seen several times to make rapidly repeated move-
ments of the mouth. The fish then passed out of sight on their
way upstream, the female still leading.

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS.

<c =
I) —

to

-I

a

•5 p.
. ii

s d

2 S

I'o
o

J3

"o -

GJ ij

P-H C

rt „_

n «

0)

15

S

24 JACOB REIGHARD.

Although the absence of the vibratory movements character-
istic of spawning suggests that the fish whose behavior is de-
scribed did not actually spawn, the attitudes can be no other
than those of spawning fish. These attitudes can be the more
readily maintained owing to the general distribution of pearl
organs over the surfaces of both sexes. Not only is the occurrence
of these organs on the female exceptional, but their distribution
in the male is unusual in that they are found on the lower surfaces
of the pectoral fins. Males which are above the female presumably
have the lower surfaces of the pectorals in contact with her back
and sides while those at the sides of the female have the upper
surfaces of their pectorals in contact with her belly.

Assuming that the spawning behavior was observed it differs
from that of the white sucker in the following particulars,
(i) The act is not preceded by change of color; the fish show the
same color characters as at other seasons. (2) The act was
participated in by more than two males. If it be objected that
all but two of the males present were supernumerary, and that
the spawning was aborted by their presence as in the case of the
white sucker, it may be replied that the fish remained in position
long enough to spawn and that those above maintained their
positions quietly and did not attempt to displace those below.
The whole appearance was that of spawning except for the
single feature of lack of vibration.

SUMMARY OF OBSERVATIONS.
A. The White Slicker (Catostomus commersonii).

1. The white sucker breeds in southern Michigan in April
and early May.

2. The fish then congregate in the shallow swift water of small
streams where the bottom is gravel and sand.

3. At this time both males and females have a yellow-white
occipito-lateral stripe not known to occur at other seasons.

4. This stripe is whiter in males than in females and below
it on the sides males have a dark lateral stripe often of rosy tinge.
The backs of males are often flecked with white not observed in
females.

5. The fins of the males exceed those of females of equal

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 25

length by the following average percentages; anal and lower lobe
of caudal 31 per cent., pectoral 18 per cent., pelvic II per cent.,
dorsal 5 per cent.

6. Effective pearl organs occur in the male on the anal and
lower lobe of the caudal on the upper surface of pectorals, on
both surfaces of pelvics and dorsals and on the scales of the
sides behind the dorsal (Figs. I, 2, 3).

7. The size of the pearl organs on the fins of the males is
roughly proportional to the percentage by which these fins
exceed in length the corresponding fins of the female.

8. Females are without pearl organs.

9. In the field males may be distinguished from females by
their coloration, by .their larger anal fins, by the possession of
pearl organs and by behavior.

10. The smallest males and females observed breeding were
about six inches long.

11. The males congregate on the rapids and the females
remain apart from them in the immediate neighborhood and
come to them one or a few at a time.

1*2. When a female comes to a rapid she is pursued by two or
more males.

13. The female intermittently flees from the males but finally
stops and permits their approach.

14. The pairing is preceded by a characteristic tremor of the
head of the male and preceded or accompanied by a spreading of
the fins and a change in the color of the dark lateral stripe from
black to a brilliant red.

15. The fins of the males are without conspicuous colors or
patterns; yet they are spread by the pairing males just as in
certain other species of teleosts in which the spreading displays
conspicuous colors or patterns.

16. The female "pairs" with two males at one time (Figs. 2, 3).

17. The two males may be of unequal size and both may be
smaller than the female.

1 8. In the pairing position, one male lies on either side of the
female with his pectoral fin spread beneath her and his dorsal
elevated. The pearl organs of his sides and of his caudal and
anal fins are then in contact with the sides, caudal and anal of
the female (Figs. 2, 3).

26 JACOB REIGHARD.

19. The caudal and anal fins of the two males may project
beyond those of the female so that their distal portions are
in contact with each other by their roughened surfaces (Figs.

2,3)-

20. The roughened surfaces of the males aid them to hold the

female between them.

21. By their contact with each other the roughened surfaces
of their caudals and anals may aid the two males to maintain
their positions with reference to each other.

22. Spawning is accomplished during a rapid vibration in
unison of the bodies of all three fish. This agitates the bottom
material and causes its lighter parts to be swept downstream
by the current.

23. During spawning milt is seen to spurt from the genital
openings of the males and cloud the wrater.

24. After spawning the fish separate and each may repeat the
act many times in various places and in combination with various
individuals.

25. Small fish usually gather at once over the area in which
the spawning occurred and there crowd together in a small space.

26. These fish root in the bottom with their snouts and appear
to be eating eggs that have just been buried.

27. At each pairing act supernumerary males may be present
and may attempt to supplant those actually engaged in pairing

(Fig. 4)"

28. At the close of the breeding season, the pearl organs are
shed and both males and females become uniformly olivaceous in
color.

29. Outside the breeding season the mature fish are not seen
on the rapids in great numbers but in Michigan are abundant in
lakes in deeper water.

B. The Red-Horse (Moxostoma aureoleum).

30. Red-horses congregate on the rapids in southern Michigan
in May in situations like those occupied by white suckers, but
not in streams as small as some of those in which the white
suckers breed.

31. In both sexes at this time there is a white lateral stripe,

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 27

not observed at other seasons. It extends forward above the
eye.

32. The backs of certain males show elongated white flecks
running lengthwise, but they do not show the red lateral stripe
characteristic of the white sucker at the moment of pairing.

33. The fins of males are longer than those of females of equal
length by the following average percentages; lower lobe of caudal
33, anal 14.5, pectoral 16, pelvic 9, dorsal 9.

34. The lower lobe of the caudal of the male is about 14 per
cent, longer than the upper lobe.

35. Effective pearl organs occur in the males on all the scales
of the back and sides; on both surfaces of caudal, anal, pelvic
and dorsal; on the upper surface of the pectoral and on the sides
and top of the head to the end of the snout (Fig. 5).

36. Minute non-effective pearl organs occur occasionally in
females on the top of the head and on the first few scales of the
back behind the head.

37. Of all the fins of the male, the caudal and anal, which
exceed the corresponding fins of the female in length by the
largest percentage, bear the largest and most effective pearls.

38. The males may be distinguished from the females on the
breeding grounds by the pearl organs, by length of anal fin, by
behavior and in some cases by differences in coloration (white
flecks on the back).

39. The behavior of the fish on the breeding grounds is like
that of the white sucker in the particulars stated in paragraphs
II," 12, 13, 14 (except color change), 15, 16, 17, 18, 19, 20, 21,

22, 24, 27, 28.

40. The attitude of the pairing males during spawning differs
from that of the white sucker (par. 18) in that the backs are
more arched and the snouts with their covering of pearl organs
are pressed against the sides of the head of the female (Figs.
4 and 5).

41. In the position indicated in paragraph 40 the female is
held not only between the tails, caudals and anals of the two
males but between their heads as well.

28 JACOB REIGHARD.

C. The Hogsucker (Catostomus nigricans}.

v

42. The hogsucker has been seen to breed in early May on the
grounds used by the red-horse.

43. No difference in the coloration of the sexes has been noted
at any season.

44. The males observed were on the average much smaller
than the females.

45. The anal fins of the male are on the average about 17 per
cent, longer than these of females of equal length. The remain-
ing fins do not differ greatly from those of the female.

46. Both sexes are provided with pearl organs which are larger
and more widely distributed in males than in females.

47. Effective pearl organs occur in the males on both surfaces
of all the fins, on the upper surface of the head, on the opercle
and on every scale of body and head except those of the ventral
surface (Fig. 7, lower male).

48. Pearl organs occur in the female in the same situations as
in the male except that they are lacking on the head, sides of
body, dorsal fins and lower surfaces of the paired fins. Those of
the anal are effective.

49. Of all the fins the anal of the male exceeds that of the
female in length by the largest percentage. It also bears the
largest and most effective pearls.

50. In the field the sexes may be distinguished by difference
in average size and by behavior.

51. The breeding behavior is like that of the white sucker in
the particulars enumerated in paragraphs II, 12, 13, 15, 20, 24.

52. In the hogsucker not less than six nor more than eight
males have been seen to "pair" with a single female at one time.

53. In the spawning position one male lies on each side of the
female and others place themselves above these and against the
sides and back of the female so as to form a mantle about her,
from which her head and tip of her caudal may project (Fig. 7).

54. The distribution of pearl organs generally over body and
fins in both sexes is such that in the spawning position the
surfaces of males in contact with females and with other males as
well as certain of the surfaces of females in contact with males
are more or less roughened and the fish are thereby the better able
to maintain their positions.

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 29

CONCLUSIONS.

The breeding behavior of three species of suckers has been
studied, the white sucker (Catostomus commersonii), the red-
horse (Moxostoma aureoleum), and the hogsucker (Catostmits
nigricans). All three make use of similar breeding grounds, the
upper parts of rapids, where the water is moderately swift and
the bottom gravel and sand. They differ in that some of the
streams in which the white sucker breeds are smaller than any
frequented by the other two species.

It is possible to discriminate the sexes of all three species on
the breeding grounds either by differences in coloration, size of
body and fins, pearl organs, behavior, or by some combination
of these characters. In all three species the males are seen to
congregate on the rapids while the females linger in the neighbor-
hood and enter the rapids at intervals, usually singly, sometimes
two or three at once. When a female has come to the rapid
she is at once pursued by as many males as happen to be near.
She flees, stops, flees as the males again approach, and so con-
tinues for some time alternately fleeing and stopping. Sooner
or later she comes to rest on the bottom and permits the males
to approach. When these have come to her, in the case of the
white sucker and the red-horse, two of them pair with her at one
time, one on either side. The fins of the males are without con-
spicuous colors or patterns. Nevertheless during pairing and
just before it the males spread their fins after the manner of
fish which thereby display conspicuous colors or markings. The
display movement occurs, although there is nothing to display.
Supernumerary males may approach and by crowding those
attempting to pair may interrupt the pairing act. In the hog-
sucker six or eight males may pair with the female at one time.
All appear to take an equal part so that no supernumerary males
can be distinguished.

In all three species pearl organs occur on the male and in the
hogsucker effective organs occur also on the female. The largest
and most effective organs on the fins of the male occur on those
fins that exceed the corresponding fins of the female by the
greatest percentage of length. In all three species the pairing
attitudes are such as to bring into contact with the female those

3O JACOB REIGHARD.

surfaces of the males roughened by pearl organs. There is at
the same time more or less contact with one another of the
roughened surfaces of cooperating males and of those of the
female of the hogsucker with pairing males. Thus the pearl
organs aid the pairing fish to keep their positions with relation
to one another in the swift water during the vigorous vibrations
which characterize spawning. These vibrations are very pro-
nounced in the white sucker and red-horse but have not been
observed in the hogsucker. They continue in the white sucker
for about a second and a half and during that time the tails of the
fish agitate the bottom and the lighter bottom materials are
swept down stream by the current.

When the pairing white suckers have separated, their eggs
are left buried in the bottom. This is inferred from the behavior
of the numerous minnows which congregate over the spot.
The similarity of spawning behavior indicates that the eggs of
the other species are buried in similar fashion.

The females of the species studied deposit eggs in various parts
of the breeding ground and in doing so each pairs with many
males. It results from the breeding activities that the eggs of a
single female are widely scattered and are fertilized by many

la-Az

yBz

zCw

xDz

wCx

yDx

wAy

xBz

yBx

yAx

zDy

•Ic't'V

n'Dz

wBx

zCy

xAz

xAy

xDw

n'By

yCx

xCz

zBw

xAw

wDy

FIG. 8. Showing the distribution of pairings and their character over the
spawning area in the case of suckers in which two males pair with one female.
A, B, C, D, females; -w, x, y, z, males.

males. It results further that the sperm of a single male fertilizes
the wide-scattered eggs of many females. The cooperation of at
least two males in pairing with a single female makes it impossible
to know the male* parentage of a given embryo. This same
cooperation makes it impossible that the eggs deposited in any
small, continuous bottom area should be fertilized by one male.
They are in fact not fertilized by one pair of males. The rela-

BREEDING BEHAVIOR OF SUCKERS AND MINNOWS. 3!

tions of the sexes are as indiscriminate as they well can be. We
may represent four females of the white sucker or red-horse by
the letters A, B, C, D, and four males by the small letters w, x,
y, z. Six pairs may be formed with the four males, wx, wy, wz,
xy, xz, yz. Assuming that each female spawns once with each
pair of males and that her spawnings are distributed at random
over the breeding ground we should have some such space-
distribution of pairings as shown in Fig. 8. In the hogsucker in
which one female pairs with more than two males the relations
would be still more complicated.

No male occupies any particular locus of the spawning ground
and attempts to defend it against other males. On the contrary
each male is free to wander over the whole spawning ground. He
may "pair" in any part of it, for he does not enter into combat
with other males but cooperates with them. The female does
not restrict her activities to any part of the spawning ground.
Were she to do so she would be finally beset by so many males
that normal pairing might be difficult or impossible. She does
not actively reject any of the males, but when beset by so many
that spawning is difficult, she seeks a new locus. This she con-
tinues to do until normal pairing becomes possible. She may
be said to try various situations of the spawning ground until
she finds one in which the conditions permit spawning. This
relation of the sexes is neither polyandry nor polygamy. It is
promiscuity, corresponding to the hypothetical communal
marriage of primitive man. It occurs along with lack of combat
amongst the breeding fish. .Where combat occurs, as in min-
nows, promiscuity gives place as will be shown in later papers,
to a sex relation that approaches polygamy.

LITERATURE CITED.
Boulenger, G. A.

'04 Teleostei, systematic part, in The Cambridge Natural History, Vol. VII..

pp. 541-727. London.
Forbes, S. A. and Richardson, E. R.

'08 The Fishes of Illinois, pp. cxxxi-357, text and atlas of 103 maps, with 40
colored plates, 15 black and white plates, and 76 figures in the text. Urbana.
Hankinson, T. L.

'08 A Biological Survey of Walnut Lake, Michigan. A Report of the Biological
Survey of Michigan, published by the State Board of Geological Survey-
as a part of the report for 1907, pp. 157-288, pi. XIII.-LXXV., 6 text figs.
Lansing.

32 JACOB REIGHARD.

Parker, G. H.

'17 The Reception of Mechanical Stimuli by the Skin, Lateral-line Organs
and Ears in Fishes, Especially in Amlurus. Am. Jour. Phys., XLV., 463-
489. Published also as Contribution from the Zoological Laboratory of
the Museum of Comparative Zoology at Harvard College, no. 298. Cam-
bridge.
Reeves, Cora D.

'07 The Breeding Habits of the Rainbow Darter (Etheostoma cceruleum Storer),
a Study in Sexual Selection. BIOL. BULL., XIV., 35-57. 3 text figures
Lancaster.
Reighard, Jacob.

'04 Further Observations on the Breeding Habits and on the Function of the
Pearl Organs in Several Species of Rventognalhi. Science, Vol. XIX.,
p. 212.

"15 An Ecological Reconnoisance of the Fishes of Douglas Lake, Michigan, in
Midsummer. Bulletin of the U. S. Bureau of Fisheries, Vol. XXXIII.
(for 1913), pp. 219-249, map and ten tables. Document no. 814, U. S.
Bureau of Fisheries. Washington.

ROUGHOID, A MUTANT LOCATED TO THE LEFT OF

SEPIA IN THE THIRD CHROMOSOME OF

DROSOPHILA MELANOGASTER.

LEONELL C. STRONG,
COLUMBIA UNIVERSITY.

ORIGIN OF THE CHARACTER, ROUGHOID (RU).

In March, 1919, Dr. A. H. Sturtevant handed over to me a
culture of Drosophila melanogasler, containing a new mutant,
named by him "roughoid," owing to its similarity with the old
character "rough."

In December, 1918, he had collected a single pair of wild
flies near Columbia, S. C. The F2 generation from these was
found, two months later (February 14, 1919), to contain many
roughoids (culture 5570 A.H.S.). Hence one of the wild flies
must have been heterozygous for the mutant gene. The point
of special interest in connection with roughoid is that its locus
is to the left of that of sepia and therefore establishes a new
left end of the map of the third chromosome.

DESCRIPTION OF THE ROUGHOID CHARACTER.

This gene produces somatic effects mainly upon the size and
texture of the normal eye; (a) it materially decreases both
the length and width of the eye as well as making it more convex;
(6) the ommatidia lose their hexagonal shape, and become
crowded together irregularly; (c) due to this irregularity, the
hairs project in all directions from the surface; (d) the hairs are
somewhat thickened and strikingly longer than in the normal
eye. These effects are very constant in appearance and have
proven to be favorable for purposes of classification. Associated
with these effects there are found sometimes, (e) several black
ommatidia distributed over the surface of the eye, but more
commonly in the posterior ventral region. They tend to darken
the normal red eye color. They are more than likely a part of
the roughoid character since I have never seen this effect in any

33

34 LEONELL C. STRONG.

fly other than one that was also roughoid. However, their
occurrence is very inconstant, most roughoid eyes being as light
red as the normal wild type.

CHROMOSOME CARRYING ROUGHOID.

The first experiment (culture No. 142) consisted in out-
crossing a roughoid male to a female showing the second-chromo-
some recessive vestigial.1 The FI generation flies were all wild
type, thus proving roughoid to be a recessive character. By
inbreeding the FI flies, F? was obtained. One quarter of the F2
vestigials showed the roughoid character (culture No. 1426).
The presence in F2 of the double recessive class proves that
roughoid cannot be located in the second chromosome. At the
same time, I out-crossed a roughoid female to a dichaete hairless
male from stock (culture No. 141). These two characters are
thirdrchromosome dominants. By back-crossing the FI dichaete
hairless females to roughoid males from stock, evidence was ob-
tained that roughoid shows linkage relations with these mutants
in the third chromosome. The size relation of the various classes
indicated that roughoid was considerably to the left of dichaete.
As the back-cross flies were not hatching out as well as expected,
I did not complete the count of the cultures of this experiment.

A cross between roughoid and the original rough gave in FI
only wild-type flies, showing that roughoid was neither rough nor
an allelomorph of rough.

LOCATION OF ROUGHOID IN THE THIRD CHROMOSOME.
The next step was to make up a stock containing both the new
character, roughoid, and sepia (located at the extreme left end of
the then known third chromosome). In the third generation
from a cross between roughoid and sepia, I obtained such a stock.
Since I had obtained roughoid and sepia together in such a short
time, I concluded that the two were not very closely linked.
By using the back-cross method, I obtained the following results:

1 This female was from the stock known as "5 pie" and carried in addition to
vestigial, the second-chromosome recessives black, purple, arc and speck. These
other characters were disregarded in the Fs classifications.

ROUGHOID.

35

ru se ru se

X <7

ru se

July 24, 1919.

Culture No.

ru se. +

ru. se.

Total.

Crossovers.

Crossover

Value.

601

177 79

31 27

270

58

21.5

602 . ...

99 98

^8 77

272

7^

27.5

606

106 87

77 77

267

70

26.6

607

56 68

26 IS

165

41

24.8

608

71 67

22 26

186

48

25.8

Total .

46 <; 700

mo 142

n<;6

202

2^.2

ru +

ru se

X c?

+ se

ru se

May 24, 1919.

Culture No.

ru. se.

ru se. +

Total.

Crossovers.

Crossover
Value.

4CK

71 77

16 36

20O

?2

26.O

The total of these cultures (1,356 flies) shows the cross-over
value for roughoid and sepia to be 25.3.

Again, a dichaete female (known to be heterozygous for rough-
oid) was backcrossed to a roughoid male from stock. From
that experiment, I found the cross-over value for roughoid and
dichsete to be 35.9, as follows:

ru

D

ru

X c?

ru

July 10, 1919.

Culture No.

ru D. +

ru D.

Total.

Cross-overs.

Cross-over
Value.

605

59 64

41 28

192

69

35-9

This indicates that roughoid is situated to the left of sepia,
since the normal distance between sepia and dichaete is known to
be about 11.7 units.

LEOXELL C. STRONG.

Three-Point Back-Cross — ru se X D.

In order to place the gene more accurately, a three-point
back-cross was undertaken. A sepia roughoid female was
crossed to a dichaete male, the FI dichaete females were then
back-crossed to roughoid sepia males from stock (pair matings
being made). Results:

ru se

ru se

X

D

ru se

July 7, 1919.

o.

J-

2.

I, 2.

Culture No.

ru se. D.

ru D. se.

ru se D. +

ru. se D.

Total.

604

82 82

17 3O

16 18

I O

246

6ll

4S 51

15 24

14 12

2 O

163

Total

127 137

72 S4

30 30

3 °

400

26O

86

60

3

The results give a cross-over value for roughoid and sepia of
19.5 and 25.1 respectively, or a mean of 21.7; a cross-over value
for sepia and dichaete of 14.2 and 17.2 or a mean of 15.4.

Four-Point Back- Cross — ru se X D H.

One other back-cross experiment was performed, using the
four characters roughoid, sepia, dichaete and hairless. From
this, 980 flies were counted, giving the cross-over value for
roughoid and sepia as 25.6.

+ +

ru se

+ +

ru se

August 12, 1919.

+ + D H

X

ru se

o.

i.

2.

3-

I, 2.

!. 3-

2^.

I, 2, 3.

Culture No.

ru D
se. H.

ru
D se.
H.

ru
se +.
D H.

ru
se D.
H,

se
ru. D
H.

ru se
D. H.

ru
se H.
D.

ru se
H. D.

600 .

^2 (>A

2T, 24

17 14

1 1 14

2 O

10 6

I 6

2 I

6lO
6l2

53 64
30 54

21 1 8
14 19

16 9
10 16

22 18

7 18

0 0
I 0

3 5
6 7

4 6
4 3

0 2
0 0

Total ....

135 182
317

58 61

119

43 39
82

40 50
90

3 o
3

19 18
37

9 15
24

^

5

ROUGHOID.

37

May 24, 1919.

ru + D H

+ se + +

X

ru se + +

ru se + +

0.

i.

2.

3-

I, 2.

i. 3-

2, 3.

I, 2, 3.

Culture No.

ru
D se.
H.

ru D
se. H.

se
ru. D
H.

ru se
D. H.

ru
se +.
DH.

ru
se D.
H.

ru se
H. D.

ru
se H.
D.

AOA

6l 7=;

24 J2

2^ 'O

2O 0

0 O

4. 12

I 4

I ?

136

66

45

29

0

16

5

6

Summing up the data in the previous tables:

Culture No.

Total Flies.

ru se Value.

se D Value.

D H Value.

404 .

"?CH

2O. O

18.4

18.4

600 . .

247

28.1

17.4

20. 6

6lO

241

2O.^5

I S.^

24.9

612 .

180

24 8

17. Q

2^.8

Total . .

080

2^.6

17. ^

21.6

The value for dichaete hairless (21.6) is approximately the
expected result (24.8). Sepia dichaete crossing over is noticeably
high, since the average normal value is 11.7 units. This is
possibly due to a complicating factor, linked to the new factor,
roughoid, and "stretching out" that part of the third chromo-
some. More work will have to be done before this can be
certain. It will be noticed that the double cross-over class (i, 2)
is lower than the triple cross-over class (i, 2, 3). This is un-
expected and remains unexplained at the present time.

SUMMARY.

(a) A new mutation has occurred in the third chromosome of
Drosophila melanogaster to the left of sepia, which for several
years has been the leftmost of the known loci.

(b) There was a roughoid sepia cross-over value of 24.9, based
on a total of 2,748 flies of which 685 were cross-overs.

(c) Associated with roughoid, there is possibly a linkage mod-
ifier increasing the crossing over between sepia and dichaete, and
perhaps between roughoid and sepia as well.

Vol. XXXVIII. February, 1920. No. 2.

BIOLOGICAL BULLETIN

NOTES ON THE BIOLOGY OF SOME COMMON

LAMPYRID^:.1

WALTER N. HESS.

CONTENTS.

PAGE
Introduction 39

History of Biological Work on Lampyrids 40

Photinus consanguineus 41

Photinus scintillans 44

Photurus pennsylvanica 49

Pyropyga fenestralis 67

Purpose of Luminosity 72

Economic Importance , 73

Summary 73

Bibliography 75

INTRODUCTION.

The fireflies (family Lampyridse) are among the most common
of insects, yet because of the larval habits of most species, com-
paratively little is known regarding them except what has been
learned from a study of the adults. The larval forms are rarely
seen, as most of them are active only at night, and usually are
found on, or in the ground, in damp or marshy regions.

The insects were observed both in the field and in specially
prepared large plant-pots at the insectary. These pots were
filled about half full of rich mellow earth on which was placed a
small amount of moss. Some of the pots were covered with
glass, while others were covered with cheese cloth. Since ovi-
position occurred very readily in captivity, eggs were obtained
for a study of the incubation periods of the different species, by
confining ripe females in small jars that had been partially filled
with earth. For a study of the feeding habits of the larvee and
adults, these insects were confined in glass jars, some of which

1 Contribution from the Entomological Laboratory of Cornell University.

39

4O WALTER N. HESS.

contained only a small amount of moist filter paper, while others
contained earth and moss. Since no reference was found to a
published account of the nature of the mouth-parts of any of our
native firefly larvae, these structures were figured in order to
better illustrate the method of feeding of these larvae.

The following species were studied: Photinus consanguineus
Lee., Photinus scintillans Say, Photurus pennsylvanica DeGeer
and Pyropyga fenestralis Mels.

The author is indebted to Dr. James G. Needham, Dr. William
A. Riley and Dr. O. A. Johannson, under whose supervision the
greater part of this study was made, for their helpful suggestions
and criticisms.

HISTORY OF BIOLOGICAL WORK ON LAMPYRIDS.

Although many workers, both in this country and in Europe,
have studied the light-organs of the fireflies, comparatively little
has been done on the biology of this group of insects.

Newport (1857) studied the life-history of the glow-worm, or
larva, of Lampyris noctiluca. He not only discussed the de-
velopment and general habits of these larvse, but he also per-
formed several experiments to determine the nature of their
feeding habits.

Hudson (1891) published an interesting account of the habits
and life-history of the New Zealand glow-worm.

Barber (1905, 1914) described the egg-laying habits of Phen-
godes, together with certain habits of this group of fireflies.

Bongardt (1904) published a brief account of the biology of
certain European Lampyridse.

Knob (1905) described the habitat, flight and light-emission
of Photinus scintillans and Photurus pennsylvanica.

McDermott (1910, 1911, 1912, 1914) studied especially the
nature of the flashing of fireflies, together with the attraction
of the sexes by means of light-emissions.

Olivier (1911) described the distinguishing structural char-
acters of the common species of fireflies. In addition, he dis-
cussed their general habitats and distribution.

Mast (1912) studied the sexual attraction of fireflies (Photinus
pyralis ?) with special reference to their orientation. He found

BIOLOGY OF SOME COMMON LAMPYRID.E. 41

that the female, in responding to the flash of the opposite sex,
always turned the ventral side of her abdomen, so that it would
emit light in the direction of the male.

Fabre (1913) maintained that the glow-worm, in feeding on
snails, injected a substance in the nature of an anaesthetic which
paralyzed its host.

Vogel (1915) did by far the best biological work that has been
done on this group of insects. He described the external and
internal anatomy of the larva of Lampyris noctiluca, together
with its life-history. He observed that the larva lived in the
ground, and that it fed on snails. By making a careful study of
the structure of the mouth-parts, pharynx and gizzard he found
that the digestive juices of the mid-intestine were emitted through
the hollow mandibles. By this means the larva was able to
paralyze its prey, and to digest the tissues before eating them.

Haddon (1915) described also the process of feeding and the
nature of the mouth-parts of Lampyris nocliluca.

Blair (1915) and Morse (1916) reported the interesting phe-
nomenon of the synchronous flashing of fireflies, in which the
fireflies in a given locality were found, at times, to flash in unison.

Williams (1917) described the life-history of several of our
common Lampyrids. His discussions of the biology of Phoiinus
consanguineus and Photurus pennsylvanica are especially valuable.

Photinus consanguineus Lee.

The insects of this species are elongate and slender with the
head covered by the prothorax. The prothorax is rounded on
the anterior and lateral sides, truncate behind with the angles
acute. It is light yellow with a black median bar, which is
bordered with pink on either side. The elytra have wide side
margins and bear two or three sub-obsolete carinae. The suture
and side margins are pale yellow, while the remainder of the
elytra is grayish in color. They are granulate and rather pilose.
The abdomen in the male is depressed, but in the female it is
often rounded, due to being distended with eggs. The eyes
of the male are larger and better developed than those of the
female. The light-organs of the male cover the entire sternites
of the sixth and seventh abdominal segments, while in the

42 WALTER N. HESS.

female the organ occupies only a small area on the sixth abdominal
segment. They measure from 8 to 12.5 mm.

LeConte reports them from Massachusetts, Pennsylvania
and Virginia. Blatchley lists them for Indiana, and Williams
found them abundant in Massachusetts.

The biology of this species has been well described by Williams
(1917), so an attempt will be made here to discuss only a few of
the more important features.

The adults begin to emerge about June I and can be found
along moist areas until about the first of August. They are
frequently found in association with Photinus scintillans and
Photurus pennsylvanica. The males are active fliers, and though
the females have well developed wings they were never found in
flight. The flight of the males begins about 8:15 P.M. and
continues until about 10:00 P.M. (old time). The light of the
male is a single bright flash, though at times he emits two or
three flashes in rather close succession, but in every case there is a
considerable interval between each flash. The female of this
species crawls up a stem of grass, or some similar object, and
emits a faint flash in response to the flash of the male. Just
before emitting the light, however, she turns her abdomen so
that the ventral side is in the direction of the male, thus in part,
at least, obviating the necessity for larger light-organs.

What was discribed as the synchronous flashing of fireflies
was first discussed by Blair (1915), who reported observing
fireflies, in a certain locality, flashing in unison. Later his
observations were confirmed by other writers. Morse (1916)
reported an observation in which the light emitted by these
little creatures pulsated in a regular synchronous rhythm, so
that at one moment the tree, about which they were flying,
would be one blaze of light, while at another the light was dim
and uncertain.

According to Blair (1915) and McDermott (1916), this phe-
nomenon does not occur among the American species of Photinus
and Photurus. The writer, however, observed the flashing of
fireflies in unison on two very dark evenings during the present
summer while collecting eggs and larvae at Ithaca, New York.
Toward the south side of the City Cemetery is a small valley,

BIOLOGY OF SOME COMMON LAMPYRID/E. 43

and on both occasions the entire valley, for a moment, was a
blaze of flashing lights, and then for a moment it was in darkness,
except for an occasional flash, which seemed to come from fire-
flies of different species than those that were flashing in unison.
The fireflies in this particular locality were almost entirely of the
species Photinus consanguineus , and at each period of flashing
both males and females were observed to emit light. On both
nights this phenomenon occurred shortly after it became dark,
at approximately 9:00 P.M. On the first night the phenomenon
was observed for approximately fifteen minutes. How much
longer it continued after that was not determined. By very
careful observation it was discovered that each period of flashing
started on the crest of the hill at the south side of the valley,
by one, or only a very few flashes, and that the impulse stimu-
lated by these few insects instantly appeared to sweep over the
valley, resulting in the great mass of flashing lights. On the
second night an experiment was performed in which it was
discovered that, by standing on the side of the valley and causing
short flashes with a pocket flash-light, the fireflies of the entire
valley responded. At first, after estimating the length of the
latent period, the flash-light was flashed just before the normal
time for the fireflies to flash, with the result that the entire mass
of fireflies responded. Then two flashes were emitted from the
pocket flash-light with the interval between flashes reduced to
about three fourths that of the normal flashing period, and the
fireflies responded with apparent equal results. Finally, the
period was reduced to approximately one half the normal time.
The fireflies as a mass appeared to respond to the first short
flash. The second time a large per cent of them responded,

»•

but after this second short period, the flashing in unison was so
disturbed that each insect flashed independently of the flashing
period of the others. Blair (1915), in commenting on the
reason for the synchronous flashing of fireflies, states: 'The
flashing in unison is too regular to be caused by chance puffs of
wind. A more probable explanation of the phenomenon is
that each flash exhausts the battery, as it were, and a period of
recuperation is required before another flash can be emitted.
It is then conceivable that the flash of a leader might act as a

44 WALTER N. HESS.

stimulus to the discharge of their flashes by the other members
of the group, and so bring about the flashing of the whole family."
From the observations and experiments performed above, it
seems evident that the theory of a leader is the most probable.
Although the author has done extensive collecting of these
insects during the past four summers, at no time was this phe-
nomenon observed on the part of any of our other native species.

The small, smooth, spherical eggs are laid on the soil at the
base of the roots of grass and moss where they hatch in about
twenty to twenty-two days. The larvae are slender, elongate,
and of a rather uniform dark grayish color. They were found
chiefly a short distance below the surface of the soil, though a
few specimens were taken at the surface. It seems, however,
that its habitat is subterranean, rather than terrestrial, in con-
trast to that of Photurus pennsylvanica. They are predacious,
feeding on snails, etc., similar to the other species studied.

It seems very evident that the insects have a two-year life
cycle, as both mature and half-grown larvae were taken at the
time of pupation.

Pupation takes place in the soil near the surface of the ground.
Here the mature larvae excavate a little chamber, in which the
period of transformation is spent. After transforming to pupae
they lie on their backs in an arcuate position. In this condition
they measure about seven mm., but when straightened out they
are about nine mm. long. The pupae are yellowish white in
color with the pleural regions somewhat pinkish. The pupal
period lasts from twelve to fifteen days.

Photinus scintillans Say.

The adult beetle of this species is rather elongate and slender'^
somewhat flattened with the head completely covered by the
prothorax. The antennae are eleven-segmented, the second seg-
ment being short and transverse. The prothorax is rounded an-
teriorly and along the sides, truncate behind with the angles
acute. The elytra have wide side margins. The head is black.
The prothorax is pale yellow, except the small black median bar
on its central posterior half, which is bordered with pink. The
elytra are pale yellow, except the side and suture margins which

BIOLOGY OF SOME COMMON LAMPYRID^.

45

are a lighter yellow. They are finely granulate and pubescent.
The females resemble the males except that they are slightly
larger and their abdomens, frequently distended with eggs,
project considerably beyond the elytra. The eyes of the male
are larger than those of the female. In the male the light-organs
cover the entire sternites of the sixth and seventh abdominal

FIG. i. FIG. 2.

FIG. i. Photinus scintillans male, ventral view of abdomen. The shaded
portion on the sixth and seventh abdominal segments represents the adult light-
organ (AO).

FIG. 2. Photinus scintillans female, ventral view of abdomen. The shaded
area on the sixth abdominal segment represents the adult light-organ (AO).

segments, while in the female the organs occupy only a small
area at the center of the sixth sternite (Fig. i and Fig. 2). The
beetle measures from 5.5 to 8 mm.

These insects were found by LeConte in Massachusetts,
Pennsylvania and Kansas. Blatchley found them in Indiana.

The natural habitat of this species appears to be very similar
to that of Photurus pennsylvanica. Yet, however, they are
much more widely distributed, and are often found in drier
localities than is characteristic of this species. They are very
common about central New York, being our most common
firefly. They were found very abundant at Ithaca, New York
in the City Cemetery on Stewart Avenue, and, in fact, on the
campus, and on the lawns near many of the private houses.

The period of emergence for this species varies from about
June i to July 20, depending on the season.

46 WALTER N. HESS.

During the daytime the adults are rarely found, as they
remain hidden underneath the leaves of low herbage, or near
the ground covered by grass or moss.

Neither sex was observed feeding, so it is not known whether
these insects eat as adults or not, but it is probable that they do
since they live for two or three weeks.

The flight of these insects begins very early in the evening,
considerably earlier than that of the other native luminous
species. Almost with the first sign of twilight they are flying
about, and because it is comparatively daylight the beetles
themselves can be distinctly seen. In fact, when they begin
flying, their flashes appear rather faint, due to the daylight ob-
scuring their brilliancy. Before total darkness sets in, the
flight of these insects has ceased, unless for an occasional one
here and there. These few stragglers, however, soon stop.
Among the fireflies there seems to be, to a certain extent, an
evening periodosity of flight. This species of fireflies flies only
during twilight, after which other species as Photinus consan-
guineus and Photurus pennsylvanica take its place. The flight
begins shortly after 7:30 P.M. and continues until about 8:30
P.M.

In comparison with most of our luminous species of fireflies,
the males of this species fly very low and slowly, and emit one
distinct flash at each period of flashing. Females were never
taken in flight and it seems probable that they never fly. In-
stead, they climb to near the top of some projecting blade of
grass, or some similar object, and apparently remain quiet until
a male comes near. Like the male, the female emits only one
flash, but it is much less distinct.

There can be little doubt but that the light-emissions among
these insects serve as definite signals between the sexes, by
which means the male is able to find the female of his species.
Numerous observations confirming this were made, which can
be well illustrated by the following example : on the evening of
June 30, 1916, a male Photinus scintillans was observed flying
about two feet above the ground, and a second after he flashed
a female that was almost underneath him flashed. He appeared
to drop directly to the ground, but his velocity of flight carried

BIOLOGY OF SOME COMMON LAMPYRID^. 47

him about a foot beyond where she was resting. He at once
quickly ascended the nearest stalk of grass, seemingly expecting
to find her near its top. He then began to descend, at which
time he emitted a flash of light and she responded. He then
rushed up another stalk of grass that was in her direction and,
not finding her, he flashed again. She responded, and he flew
in her direction but alighted nearly a foot beyond her, and he
then immediately ascended the nearest stalk of grass. Flashes
were exchanged for a period of twelve minutes, during which
time he ascended about twenty stalks of grass, and he flew in
her direction five different times before he finally found her.
Each time she responded to his flash, and each time she orientated
her abdomen so that the light was emitted in his direction.
It was 8:13 P.M. when he found her, and at once copula-
tion took place. This lasted until 8:58 P.M. when they separ-
ated, and at once crawled down the blade of grass to the ground
where they were concealed by the vegetation. On another
occasion, while the male was in the grass about a foot from a
female in his search to find her, a small pocket flash-light was
used, in which it was found that she would readily respond to a
very short flash. As soon as the pair was in copulation the
flash-light was again flashed several times, at different intervals,
but with no response. After the pair separated the flash-light
was again flashed, but with no results. After copulation took
place neither one of the pair was observed to flash, and after
separating they concealed themselves in the grass, without
emitting light, all of which seems still further to prove that the
light-organs serve to bring the sexes together, and having accom-
plished this end, they are no longer functional until at some
possible later date. In a very few instances flashing was ob-
served on the part of one of the copulating members, but it
seemed to occur only when they were disturbed, which was the
exception rather than the rule.

The pocket flash-light referred to above was used with good
success in collecting females. They usually responded when
the light was at a distance of at least eight feet, but rarely
responded when it was nearer.

Oviposition usually took place about one week after emergence

48 WALTER N. HESS.

and continued for a period of two or three weeks. As the eggs
of the females do not all ripen together, they are deposited over a
considerable period of time. Like the other species studied
the small, wrhitish eggs were deposited on the ground, at, or
near, the base of moss and grass. They were usually laid singly,
though in a few instances they were found in masses. When
deposited they were covered with an adhesive substance, which
caused them to adhere to the object on which they were placed.
The period of incubation occupied from eighteen to twenty-one
days.

In this species, as in all the others- studied, the eggs appeared
very slightly luminous at the time of laying, but this faint lumin-
osity disappeared in about a day and there was no more light
emitted from the eggs until the larval light-organs became func-
tional shortly before hatching.

The newly hatched larva is whitish, except for the black lateral
eyes and brownish mouthparts. It soon becomes pigmented,
appearing dull gray in color. At this stage it measures about
2.4 mm.

The mature larva resembles in general shape that of the first
instar. It is elongate and narrow, varying from 12 to 13 mm.
in length. Its head is small, being about half as wide as the
prothorax. Like the other species studied, the head can be
withdrawn into the thorax. The body is widest in the region
of the thorax and tapers gradually posteriorly. The head is
black, the tergites dull gray, and the pleural regions slightly
pinkish. The habitat of the larva is largely subterranean,
though it is usually found near the surface. Its feeding habits
resemble very closely those of Photurus pennsylvanica. It is
not active during the day.

The larvae were not reared from eggs to adults, yet while
collecting these insects each spring, larvae of two sizes were found :
some that were mature and others that were about half as large,
indicating that the insect, probably, has a two-year life cycle.

Pupation usually takes place near the surface, although it
sometimes occurs under stones. The pupal period is rather
brief, taking from nine to twelve days. The pupa assumes an
arcuate position, lying on its back within the pupal cell. It

BIOLOGY OF SOME COMMON LAMPYRID.E. 49

j

measures about 8 mm. in this arcuate position, but when straight-
ened out it is about 10 mm. long. The lateral tergites project
slightly and each bears a group of short setae. The body is
somewhat flattened, being yellowish white in color except along
the lateral sides of the thorax and adbomen, which are slightly
pink. The larval light-organs function throughout the pupal
period and do not degenerate until shortly after the emergence
of the adult. The adult light-organs, which develope indepen-
dently of the larval organs, become functional shortly before
the adults emerge.

Plwturus pennsylvanica DeGeer.

This is one of the largest of our native fireflies throughout
central New York, and though very common in certain moist
localities, it is by no means our most common species.

The adult insect is elongate, somewhat flattened, with the
head partially covered by the thorax. The head is rather
rounded, and slightly narrowed behind the large .convex eyes.
The antennae are eleven-segmented, slender and tapering, ex-
tending about half the length of the body. The prothorax is
rounded anteriorly and along the sides, and subtruncate pos-
teriorly. It is dull yellow with a central and basal dark stripe,
while the disc at each side of the dark area is red. The surface
is rather coarsely punctate. The elytra are elongate, extending
considerably beyond the end of the abdomen. They are brownish
except the lateral margins and a narrow tapering area extending
from the anterior part to beyond the center, which is a dull
yellow. The body is covered with short yellowish pile. The
length ranges from 12 to 15 mm. An illustration of the male is
shown on Fig. 3.

The sexes are similar in form except that the female is slightly
larger than the male. The light-organs of the male cover the
entire sternites of the sixth and seventh abdominal segments
(Fig. 4), while these organs occupy only about two-thirds of
the corresponding region in the female (Fig. 5). In the male
the abdomen ends in a point while that of the female is truncate
on the tip.

This insect is widely distributed throughout North America.

5O WALTER N. HESS.

According to LeConte (1851) it is abundant in every part of the
United States. Blatchley states that it is the most common
firefly in Indiana. Williams also makes a similar statement for
Massachusetts.

As is characteristic of our luminous fireflies, the adults of this

FIG. 3. Phoiurus pennsylvanica male.

insect are usually found active only at night. Like many other
insects, this species has well-defined centers of distribution,
being rarely found except along marshy or moist localities.
This, however, is more characteristic of the larvae, but as a rule
the adults are found comparatively rarely outside of such regions.
They were found most abundant in the City Cemetery, on Stewart
Avenue, and in the Renwick marshes, at Ithaca, New York.
The cemetery, while not in any sense marshy, has been filled in

BIOLOGY OF SOME COMMON LAMPYRID^.

with dark loamy soil, which is usually moist in the depressions
along the walks and between the graves.

The period of emergence for this species, in this locality,
extends from June 5 to July 15, depending on weather conditions.

During the daytime the adults remain in seclusion, usually
at the base of moss or grass, although occasionally specimens
were found clinging to the underside of leaves of low vegetation.

Many insects during their adult life eat little or no food, but
the adults of this species, especially the females, are very vora-

FIG. 4. FIG. 5.

FIG. 4. Photurus pennsylvanica male, ventral view of abdomen. The shaded
portion on the sixth and seventh abdominal segments represents the adult light-
organ (AO).

FIG. 5. Photurus pennsylvanica female, ventral view of abdomen. The shaded
portion on the sixth and seventh abdominal segments represents the adult light-
organ (AO).

cious in their feeding habits. During the evening, while they
are active, they are either found flying, or on the ground, usually
about the base of grass, actively in search of food. The females
were commonly observed devouring other species of fireflies
(Photinus scintillans, Photinus marginellus and Photinus con-

52 WALTER N. HESS.

sanguineus). In each case, however, the firefly being devoured
was a male, which had probably attracted the Photurus by means
of its frequent flashing, as the females of these species flash very
rarely, except in response to the flash of the male. In captivity
the female was often found devouring the males of her own species,
and occasionally a member of her own sex. The males were not
found feeding, though it seems probable that they also are
predatory at times.

Though both sexes fly readily, comparatively few females were
found on the wing during the early part of the season, although
later (about July first to fifteenth), when it seemed evident that
there was a less abundance of males, the females were found
flying fully as much as those of the opposite sex. The flight
began about 8:15 P.M. and continued until after 10:00 P.M.,
although by that time comparatively few were flying. Some
have been observed on the wing as late as I :oo A.M. Unlike
our other native species, these fireflies frequently fly high and
their flashes can often be seen in the tops of the highest trees.

The flashing of this firefly is very distinctly different from that
of any of our other native species. Like the other fireflies, these
have definite periods of flashing. The male flashes three, four,
and even five times in rapid succession at each period. The
flashes are bright, although they become less distinct at the
end of each period. In the case of the female the number of
flashes at each period is reduced to three, two or one. There
is usually a longer period between the flashes than in the male
and they are less distinct. These periods of flashing occur at
rather regular intervals of about eight to ten seconds. The
male flashes slightly more frequently than the female. When
on the ground the brilliancy and regularity of flashing, on the
part of both sexes, seems to be the same as when they are flying.

It is agreed by most students of fireflies that the light-emissions
serve to bring the two sexes together, although McDermott
(1911) and Williams (1917) seem to doubt that they can serve
such a function in this species. In the other native luminous
species that were studied there is a definite interchange of flashes,
in which the female responds to the flash of the male. In
Photurus the female is an active flier and flashes frequently,

BIOLOGY OF SOME COMMON LAMPYRID^i. 53

whether in the presence of the male or not. In no case was there
observed a definite exchange of flashes between the sexes, yet it
seems very evident that the light-emission functions in bringing
the sexes together. On several occasions, while holding females
in my hand, males flew to them and they would have alighted
had my presence not scared them away, and on two occasions,
while holding males, females flew and alighted beside the cap-
tured males. This would lead one to believe that there is a
definite sexual attraction by means of light-emissions between
the sexes of this species, and that the female, having become an
active flier, is also attracted to the male. In each case observed
where the two sexes were attracted to one another, both con-
tinued to flash actively, but in no case was there any evidence
that one was responding to the flash of the other. In no case
did the female of this species assume a vertical position or expose
the abdomen so that the light would be flashed in the direction of
the male. It is possible that the brilliancy of the light-emissions
on the part of the female has obviated the necessity for such
exposure. Even while the females were on the ground the
flashes were easily perceptible for a considerable distance.

Copulation was observed in the field on three occasions.
Unlike our other native fireflies, these beetles were never observed
in copula on, or near, the ground, but while clinging to the leaves
of trees, often at considerable height. While in this state the
flashing of the light-organs apparently ceased, though one female
was found emitting a rather dim continuous glow during this
period.

Egg laying usually began about one week after emergence
and continued at intervals for a period of about two or three
weeks. Several females, which were captured early in the
season, were dissected to determine their egg laying capacity.
On the average, each female contained about fifty mature eggs,
with from seventy-five to a hundred smaller ones, indicating
that the eggs were at varying stages of maturity.

The characteristic place for the oviposition by the female is at
the base of grass or moss in damp loamy soil. Oviposition was
not observed in the field, yet on several occasions eggs were
found which had been deposited at the base of the roots of grass

54 WALTER N. HESS.

and moss. In confinement, at the insectary, oviposition was
observed on several occasions. The female walked slowly over
the soil, thrusting out her long ovipositor into the depressions in

the earth, where the light yellowish eggs
were deposited. Eggs laid by these
females were usually deposited singly,
though sometimes in masses. Some were
placed from one eighth to a quarter of
an inch underneath the surface, others
on the surface, and some were deposited
on the roots, basal stems and leaves of
moss and grass. The egg (Fig. 6) is

FIG. 6. Photurus pennsylva-

nicaegg. small, nearly spherical, about .7 by .8

mm. in diameter. It is without surface

markings, though at the time of laying it is covered with an
adhesive surface.

It is frequently stated in literature that the eggs of fireflies
are luminous, and Williams (1917) states that the eggs of this
species glow when deposited and probably continue to emit light
until the time of hatching. The certainty of this statement
seems in doubt. At the time of laying, the eggs were found to
be very slightly luminescent for a period of two days, but in no
case did they, at this period nor until the light-organs of the
larva were developed, definitely emit light. Eggs removed
from the ovaries of a ripe female showed no evidence of luminosity,
though Williams thought he saw a light in some of the eggs
that he so removed. Fabre (1913) states that the eggs are lumin-
ous even before leaving the body of the mother, but this seems
very much in doubt. The so-called luminescence of the eggs
at the time of laying is probably due to the substance with which
they are covered, rather than to any internal property of their
own, and as this becomes dry the slight luminescence disappears.
At a period about four days before the eggs hatch the larval
light-organs become functional, and from this time until hatching,
the eggs emit a distinct light.

In the breeding cages at the insectary where normal outdoor
conditions were maintained, the eggs of this species hatched in
from twenty-five to twenty-seven days, depending on weather

BIOLOGY OF SOME COMMON LAMPYRID.E.

55

conditions at different periods. The largest number hatched
on the twenty-sixth day. The newly hatched larvae are not
pigmented for a period of a few hours, but they soon resemble
the mature larva in shape and appearance, except for their mina-
ture size. At this stage they meas-
ure about 2.25 mm. in length. As
nearly as could be determined the
larvae were in the fourth instar by
winter. The first two molts take
place rather early in the life of the
larva, the first occurring at about
the age of two or three weeks.
The mature larva (Fig. 7) is elon-
gate, rather narrow, varying from
16—19 mm. in length, and it is
about three times as long as it is
wide. The head is small, a little
less than one third as wide as the
prothorax, and it can be withdrawn
into a pouch within the thorax.
The Antennae are three-jointed, the
mandibles are arcuate and notched
near the middle. The legs are some-
what spinose and of nearly uniform
length. The body is much flat-
tened. The prothorax is rounded
on the anterior and lateral sides
and subtruncate behind, much the
same as in the adult. This is the
largest of the body segments.
From the metathoraeic segment to
the caudal end of the larva the ter-
gites are concave posteriorly espec-
ially the caudal ones. Each bears a spine on its caudo-lateral
margin. The head, mouth-parts and tergites are colored a dark
brown, except for a few irregular pale yellowish areas. The
dorsal surface is more or less irregularly coarsely punctate. The
last segment of the abdomen is provided with numerous retrac-

FIG. 7. Photunis pennsylvanica
larva, full grown, dorsal view.

WALTER N. HESS.

LO—

tile elements, the caudal filaments (Fig. 8, C, F), which are used
in propelling the body forward. On the lateral sides of the
eighth abdominal sternite are two luminous areas, the larval
light-organs (LO}.

Resembling to a certain extent the habits of the adults, these

larvae unless disturbed were not
found active during the daytime.
At this time they are usually under-
neath stones or concealed in de-
pressions in the ground. At night
they become active and their light
can be frequently seen as they
wander about in their natural hab-
itat. When the ground is smooth
a small garden rake is of advantage
in collecting them, as they can be
easily disturbed by it, thus expos-
ing their light-organs to view. The
glow of their light-organs is not
visible when the larva? are lying
on the ground, as it can only be
seen when the ventral sides of their

FIG. 8. Photurus pennsylvanica
larva, ventral view of abdomen, abdomens are exposed to view. { A

LO, larval light-organ; CF, caudal flash-light is also of assistance in
filaments' collecting them.

While in the field, collecting on numerous occasions during the
evening, larvae were found wandering about apparently in search
of food. On two occasions they were taken while feeding on
snails, which they had evidently killed a short time before being
discovered. As the larvae crawled about, their heads were fully
distended, the maxillary palpi and antennas were constantly in
motion, and it appeared as if they were feeling their way by
means of these organs. From observations on their movements,
even when a snail was very near, it seemed very probable that
the larvae find their food by chance, and having found it they
tap it several times with their maxillary palpi and antennae before
beginning to feed. Since newly hatched larvae and those one-
year-old were abundant, efforts were made to determine more in
detail the methods of feeding and the nature of their food.

.--C F

BIOLOGY OF SOME COMMON LAMPYRIDJE. 57

Larvae taken into the laboratory were placed under as normal
conditions as possible, where various experiments were per-
formed to determine the possible nature of their food. In no
case were newly hatched larvae found in the act of killing their
prey, though they were observed feeding on bits of snail that
was cut up and placed near them. They fed much the same as
the older larvae, and, in fact, there can be little doubt but that
their food habits are similar. As the larvae were not active
during the daytime, and as they were disturbed by artificial
light, my observations on the feeding habits were largely limited
to the larvae of Pryopyga fenestralis which is active during the day.
On six different occasions a slug (Agriolimax campestris Binney)
was placed with six larvae of Photurus pennsylvanica and in
every case it had been eaten before morning. A slug (Agriolimax
agrestris L.) and a snail (Succinea avery Say) were put in with
six larvae. The snail was eaten during the first night, but the
slug was not killed and eaten until the third night. On two
occasions a small earthworm (Lumbricus terrestris L.) was placed
in a jar, without earth, which contained eight larvae. One was
killed and eaten the second night, and the other on the fifth
night. On two occasions a very large specimen of Limbricus
lerrestris was placed with twelve larvae. In each case the earth-
worm was not disturbed, though it remained with the firefly
larvae for over a week, and they received no other food during
that time. On two occasions a potato-beetle larva (Leptinotarsa
decemlineata Say) was placed in a jar with six larvae and each
time it was eaten the first night. On two occasions cutworm
larvae (Paragrotis messoria Harris, Paragrotis tessellata Harris
and Peridroma margaritosa Haworth) were each placed in jars
with six larvae and in every instance they were eaten the first
night. Finally, on four different nights, two second and two
third stage squash-bug nymphs (Anasa tristis DeGeer) were
placed with six larvae, and in each instance they were eaten
before morning. Sowbugs (Oniscus asellus Paulmeier), wire-
worm larvae (Agriotes mancus Say), ants (Formica sp.) and coleop-
terous beetles including the common ground beetles (Nebia
pallipes Say and Chcelnius pennsylvanicus Say) were placed in
with these larvae, but they were never eaten, indicating that the

58 WALTER N. HESS.

larvae require a soft-bodied animal into which they can pierce
their mandibles and inject the poisonous secretion.

These results, while giving no definite data as to the exact
food of these larvae, lead one to conclude that they probably eat
any soft-bodied insect larva, mollusca or annelid, that they happen
to find in their nocturnal wanderings. Snails are probably one
of their chief foods, and though these animals are not supposed
to be very abundant, they were found abundantly at night in
the damp regions where these larvae live. Cutworm larvae were
also abundant. Earthworms, except very small ones, were not
eaten until they had been with the larvae for a considerable length
of time, while other food was eaten very readily. This would
seem to indicate that probably snails and small insect larvae, expe-
cially cutworms, are their natural foods.

Among the larvae of many members of the Lampyridas, as
well as among certain other more or less widely separated groups
of insects, digestion takes place entirely, or partially, outside of
the body. This is accomplished by the digestive juices being
exuded from the mouth upon the food which is later eaten by the
larvae in a more or less completely digested condition.

'It is characteristic of most insects that feed in this way to have

FIG. 9. Photurus pennsylvanica larva, labrum, dorsal view.

very small heads, so that the mouth is not sufficiently large to
take in only very small pieces of food. These insects, like the
larvae of the Dytiscida?, are predaceous, feeding on living animal
food. The nature of this food is such that it could not be easily

BIOLOGY OF SOME COMMON LAMPYRID.E.

59

AO

chewed, like vegetable food or decayed animal tissues, and hence
it seems that some such method of feeding became necessary.
Some have the mandibles grooved, as the larvae of the Chry-
sopidse and Myrmeleonidae, while others have them pierced by a
small canal, such as certain of the larvae of the Dytiscidse and
Lampyridae.

The most extensive work on this subject was done by two Euro-
pean workers, Vogel (1912, 1915) and
Haddon (1915), although earlier work-
ers made a less detailed study of the
problem.

When the head of this insect is
withdrawn into the thorax, only the
tips of the mandibles and other mouth-
parts are visible, but when it is ex-
tended the large mandibles surrounded
by the other mouth-parts can be dis-
tinctly seen.

As the anterior half of the mandib-
les are exposed on the dorsal side, the
labrum lies considerably caudad on
the head, extending across the basal

FIG. 10. Plioturus pennsyl-

portions of these large jaws. It is mw/ca larva, left mandible, dor-
rounded On the lateral margins, while sal view. AO, anterior opening
on its cephalic border are three prom- of mandibuiar canal; c, con-

r . . ... dyle; PO, posterior opening of

inent forward projecting portions with .... . ~ . ,,

mandibuiar canal; T, tooth on

rather acute terminations. Its dorsal inner edge of mandible.
side (Fig. 9), shows numerous long

projecting bristles while the ventral side is covered with a rather
dense mass of small setae which project forward.

The mandibles (Fig. 10) are very strong. Each has a large
curved, anterior tapering tooth and they meet in a median line
slightly anterior of the head. On the inner median margin of
each mandible is a secondary tooth (7"). At the base is a knob-
like condyle (C) by which the mandible articulates with the head.
Both mandibles are covered for their entire extent with setae of
varying length, except at the distal end, and around the condyle.
About the base is a dense brush of short setae which project for-

po

6o

WALTER N. HESS.

ward. Between this area and the secondary tooth there are
larger and stiffer setae. Those covering the remainder of the
surface are rather short and dense. Extending from the base to
near its tip is a tubular canal, through which digestive juices

FIG. ii. Photurus pennsylvanica larva, labium, maxillae and lacinia, ventral
view. LP, labial palpus of labium; M, mentum of labium; SM, submentum of
Jabiuni; MP, maxillary palpus; ST, stipes of maxilla; Cd, cardo of maxilla; L
lacnia; G, galea.

pass, while the larva is feeding. It does not open at the tip but
slightly caudad on the outer margin.

The maxillae, labium and lacinia lie on the ventral side of the
head, and in these larvae they are fused into a flat fleshy plate
(Fig. n). When examined on its ventral side, the maxilla has
at its caudal portion a small triangular plate, the cardo (Cd),
which bears several short setae. Anterior to it is the large elon-

BIOLOGY OF SOME COMMON LAMPYRID^. 6l

gated stipes (St), which also bears several long bristles and many
shorter setae. Anterior of the stipes on the external side is a
stout four-segmented palpus, the maxillary palpus (MP),
bearing several forward projecting setae. At the side of the
maxillary palpus, internally, is the two-segmented galea (G) of
the maxilla, which resembles very much in appearance a two-
segmented palpus. Beside the galea, internally and also extend-
ing a short distance along the stipes, is the rather dense chitinized
flattened lacinia (L). Both the galea and lacinia bear many
setae, and along the inner margin of the lacinia is a row of rather
stiff bristles. The tip of the galea ends in short setae, which
probably function as specialized sense organs.

On the ventral side of the labium, the submentum (SM)
appears much elongated and it lies in
the central region of the mouth-parts.
It bears numerous short setae and a few
long bristles. The mentum (M) is in
the form of a thickened bi-lobed struc-
ture, from which project anteriorly the
two-segmented labial palpi (LP). Both
structures bear a considerable number
of anteriorly projecting setae and bris-
tles. On the underside, the sclerites of FIG. 12. Photums pennsyi-
the maxillae, labium and lacinia are in- vanica larva- hypopharynx.

... ... . dorsal view.

distinct, though in this region there are
numerous small anteriorly projecting setae.

The hypopharynx (Fig. 12) is triangular in shape with the
anterior portion bi-lobed. It is entirely covered with a dense
mass of anteriorly projecting setae, except along its basal portion.
The setae of the anterior half are much denser and longer, and
many of them are branched.

From the previous description of the mouth-parts, it is evident
that there are numerous setae projecting forward inside of the
mouth region. These setae, according to Haddon (1915) and
Vogel (1915), function as a strainer in preventing all except
liquid food from entering the mouth.

In the pharynx region there are several heavily chitinized
plates to which numerous muscles are attached, and which,

62 WALTER N. HESS.

according to these same authors, function as a suction pump in
drawing the liquid food into the mouth. Vogel also described
how the mid-intestinal secretions are forced forward, by means
of heavy oblique and circular pro-ventricular muscles, into the
mouth and out through the mandibular canals.

Though the intestinal juices were not observed passing through
the canals of the mandibles, a rather dark-colored liquid was seen
suspended from the ends of the long curved teeth and on the
bodies of snails and earthworms in the region where the larvae
were feeding. It, therefore, seems very probable that a certain
liquid, which, since the larvae have no salivary glands, must
come from the region of the mid-intestine, is exuded through the
canals of their mandibles and out through the mouth, and that
this liquid functions in paralyzing and digesting the tissues of
their prey.

The larvae were observed, however, to take into their mouths
portions of food of considerable size. The fact that they can
take up mouthfuls of earth and masticate it in the construction
of their pupal-cells is evidence that the larvae are able to take into
their mouths small masses of food before it is completely digested.
Yet there can be little doubt but that the-greater part of the food
is digested outside of the body and taken in through the mouth,
in the liquid state. Whether the mandibles function in the
intake of food was not determined, but the greater part of it
was apparently taken in through the mouth. The portions of
undigested food that were taken into the mouth, were no doubt
largely digested here before passing en into the intestine, as
the larvee masticated these masses for a considerable time before
they disappeared.

The larval light-organs are fully developed at birth, so the
larva is luminous from the time that it hatches until it finally
enters the pupal state. As stated before, the light is emitted
from two elliptical areas on the ventral side of the eighth ab-
dominal segment. The larval light-organs do not emit light in
flashes, as the organs of many adults do, but on the other hand,
the glow is nearly uniform. While the larvae were active, the
light was found to glow continuously. During the dormant
periods, as during the day, but especially during hibernation,

BIOLOGY OF SOME COMMON LAMPYRID^. 63

the glow becomes very faint and it frequently is not perceptible
even while holding the larvse ventral side up in a dark room.
By moving the insects about so as to agitate them the lights
usually become visible. It seems probable that the brilliancy
of the glow is in direct proportion to the activity of the larvse.

The larvae which were kept in confinement at the insectary
were found to go underneath stones, or enter cracks in the soil,
late in October, in preparation for hibernation. Some constructed
about themselves earthen chambers, while others occupied natural
depressions in these protected places. In no instance were they
found lying on their dorsal side, such as is characteristic of the
pupae.

During the warm nights of April the larvae leave their winter
quarters and go about in search of food. At this season their
little lights can again be seen as they wander about in their
natural habitats at night.

Since this species of insect lives as a larva at the base of grass
in moist loamy soil, and since it does not enter the ground, or
seek other natural means of concealment in which to pass the

•• l-v...

FIG. 13. Photuriis pennsylvanica partially constructed pupal chamber.

pupal period, as many members of the Coleoptera, as well as
some Lampyrids do, it constructs a small earthen chamber in
which to pass this period of transformation.

A suitable spot on the surface of the ground, usually at the
base of moss or grass is chosen, and at once the larva begins
building a lattice work of soft earth over itself, in the shape of a
small dome (Fig. 13), by which means it conceals itself , in about a
day. In the construction of this cell the larva removes earth
from underneath itself by means of its mandibles. This it
masticates and mixes in its mouth for a period of about half a
minute. It then extends its head to the lattice work of the

64 WALTER N. HESS.

dome and regurgitates the moist earth in the form of a short
ribbon-like mass, which it applies to the walls of the chamber.
By the frequent repetition of this process, the lattice-like frame-
work finally entirely covers the larva (Fig. 14). Even after it
forms a complete dome, the larva can be seen for several hours
between the meshes, before it is entirely concealed. By re-
peatedly removing the earth from the bottom of the chamber

FIG. 14. Pholurus pennsylvanica pupal chamber completed, lateral view.

and adding it to the inside of the dome-like wall, the chamber
is deepened, and its covering is strengthened and made thicker.

The completed chamber is in the form of an elliptical depres-
sion in the ground, about one half of an inch in length and about
seven sixteenths of an inch in width. Even after the larva con-
ceals itself it continues to add to the walls of the cell until they
are from one eighth of an inch, to as much as one half of an inch
in thickness.

The time spent in building the pupal-cell is about two days,
though larvae sometimes continue to excavate for three or even
four days, making a firmer and thicker covering for their cells.

That the intestinal secretions of the larva are used for moist-
ening the earth, which is used in constructing the cell, there can
be little doubt, yet it evidently has no special adhesive content,
for the pupal-cells are easily broken, and they seem to offer no
more resistance than ordinary earth which has dried, after
having been mixed with water. This liquid seems to serve simply
as a fluid in which to mix the earth and make it plastic.

From the examination of several of these completed pupal-
cells, it seems evident that the method of construction is such as
to allow a small amount of air for respiration to pass in and out
between their meshes. Some of the domes of the completed

BIOLOGY OF SOME COMMON LAMPYRID/E. 65

cells, when held against the light, allowed small rays to pass
through, giving evidence of their slight porous nature. The
pupal-cell, however, makes a sufficiently well-constructed cham-
ber to protect the pupa from drying or other injury. The
reason that the larva usually seeks a damp locality previous to
pupation is probably for the purpose of choosing a place where
excessive drought will not be liable to affect it during trans-
formation.

By the time the pupal-cell is constructed the larva becomes
very sluggish, its body becomes distended and in from one to
three days the cuticula splits down the anterior half of the
back, and the pupa gradually comes forth. From this time,

FIG. 15. Photurus pennsylvanica pupal chamber completed, internal view with

pupa in position.

throughout the entire pupal stage, it lies on its dorsal side within
its pupal-cell, and is largely supported by the long lateral setae,
which project from each of the thoracic and abdominal tergites
(Fig- 15)-

The straightened out pupa measures about twelve mm. and
in the arcuate position ten mm. The body is somewhat de-
pressed, with the appendages and wing-pads rather long, and
with the lateral tergites drawn out at considerable length. At
the end of each of these appendages is a mass of coarse bristles.
The color of the pupa is yellowish white. It is quite active and
can move about considerably within its pupal-chamber.

Throughout the entire pupal period the light-organs, that
were functional in the larva, can be distinctly seen to emit light,

66

WALTER N. HESS.

although they do not shine with a bright luminescence unless
the pupa moves, or is disturbed (Fig. 16).

The head, thorax, and even the abdomen of the pupa, as well
as the newly emerged adult, have been described as luminous
(Williams, 1916). It must be admitted that the entire insect
appears faintly luminous at these periods, but the cause for it

LO —-

LO

FIG. 16. FIG. 17. ,

FIG. 16. Photurus pennsylvanica pupa, ventral view of abdomen. LO, larval
light-organ.

FIG. 17. Same as Fig. 2, except taken one day after emergence when larval
light-organs (LO) were still visible. AO, adult light-organ.

seems to be due more to the light of the abdominal light-organs
shining through the non-pigmented coverings of the insect's
body, than to the luminosity of the fat or other internal struc-
tures. A freshly molted larva appears much the sam«. as the
two stages refened to above.

The date of pupation for this species, at Ithaca, New York,
ranges from May 20 to June 15, depending largely on the season.

BIOLOGY OF SOME COMMON LAMPYRID^E. 67

The extent of the pupal period was found to vary from sixteen
to eighteen days under outdoor conditions.

From the fifth of June to about the first of July the pupae may
be found transforming to adults within their pupal-cells. On the
first or second night, after transformation, the adult ruptures
the pupal-cell and comes forth as a mature insect.

The larval light-organs which are functional during the pupal
stage continue to glow until the end of the second day of adult
life, when they become fainter and fainter and cease to function.

One species that was reared did not show evidence of the adult
organ until after emergence, and then only in the sixth abdominal
segment, although the organ in segment seven began to glow a
few hours later.

Since the adult light-organs are functional at the time of
emergence and since the larval light-organs function for a period
of about two days of adult life, there is a brief period during which
both organs are luminous (Fig. 17).

Pyropyga fenesiralis Mels.

This beetle is elongate, oval and slender. It is entirely black
or blackish except for the large pinkish subtriangular space on
each side of the black central disc of the prothorax. The head
is completely covered by the prothorax. The eyes are small in
both sexes. The antennae are eleven-segmented and slender.
The elytra are costate. The sexes are similar except that the
female is considerably larger than the male, and her abdomen
which is usually distended extends beyond the elytra. The
length varies from 6.5 to 10 mm.

LeConte reports these insects from Pennsylvania, Lake
Superior Region, Colorado and California. Blatchley found
them plentiful in Indiana.

The adults of these fireflies differ from those of the species
previously referred to in this paper, in that they are active only
during the daytime and not at night. This change in habit can
probably be accounted for by the absence of light-organs in the
adults. With this one exception, their general habits are
similar to those of the other species studied. They are never
found in dry localities, being chiefly found in low meadows along
streams or marshy areas.

68 WALTER N. HESS.

The period of emergence is more prolonged than is charac-
teristic of most species, varying from June 15 to August 10.

During the day these insects can usually be found clinging to
grass and weeds by the side of streams. The males are active
fliers but the females were not taken in flight, yet it is possible
that they fly, for they are sometimes found at a considerable
distance from their natural habitat.

The adults were never observed feeding. If they feed at all
it is probably very sparingly, for specimens were kept in the
laboratory without food from the time of emergence through
the period of oviposition.

Copulation was observed to take place from two to ten days
after emergence, with frequent repetitions. In the field the
copulating pairs were always found about a foot from the ground
on a blade of grass or some similar object. The female, early
in the season, is usually distended with eggs in varying stages of
maturity.

Egg laying starts about four to five days after emergence,
and continues for a period of two to three weeks. The females of
this species also lay their eggs on the ground at the base of vege-
tation in damp ground. In every case observed, the eggs were
deposited singly.

The egg is small, spherical, whitish in color, measuring about
.65 mm. in diameter. It is without surface markings and when
first deposited it is covered with an adhesive substance. These
eggs, while they are in no sense luminous, give off a slight lumines-
cence when first laid.. Under normal conditions they hatch in
from nineteen to twenty-one days into little whitish, elongated
larvae.

The larvae, when first hatched, resemble very closely in appear-
ance those of Photinus scintillans, in that they are whitish with
dark eyes and brownish mandibles. At this stage they measure
about 2.3 mm. in length. On becoming pigmented, however, they
appeared dark gray in color.

The mature larva is elongate, narrow, varying from 12 to 14
mm. in length. The blackish head is small, being about half
as wide as the prothorax. The body is widest in the region of
the thorax and tapers posteriorly. The tergites are dull blackish

BIOLOGY OF SOME COMMON LAMPYRID.^. 69

in color, while the pleural regions are distinctly pinkish. On the
ventro-lateral sternite of the eighth abdominal segment are
two luminous areas, the larval light-organs.

Similar to the adults of this species the larvae are active during
the day, though as a rule they are more active at night, when
they can be found wandering about on the ground at the edge of
streams apparently in search of food. Frequently, however,
while searching for them during the day they were found con-
cealed under stones and in the ground to a depth of from one to
two inches. On several occasions numbers of them were found
assembled together, where they were feeding on a captured
snail or earthworm. They were found most abundant at Ithaca,
N. Y., on the gravel at the edge of Cascadilla Creek a short dis-
tance below the new fish hatchery.

Since these larvae were active during the day-time, and as
they readily took food, it seemed advisable to study more in
detail their food habits.

Newport (1857), Meinert (1886), Fabre (1913), Haddon
(1915) and Vogel (1915) each studied the food habits of Lampyris
nocliluca. Newport evidently did not observe closely the man-
dibles, for he did not mention their hollow nature, which was
observed by the other four workers. Newport maintained that
the bite of the larva definitely injured the snail upon which it
was feeding, while Fabre took a different view, maintaining that
the larva injected a substance into its host in the nature of an
anaesthetic which paralyzed it, thus making it possible for the
larva to feed without being disturbed by the efforts of the snail
in trying to escape.

So far as my observations go, I am inclined to believe that the
larva does inject a definite substance which serves to paralyze
and finally to kill its host. On one occasion a rather small
earthworm, about three inches long, was placed with six larvae.
In about two minutes one of the larvae, as it moved about,
feeling its way with its maxillary palpi and antennae, came in
contact with the worm. It touched it several times with these
structures, which undoubtedly were supplied with sense organs,
and then pierced the worm with its mandibles. The earthworm
quickly moved and was evidently slightly injured, but it soon

7O WALTER N. HESS.

became quiet. The larva soon released and bit the worm again
in the same region. This time the worm moved much less.
The process was repeated several times, and each time the worm
was less disturbed, until at the sixth or seventh bite the earth-
worm was not aware of the larva's presence. At about this
time another firefly larva bit the worm about an inch away from
where the first one was feeding. This part of the earthworm
was sensitive, and the worm, though much less active, responded
much the same as before. About two minutes later a third
larva bit the worm in a still different region, with the result that
the worm moved a little but much less than for the other two
larvae. Ten minutes from the time that the first larva bit the
worm it apparently was perfectly paralyzed, and so far as could
be determined it was dead. Several slugs and snails were
observed in the process of being killed by these larvae. As the
reaction of these snails is similar to that of the earthworm,
it is sufficient to state that at first the snail appeared to be
slightly injured, for in every instance it contracted. In each
case, a few minutes after the larva began feeding, the snail was
evidently paralyzed for it no longer moved. One of the largest
specimens of earthworms obtainable (about seven inches long)
was placed in a jar containing moist filter paper with twelve
larvae. Although the worm was left with the larvae for six days,
and though they were given no other food there was no evidence
of any effort being made to eat it.

While feeding on its prey, the larva keeps its jaws actively
moving back and forth, apparently ejecting digestive juices
from its mid-intestine by way of the mandibular canal, and with
the two inner teeth on the mandibles it tears the worm to pieces
and draws small portions into its mouth. At times the larva
would almost bury itself inside the body of its prey, apparently
preferring the softer internal tissues to those of the exterior.
From an examination of the mouth and intestinal contents of
larvae, which were killed while they were feeding, and from careful
observation of their feeding habits, it seems evident that a certain
amount of digestive juices are exuded through the mandibular
canals into the host, when it is first attacked, which serve to
paralyze it. It is also probable that a certain amount of the

BIOLOGY OF SOME COMMON LAMPYRID^. 71

digestive juice is constantly being exuded through the mouth
as well as by these canals, while the larva is feeding, which serves
to break down and partially digest the tissues of the host. Fur-
ther, it was evident that small portions of undigested flesh, as
well as food that was nearly digested, was taken into the mouth,
where it was bathed in the digestive juices and worked back
and forth by means of the two mandibular teeth, and that liquid
food together with very small portions of partially digested food
passed into the crop and intestinal region where digestion was
completed. As far as could be determined, the mandibular
canals did not serve for the intake of foods, as in the Chrysopidae,
but they appeared to function, as far as digestion is concerned,
simply as canals for exuding digestive fluids into the host.

At frequent intervals while feeding, certain of the larvae were
observed to extend the caudal filaments, twist the body around,
and apply them to the portion of food that was being eaten.
It appeared as if the larva was placing some glandular secretion
upon the food, or possibly helping to push it into its mouth.

Other experiments were performed to determine whether or
not these larvae limited their food to snails and earthworms.
On successive days cutworm larvae (Peridroma margaritosa
Haworth, Paragrotis tessellata Harris), potato-bettle larvae
(Leptinotarsa decemlineata Say), squash-bug nymphs (Anasa
tristis DeGeer) and wireworms (Agriotes mancus Say) were
placed in the pots with these firefly larvae. They readily ate all
of the cutworm larvae, but in no instance were any of the other
insects eaten.

From the above experiments it seems evident that these
firefly larvae feed chiefly on snails, cutworms and small earth-
worms. The other insects that were offered them as food were
chosen because they are sometimes found on, or near the ground,
and it was desired to see how wide a range of food they would eat.

Although the adults have no light-organs, the larvae have
well developed light-organs which resemble very closely, in
appearance, those of the other species studied. It, however,
appears to be somewhat smaller and emits a less distinct light.

Larvae were not reared from the egg to the adult, so the length
of their life-cycle cannot be given with certainty. Yet, since half-

72 WALTER N. HESS.

grown larvae were found during July and August in association
with the mature ones, it seems very probable that the larvae
live for two years before transforming to adults.

Pupation takes place from about June 10 to August i. The
larvae do not build an elaborate pupal-cell, as the larva of Pho-
turus pennsylvanica does, but on the other hand, they crawrl
back away from the edge of the water, three to ten feet, where
they go underneath stones, and there excavate little cells in
which the pupal period is spent. The mature larva shortens up
slightly and assumes an arcuate position. The cuticula splits
down the middle of the dorsal thoracic region and gradually
liberates the pupa.

The pupa, except for the yellowish white head, appendages
and tip of the abdomen, is of a delicate roseate color. The
pleural regions, however, are decidedly pinkish. The abdominal
and thoracic tergites are drawn out ventro-laterally into rather
acute projections. At the end of each is a mass of setae. The
straightened out pupa measures 8 mm., and in an arcuate position,
6.5 mm.

The pupal period is rather brief, extending for only seven or
eight days.

The larval light-organs function throughout the pupal period,
but as the time for emergence approaches, the glow usually
becomes very faint. In a few specimens the light was observed
in the adults for a brief period after emergence.

PURPOSE OF LUMINOSITY.

There can be little doubt but that the chief function of light-
emission in insects is to assist in securing the mating of the sexes.
It is evident that this has come about as a secondary character
when one considers the varying degrees to wrhich the light-organs
are developed among Lamyprids. It was suggested by Blair
(1915) that possibly the light may be an indication of impalata-
bility. In the case of the adults of Photinus scintillans, this
does not seem probable, as one frequently finds the females of
Photurus pennsylvanica and numerous species of spiders feeding
upon them. Many adult fireflies when captured emit a pale
yellow fluid from between the last coxal joints and from the

BIOLOGY OF SOME COMMON LAMPYRID/E. 73

pygidium, which it seems may be impalatable to many of their
enemies. In the case of luminous larvae there seems to be little
possible use for light-organs, unless it is to warn any possible
enemies not to eat them. There are many species of insects
outside of the family Lampyridae that resemble very closely in
appearance certain species of fireflies. If the explanation of
this mimicry is that of protection, it would seem that the theory
of impalatability must have some basis. The exact purpose
for the presence of light-organs in larvae is still indefinitely
understood.

ECONOMIC IMPORTANCE.

Practically no work has been done on the possible economic
value of these insects, and though the data here given are little
more than suggestive, every evidence seems to indicate that
these insects are of considerable economic importance. The
adults are of little value, as most of them probably eat compara-
tively little, but the larvae are voracious little creatures which
live on and in the ground, and feed on snails, earthworms,
cutworm larvae, and, in fact, on larvae of many injurious insects.
Most of the soft-bodied animals living on the ground are injurious
(unless it is the earthworms), and as the food of firefly larvae is
probably limited to these small animals, they necessarily do much
economic good in killing them. The slugs and cutworm larvae
are among our worst economic pests, and it seems evident that
they furnish a large part of the food of these larvae. Since most
of the fireflies live two years as larvae, the number of larvae that
are feeding on the ground during any season is approximately
twice that of the adult fireflies. Anyone who has been out during
a June or July evening knows that the fireflies are one of our
most abundant insects, which, together with the voracious
habits of these larvae leads to the belief that they are of much
more economic importance than has been attributed to them

heretofore.

SUMMARY.

1. The fireflies studied are luminous both in the larval and
adult states, except Pyropyga fenestralis, which is luminous only
during larval life.

2. In the luminous species the light-organ of the male is

74 WALTER N. HESS.

better developed and more brilliant than that of the female.
In Photurus pennsylvanica, however, the light-organ in the female
is nearly as large and as brilliant as that of the male, but in the
other luminous species the organ of the female is limited to a
small area on the sixth abdominal segment and it emits a rather
faint light.

3. The light-organs of the adults undoubtedly function chiefly
in bringing the sexes together. The nature of the flash differs
for each species, so that members of the same species can readily
recognize the flash of the opposite sex. The males and females
of certain species, at times, respond to the light of a small flash-
light, indicating that this is not always true.

4. In all these luminous species, except Photurus pennsyl-
vanica, the female orientates her abdomen so that the light
emission is in the direction of the male.

5. In all the species studied the larval light-organ degenerates
and, in the luminous species, separate light-organs are developed,
which function during adult life.

6. In the case of Photinus consanguineus, synchronous flashing
was observed in which a few leaders, by flashing, acted as a
stimulus to the discharge of the flashes by the others, thus bring-
ing about the flashing in unison of the whole group.

7. The adult fireflies were not observed feeding, except the
female of Photurus pennsylvanica. She was found to be very
voracious, feeding chiefly on the adults of smaller fireflies, but
at times she was found to be cannibalistic.

8. The larvae of all species were found to be predaceous,
feeding on snails, earthworms and numerous species of insect
larvae, especially cutworms. They are at times quite voracious,
and it seems evident that they are of considerable economic
importance.

9. By means of their hollow mandibles, the larvae eject a
portion of the mid-intestinal juices into its host, thereby paralyz-
ing it, and later digesting it so that when the food is taken into
the body it is in a digested or nearly digested condition.

10. The larval life, at least, in most of our native fireflies,
extends over a period of two years, during which time the larvae
live on or near the surface of the ground.

BIOLOGY OF SOME COMMON LAMPYRID/E. 75

1 1 . All species that were studied hibernate as larvae underneath
stones, a short distance under the ground or near the surface,
often in specially constructed chambers.

12. Pupation takes place in moist earth under stones, or in
specially constructed pupal-cells at the surface of the ground.
The pupal cell of Photurus pennsylvanica is made of short,
ribbon-like pieces of earth, which the larva masticates and con-
structs into a lattice-like dome.

13. The pupae retain the larval light-organs which function
throughout the pupal period. This period was found to be
from seven to eighteen days according to the species.

14. The adults live for a period of two to four weeks, during
which time they deposit their eggs on the ground about the roots
of grass and moss.

BIBLIOGRAPHY.
Annandale, N.

'oo Insect Luminosity. An Aquatic Lampyrid Larva. Proc. Zool. Soc.

London, pp. 862-865.
Barber, H. S.

'05 Note on Phengodes in Vicinity of Washingtcn, D. C. Proc. Ent. Soc.

Wash., VII., pp. 196-197.
'14 On Interspecific Mating in Phengodes and Inbreeding in Eros. Proc.

Ent. Soc. Wash., XVI., pp. 32-34.
Bethune, C. J. S.

'68 Luminous Larvae. Can. Ent., I., pp. 38-39.
Blair, K. G.

'15 Luminous Insects. Nature, XCVL, pp. 411-415.
Bongardt, J.

'04 Zur Biologic unserer Leuchtkafer. Natw. Wochenschr., Jena XIX., pp.

305-310.
Dahlgren, U.

'17 The Production of Light by Animals — The Fireflies or Lampyrids. Jour.

of Franklin Inst., CLXXXIIL, pp. 323-348.
Dubois, R.

'86 Contribution a 1'etude de la production de la lumiere par les etres vivants.
Les Eletreides lumineux. Bull. Soc. Zool. France, IX., pp. 1-275, P^s-
I-IX.
Fabre, Henri.

'13 The Glow-worm. The First User of Anaesthetics. The Century Magazine,

LXXXVIL, pp. 105-112. (Translated by A. T. de Mattos.)
Haddon, K.

'iS On the Methods of Feeding and the Mouthparts of the Larva of the Glow-
worm (Lampyris noctiiuca). Proc. Zool. Soc. London, pp. 77-82, pi. i.
Hudson, G. V.

'91 The Habits and Life-history of the New Zealand Glow-worm. Trans, and
Proc. New Zeal. Inst., XXIII. , pp. 43-47. pi- VIII.

76 WALTER N. HESS.

Knab, F.

'05 Observations on Lampyridae. Can. Ent., XXXVII., pp. 238-239.
LeConte, J. L.

'81 Synopsis of the Lampyridae of the United States. Trans. Am. Ent. Soc.,

IX., pp. 15-72.
Mast, S. O.
'12 Behavior of the Fireflies (Photinus pyr'alis) ? with Special Reference to the

Problem of Orientation. Jour. Anim. Behav., II., pp. 256-272.
McDermott, F. A.

'10 A Note on the Light-emission of some American Lampyridae. Can. Ent.,

XLIL, pp. 237-363.

'n Some Further Observations on the Light-emission of American Lampyridae:
the Photogenic Function as a Mating Adaptation in the Photinini. Can.
Ent., XLIIL, pp. 399-406.
'12 Observations on the Light-emission of American Lampyridae. Can. Ent.,

XLIV., pp. 309-312.
'14 Ecologic Relations of Photogenic Function Among Insects. Zeit. f. wiss.

Insecten Biol., X., pp. 303-307.

'16 Flashing of Fireflies. Science, XLIY., p. 610.
Morse, E. S.

'16 Firefly Flashing in LTnison. Science, K.S., XLIIL, pp. 169-170; XLYI ,

PP- 387-388.
Needham, J. G.

'03 Button-bush Insects. Psyche, X., pp. 22-30.
Newport, G.

'57 On the Natural History of the Glow-worm (Lampyris noctilnca). Jour.

Proc. Linn. Soc. Zool. London, I., pp. 40-71.
Olivier, E.

'n Contribution a 1'histoire des Lampyridje. Mem. Congr. Internat. Brussels

(l9IO), II., pp. 273-282, 2 CUtS.

Vogel, R.
'12 Beitrage zur Anatomic und Biologie der Larva von Lampyris noctilnca.

Zool. Anz., XXXIX., pp. 515-519.

'15 Beitrage zur Kenntniss des Baues und der Lebensweise der Larve von
Lampyris noctiluca. Zeit. f. wiss. Zool., CVIL, pp. 291-432. 35 text
figures, pi. 4.
Wenzel, H. W.

'96 Notes on Lampyridae, with the Description of a Female and Larva. Ent.

News, VII., pp. 294-296, pi. XI.
Williams, F. X.

'17 Notes on the Life-history of some North American Lampyridae. Jour.
N. Y. Ent. Soc , XXV., pp. 11-33.

AN EXCEPTION TO BATESON'S RULE OF SECOND-
ARY SYMMETRY.

A. B. DAWSON,
DEPARTMENT OF ANATOMY, LOYOLA UNIVERSITY SCHOOL OF MEDICINE.

In the summer of 1917, while employed by the Biological
Board of Canada in a Dominion Lobster Hatchery at Bay View,
N. S., my attention was called to an abnormal lobster caught by a
local fisherman in the adjoining waters of Pictou Harbor. The
lobster was a male and measured 7>^ inches from rostrum to
telson. The abnormality consisted of a double extra claw on
the right cheliped, resulting in a condition of incomplete tripli-
cation. The presence of the extra parts on the right side did not
appear to greatly handicap the animal which, when placed in
one of the hatchery tanks, moved about freely. Other abnor-
malities somewhat similar have been described for the lobster
by Faxon ('81), Emmel ('07) and Cole ('10).

DESCRIPTION.

The "triple" chela of the right side consisted of a small "nip-
per" and a double extra "crusher," while the corresponding
appendage of the left side was large, normally developed and of
the nipping type. The small "nipper" of the monster claw
apparently represented the primary member of the group and,
with the exception of the meropodite (M.), was normal in all its
segments. Arising from the posterior (morphologically ventral)
surface of this meropodite was the double extra claw.

The first segment of the right chela, the ischiopodite (Is.}
was not exceptionally large and presented no evidence of dis-
tortion; but in the meropodite, as already noted, the effect of the
abnormality in the appendage was very evident. This segment
was of normal width proximally but broadened out rapidly
distally and terminated in two diverging branches (M., M.' + ")•
On M was borne the small primary claw while M.' + " carried
the much larger abnormal structure.

77

7o A. B. DAWSON.

On examining the more posterior prong of the compound
meropodite, we find that the next distal segment, i.e., the extra
carpopodite (C. ' + ") is morphologically double. It is more
than twice as large as the corresponding segment (C.) of the
primary claw and bears on it supper surface two groups of spines
separated by a shallow longitudinal groove. On the end of the
massive carpopodite is the large double protopodite (P. ' + ").
This segment is incompletely divided, with the separation
extending only as far as the region on a level with the bases of

D. I.

D"

FIG. i. (One half natural size.) Male lobster possessing an abnormal right
chela consisting of a primary nipping claw plus a double extra "crusher." C.,
primary carpopodite; C.' * " , double extra carpopodite; D., primary dactyl;
D.', D.", two extra dactyls; /., primary index; /.', /.", two extra indices; Is.,
ischiopodite; M., primary portion of meropodite; M .' + ", extra portion of
meropodite; O., conical protuberance on double protopodite; P., primary proto-
podite; P' + " (P omitted in reproduction), double extra protopodite; S., scar on
left chela.

the indices (/. ', /.")• Opposing the indices are two well de-
veloped movable dactyls (D/, D."). The two partially sepa-
rated chelae are almost exact mirror images of each other. The
dactyls and indices are practically identical and the dentition of
both consists of heavy tubercle-like teeth. There are no tactile
hairs.

Each segment of the double extra claw was decidedly larger
than the corresponding portion of the primary claw, but the
entire "triple" appendage barely equalled the left nipping claw
in weight. The long axes of the double appendage and primary

EXCEPTION TO BATESON's RULE OF SECONDARY SYMMETRY. 79

appendage were in one plane and all three dactyls moved to meet
their opposing indices in this plane.

Another point of considerable interest is the color relation of
the extra portions. In the double extra protopodite and the
two extra dactyls the pigmentation was completely reversed,
giving as a result a light colored upper and a densely pigmented
lower surface. A peculiar conical protuberance present on the
light upper surface of the double protopodite was practically
free from pigment.

THEORETICAL CONSIDERATIONS.

The present example, like most abnormal crustacean appen-
dages, falls into the category established by Bateson ('94),
"in which the extra limb or extra parts of a limb are themselves
morphologically double," but unlike the others it does not
conform to the rules of secondary symmetry formulated by the
same author. According to these rules the normal appendage
and the extra parts lie in the same plane and "the nearer of the
two extra appendages is in structure and position formed as the
image of the normal appendage in a plane mirror placed between
the normal appendage and the nearer one, at right angles to the
plane of the three axes; and the remoter appendage is the image
of the nearer in a plane mirror similarly placed between the two
extra appendages" (p. 479).

In the case under discussion the pair of extra chelae are mirror
images of each other but the appendage nearer the primary claw
is not, as the rules provide, a mirror image of the latter. More-
over, it does not appear possible to explain this exception to the
rules of secondary symmetry by a reference to torsion. Emmel
('07) and Cole ('10) were unable to explain apparent exceptions
to Bateson's rules by making allowances for changes in position
due to possible torsion. It has previously been pointed out that
the coloration of the abnormal double structure was completely
reversed, but even if the extra segments were rotated so that
the dark surface became uppermost the relations of the claws
as regards secondary symmetry would remain unchanged since
the double extra claw is bilaterally symmetrical.

The specimen here described while exhibiting many points of

8O A. B. DAWSON.

resemblance to the abnormal chelipeds figured by Emmel
('07, PI. 2, Fig. 5) and Cole ('10, Figs, i, 2) also presents several
novel features. The degree of "triplication" of the claw is less
than that seen in Emmel's specimen but greater than that de-
scribed by Cole. In the former, the abnormal processes, con-
sisting of a double carpopodite, two protopodites and two dactyls
arose from the meropodite. In the latter, the abnormal struc-
ture, two extra indices and a double extra dactyl, was borne on
the normal protopodite. In Faxon's specimen ('81) the mor-
phological character of the extra branch which is borne on the
forked meropodite is questionable. Faxon himself does not
believe that the structure is double but Bateson ( '94) is inclined
to regard it as being morphologically double. Furthermore
both Emmel and Cole found that the conditions in their "triple"
claws illustrate the rules of secondary symmetry almost diagram-
matically, i.e., when allowance was made for shifting due to
torsion.

In the two chelae described by Emmel and Cole the two extra
claws were of the same character as the primary claw, "crushers."
In my specimen the primary claw is a "nipper" while the two
extra claws are of the crushing type with well-developed tubercle-
like teeth. This condition is of special significance, when we
recall that the claw of the opposite side is also of the nipping
type. In other words, we have a lobster with the "great" claws
symmetrical with reference to each other, besides bearing on the
meropodite of the right a double extra "crusher."

Emmel ('07) finds the pigmentation reversed on one of the
extra claws but is able to explain this abnormal condition by a
reference to torsion. The case which I am describing does not
admit of such an explanation. The cause of the abnormality
is unknown. No scars were to be found on the "triple" claw,
but on the protopodite of the appendage of the opposite side a
definite scar (S) was present. There were no further evidences
of mutilation.

The results obtained by Harrison ('17) with transplantation
of limb buds in Amblystoma larvae suggest that there may possibly
be some direct relation between the reversal of pigmentation and
the doubling of the extra appendage. Harrison found that in

EXCEPTION TO BATESON's RULE OF SECONDARY SYMMETRY. 8 1

transplantations in which the limb buds were inverted, a certain
percentage gave rise to double or twin limbs, one being a
mirror image of the other. In a few cases there were still further
duplications so that more or less complete triple limbs resulted,
having approximately the same relations as found by Bateson in
the Arthropoda.

In the lobster just described, two facts were noted; first, the
extra parts are double and second, their pigmentation is reversed
indicating an inversion of the double portion of the triple appen-
dage. Furthermore, it is generally conceded that most abnormal
and duplicated appendages among Crustacea are the result of
regenerative processes. Both Reed ( '04) and Emmel ( '07) have
found that abnormalities can be produced experimentally by
mutilating either the proximal stump or the developing limb bud.
It is conceivable then that the triplication found in the present
instance may be due to regeneration following injury. In the
course of regeneration the growing but may have been also
injured so as to cause the development of an extra process
(Emmel, '07, pp. 114-115), and this process may have had its
dorso-ventral orientation reversed either at the time its develop-
ment was initiated or at some later date.

Keeping the above statements in mind it therefore becomes
possible to elaborate a more or less satisfactory hypothetical
explanation of the triple chela. First, there may have been an
injury to a normally developed crushing claw, followed by autot-
omy. Later the developing bud may have been injured, resul-
ting in the appearance of an extra bud on the surface of the
primary one. As a consequence of the injury, the primary bud
did not develop as a crushing claw but as a small "nipper."
(Sufficient evidence has been adduced by Emmel in 1907, Figs.
24 to 31, to prove that abnormal symmetrical claws in lobsters
do arise through regeneration following mutilation.) The
extra bud, due to mechanical displacement of its tissue at the time
of the first injury, or as a result of a subsequent accident, has
had its dorso-ventral orientation reversed, causing the develop-
ment of twin crushing claws which are mirror images of each
other.

82 A. B. DAWSON.

SUMMARY.

1. An abnormal lobster cheliped is described in this paper.
The abnormality consists of a double extra crushing claw arising
from the meropodite of a claw of the nipping type.

2. The conditions found here are shown to present an exception
to Bateson's rules of secondary symmetry.

3. The primary member of the so-called triple claw was found
to be of the same character as the corresponding appendage of
the opposite side, furnishing an exception to the normal condition
of asymmetry.

4. Pigmentation was reversed in the double extra claw.

5. There is some evidence in favor of the view that there is
a definite relation between the reversal of pigmentation and the
doubling of the extra appendage.

LITERATURE.
Bateson, W.

'94 Materials for the Study of Variation, Treated with Especial Regard to

Discontinuity in the Origin of Species. London.
Cole, L. J.

'10 Description of an Abnormal Lobster Cheliped. BIOLOGICAL BULLETIN,

Vol. 18, pp. 252-268. 9 figs.
Emmel, V. E.

'07 Regenerated and Abnormal Appendages in the Lobster. _3?th Annual

Report Com. of Inland Fisheries, R. I., pp. 99-152, PI. 1-9.
Faxon, W.

'81 On Some Crustacean Deformities. Bull, of Mus. Comp. Zool., Vol. 8,

No. 13, PP- 257-275, Pis. 1-2.
Harrison, R. G.

'17 Transplantation of Limbs. Proc. Nat. Acad. Sci., pp. 245-251.
Reed, Margaret A.

'04 The Regeneration of the First Leg of the Crayfish. Arch. Entw.-Mech.
Bd. 18, p. 307.

A COMPARATIVE STUDY OF THE CHROMOSOMES
OF LACHNOSTERNA (COLEOPTERA).

E. L. SHAFFER,
DEPARTMENT OF BIOLOGY, PRINCETON UNIVERSITY.

A. INTRODUCTION.

The studies presented here were begun in the spring of 1916 at
Princeton University and were continued until the spring of 1917.
Most of the material used was collected at Cold Spring Harbor
during the summer of 1916. l The studies were discontinued in
the summer of 1917 owing to the war. On returning to Princeton
last spring, it was thought advisable to assemble the observations
previously made despite the fact that they did not represent as
complete a study as had been intended.

I wish to express my sincere thanks to Professor E. G. Conklin
for much valuable assistance and encouragement.

B. MATERIALS AND PLAN OF STUDY.

It was originally intended to make a detailed study of the
process of synapsis as well as a comparative study of the chro-
mosomes of four selected species of May beetles, genus Lach-
nosterna. While the material was not entirely favorable for these
purposes, some interesting facts were brought to light.

The four scarab beetles of the genus Lachnosterna which were
selected for study were the species delata, fusca, gracilis and
tristis. Besides these, for comparative purposes, two other
scarab beetles were .studied, Pelidonota pimctata and Cotalpa
lanigera. The form most studied was L. delata and since the
other forms showed no essential differences from delata, the latter
will be used as the basis of description in the present paper.

Comparatively little detailed study of spermatogenesis in the
Coleoptera has been done. The work of Miss Stevens ('05, '06),

1 The writer wishes to express his thanks to the Brooklyn Institute of Arts anP
Sciences for the privileges of a research fellowship at the Laboratory of the Institute
at Cold Spring Harbor, L. L, during the summer of 1916.

83

84 E. L. SHAFFER.

while it covered a large number of species, was concerned only
with chromosome counts, and especially with reference to the
sex chromosomes. In only two species of beetles has there been
anything like a detailed study of the chromosomes in synapsis,
Voinov (1903) on Cybister roselii and Schafer (1907) on Dytiscus
marginalis. Both these authors after a detailed study of the
growth stages of the spermatocytes describe parasynapsis,
while Miss Stevens claims telosynapsis in the forms she studied.

FIG. i. Single testis of Lachnosterna (side-View), with its efferent duct.

The material used in this study was fixed in Flemming's, Her-
mann's, Bouin's, Gilson's and Carnoy's fixing fluids. In general,
the Flemming and the Hermann material was best for the growth
stages of the spermatocytes, while the Bouin material was best
for the chromosomes. Iron-haematoxylin, with and without a
counterstain, was employed entirely for staining. Aceto-carmine
smears were valuable in checking the observations on the fixed
material. All the testes, except those of L. ftisca, were taken
v from the adult beetles. The material was gathered in midsummer
and showed all stages from spermatogonia to ripe spermatozoa.
In the case of L. fusca, the adult testes showed few favorable
stages and it was necessary to study the larval gonads.

C. DESCRIPTION OF TESTES AND SERIATION OF STAGES.

The testes consist of twelve mushroom-like bodies, three pair
in each side of the abdomen. Each testis has its duct (Figs.
i and 2) and the ducts from each group of testes unite to form
two larger ducts; these four larger ducts in turn unite to form
the single median vas deferens.

STUDY OF CHROMOSOMES OF LACHNOSTERNA. 85

The testes, although of an unusual shape, show the seriation
of the stages clearly, being but a modification of the simple
straight (orthopteran) type with a linear seriation of the cells.
In the species studied, the testes consist of a great many follicles
radiating from the center. Fig. 2 represents a diagrammatic
section through the center of the testis and perpendicular to its
broad surface. In the center of the testis from which the follicles
radiate (Fig. 2, A), one finds all the spermatogonia and here new
cysts are in the process of formation. On each side of this

FIG. 2. Diagrammatic section through testis at right angles to its broad surface,
to show seriation of stages. .4, region in which are all spermatogonia; B, early
growth stages and synezesis; C, pacyhtene and later growth stages; D, maturation
divisions; £, spermatids and spermatozoa; F, cavity in testis where spermatozoa
are retained prior to discharge from testis; G, efferent duct.

region (B, B} are the early synaptic stages (synezesis). In the
regions C-C one finds the later synaptic stages (pachytene and
diplotenes), while in regions D-D show most of the spermatocyte
divisions; regions E-E contain most of the spermatids and
spermatozoa. The chamber (F) shown in the figure serves as a
place where the spermatozoa are collected and stored prior to
discharge from the testis; the duct (G) leads from the storage
chamber. Of course the stages above seriated overlap and
there is no sharp delimitation as is diagrammatically shown in
the figure. The formation of the cysts in the region A was
followed out and my observations confirm those of Wieman ('10)
and Hegner ('14) that each testicular cyst is derived from a single
spermatogonium. There is, however, no evidence that cell
division is by amitosis as Wieman found.

86 E. L. SHAFFER.

D. OBSERVATIONS.
i. Spermatogonia and Diploid Chromosome Groups.

In all of the species of Lachnosterna studied, as well as Peli-
donota and Cotalpa, the diploid number of chromosomes as shown
in the spermatogonia is twenty, including an unequal (sex) pair
(Figs. 1-6). Dividing follicle cells in the ovaries show ten equal
pairs of chromosomes (Fig. 7). There are three pairs of J- or
U-shaped chromosomes, one pair of which is considerably larger
than the others (Figs. 1-5, A A). The sex chromosomes are the
smallest in the complex, consisting of a very small round chro-
mosome (;y) and a somewhat larger rod-shaped chromosome (x).
In comparing the size relations of the chromosomes in the
several species studied, one finds no marked differences. In
many cases the chromosomes in the diploid complexes are
arranged in pairs, homologous chromosomes lying beside each
other. In the Diptera, Metz (1916) has found that pairing of
chromosomes is not confined to the maturation stages, but at
each cell division homologous chromosomes come together. In
the Diptera the diploid chromosome number is relatively low; in
species where the. chromosome number is high, pairing of homo-
logous chromosomes is usually not found to be so complete. It
therefore seems that chromosome pairing, outside of the matur-
ation stages, is related to chromosome number.

In the spermatogonial telophases, the chromosomes spin out
into fine chromatic threads (Fig. 13) and as the nucleus grows
the threads become more and more complex forming a chromatic
reticulum or typical resting nucleus. This "resting" nucleus
is of relatively short duration, for soon the chromatin begin to
condense into heavier threads (Fig. 15), and as condensation
continues, all the chromatin of the nucleus becomes confined into
large chromatic blocks of a granular nature (Figs. 16, 17).
Counts of these chromatic blocks in uncut nuclei always approxi-
mate the diploid chromosome number and these blocks may be
considered as the anlages of the future spermatogonial chromo-
somes. The blocks consist of a linin-core on which are imbedded
the chromatin granules; they are connected to each other by a
fine net-work of linin which seems to be continuous with the linin
forming the core of the blocks. Most of the cells in the spermato-

STUDY OF CHROMOSOMES OF LACHNOSTERNA. 87

gonial area of the testes are in this stage and apparently it is of
of much longer duration than the reticular or "resting" stage.

In some respects the chromatic blocks above described cor-
respond to the " prochromosomes " which -have been described
by Overton ('09) in Podophyllum, Arnold ('08) in Hydrophilus
piceus, Goodrich ('16) in Ascaris incurva and other workers.
In these cases, however, the chromatic bodies appeared at the
beginning of the growth period and, according to the above
workers, these bodies arranged themselves in pairs, thereby
accomplishing the synaptic process. In Lachno sterna there is no
such paired arrangement of the chromatic blocks; they merely
represent stages in the formation of the spermatogonial chro-
mosome groups and might really be called prophases, except
that they are of relatively long duration. In some cases a
precocious longitudinal split can be detected, preparing the
chromosome for the next cell division.

2. The Synaptic Stages and Maturation Divisions.

Following the telophase of the last spermatogonial division
(Fig. 13), the chromosomes spin out in the form of very fine
(leptotene] threads (Fig. 18, 19) which entirely fill the nucleus and
prevent a minute analysis of this stage. The actual pairing of
the homologous chromosomes could not be followed in detail,
but observations on a few favorable cells (Fig. 19) indicate that
the union is side-to-side (parasynapsis) . Stevens ('06) has
described telosynapsis in the Coleoptera, but she did not make a
study of the early growth stages.

The leptotene stage gradually merges into a definite contrac-
tion stage (synezesis) with all the chromatic threads polarized
at one side of the nucleus (Figs. 20, 44). These stages are
always found in a definite part of the testes, namely in region B
(Fig. 2), and are found nowhere else. McClung ('05) used the
word "synezesis" to describe that "condition of the nucleus in
which the chromatin is found massed at one side of the vesicle,
without regard to whether it is a normal phenomenon or not."
McClung and recently some of his students, Whiting ('17) and
Hance ('17), have maintained that a unilateral massing of the
chromatin or synezesis is an artifact and is due to improper

88 E. L. SHAFFER.

fixation methods. With this in mind, the writer has sought with
most careful technical methods to obtain fixed material which
might not show these contraction figures; but without exception
cells in the contraction phase were always found in the definite
region of the testes mentioned above. It is quite true that
poorly fixed material shows an abundance of contraction figures,
but in these cases, as will be shown later, they are just as likely
to appear in other regions of the testes than in the very definite
location above mentioned. There is no doubt that even the
best fixation will tend to emphasize the contraction of the
chromatin just as it does in the case of the other cell structures,
but synezesis is unquestionably a normal process in the beetles
studied here. Whiting ('17) has advanced the idea that the
chromatic elements during synapsis are in an unstable condition
and that "any shock is likely to cause them to clump together."
It is questionable whether good fixation is much less of a " shock"
than indifferent fixation. It is conceivable that true synezesis
may not occur as a normal phase of the maturation possesses in
some animals (e.g., Orthoptera), but the fact that it has been
described by many workers using a variety of fixing methods
supports the fact that it is a normally occurring phase in some
cases. In any event it proves that the nuclear condition is
peculiar in cases of synezesis.

Following the stage of synezesis, the chromatin threads are
released from the polarized bouquet in the form of thick ragged
looking pachytene threads (Fig. 21). Usually a longitudinal
split can be seen in the threads, which marks the point of synapsis
of the homologous threads. The chromomeres are imbedded in a
linin base, chromomeres of the same size lying opposite each
other and being connected with each other by fine linin threads.
In the later stages the threads become more widely separated
from each other (Fig. 23) assuming the diplotene form. In these
stages and in still later ones, the threads show a variety of twisting
about each other forming rings with and without crossed ends,
figures 8, double and even triple crossing-over of the threads.
In no case could a secondary split be seen. The strepsistene
threads continue to become more widely separated and it soon
becomes impossible to trace the individual threads (Fig. 22).

STUDY OF CHROMOSOMES OF LACHNOSTERNA. 89

This unanalyzable stage is of relatively short duration and is
followed by a condensation of the chromatin in the form of heavy
threads (Fig. 25). Condensation of the chromatin continues
and the definitive maturation tetrads begin to make their appear-
ance. In these early prophases one often finds cells in which
all the chromatin is massed at one side of the nucleus (Fig. 26),
resembling very much a synezesis figure. Gross ('07) has
described a second synezesis in Pyrrhocoris and Mottier ( '07) be-
lieves that in the plants it is a regularly occurring phase in the
maturation processes. In Lachnosterna these contraction figures
are most abundant in material which is poorly fixed and I
consider them as artifacts. Fig. 27 represents this stage from
well fixed material as contrasted with Fig. 26 from poorly fixed
material.

All the first spermatocyte metaphase plates of the four species
of Lachnosterna studied, as well as Pelidonota and Cotalpa, show
ten bivalent chromosomes the smallest of which represents the
sex pair (Figs. 8, 9, 10, n, 12). These are usually arranged in
characteristic groups with fine linin threads connecting the various
members of the complex to each other. A comparison of the
tetrads of L. delata with those of L. fusca (Figs. 28, 29) shows no
marked differences either in form or in size of the tetrads. Using
Miss Carothers' ('17) nomenclature, there are five atelomitic
tetrads (non-terminal spindle fiber attachments) and five telom-
itic tetrads (terminal spindle fiber attachments). The atelom-
itic tetrads are the largest in the complex and are derived from
the three pairs of J-shaped spermatogonial chromosomes and
two pairs of the bent rod-shaped ones. In sideview metaphases,
the largest of the tetrads (Figs. 28, 29) has a sub-terminal
spindle fiber attachment, and is derived from the AA pair
(Fig. 5) of the diploid chromosome group which also have sub-
terminal fiber attachments. The other atelomitic tetrads
consist of two typical crosses and two annular tetrads 'of the
Stenobothrus type. The other four autosome tetrads are of the
ordinary dumb-bell type while the x and y elements (sex pair)
are fused end to end (Figs. 28, 29) ,

The types of tetrads above described are found in all four
species of Lachnosterna studied. On the other hand, in Cotalpa

9O E. L. SHAFFER.

and Pelidonota no cross-shaped tetrads and only one ring tetrad
are found. The question of reduction division is difficult to
analyze here, with the exception of the ring tetrads. The latter
are always arranged on the spindle in the direction of the spindle
axis and the spindle fiber attachment is median. Consequently
the separation of the dyads occurrs at the point of the synaptic
union and the division is reductional.

3. Sex Chromosomes.

The earliest work on the sex chromosomes was done by Miss
Stevens ('05, '06) on the Coleoptera. She found the so-called
sex chromosomes in over forty species and her work and that
of Wilson's on the Hemiptera and McClung's on the Orthoptera
have been the basis of the later work correlating sex determina-
tion with the chromosomes. In the Coleoptera the sex chro-
mosomes are found as unpaired "accessory" and as unequal
elements which separate in one of the maturation divisions and
divide equationally in the other maturation division. Arnold
('08) has maintained that in Hydrophilus piceus there are no
sex chromosomes. There is present in the growth stages a
chromatin nucleolus which may even persist up to the first
maturation division and may even pass undivided to one cell.
However, it disappears and cannot be found in any of the second
spermatocytes.

In Ladmosterna, Pelidonota and Cotalpa the sex chromosomes
are of the xy type the y element being the smaller of the unequal
pair (Figs. 1-6). There are no marked differences in the size
and form of the sex chromosomes in the four species of Lach-
nosterna studied, but in Pelidonota the x element is considerably
larger than in the Lachnosterna material. In all cases the sex
pair separate in the first maturation division and divide equa-
tionally in the second, thus yielding two types of spermatozoa.
(Figs. 31, 32, 33, 35, 36). In a single case the sex chromosomes
failed to separate in the first maturation division, both chromo-
somes going into one of the daughter cells. This is undoubtedly
a case of non-disjunction similar to that which has been found
genetically and cytologically by Bridges ('16) in Drosophila.

The sex chromosomes presist throughout the entire growth

STUDY OF CHROMOSOMES OF LACHNOSTERNA. QI

period as definite compact chromatic bodies. They are always
contained within a chromosomal vesicle such as has been de-
scribed by Wilson ('12) in Oncopeltus and Lygeeus (Figs. 21, 25).
In Lachnosterna the sex elements usually remain separate from
each other, each enclosed in a separate vesicle. In Pelidonota
and Cotalpa, the sex pair remain fused during the synaptic
period, the smaller (y) element usually being imbedded along the
side of the larger (x) element.

E. GENERAL CONSIDERATIONS.
i. Chromosome Number and Species.

The intensive work of McClung and his students on one family
of Orthoptera, the Acrididse, has shown that the chromosome
number in all the species studied of this group is the same,
namely 23 in the male. This has led McClung to the generali-
zation that species closely related taxonomically might show
similarity in their chromosome groups. It is very evident that
this generalization cannot apply to all groups since, in some
cases there is a wide divergence in chromosome number among
members of the same genus. It is possible that in some cases
this difference in chromosome number between closely related
species may be due to a fusion of several chromosomes or else a
breaking up of one or more chromosomes into several distinct
components. In the case of Hesperotettix, McClung ('17) has
shown that the chromosome number may vary from 17 to 23.
He has shown that these variations are due to a fusion of chromo-
somes resulting in the formation of "multiple chromosomes."
In one species, Hesperotettix viridis, he has found the haploid or
reduced number to vary from 9 to 13. On the other hand, the
work of Stevens on the Diabrolicas (Coleoptera) has shown that
the species vitatta has 21 chromosomes, while the species soror
and 12-punctata have but 19. However, in the latter two species
there may be present from i to 4 additional or "supernumary "
small chromosomes. It is quite possible that the supernumary
chromosomes of the species soror and 12-punctata represent the
fragments of a pair of chromosomes, which would therefore make
an agreement in chromosome number between these two species
and the species vitatta. As McClung ('17, p. 545) has pointed

92 E. L. SHAFFER.

out, he has confined his idea of this similarity of chromosome
number in closely related species, only to the family Acrididse.
It is possible that in other forms correspondence in chromosomes
may extend only to the subfamily or genus. In the Hemiptera
and Coleoptera certainly there is no such uniformity of chrom-
osome number in the various families as is found in the Acrididse.
The four species of Lachnosterna studied here differ from each
other very much as far as taxonomic characters are concerned,
nevertheless the chromosome groups show no difference either
in form or number. The two other forms studied, Pelidonota
and Cotalpa, differing generically, have the same chromosome
number (20 in the diploid groups), but there are some differences
in the form of the maturation tetrads. Only one other scarab
bettle has been studied, Euphoria inda by Stevens ('06), and it
corresponds with a diploid group of 20 chromosomes, so that all
the species of the family Scarabidce thus far studied correspond
in chromosome number. The genus Lachnosterna embraces
over one hundred species, some very much alike so that it is
difficult to separate them taxonomically, others differing markedly
from each other. The most constant difference is found in the
male copulatory organs, which probably prevents the inter-
breeding of species in nature. Perhaps further cytological
studies in this genus will yield results similar to those in the
Acrididse. Certainly there is a wealth of material for such a
comparative study.

2. Cyst Formation and Cell Polarity.

Hegner ('14) has studied the formation of spermatogonial
cysts in the testes of Leptinotarsa; the facts concerning cyst
formation in the beetles studied here show results essentially
similar to those of Hegner. The primary spermatogonia are
not arranged in cysts and are more or less polygonal in shape,
with the nucleus usually located in the center. Cyst formation
begins by the rapid division of a single primary spermatogonium,
together with an adjacent epithelial cell which forms a follicular
membrane around the cyst. Consequently we can say that all
the cells within any one cyst are the descendants of a single
primary spermatogonium. With the formation of the cyst,

STUDY OF CHROMOSOMES OF LACHNOSTERNA. 93

the spermatogonia are arranged in the form of a rosette, and are
now triangular or wedge-shaped with the nucleus at the base and
the rest of the cytoplasm extending toward the cyst cavity.
Thus, with the formation of the cyst, there is a polarity estab-
lished in the spermatogonia which is maintained up to the for-
mation of the ripe spermatozoa, for, the side where the nucleus
is located is destined to form the head of the spermatozoon, and
the cytoplasmic portion extending toward the cyst cavity is
destined to form its tail. Hegner. ('14) has homologized the
process of cyst formation with the differential divisions in insect
oogenesis which establish nurse cells and oocytes. It has long-
been known that the insect egg possesses a remarkable polarity
besides being highly organized. Since, as it has been above
shown, the polarity of the sperm cells are established at the time
of cyst formation, and since this process is homologous to nurse
cell-oocyte differentiation, it is probable that the polarity of the
egg may have its origin at the time of the differentiation of
nurse cells and oocytes.

3. Linin and Chromosome Structure.

When one studies the history of the chromatin of the nucleus
from the resting stage through the synaptic period up to the
reconstitution of the definitive maturation chromosomes, one
begins to seek for some of the underlying mechanisms concerned
in the movements of the chromatin particles. From the diffuse
granular state of the chromatin up until the formation of the
chromosomes, the linin network of the nucleus plays an active
part. Chromatin granules in the nucleus are never isolated
as such, but always have linin connections with other granules.
The synaptic threads consist of linin threads with the chro-
momeres embedded along them. As has been before stated,
homologous chromomeres have linin connections running between
them. Conklin ('17) has shown that the ground-work of the
cytoplasm is the relatively stable and elastic spongioplasm, and
he attributes to it the maintenance of cytoplasmic organization
and the movements and localization of cytoplasmic substances.
Similarly in the nucleus it seems that the linin is a relatively
elastic substance which forms the ground-work of the nucleus

94 E. L. SHAFFER.

and maintains the organization of the nuclear elements. Wenrich
('16) has shown how remarkably constant the organization and
"architecture" of the chromosomes are. By means of certain
structural peculiarities which his "selected" chromosomes
presented, he was able to recognize and trace them through all
the stages of spermatogenesis. The tendency has been noted
in many forms for the chromosomes to appear in the metaphase
always in a definite configuration. It is possible that the linin
connections between the chromosomes which have often been
figured (Figs. 8-12) are responsible for the definite patterns as-
sumed by the chromosomes in the metaphase plate. The
uniting in pairs of the homologous leptotene threads may be due
to the contractility of the linin connectives running between
the homologous chromomeres. In short, the morphological
stability of the nuclear elements and the constancy of their
form, arrangement and organization is in the last analysis
referable to the linin.

F. SUMMARY.

1. The diploid chromosome groups of four species of Lach-
nosterna, namely delala, fusca, gracilis and tristis, as well as
Pelidonota punctata and Cotalpa lanigera, show twenty chromo-
somes, one pair of which is composed of two unequal elements
(sex chromosomes) .

2. There are no essential differences in the form and arrange-
ment of the chromosomes in the species studied.

3. The growth period of the spermatocytes is marked by the
appearance of delicate leptotene threads which are derived from
the chromosomes of the last spermatogonial division. These
threads become polarized and there is evidence that they are
arranged in pairs parasynaptically.

4. There is a definite contraction stage which does not seem
to be caused by fixation, but is a normally occurring phase in
the growth period.

5. The sex chromosomes persist through the entire growth-
period in the form of definite compact bodies, sometimes being
contained within chromosomal vesicles. The unequal sex
elements separate in the first maturation division and divide
equationally in the second maturation division.

STUDY OF CHROMOSOMES OF LACHNOSTERNA. 95

6. There are five atelomitic tetrads in the first maturation
division and five telomitic tetrads (including the sex pair).

7. Cyst formation in the testis begins by the rapid division of a
single primary spermatogonium, so that all the cells within any
particular cyst are the descendants of a single cell. The visible
polarity of the cells seems to be established at the time of cyst
formation.

G. ADDENDA.

Since the manuscript of the foregoing study was written, the
work of Goldsmith1 has appeared on the chromosomes of the
Cicindelidae. He has studied in all five species of this family
and finds that they agree in chromosome number. He has
described a double odd-chromosome which passes undivided
to one pole of the spindle in the first maturation division and
divides in the second maturation division giving rise to sper-
matozoa with ten and twelve chromosomes respectively. In
his study of the growth stages of the spermatocytes he has been
unable to find that the leptotene threads actually pair. His
figures of the synaptic stages are not clear and he makes no
decision as to the method of synapsis (parasynapsis or telosyn-
apsis).

He describes the "early" spermatogonia as being arranged in
syncytia without any discernible cell-walls. He describes the
appearance within the syncytial cytoplasm of "cytoplasmic
fibrillar bridges." "With the increase in age and size of the
cells, these bridges become more dense and assume a definite
arrangement about a number of cells. This continues until the
entire tubule is subdivided into a large number of syncytia—
cysts containing cells without perceptible cell walls" (p. 445).
Both his descriptions and figures of this peculiar method of cyst
formation lack in clarity. From his Fig. 5, I interpret the
"cytoplasmic fibrillar bridges "as being probably the persisting
spindle remains or mitosome of the previous division, and it is also
probable that the deeply staining bodies in the cytoplasm are
the mid-bodies (cell-plate) persisting with the mitosome. That
these "fibrillar bridges" are really spindle remains is further

1A comparative study of the chromosomes of the tiger beetles (Cicindelidae).
Jour. Morph., Vol. 32, No. 3, 1919.

96 E. L. SHAFFER.

indicated by the fact that they" become more dense and assume
a definite arrangement about a number of cells." This is exactly
the behavior of the spindle remains which Hegner1 has described
in Leptinotarsa and which I have described in Passalus.2 It is
difficult to see how these "fibrillar bridges" are concerned in
dividing the syncytia into a number of cysts. Furthermore,
it is difficult to believe that a true syncytium of spermatogonia
actually does exist, for the later stages certainly do possess
cell-walls which must have been preexisting.

LITERATURE CITED.
Arnold, G.

'08 The Nucleolus and Michrochromosomes in the Spermatogenesis of Hydro-

philus piceus. Arch. f. Zellforsch., Bd. 2.
Bridges, C. B.

'16 Non-disjunction as a Proof of the Chromosome Theory of Heredity.

Genetics, Vol. i.
Carothers, E. E.

'17 The Segregation and Recombination of Homologous Chromosomes as
Found in Two Genera of Acrididae (Orthoptera). Jour. Morph., Vol. 28,
No. 2.

Conklin, E. G.
'17 Effects of Centrifugal Force on the Structure and Development of the

Eggs of Crepidula. Jour. Exp. Zool., Vol. 22, No. 2.
Goodrich, H. B.

'16 The Germ Cells in Ascaris incurva. Jour. Exp. Zool., Vol. 21, No. i.
Gross, J.

'07 Die spermatogenese von Pyrrhocoris apterus L. Zool. Jahrb Abtheil. f.

Ariat. u. Ontog., Bd. 23.
Hegner, R. W.

'14 The Germ Cell Cycle in Animals. Macmillans.
Metz, C. W.

'16 Chromosomes Studies on Diptera, II. Jour. Exp. Zool., Vol. 21, No. 2.
Mottier, D. M.

'07 The Development of Heterotype Chromosomes in Pollen Mother Cells.

Ann. of Bot., Vol. XXL
McClung, C. E.
'05 The Chromosome Complex of Orthopteran Spermatocytes. BIOL. BULL.,

Vol. IX., No. s-
'17 The Multiple Chromosomes of Hesperotettix and Mermiria (Orthoptera).

Jour. Morph., Vol. 29.
Overton, J. B.

'09 On the Organization of the Nuclei in the Pollen Mother Cells of Certain
Plants with Especial Reference to the Permanence of the Chromosomes.
Arch. f. Mikr. Anat., Bd. 70.

1 Studies on the Germ Cells, I and II. Jour. Morph., Vol. 25, No. 3, 1914.

2 Mitochondria and Other Cytoplasmic Structures in the Spermatogenesis of
Passulus cornutus. BIOL. BULL., Vol. XXIL, No. 5, 1917.

STUDY OF CHROMOSOMES OF LACHNOSTERNA. 97

Schafer, F.

'07 Spermatogenese von Dytiscus. Zool. Jahrb. Abtheil. f. Anat. u. Ontog.,

Bd. 23.
Stevens, N. M.

'06 Studies on Spermatogenesis, II. Carnegie Inst. Pub. No. 36.

'05 Studies in Spermatogenesis with Especial Reference to the Accessory

Chromosome. Carnegie Inst. Pub. No. 36.
'08 The chromosomes in Diabrotica vitatta, Diabrotica soror and Diabrotica

i2-punctata. Jour. Exp. Zool., Vol. V., No. 4.
'09 Further Studies on the Chromosomes of the Coleoptera. Jour. Exp. Zool.,

Vol. VI., No. i.
Voinov, D.

'03 La spermatogenese d'ete chez le Cybister roeselii. Arch, de Zool. Exp.

et Gen., 4e Series, T. i.
Wenrich, D. H.

'16 The Spermatogenesis of Phyronotettix magnus with Especial Reference to
Synapsis and the Individuality of the Chromosomes. Bull. Mus. Comp.
Zool., Harvard Coll., Vol. IX., No. 3.
Whiting, P. W.

'17 The Chromosomes of the Common House-mosquito, Culex pipiens. Jour.

Morph., Vol. 28, No. 2.
Wilson, E. B.

'12 Studies on the Chromosomes, VIII. Jour. Exp. Zool., Vol. 13, No. 3.
Wieman, H. L.

'10 A Study in the Germ Cells of Leptinotarsa signaticollis . Jour. Morph.,
Vol. 21.

q8 E. L. SHAFFER.

EXPLANATION OF PLATES.

All drawings (except Figs. 28, 29, 30) were made with camera lucida using Nc.
12 ocular and 1/16 mm. oil immersion objective. Figs. 28, 29, 30 were made
using No. 18 ocular and 1/16 mm. oil objective.

PLATE I.
(Figs, i to 12.)

FIGS, i to 6. Metaphase plates of spermatogonia in the six beetles studied all
showing 20 chromosomes including an unequal pair (XY).

FIG. i. L. delata.

FIG. 2. L. fusca.

FIG. 3. L. tristis.

FIG. 4. L. gracilis.

FIG. 5. Pelidonota.

FIG. 6. Cotalpa.

FIG. 7. Metaphase plate of follicle cell from ovary of L. delata showing ten
equal pairs.

FIGS. 8 to 12. Metaphase plates of ist spermatocytes showing ten bivalent
chromosomes.

FIG. 8. L. delata.

FIG. 9. L. fusca.

FIG. 10. L. gracilis.

FIG. ii. Pelidonota.

FIG. 12. Cotalpa.

BIOLOGICAL BULLETIN, VOL. XXXVIII.

PLATE I.

4

A B

6

9

10

11

12

E. L. SHAFFER.

IOO E. L. SHAFFER.

PLATE II.
(Figs. 13 to 27.)

FIG. 13. Telophase nucleus of spermatogonium, showing chromosomes spinning
out into delicate threads.

FIG. 14. Characteristic resting nucleus of spermatogonium.

FIGS. 15, 16, 17. Stages in the condensation of the chromatin from the resting
stage to the formation of the chromatic blocks.

FIGS. 18, 19. Early growth stages. Evidences of parallel pairing of leptonete
threads.

FIG. 20. Contraction (synezesis) stage.

FIG. 21. Pachytene threads released from synezesis stage.

FIG. 22. Strepsistene nucleus. Chromatin threads unanalyzable.

FIG. 23. Various forms of diplotene and strepsitene threads.

FIG. 25. Early prophase of ist spermatocyte. Tetrads beginning to form.

FIG. 26. Cell in prophase simulating synezesis; due to faulty fixation.

FIG. 27. Cell in stage similar to Fig. 26 from well-fixed material.

BIOLOGICAL BULLETIN, VOL. XXXHII.

PLATE II.

13

17

V

21

E. L, SHAFFER.

IO2 E. L. SHAFFER.

PLATE III.
(Figs. 28 to 36.)

FIG. 28. Side view of maturation tetrads of L. delata.
• FIG. 29. Side view of maturation tetrads of L. fusca.
FIG. 30. Side view of maturation tetrads of Pelidonota punctata.
FIG. 31. Early anaphase of first maturation division in L. fusca showing
separation of sex pair.

FIG. 32. Anaphase of first maturation in Pelidonota.
FIG. 33. Telophase of first maturation in L. delata.

FIG. 34. Telophase of first maturation in L. delata in which the sex elements
have failed to disjoin, and have passed to one daughter cell.

FIGS. 35, 36. Daughter plates of second spermatocyte of L. delata and Pelido-
nota, respectively.

BIOLOGICAL BULLETIN. VOL. XXXVIII

PIAIE Ml.

4100**"'

28

-. . . .

«

29

33

32

35

\

E. L. SHAFFER.

Vol. XXXVIII. March, 1920. No. 3.

BIOLOGICAL BULLETIN

CARBON DIOXIDE PRODUCTION IN RELATION TO

REGENERATION IN PLANARIA

DOROTOCEPHALA.1

HARRIET L. ROBBINS AND C. M. CHILD.

Numerous lines of evidence indicate that the body of Planaria
dorotocephala consists physiologically of more than one individual
or "zooid" after a certain limit of size is exceeded in the course
of growth. The first or chief zooid of the series includes the
region from the head to a level slightly posterior to the mouth,
the level at which fission usually occurs, and the region posterior
to this consists of one or more short zooids, the limits of which
can be distinguished physiologically, but not morphologically
(Child, '10, nc, '15, chap. VI). The susceptibility of the
body to a large number of chemical and physical agents in con-
centrations or intensities too high to permit acclimation or
tolerance, decreases from the head posteriorly through the
length of the first zooid, increases sharply at the level of fission
and in long animals shows one or more rises posterior to that
level (Child '13^). The process of regeneration also shows
graded differences at different levels, corresponding to the dif-
ferences in susceptibility (Child 'lib).

In the papers referred to, as well as in many others evidence
has been presented to that the susceptibility differences in general
are in some degree an indicator of quantitative differences in
metabolic condition, particularly in rate of oxidations, though
susceptibility has never been regarded as an exact quantitative
measure of oxidation. According to this point of view the
gradations of susceptibility in the body of Planaria dorotoceph-
ala indicate that the rate of oxidation decreases gradually, at

1 From the Hull Zoological Laboratory, University of Chicago.

103

104 HARRIET L. ROBBINS AND C. M. CHILD.

least in the ectoderm and body wall, from the head to the level
of fission and rises suddenly at the anterior end of the second
zooid and of each further zooid, if such are present.

It has also been shown that in isolated pieces of the body of
Planaria, particularly in those below a certain fraction of the
body length, susceptibility is increased during the first few hours
following section. This temporary increase of susceptibility,
apparently associated with the stimulation of the pieces by
section, is least in pieces nearest the head, and increases in suc-
cessive pieces posteriorly to the level of fission, where it again
decreases (Child, '14). In general the regions of highest sus-
ceptibility in the intact animal show the least increase in sus-
ceptibility in isolated pieces, and vice versa. Within twelve to
twenty-four hours after section the temporary increase in sus-
ceptibility has disappeared and the susceptibility of each piece
is about the same as, or slightly lower than before section.
From twenty-four to forty-eight hours after section a gradual
rise in susceptibility begins as the regulatory processes leading
to the development of a new individual make themselves evident,
and the susceptibility after regeneration is completed is much
higher than that of the region of the body of the original animal
which a given piece represents. This is true not only for the
newly developed head and posterior end of the regenerated animal
but for the whole body (Child, '15, Chap. V). In the light of
various facts, these changes in susceptibility have been inter-
preted as indicating and in some way associated with changes
in rate of oxidation. The present paper constitutes additional
evidence for this conclusion. Carbon dioxide production was
chosen instead of oxygen consumption as the subject of investi-
gation first because investigation of the temporary changes,
immediately following section involves the technical difficulty
of preparing very large numbers of pieces within a very short
time in order that oxygen consumption may be sufficient for
determination in short periods, and second because work on
oxygen consumption during the course of regeneration was
already in progress in this laboratory. The colorimetric method
was used since it is well adapted for obtaining comparative data
with small amounts of material.

CO2 IN RELATION TO REGENERATION IN PLANARIA. 1 05

A few experiments similar to those recorded below were per-
formed by Child from time to time during several years past,
and these experiments, although they brought to light certain
difficulties as regards technique and were too few in number to
constitute conclusive evidence, indicated very clearly that the
differences and changes in susceptibility following section and
after regeneration are paralleled by differences and changes in
rate of carbon dioxide production. A more extensive investi-
gation of the subject by another person was desirable, and this
was undertaken by Miss H. L. Robbins in candidacy for the
degree of M.S. The data tabulated in the present paper are
those obtained in this investigation.

The special technique for the different experiments is described
in the following sections, but the general technique used in
connection with the colorimetric method is very similar to that
employed in the study of CO2 production during starvation
(Child, '190). The weighing of the animals or pieces is the
most difficult feature of the preparation, since the presence of
superfluous water introduces an error in weight, and since
weighing must be done as rapidly as possible to avoid injury or
death from drying. After considerable practice, involving
repeated weighings of the same lots and determination of the
length of time which could be allowed without injury for drainage
on filter paper and exposure to the air while weighing, a satis-
factory method of procedure was developed, and this was fol-
lowed in all the tabulated experiments. The container was
first weighed and both container and weights were left on the
balance pans : next weights equal to the estimated weight of the
lot of worms or pieces were placed on the weight pan in order to
reduce the time necessary for weighing the animals: the worms
were then brought together at the tip of a funnel of well washed
filter paper, drained for a certain length of time, transferred on a
slightly vaselined scalpel blade to the container and weighed,
each of these operations being performed in as nearly as possible
the same length of time in each case. Equal or approximately
equal weights of the lots to be compared were used (see tables).
For the colorimetric estimations the Hynson Westcott and
Dunning H ion outfit with phenolsulphonephthalein as indicator

IO6 HARRIET L. ROBBINS AND C. M. CHILD.

was used. After weighing, the worms were returned at once to
water and then transferred to pyrex tubes fused at one end and
of the same diameter as the standard tubes. After washing
twice in aqueous indicator solution of the same concentration
as that in the standard tubes, each tube was filled with indicator
solution to a 3-c.c. level previously marked. The tubes were
then sealed without air bubbles by running in on the surface of
the fluid about i c.c. of soft paraffin at a temperature just above
melting point, the worms being kept at the bottom of the tube
to avoid injury from change of temperature. Since leakage
past the paraffin plug is difficult to avoid when changes of tem-
perature occur and since it was found desirable to reduce the
temperature slightly as a means of keeping the animals quiet
(see below), the following method of providing for changes in
volume of the fluid was used. A piece of closely fitting soft
rubber tubing, previously coated with soft paraffin was drawn
over the open end of the pyrex tube, leaving 2 to 3 cm. of the
tubing beyond the end of the glass. Indicator solution was
then added to fill both the glass tube above the paraffin plug and
the rubber tube, and the latter is then closed by a screw clasp.
This procedure makes impossible the entrance of air past the
paraffin plug when the temperature is lowered. Instead of air
a small amount of the indicator solution above the plug may be
drawn below, but this occurs within the first five minutes or
less of the experiment, and with the changes of temperature
involved the amount of fluid passing the plug is negligible, so
far as the results are concerned. The indicator solution between
the paraffin plug and the rubber tubing is visible and serves as a
control for the color change below the plug. Closure by means
of the rubber tubing and clamp alone was found to be unsatisfac-
tory because some of the worms creep into the rubber tubing,
where they cannot be seen, and it is therefore impossible to
determine whether motor activity is going on and whether all
are in good condition. Moreover, pieces in the rubber tubing
are often overlooked when the lots are removed from the exper-
imental tube and the whole lot becomes valueless for further
experiment, unless substitution for the lost pieces is made, but
this is at best an undesirable procedure.

CO2 IN RELATION TO REGENERATION IN PLANARIA. IO7

The pH at the beginning of the experiment was of course the
same in all lots to be compared, but the starting point differed
somewhat in different experiments, the extreme range being 7.8
to 7.95. Observations were made at least every half hour with
few exceptions, but the tables, instead of recording all the pH
readings, give the times required to reach an arbitrary end point,
Hp 7.3 being selected as this end point. As each lot approached
this point, observation was more or less continuous. A daylight
lamp was used for all color comparisons.

A serious difficulty in the earlier experiments, particularly
with the pieces after section, was the occurrence of motor activity,
which of course increased COs production and introduced a
source of error. In these experiments as in earlier work, it was
observed that pieces from regions near the head are much more
likely to show apparently spontaneous motor activity during
the first few hours after section, than pieces from the more pos-
terior levels of the first zooid. After attempting in various ways
to eliminate motor activity, it was found that decrease of a few
degrees in temperature was usually effective for the length of
time necessary. In all the experiments tabulated below the
animals were kept at 21° to 22° C in the stocks and during prepar-
ation, but as soon as the tubes were sealed they were placed in
water at a constant temperature of 18° in very dim diffuse
daylight. Under these conditions motor activity occurred only
rarely, but its occurrence was always noted in the record of the
experiment. The procedure adopted in these experiments has
been described at some length because in experimental work with
animals of such small size it is extremely easy to go astray, if
the various sources of error are not carefully controlled as far as
possible.

THE STIMULATION OF PIECES FOLLOWING SECTION.

The temporary increase in susceptibility following section
and its characteristic relation, both to size of piece and region of
body (Child, 14, also p. 104 above) are so clearly shown by the
susceptibility method that they are often used as laboratory
experiments. Because they are temporary and -apparently
excitatory in character and so definitely related to size of piece

108 HARRIET L. ROBBiNS AND C. M. CHILD.

and region of body these changes in susceptibility are of special
interest in connection with the question of the relation between
susceptibility and metabolism. If changes in the rate of funda-
mental metabolism or of certain fundamental reactions are found
to parallel these changes in susceptibility, it is evident that the
susceptibility method, when properly used, is a rather delicate
indicator of at least certain aspects of metabolic condition.

In the earlier experiments of Child, as well as in the preliminary
work of Robbins, it was found that the changes and differences
in CO2 production following section appeared more clearly in
animals which had been starved for a few days before experiment,
than in those which had been more recently fed, although in the
former the total CO2 production and oxygen consumption are less
than in the latter (Child, 'iqa, Hyman, 'IQ&). It has been pointed
out elsewhere (Child, '196, 'igc and various earlier papers), that
the susceptibility method as used in these experiments gives in-
formation primarily concerning conditions in ectoderm and body
wall, though with certain precautions it may be used to show
differences in condition in the alimentary tract of Planaria and
other forms. The susceptibility data, as well as other facts,
indicate that the changes following section are at least in large
measure confined to ectoderm and body wall, the alimentary
tract not being affected to any great degree. The CC>2 production
of the alimentary tract in fed animals constitutes, however, a
large proportion of the total CO2 production, moreover, the
volume of the alimentary tract as compared with that of other
organs is greater and a larger amount of food and reserves is
usually present in regions near the mouth than in regions near
the head, therefore it is desirable to decrease this alimentary CO2
production as far as possible, in order that changes in other parts
of the body may appear more clearly. It has been shown
(Child, 'iga, Hyman, '196) that a rapid decrease in both COa
production and oxygen consumption occurs in Planaria doroto-
cephala during the first few days of starvation and that an in-
crease in both follows so rapidly after even a single feeding that it
cannot be due to oxidation of the food following assimilation,
but must be due to stimulation of the alimentary tract by the
food. Allen ('19) has recently recorded the occurrence of similar

CO2 IN RELATION TO REGENERATION IN PLANARIA. 1 09

changes in oxygen consumption in two other species of Planaria
during the early stages of starvation. It is evident that the
rapid decrease in CC^ production and oxygen consumption during
the early stages of starvation is due in large measure, if not
wholly, to the decrease in functional activity of the alimentary
tract in the absence of food newly ingested. In the light of all
these facts the reason for the use in these experiments of animals
which have been starved a few days is evident. The length of the
period without food is given in each experiment: in no case is it
long enough to produce any marked reduction in size or other
changes except those in the alimentary tract.

Since the purpose of these experiments is to determine whether
differences and changes in susceptibility following section are
paralleled by differences and changes in CO2 production, the
size of animals and pieces used and the regions of body included
are those which show the most definite and characteristic differ-
ences and changes in susceptibility. The experimental material for
the data presented in Table I. was prepared as follows: animals
sixteen to eighteen mm. were selected from well fed laboratory
stock, those which had recently undergone fission being excluded,
were kept without food for several days (see Table I.) and were
then cut into pieces as indicated in Fig. I, piece C being cut so
that the greater part of the pharynx is separated from its attach-
ment and is extracted from the pharyngeal pouch, i.e., piece C
contains a part of the pharyngeal pouch, but no portion of the
pharynx. Pieces A and C were used in the experiments as repre-
senting respectively the most anterior and the most posterior
portion of the first or chief member or zooid in animals of this
size (Child, 'lib). In the intact animal the susceptibility of the
region corresponding to piece A is very much greater than that of
the region corresponding to piece C. Immediately after section
the susceptibility of piece A shows either a slight increase or no
marked change, while the susceptibility of C is increased to such a
degree that it is equal to or even greater than that of A (Child,
'14). The region B between A and C is intermediate both as
regards original susceptibility and the changes following section
and is not used in these experiments. The increase in suscepti-
bility following section, which is most marked in the C-piece is

no

HARRIET L. ROBBINS AND C. M. CHILD.

A

B

C

temporary and after six to eight hours is in course of disappearance
and sooner or later (12 to 24 hours under ordinary conditions)
the susceptibility of the piece becomes about the same as or a
little less than that of the corresponding regions in the intact
animal. These temporary changes in susceptibility indicate
that the pieces have been stimulated by section, anterior pieces
least, posterior pieces most, and that the condi-
tion of excitation disappears after a few hours.
It has been suggested elsewhere (Child, '14)
that the difference in the degree of stimulation
in anterior and posterior pieces results from the
greater dependence of the more posterior regions
upon impulses from regions anterior to them, so
that when the paths of these impulses are cut
the condition of the more posterior levels is
more affected than that of the anterior levels.
This is in agreement with the observations of
various investigators concerning the difference
in intensity of reaction in many animals between
levels anterior and those posterior to a trans-
verse cut.

The region of the body posterior to C in Fig.
I is not used in the tabulated experiments be-
cause it consists of one or more zooids indicated
by differences in physiological condition (Child,
'10, 'nc). Since the number of these posterior
zooids differs in different animals and no mor-
phological boundaries are visible, pieces from
the same levels of this region in different ani-
mals are not always strictly comparable physio-
logically and therefore merely complicate the
experimental data.
Each experiment on CO2 production in the pieces A and C
consists essentially in determining the rate of change in pH,
once as soon as possible after section and again several hours
later, of one lot of ^.-pieces and one of C-pieces, consisting of
thirty to fifty pieces each and of as nearly as possible equal weight,
in equal volumes of indicator solution at a constant temperature

FIG. i.

CO2 IN RELATION TO REGENERATION IN PLANARIA. Ill

of 18° C. As soon as £#7-3 is reached in the first determination,
the tubes are opened, the pieces are returned to water and left un-
disturbed for the desired number of hours, i.e., at least long enough
as determined by the susceptibility method for disappearance in
large measure of the temporary changes following section. At
the end of this time the rate of pH change is again determined
for each lot under as nearly as possible the same experimental
conditions as the first determination.

Fifteen such experiments were performed before the technique
used in the experiments recorded in Table I. was fully worked
out. In eight of these fifteen experiments the A -pieces showed
marked motor activity and a more rapid change in pH than the
C-pieces immediately after section, in three there was either death
of some of the pieces from drying or an error in weighing, and
in four none of these sources of error was involved and the result
were similar to those in Table I.

Table I. gives the data for thirteen experiments, in all of which
the technique was as nearly as possible the same. The only
feature of the table which requires explanation is the column
'Time after section." The times given in this column are the
times of sealing of the tubes containing the animals in indicator
solution. The first time for each lot, "Immed.," i.e., imme-
diately, means simply that the pH determination was begun as
soon as possible after the pieces were sectioned. Since it requires
a half hour or more to cut lots consisting of forty to fifty A and
C-pieces, the time between section and sealing may be as much
as an hour for the pieces cut first and only a few minutes for
those cut latest.

In eleven of the thirteen experiments recorded, 84 per cent.,
the CO2 production of the C-pieces immediately after section
is about equal to, or greater than that of the A -pieces. As
regards the other two experiments, Nos. 7 and 8, in No. 7 three
C-pieces died during the first determination, and in No. 8,
motor activity occurred in the ^.-pieces, perhaps because of a
slight rise in temperature of the water in which the tubes were
kept. The second pH determination, begun at various lengths
of time ranging from nine to forty-two hours after section, shows
in every case a lower rate of CC>2 production in C than in A .

112

HARRIET L. ROBBINS AND C. M. CHILD.

TABLE I.

A COMPARISON OF ANTERIOR AND POSTERIOR PIECES IMMEDIATELY AFTER
SECTION AND A NUMBER OF HOURS LATER.

Expt.
No.

No. of Wt. in
Pieces. Grams.

No. of Days
Without
Food.

Tempera-
ture in
Degrees C.

pH Start-
ing Point.

Time After
Section in
Hrs. and
Min.

Time in Hrs
and Min. to
Reach pH

7-3-

A 37 0.023

II

17-5

7.8

immed.
22:30

5:40
5:40

I

C 31 0.0231

II

17-5

7-8

immed.
22:30

5:25

5:55

A 49

0.0255

24 IS

7-9

immed.
9:00

6:25
6:30

2

C45

0.0257

24 18

7-9

immed.
9:00

6:15

6:45

A 56

0.0245

21

18

7.82

immed.
19:40

6:05

3

C 48

0.0243

21

18

7.82

immed.
19:40

5:05
6:23

A 20

0.0182

8 18

7.83

immed.

18:40

6:10
6:22

4

C 23 0.0185

8 18

7.83

immed.
18:40

6:OO
6:45

A 51 0.0208

9 18

7.83

immed.
20:15

5:30
5:35

5

C44

0.0207

9

18

7-83

immed.
20:15

5:05

5:55

A 37 0.0174

12

18

7.83

immed.
18:00

6:30
6:35

C 42 0.0176

12

18

7-83

immed.
1 8:00

6:10
6:45

71

A 37 0.0143

13

IS

7-83

immed.
21:30

7:20
7=45

C 37 0.0142

13

18

7-83

immed.
21:30

7:35
8:00

K~

A 40

0.0216

7

18

7.83

immed.
17:45

4:50
5:05

o

C4i

0.0213

7

18

7-83

immed.
17:45

4:55
5:10

A 36

0.0238

8

18
18.5

7.83

immed.
19:00

4:50
438

93

C 36

0.0239

8

18
18.5

7.83

immed.
19:00

4:35
4:50

A 37 0.0242

9

18

7-83

immed.
22:00

4:40
4:55

10

C 38 0.0241

9

18

7-83

immed.

22 :oo

5 • ^ 5

A 44 0.0242

7

18

7-95

immed.
19:00

5:30
5:35

1 1

C45

0.024

7

18

7-95

immed.
19:00

5:10
5:50

A 44

0.0306

8

18

7-95

immed.
42:00

4:45
4:40

12

C 50

0.0305

8

18

7-95

immed.
42:00

4:50

A 29

O.02I

13

18

7-8

immed.
25:00

4:05
4:10

13

C 21

O.0208

13

18

7.8

immed.

3:50

25:00

4:25

CO2 IN RELATION TO REGENERATION IN PLANARIA. 113

As regards the two records for the same lot, the A -pieces show
in general a slightly lower CO2 production in the second period
than in the first, or in some cases about the same. Experiment
9, however, shows a slightly higher rate in A during the second
period, this being due probably to the slightly higher water
temperature during the second period. In the C-pieces the rate
is distinctly lower, in many cases much lower in the second period
than in the first. Differences in time of five minutes between
the records of two lots made at the same time mean only that
the two were barely distinguishable as different. Differences
of this mangitude between the two records of the same lot can
mean no more than that the rate is essentially the same in the
two periods. Differences of ten minutes or more are however,
unquestionably significant and differences of much greater
magnitude appear in the table.

The table show's then that immediately after section there
is in some cases a slight temporary increase in CO2 produc-
tion in the ^.-pieces and a very marked increase in the C-pieces,
that is to say the differences and changes in CO2 production
following section are in general parallel to the differences and
changes in susceptibility. The objection may be raised that
the differences in CO2 production are less than would be expected
from the stimulation by section, but when it is remembered that
there is no reason to believe that the alimentary tract shares in
this stimulation except locally at the cut surface and that the
pieces do not undergo motor activity during the determination
this objection has little weight. As a matter of fact the close
parallelism between the data on CO2 production and those on
susceptibility indicates that even as regards the temporary
changes following section of pieces, susceptibility is in some degree

1 During the first determination in No. 7 three C-pieces died. When the
determination was repeated the following day, two C-pieces, which had been cut
with the others but not used before, were added to lot C and one piece was removed
from lot A to make weights as nearly as possible equal. Only two extra worms
had been sectioned, therefore three pieces could not be added to take the place of
those dead.

2 Slight motor activity occurred in the A-pieces during the first determination.

3 Here the temperature was half a degree higher in the second determination
than in the first, and the A -pieces show a slightly increased rate in the second
determination. Footnotes for Table I.

HARRIET L. ROBBINS AND C. M. CHILD.

a measure of physiological and particularly of respiratory con-
dition.

CARBON DIOXIDE PRODUCTION AFTER REGENERATION.

In these experiments stocks of pieces of the size of A, B, C,
Fig. I, were cut from animals 16— 18 mm. and allowed to undergo
regeneration until the development of the new individuals was
essentially complete, usually about two weeks. From these
stocks the experimental lots were selected. These consisted
only of normal individuals, or in some cases where the number
of normal animals in the regenerated stock was not sufficient,
mostly of normal with a few teratophthalmic individuals (Child,
'na) added. Since the pieces undergoing regulatory develop-
ment cannot feed until they attain an advanced stage, the animals
representing the condition before regulation, with which the
regenerated individuals are to be compared, are kept without
food for the same length of time as the pieces undergoing regener-
ation. Usually a stock of intact worms of the same size, 16-18
mm., as those from which the pieces were taken is isolated at

TABLE II.

A COMPARISON OF SMALL ANIMALS REGENERATED FROM PIECES WITH LARGE
ANIMALS THAT HAVE NOT RECENTLY UNDERGONE REGENERATION OR FISSION,
OF THE SAME SIZE AS THOSE FROM WHICH THE PIECES WERE TAKEN. BOTH
UNFED. DETERMINATIONS OF pH AT + 18° C.

Expt. No.

No. of Worms.

Wt. in Grams.

Period of Re-
generation or
Starvation in
Days.

pH Starting
Point.

Time in Hrs.
and Min. to
Reach pH 7.3.

I

3 large
1 6 regen.

O.OII4
O.OII3

IO
10

7.89
7.89

10:25
9:00

2

4 large
22 regen.

0.0163
0.016

II
II

7.89
7.89

8:55
6:45

3

4 large
27 regen.

0.0174
0.0176

10
IO

7.89
7.89

8:00
6:15

4

4 large
19 regen.

0.0117
0.0115

14
14

7-9
7-9

14:35
12:35

5

5 large
34 regen.

0.0205
0.0204

14

14

7-9
7-9

9:35
6:50

3 large

O.OIOI

21

7-8

10:00

6

7 starved

26 regen.

0.099

14 regen.

7-8

7:45

3 large

0.0109

20

7.8

8:30

7

7 starved

19 regen.

O.OII

13 regen.

7-8

7:50

8

3 large
28 regen.

0.0166
0.0164

10
10

7.8
7-8

6:30
3:50

CO2 IN RELATION TO REGENERATION IN PLANARIA. 115

the same time the pieces are cut. This stock merely undergoes
a slight degree of starvation, while the pieces undergo starvation
for the same period and in addition the regulatory changes.
This is the procedure in experiments 1-5 and 8 in Table II.,
but in experiments 6 and 7 the stock from which both pieces
and whole animals were obtained was starved seven days before
the pieces were cut.

Each experiment in Table II., includes one lot of worms about
5 mm. in length which have developed from pieces cut ten to
fourteen days earlier ("regen." in table), and one lot of as nearly
as possible the same weight of worms 16-18 mm. in length, the
same size and from the same general stock as the worms from
which the pieces were cut, and kept without food for the same
length of time ("large" in table). In all cases the pH deter-
minations are made before feeding is resumed.

Examination of the last column of the table shows that the
rate of CC>2 production is much higher in the small regenerated,
than in the large old animals, i.e., the regulatory processes have
been accompanied by an increase in rate of CC>2 production.
Moreover the rate is in general higher in the regenerated animals

TABLE III.

A COMPARISON OF SMALL ANIMALS REGENERATED FROM PIECES WITH LARGE

ANIMALS OF THE SAME SIZE AS THOSE FROM WHICH THE PIECES WERE TAKEN.

Fed three times. Determinations of pH at 18° C.

Expt. No.

No. of Worms.

\Vt. in Grams.

pH Starting Point.

Time in Hours and
Minutes to Reach
pH 7.3.

j

3 large

O.OI27

7.82

5:00

10 regen.

O.OI25

7-82

4:30

4 large

0.0193

7.82

3:45

2

15 regen.

0.0191

7.82

2:55

4 large

0.0154

7.82

4:50

3

13 regen.

0.0155

7-82

3:35

4 large

0.0135

7-9

7:15

4

i 8 regen.

0.0134

7-9

5:50

5 large

0.0233

7-9

5:05

5

28 regen.

0.023

7-9

3:45

6

2 large heads off

O.OIS5

7.8

8:15

1 8 regen.

O.OII3

7.8

4:25

than in the pieces of Table I. although the period without food
is in most cases longer in the latter than in the former and CO^
production decreases during the early stages of starvation.

Table III. records experiments similar to those of Table II.

Il6 HARRIET L. ROBBINS AND C. M. CHILD.

except for the fact that both lots were fed with beef liver three
times before weighing and pH determination, the first two feed-
ings being on successive days, the third after an interval of one
day. In the first five experiments of Table III. worms from
the first five experiments of Table II. were used, but in smaller
numbers because of the increased weights after feeding, partic-
ularly in the regenerated animals. In Table III., as in Table II.
the "large" animals are those which have not undergone regener-
ation and represent as nearly as possible the animals from which
the pieces were taken, and the "regenerated" animals are those
which have developed from the pieces. Table III. agrees with
earlier work (Child, '190) in showing that the rate of change in
pH is increased in all animals by feeding after a period of star-
vation, but it also shows that the difference in rate between the
regenerated and the large animals persists after feeding. Here
again the data on CC>2 production agree with the results of the
susceptibility method (Child, '15, Chap. IV.). Data on oxygen
consumption recently published by Allen ('19) and by Hyman
('196) also agree with these results.

ADDITIONAL DATA.

Table IV. includes a number of miscellaneous experiments
of some interest.

In the course of regulatory development an outgrowth of
new tissue occurs at anterior and posterior ends of each piece.
All the facts indicate that this tissue, which forms the new
head and posterior end, is more or less embryonic in character
when it arises and possesses, at least at first, a higher rate of
metabolism than the remainder of the piece. In order to deter-
mine whether the higher rate of CO2 production in regenerated
animals is due solely to the more intense activity of this new
tissue or whether the rate is also increased in other parts, the new
heads and posterior ends were removed from regenerated animals
leaving only the so-called old or less extremely altered tissue of
the middle regions. Lots of such pieces were then compared
with lots of equal weight of freshly cut ^4-pieces (Fig. i) and of
animals 16-18 mm. like those from which the pieces were taken,
the heads being removed from these large animals in order to

C02 IN RELATION TO REGENERATION IN PLANARIA.

make them more nearly comparable in condition with the ^.-pieces
and the headless regenerated animals.

Experiment I of Table IV. includes one lot of each of the three
groups and it is seen that the "old" parts of the regenerated

TABLE IV.

MISCELLANEOUS DATA.

Regenerated animals from which anterior and posterior new tissue has been
removed, compared with A-pieces and with large headless animals: whole re-
generated animals compared with /1-pieces and with growing worms of same size
from stock. Fed or unfed. Determinations of pH at 18° C.; pH at beginning
of experiment 7.8.

Expt.No.

No. and Condition
of Animals.

Wt. in Grams.

Nutrition.

Time in Hours ar.d
Minutes to Reach
pH 7.3.

3 large, heads off.

0.093

Starved 19 days.

8:25

I

12 A-pieces.

0.094

Starved 19 days.

7:i5

ii regen. new

0.095

Starved 18 days.

6:25

tissue off.

9 A -pieces.

0.08

Fed three times.

8:15

2

22 regen. new

O.O82

Fed three times.

7=30

tissue off.

3 large.

O.OIII

Fed three times.

7=35

3

15 regen. new

0.0108

Fed three times.

6:05

tissue off.

22 A -pieces.

0.0167

Starved 10 days.

4:10

4

28 regen.

0.0164

Starved 10 days.

3:50

ii small.

0.0096

Fed three times.

6:25

1 8 regen.

0.0095

Fed three times.

5:25

animals show a higher rate of change than the ^4-pieces, while
the headless large animals show the lowest rate of all.

In Experiment 2 the "old" parts of regenerated animals are
compared with A -pieces, both lots being fed three times before
sectioning. The result is the same as in experiment I, the rate
being distinctly higher in the parts of regenerated animals than
in the A -pieces.

In experiment 3 the "old" parts of regenerated animals are
compared with large old animals from which the heads have not
been removed, both lots being fed three times after a star-
vation period of about two weeks. The result is the same as in
Experiment I, the parts of regenerated animals showing the
higher rate.

Experiment 4 is a comparison of A -pieces with entire regen
crated animals, i.e., including the new heads and posterior

Il8 HARRIET L. ROBBINS AND C. M. CHILD.

ends. Here the difference in rate is proportionally about the
same as in Experiment 2, but somewhat less than in Experiment I .

In Experiment 5 regenerated animals are compared with
stock animals of slightly larger size (the smallest in the stock at
the time) which were kept without food while the pieces were
regenerating, both lots being fed three times before the experi-
ment, and both consisting of entire animals. Here again the
regenerated animals show a higher rate of CO% production than
the slightly larger stock animals.

A few other experiments performed by one of us and in some
cases also by students in the laboratory, are briefly mentioned
here without tabulation of the data. It has been found, for
example, that the degree of increase in both susceptibility and
COz production occurring in regeneration depends upon the
degree of reorganization which occurs. Consequently the smaller
the piece in relation to the size of the body from which it is taken,
the greater the amount of increase in susceptibility and CO%.
Similarly in natural fission the posterior piece is not only smaller
than the anterior but develops a new head at the anterior end
and a prepharyngeal and pharyngeal region by reorganization
and redifterentiation within the piece, while the anterior fission
piece develops merely a new posterior end. In the animal
developed from the posterior piece susceptibility and COz
production show a marked increase while in the anterior animal
the only marked change in susceptibility is a slight increase in
the posterior region, where reorganization and growth have
occurred and the increase in CO2 production is either slight or
inappreciable. Allen ('19) has recently recorded somewhat
similar results as regards oxygen consumption, an increase
occurring in the posterior, but not in the anterior product of
fission.

It has also been determined by one of us that susceptibility
to lack of oxygen increases during the development of a new
individual from a piece, the susceptibility of the new individual
about two weeks after section of the piece, being distinctly higher
than that of well fed animals of the same size and about the
same as, or slightly than that of animals of the same size, kept
without food for the same length of time as the regenerating

CO2 IN RELATION TO REGENERATION IN PLANARIA. IIQ

pieces. In other words, the new individuals developed from
pieces show a susceptibility to lack of oxygen equal to or greater
than that of much smaller younger animals than those from
which the pieces were taken. These observations concern
primarily the susceptibility of ectoderm and body wall.

CONCLUSION.

These experiments constitute a new test of the validity of the
susceptibility method as a rough comparative means of deter-
mining physiological or metabolic condition and at every point
the differences and changes in susceptibility, as determined by
KNC and in many cases by various other agents also, are paral-
leled by differences and changes in rate of CC>2 production.
Even the temporary stimulation of the pieces after section, which
is slight or absent in the A -pieces and very marked in the C-pieces,
appears in the data on CO2 production and the increase in rate,
at least of respiratory metabolism associated with regulation, is
evident in the marked increase in rate of CO2 production even in
the "old" parts of the regenerated animal. This work may per-
haps be regarded as in some respects the most delicate test of the
relation between susceptibility and respiration which has been
made up to the present.

As has been repeatedly stated, the susceptibility method is
not an exact quantitative method, but a rather crude means of
indicating differences of some sort in physiological condition
and the differential susceptibility of different regions of the
same individual affords a means of modifying and controlling
various developmental and other processes. The facts at hand
concerning susceptibility, e.g., the lack of specificity, the close
relation between susceptibility and physiological activity in
development, growth and function as well as the positive evidence
already obtained concerning the parallelism between suscepti-
bility, oxygen consumption and CO2 production indicate very
clearly that a more or less definite relation exists between the
susceptibility of living protoplasm, to at least many external
agents and conditions within certain ranges of concentration or
intensity, and the rate or intensity of certain fundamental
physiological processes, particularly those which liberate energy.

120 HARRIET L. ROBBINS AND C. M. CHILD.

This is all that is meant when susceptibility is interpreted in
terms of metabolism or oxidation and the exact nature, degree
and extent of this relation of course remains to be determined.
This interpretation does not involve the assumption that sus-
ceptibility must always be parallel or even proportional to total
oxidation or even to total oxygen consumption or COo production.

The relation between susceptibility and oxidation is undoubt-
edly indirect in at least most cases and it is conceivable that
susceptibility may be related only or primarily to certain oxi-
dative reactions or to conditions associated with them. More-
over, it is certain that in many cases susceptibility as determined
by death and disintegration is dependent primarily upon con-
ditions or reactions in particular regions of the body, e.g., in
ciliate infusoria the ectoplasm, in Planaria the ectoderm and
body-wall. Moreover, as many investigators have pointed
out, it is by no means certain that oxygen consumption and CO2
production are exact quantitative measures of oxidation at
any given time. It is to be expected that susceptibility will not
always be proportional to total oxygen consumption or CO2
production, but even then susceptibility may prove in the long
run to be a better indicator or comparative measure of physiologi-
cal condition than the respiratory data.

From what has been said above and in earlier papers (e.g.,
Child, 'IQC) it is evident that the criticisms of the susceptibility
method recently advanced by Lund ('i8a, &) and Allen ('18, '19)
need no discussion here, since they are largely beside the point
and result from failure to grasp the conception of susceptibility,
which has developed from many different lines of investigation,
not from one alone. Even if we grant the correctness of certain
of their conclusions from experimental data which are or appear
at present to be in conflict with conclusions reached in this
laboratory (Child, 'iqa, b, c, Hyman, '190, b) on the basis of
more extensive investigation, with more satisfactory technique
and several different methods instead of one they do not con-
stitute adequate grounds for denying the physiological signifi-
cance of susceptibility, but rather merely a starting point for
the further analysis of the particular cases in question.

As regards the real significance of susceptibility, it makes

CO2 IN RELATION TO REGENERATION IN PLANARIA. 121

little difference whether or not it shall be found to run exactly
parallel to the other indices of total respiration in any particular
case. It cannot, however be denied that a wide range of facts
determined by many different lines of investigation do indicate
clearly the existence of a more or less definite relation between
susceptibility and oxidation in at least many cases, and it is of
interest to determine range, degree and nature of this relation.
The present paper like several others which have recently ap-
peared from this laboratory is a contribution to this problem,
but it must be remembered that data such as these are not the
only criteria of the physiological significance of susceptibility
and its relation to the energy-liberating reactions in the metabolic
complex.

SUMMARY.

1. The colorimetric estimation of CO2 production shows that
the changes in CC>2 production following section in pieces of the
body of Planaria dorotocephala run parallel with changes in sus-
ceptibility. Immediately following section CO2 production is
markedly increased in pieces cut from near the mouth region,
while in pieces from regions near the head it is only slightly if at
all increased. These changes are temporary excitations fol-
lowing section and disappear after a number of hours.

2. The development of a new individual from a piece is accom-
panied by a very considerable increase in COs production which
involves not only the new outgrowths at the two ends of the new
animal but the "old" parts as well. This increase in CO2
production is found both before and after feeding is resumed
following the development of the piece. In these respects also
the changes in CC>2 production parallel changes in susceptibility,
both series of data indicating that the animal developing from
an isolated piece becomes in the course of this development,
physiologically younger than the animal from which the piece
originated.

REFERENCES.
Allen, G. D.

"18 Quantitative Studies of the Rate of Respiratory Metabolism in Planaria,

I. Amer. Jour. Physiol., XLVIII.

'19 Quantitative Studies, etc., II. Amer. Jour. Physiol., XLIX.
Child, C. M.

'10 Physiological Isolation of Parts and Fission in Planaria. Arch. f.

Entwickelungsmech., XXX., II. Teil.

122 HARRIET L. ROBB1NS AND C. M. CHILD.

'ua Experimental Control of Morphogenesis in the Regulation of Planaria,

BIOL. BULL., XX.

'ub Studies on the Dynamics of Morphogenesis, etc., I. Jour. Exp. Zool., X.
'uc Studies, III. Jour. Exp. Zool., XL
'133 Studies, etc., V. Jour. Exp. Zool., XIV.

'i3b Studies, etc., VI. Arch. f. Entwickelungsmech., XXXVII.
'14 Studies, etc., VII. Jour. Exp. Zool., XVI.
'15 Senescence and Rejuvenescence. Chicago.
'iga. A Comparative Study of Carbon Dioxde Production During Starvation.

in Planaria. Amer. Jour. Physiol., XLVIII.
The Effect of Cyanides on Carbon Dioxide Production in Planaria

dorotocephala. Amer. Jour. Physiol., XLVII.
Susceptibility to Lack of Oxygen During Starvation in Planaria. Amer.

Jour. Physiol., XLIX.
Hyman, L. H.

'193 On the Action of Certain Substances on Oxygen Consumption, II. Amer,

Jour. Physiol., XLVIII.

'igb Physiological Studies on Planaria, I. Amer. Jour. Physiol., XLIX.
'IQC Physiological Studies on Planaria, II. Amer. Jour. Physiol., L.
Lund, E. J.

'i8a Quantitative Studies on Intracellular Respiration, II. Amer. Jour.

Physiol., XLV.
'i8b Quantitative Studies, etc., III. Amer. Jour. Physiol., XLVII.

SUSCEPTIBLE AND RESISTANT PHASES OF THE

DIVIDING SEA-URCHIN EGG WHEN SUBJECTED

TO VARIOUS CONCENTRATIONS OF LIPOID-

SOLUBLE SUBSTANCES, ESPECIALLY

THE HIGHER ALCOHOLS.

FRANCIS MARSH BALDWIN.

INTRODUCTION.

That dividing Arbacia eggs show periods of varying suscep-
tibility and resistance when exposed to chemical substances
and to various physical conditions has been proved by numerous
investigations. When eggs, from a single fertilized lot were
placed at regular successive intervals after fertilization in cyanide-
containing sea-water (m/ioo to w/2Oo), Lyon'2 found that they
were highly resistant to poisoning fifteen or twenty minutes after
fertilization, while eggs exposed to the same solution at the time
of cytoplasmic division were promptly killed. Later, after the
first division had been completed, the resistance to poisoning
again returned, followed by a second susceptible period at the
second cleavage. Loeb3 later noted that the unfertilized eggs
show greater resistance to cyanide poisoning than the fertilized
eggs, and Mathews4 indicated that in dividing eggs, the period
of maximum susceptibility is "immediately before and during
segmentation," and that just after segmentation the egg becomes
relatively highly resistant. Similar results were obtained by
Spaulding5 in experiments with weak solutions of ether (1/64
per cent, in sea- water). The period of high resistance continued
up to the beginning of the first cleavage, and then fell during
cleavage to zero, with a sharp rise immediately afterwards.
There was a short period of susceptibility immediately following

1 From the Marine Biological Laboratory, Woods Hole, Mass., and the Depart-
ment of Zoology, Iowa State College, Ames, Iowa.

2 E. P. Lyon, Amer. Jour. Physiol., 1902, Vol. 7. p. 56.

3 J. Loeb, Biochem. Zeitschr., 1906, Vol. i, p. 200.

4 A. P. Mathews, BIOL. BULL., 1906, Vol. n, p. 137.
6 E. G. Spaulding, BIOL. BULL., 1904, Vol. 6, p. 224.

123

124 FRANCIS MARSH BALDWIN.

fertilization. He found also in acid and salt solutions (pure
isotonic KCL and NaCL) a similar but less clearly denned rhythm
of susceptibility. Eggs subjected to heat, electrical stimulation
and hypertonic sea-water behave in a similar manner. Thus,
Lyon,1 observed that the eggs were most resistant to heat at a
time previous to the first cleavage, and were most readily injured
at the time of division. A. R. Moore2 finds that the resistance
to hypertonic sea-water is least "immediately before and during
each cytoplasmic division, and that the maximal resistance is
shown 35 to 45 minutes after fertilization and just after each
division." More recently Lillie3 (1916), has made an extensive
study of the rhythmical changes in the resistance of the dividing
sea-urchin egg to hypotonic sea-water, and has discussed the
physiological significance of this rhythm. His experiments
show clearly that at or about the time of formation of the cleavage
furrow, a marked decline takes place in the resistance of the egg
to hypotony, and cytolysis is then rapid and complete. After
the cleavage furrow is fully formed the original resistance returns.
A similar reversible decline of resistance takes place at the second
and third cleavage, and is probably general for mitotic cell-
division. The minimum of resistance is found during the for-
mation of the furrow. Both the decline and the return of resis-
tance are rapid, the greater part of each phase occupying four
to five minutes. Some increase of susceptibility is apparent
ten or twelve minutes before the first appearance of the furrow.
Similar observations have been made by Herlant4 in the egg of
Paracentrotus lividus.

From such experiments it appears that the resistance of the
eggs to a variety of injurious agencies is least at the time when
they are undergoing rapid change of form. To account for
these rhythmical changes in the physiological state of the egg,
Lillie4 (1909) puts forward the hypothesis that they are essentially
the result of variations in the physical condition, especially the
permeability, of the surface-film of plasma-membrane, the latter

1 E. P. Lyon, Atner. Jour. Physiol., 1904, Vol. 11, p. 52.

2 A. R. Moore, BIOL. BULL., 1915, Vol. 28, p. 257.

3 R. S. Lillie, Jour. Exper. Zoo/., 1916, Vol. 21, No. 3, p. 401.

4 M. Herlant, Comptes rendus d. I. Societe d. Biologic, 1918, Vol. 81, p. 151.
6 R. S. Lillie, BIOL. BULL., 1909, Vol. 17, p. 207.

PHASES OF DIVIDING SEA-URCHIN EGG. 125

undergoing a reversible increase in permeability at the time of
cleavage. If a rhythm of alternate increase and decrease of
permeability accompanies the rhythm of the mitotic process, it
seems logical to infer that the entrance of solutes into the cell
would occur most readily when there is a loss of semi-permea-
bility. Accompanying this change would be a decrease of the
electrical surface-polarization, and this in turn probably would
alter the metabolic processes, especially oxidations within the
cell. Cell metabolism then is inseparably bound up with cell-
permeability; and the plasma-membrane, or semi-permeable
surface-layer is something more than a haptogen membrane (to
which it has frequently been compared). In discussing this
subject in a later paper, Lillie1 makes it especially clear that this
''general characteristic of semi-permeability (the all-essential
insulating and diffusing-preventing property) is not merely the
result of a special chemical composition and structural density,
such as determine the semi-permeability of a precipitation-
membrane, but is inseparable from the living condition, i.e., is
actively maintained by a continual process of metabolism. The
proof of this is that death — the cessation of metabolism — how-
ever caused, is invariably followed by a loss of semi-permeability,
i.e., the normal state of the membrane then ceases to be main-
tained and the unhindered processes of diffusion lead to the
disintegration of the cell. Hence destruction of the surface-
layer by artificial means — cytolytic substances, heat, extensive
mechanical injury — is quickly fatal to all cells."

In the experiments about to be described, I have studied the
behavior of fertilized Arbacia eggs when subjected for definite
brief lengths of time to various concentrations of some of the
higher alcohols — anyl, hexyl, heptyl, octyl and capryl — at differ-
ent periods of the cell-division cycle. This work was undertaken
at the Marine Biological Laboratory, at Woods Hole, Mass.,
during the past summer at the suggestion of Professor Ralph
Lillie, to whom the writer expresses his hearty thanks for many
kind suggestions and directions during its prosecution.

1 R. S. Lillie, Amer. Journ. Physiol., 1918, Vol. 45, No. 4, p. 406.

126 FRANCIS MARSH BALDWIN.

EXPERIMENTATION.

In order to procure a sufficient number of eggs for each series
of experiments, between one and two dozen large females were
opened, and their eggs collected into finger bowls. By successive
washing and settling, a uniform mass of mature eggs was obtained,
which could be inseminated and divided into two parts ; one to be
used for the control, and the other for the experiments. It was
found early in the work that the success of the experiments de-
pended upon having batches of eggs which were sufficiently
mature and uniform, so that all eggs reached successive stages in
their development at practically the same time. It was also
found that great exactness in the time-relations of the operations
was absolutely essential, and that any variation once entered
upon was sufficient to make the results worthless from a com-
parative standpoint. Usually two series of experiments were
started in a day; one in the morning to be carried over to the
gastrula stage by the following morning, and one in the afternoon,
to be examined the following afternoon. After extended pre-
liminary experimentation, it was found convenient, in any one
series, to keep the time of exposure constant and to vary the
concentration of substance used, although in a considerable
number of experiments the opposite procedure was adopted, i.e.,
the time was varied and the concentration kept constant.

Practically the same procedure was observed throughout the
entire experimentation. At each of the successive intervals
after fertilization, usually ten minute intervals, about one half
of medicine pipette containing a suspension of the inseminated
eggs was placed in a small corked Erlenmeyer flask, containing
50 c.c. of the solution of the alcohol in sea-water, and allowed
to remain for the time of exposure chosen (usually five minutes).
After the given time had nearly elapsed, the excess liquid was
poured off, and the eggs with a little of the liquid were placed in
a watch glass and the immediate results of the treatment were
observed under the low power of the microscope. At the ter-
mination of the time of exposure, the watch glass containing the
eggs was carefully immersed in a large volume of sea-water in a
finger bowl and the water was changed several times to rid it of
the excess substance. Finally the eggs were very carefully

PHASES OF DIVIDING SEA-URCHIN EGG. 127

washed with a stream of water from the medicine-dropper, and
set aside to undergo development. The proportion proceeding
with development to the free-swimming larval stage was sub-
sequently determined. It was found that the estimate of the
proportion surviving to the blastula stage was more readily and
exactly made, if the watch glass containing the eggs was removed
from the bowl of sea-water just before the free-swimming larval
stage was reached. Thus all survivors could be confined within a
small volume, and the count or estimate easily made. As a rule,
the experiments were carried only up to about the time of second
cleavage; since the evidence indicates that the same variation of
susceptibility occurs in each cell division cycle; moreover diver-
gencies between the different eggs in any lot become more pro-
nounced as time elapses, and it is important that all eggs of a lot
should be in the same physiological state at the time of treatment.
At first several preliminary experiments were necessary in
order to determine the most suitable range of concentrations to
be used, since the time of exposures determined upon were brief,
the longest being ten minutes; in some cases of exposures only
three minutes were used. In this connection, the tables given
by Lillie1 in his paper on the action of various anaesthetics in
suppressing cell-division in sea-urchin eggs, were exceedingly
helpful. For i-Amyl2 alcohol, he finds 0.45 to 0.4 vol. per
cent, a favorable anaesthetic concentration for eggs subjected
for two and one half hours, while 0.5 vol. per cent, and above are
somewhat rapidly toxic. For Capryl3 alcohol he finds the anaes-
thetizing concentrations to range between 0.012 and 0.02, and
notes that even in sub-anaesthetic concentrations this alcohol
exhibits a relatively high specific toxicity. With the help of
these data, and also Fiihner's4 observations showing that in a
series of monohydric aliphatic alcohols each member of the group
is from three to four times as effective (for equimolecular con-
centrations)5 as its immediate predecessor, it became a compara-

1 R. S. Lillie, Journ. Biolog. Chem., 1914, Vol. 17, No. 2, pp. 129-139.

2 Cf. reference just cited; Table VIII., p. 135.

3 Cf. reference just cited; Table IX., p. 137.

4 H. Fuhner, Arch. f. exp. Path. u. Pharm., 1904, LII., p. 69.

6 Capryl alcohol used in exposures of five minutes seemed not to obey this
general rule, since in practically all experiments it was used in concentrations
nearly three times its computed strength. (See p. 137.)

128

FRANCIS MARSH BALDWIN.

tively easy matter to approximate the most suitable concen-
tration of each alcohol after the first had been determined.

AMYL ALCOHOL.

Summarizing briefly the results of preliminary observations it
was found that the most satisfactory concentration of i-Amyl
alcohol when used with exposures of three to eight minutes, was
between 0.7 and 0.9 vol. per cent. Solutions of this strength
are sufficiently toxic to prevent many but not all of the eggs
thus treated from developing to a larval stage. Solutions weaker
than 0.7 vol. per cent, permit practically all eggs to proceed to

i -AMU. ALCOHOL.

PLOT OP THE CUHVB OP SUSCEPTIBLE AND BESSHm* PRASTO OF THE
MT1STHO SEA-UBCH1H BOC3 WHEtl SUBJECTED TO 0.8 VOL PEH OETfT i-AMYL
AHJOWL FOB EfGffr HIHUTB AT SUCESSIVE TEH MMDTB

the blastula stage with these exposures, with little appreciable
difference. At i.o vol. per cent, not more than 20 per cent, of
eggs form free-swimming blastulae even when exposed at the
period of highest resistance; (e.g., 30 to 40 minutes after fertili-

PHASES OF DIVIDING SEA-URCHIN EGG.

129

zation) and above this concentration the toxicity is such that
the decline in survivals is very rapid. In 1.25 vol. per cent,
solutions, all eggs are killed in all stages with exposures of nine
minutes.

Table I. summarizes the results of a typical series of experi-
ments with i-amyl alcohol. This particular series of experiments
was started in the afternoon of July 16, and the observations
noted in the third column were carried over into the morning of

TABLE I.

I-AMYL ALCOHOL.

July 16, 1:45 P.M. The fertilized eggs were placed at the intervals after
fertilization noted in column i in 50 cc. of 0.9 vol. per cent i-amyl alcohol. Al
exposures except the first (i) (6 minutes) were of eight minutes duration.

Intervals After
Fertilization.

Observed Condition of the Eggs at the
Time of Removal from Sol.

Proportion Forming Blastulse and Con-
dition of Remaining Eggs Next Day.

(1) 3- 9 m.

(2) 10-18 m.

(3) 20-28 m.

(4) 30-38 m.

(5) 40-48 m.

(6) 50-58 m.

(7) 60-68 m.

(8) 70-78 m.

(9) 80-88 m.

Fertilization-membranes well
formed. No marked cytolytic
change noted.

No marked change in appearance.
Uniform.

Membranes markedly swollen in
some cases. Slight fading of
pigment.

A few cells plasmolyzed, other
membranes markedly swollen.

No marked change. Faint indi-
cation of cleavage furrow in a
few scattered cells.

About 65 per cent, have entered
the two-celled stage. Some
show shrinkage.

About 90 per cent, in two-celled
stage. Others intact.

Practically all in two-celled stage.

A few are starting second cleav-
age furrow.

About 5 per cent, form free
swimming blastulae. Consider-
able numbers cytolyzed.

About 10 per cent, form blastulae.

Not so badly cytolyzed; most
cells intact.

Nearly 30 per cent, free-swim-
ming blastulae. Most eggs
intact.

Between 30 and 40 per cent,
free-swimming blastulae. Most
others intact but swollen.

Large majority (75-80 per cent.)
form swimming blastulae.

Relatively few (less than 10 per
cent.) form surviving blastulae.

Between 30-35 per cent, form

blastulae.
Few (15-20 per cent.) form

blastulae. Others intact.
3-5 per cent, form blastulae.

Others intact.

the following day. A similar series of experiments performed at
about the same time with the same alcohol in somewhat lower
concentration (0.8 vol. per cent.), but with slightly longer
(lo-minute) exposures yielded substantially the same results.
On the following day experiments were carried out on eggs
subjected to i.i vol. per cent, solutions with only brief (3-, 4-
and 5-minute) exposures with the results noted above.

130 FRANCIS MARSH BALDWIN.

In the controls about one half of the eggs were in the two-
celled stage at fifty-three minutes after fertilization, and at
sixty-five minutes between 85 and 90 per cent, were divided.
There is a definite period of well-marked susceptibility imme-
diately following fertilization; the susceptibility then gradually
and progressively declines up to the end of forty-eight minutes
(just before the first cleavage). There then follows a very sus-
ceptible period just at the time of cleavage. Later the resistant
phase reappears until about the time of second cleavage. If
the time intervals are plotted as abscissae, and the percentage of
surviving blastulae as ordinates, the relationships may be repre-
sented in the curve shown in Fig. I.

HEXYL ALCOHOL.

In exploring the range of suitable concentrations for hexyl
alcohol, the next higher member of the series, assuming that it
should be approximately three times as effective as i-amyl alcohol,
three preliminary experiments were performed. For these, solu-
tions of o.i, 0.25 and 0.30 vol. per cent, were used respectively.
The time of exposure was shortened to five minutes, for the
reason that it was thought the concentrations were, if anything,
a little above the optimum. The results clearly showed that
the solutions of o.i vol. per cent, was not sufficiently toxic to
demonstrate any variation of susceptibility in the eggs, since at
whatever period they were exposed practically all eggs survived
to the free-swimming blastula stage. On the other hand, the
two higher concentrations proved too toxic, so that practically
none of the eggs continued their development after subjection
to these solutions at any period. The 0.25 vol. per cent, solution,
although it suppressed further development, was not quite
intense enough in its action to cause cytolysis in the eggs, with
few exceptions. The 0.30 vol. per cent, concentration caused
very evident cytolysis, and rupture was almost universal.
Accordingly, series of experiments were carried out to test the
various concentrations between o.i vol. per cent, and 0.25 vol.
per cent. Two of these experiments are summarized in Table II.,
and may be regarded as typical.

These results show a much less definite evidence of a rhythm of

PHASES OF DIVIDING SEA-URCHIN EGG.

131

TABLE II.

HEXYL ALCOHOL.

August n, 10:15 A.M. Fertilized Arbacia eggs were placed at intervals noted
in 50 cc. of 0.13 and 0.17 vol. per cent, of Hexyl alcohol respectively, and allowed
to remain in them for five minutes. They were then placed in watch glasses
quickly observed, and treated as described in the previous experiment.

Intervals
After Ferti-
lization.

(A) 0.13 Vol. Per Cent.

(B) 0.17 Vol. Per Cent.

Observed Condition.

Proportion Forming
Blastulae.

Observed Condition.

Proportion Forming
Blastulse.

(i)i5-20m.

Fert. membrane

Majority (90 per No marked cyto-

Between 70 and

well formed. No

cent.) form

lysis. Uniform

80 per cent, form

marked cytoly-

blastulae.

batch of fertile

blastulae. Other

sis noted.

eggs.

cytolyzed but

intact.

(2) 25-3om.

No marked

Between 85—90

Fert. membrane

70-75 per cent.

change.

per cent, form

in most eggs

blastulae. Few

swimming blas-

swollen; no

ruptured.

tulae

great change

otherwise.

(3)3S-4om.

Membrane swol-

Practically all

No marked

90 per cent, form

len, few show

(90 per cent, or

change noted.

blastulae.

slight plas-

over) form blas-

molysis.

tulae. Others

intact.

(4) 45-Som.

Slight loss in igo per cent, form

Slight fading of

Between 65-70

pigmentation.

blastulae. Few pigment. No

per cent, form

No marked cy-

ruptured.

other marked

blastulae. Few

tolysis however.

change.

scattered cells

ruptured.

(5)55-6om.

About half in

80-85 per cent. All intact, 50 per Nearly 60 per

two-celled

form blastulae. • cent, or over in cent, form bias-

stage. No

Others mostly

two-celled tulae.

marked change.

intact.

stage.

(6) 65-7om.

Fully 85 percent.

About 76 per

No change.

Not more than 50

in two-celled

cent, form

per cent, form

stage ; no

swimming blas-

blastulae. Others

marked change

tulee.

badly cytolyzed

from preceding.

but mostly in-

tact.

(7)75-8om.

No marked

Between 65—70

Slight loss of

About 60 per

change.

per cent, swim-

pigment. No

cent, blastulae.

ming blastulse;

marked change.

Two-celled eggs

others badly

conspicuous.

cytolyzed.

Most others

intact.

(8)85-9om.

Few cells (ca.

About 60 per

Few cells in

Nearly 85 per

5-7 per cent.)

cent. form

second cleavage.

cent, swimming

show second

swimming blas-

No marked cy-

blastulae.

cleavage. No tulae. Most

tolysis.

marked change, others intact.

susceptibility than those just described for i-amyl alcohol.
The eggs apparently maintain throughout the cycle a relatively
high resistance to the concentrations of hexyl alcohol here
used, with only a slight increase of susceptibility at the time of

132

FRANCIS MARSH BALDWIN.

first cleavage; in the one case (0.17 vol. per cent.) there is evidence
of a slight return of resistance just afterwards, and in the other
(0.13 vol. per cent.) there is not. Why there should be this
difference in the behavior of the two alcohols is difficult to explain.
The exposures to hexyl alcohol were perhaps insufficiently

PER
CEHT
OF

HEFTYL ALCOHOL.

PLOT OF THE CUHVE OT SUSCEPTIBLE SND RESISTANT FHA3E3 OT THE
3EA-ORCHIH COOS WMJ1 SUBJECTED TO 0.07VOL.FER CENT nEFTYL ALCOHOL FOR 1TVE
MINUTES AT TO1 MJMUTE INTERVALS BEOINWHG THIRTY MIHOTE3 AFTER rERTTLIZATlOM.

SURVIVORS. THE PER CEHT OF SORYJVIflG BLASTULAC IS PLOTTED XG/UN3T THE TIME IMTESVAL3.

USING ran convEnrm<z THE WE or UCH nvr MINUTE FKHIOD.

100

90

80

50

40

\

7

10

35

45
FIG. 2.

65

85 TIME.

prolonged to bring out a well marked differential effect at the
different stages of the cycle. Again, it is well known that certain
anaesthetics are less effective than others in suppressing the
cell-division process; also many neuromuscular responses react
differently to a given anaesthetic in different animals, and in
the same animal at different ages. Thus Lillie1 found chloretone
much less effective than chloral hydrate in suppressing cleavage
in Arbacia eggs. As regards the alcohols that he tried, he lists
propyl, butyl, amyl in the order of increasing favorability, while

1 R. S. Lillie, Jour. Biol. Chem., 1914, Vol. 17, No. 2, p. 130.

PHASES OF DIVIDING SEA-URCHIN EGG. 133

ethyl and capryl show a higher toxicity than the others.1 It
may be that differences, — both qualitative and quantitative—
in the lipoid elements of the tissues, and hence in the plasmamem-
brane, form the basis of the observed physiological difference.

In the controls which were running parallel to the two exper-
iments just described, a majority of the eggs entered the two-
celled stage at about fifty-eight minutes after fertilization. At
sixty-five minutes after fertilization, between 85 and 90 per cent,
had cleaved. Practically all eggs in the controls had reached the
blastula stage the following day.

HEPTYL ALCOHOL.

A series of six experiments were performed with this alcohol
to determine the limits of suitable concentration. Reasoning
from the data in the proceeding, it was thought that the optimum
concentration would be in the proximity of between 0.04 and
0.07 vol. per cent., but to make sure, concentrations as low as
0.02 and as high as 0.08 vol. per cent, were used. As a matter
of fact, solutions of 0.06 and 0.07 vol. per cent, were found to be
the best suited to the experiments, although it is interesting to
note that even weaker solutions showed marked toxic action on
eggs at the time of formation of the first cleavage furrow, while
during the stages preceding and succeeding division, they had
relatively little influence. Table III. summarizes three exper-
iments with this alcohol, and is fairly typical of results obtained
in other experiments. By some unavoidable oversight in tech-
nical procedure, records for the first two periods (i.e., respectively
10 and 20 minutes after fertilization) were not obtained, but
from the results of other series it is evident that the eggs maintain
a rather high resistance at these times. The results given in
Table III. show that 0.07 vol. per cent, approximates the favor-
able concentration for this alcohol, while the solutions on either
side are slightly hypo- and hyper- toxic respectively; i.e., in the
one case practically all eggs survive to the blastula stage, and in
the other nearly all die. Recovery of resistance after the first
division appears to be relatively slow. When the percentage of
surviving blastulse is plotted against the time intervals, regarding

2 See ibid., p. 133.

134

FRANCIS MARSH BALDWIN.

•a

I

•d

. T3

to CL)

s? -2

> >-<

C CJ

OJ CO

-u <L>

.S "°

CU ' — '

c o

r--.

13

I B

ft
HI

^ >>

O (LI

O
I— I O

W <

M ^
•< H

H S
ffi

o I

M.s

.S E
_o >
"o ,

^ S.

O ^_;

10 o

B >

— oo

to

bo

B

II

2

5 S -o

y -M

-0

a'E

PH

«»a

n •<

•S S

JJS

•

§

g

"8 • -S

J2*J>

ft

cs ^

S „ .

.^4J

3cx2c

1

ill

C/3

-

'

ii

CO +j +J

ii

•— TO Cu

lli

G CO -+-»

III

ba

E

E

E

E

U-)

uo

l^

IO

ro

^t"

LO

O

i

i

<i,

ro

•*

to

o

^

V f

S

3

PHASES OF DIVIDING SEA-URCHIN EGG.

135

W

h-J

CM

•a -6

+1

CH * n |

N ^H

u 83 "^

>. 3

^ "5 c

.

"o o.

. 5 3

01

, i ™"

?Ti ^ ~*~*

a

^ 2

X o3 O*

m

° Hi

,0 2

d

~~rt

•c •£

r^e

.Q Cu

2 ""* &

<§'

3 °

£%6

d

^

*^

2 a
o

•3

o ^ £

"aj ~ S

^ ° >>

"S -a ^

•3

^ QJ 6/)

flj ^^

C

CJ *^^

CO ^ CH

o

U

o

.2 d ai -^

•u

£*•> ^»

.1 •*-* S

4_J ? •

•y M " a

1

~n "*-* i^ *^

I 'I ^ '^

O

U .^H CO ^-t-H

E

!l x1

o ^ e »1

^> 0) OJ >>

tn
d

N > ^ d

i.^-S-2

5

i- °w .a .13

OJ O ^J 4)

M ^ CJ

d

0 a. £ N 3

§j • ' ^ £

^ g^^> a.

-t-J (L* ^3 Cd rf

q

,| 0 0 *, ^

CJ O -4-t -4-> r^

pq

o

£ S

O 2 "O

aJ

M «

;|

"^ -y o

C

a

• "§) c

JS

o

"tn '"* "*^

O
•a

flj C/J fli

u .1

•o

u

il ? ,2r

rt

CX^ ^

E

O

U <-i-i
0^0

o

o\

^

>H <H i m j

D ^ a t. >>

0. 3 <u U

liil

O . J3 K^

^ G -d o

{A

_rt

|| §o1

CJ CJ .1

d
fc

i rt 3 : -d

§ j 8*2 S ^

CJ ^^ IH

^ g > ^ T^ /"^

JH >-i o ^

•*-* o oj ^ aJ O

? 3 ^ cJ ^3

VO

(U CJ <+H ^ J-, -J_>

4) +J X! C D

q

pq

fe

d

S c

jl"g

EC

o

ai o

3 O

•3

SJ S ^

c o

c

^ ^

0) m

o
U

•O
U

>>T3 "o
*J D *j

C > >>

2 >

ri o

£.fi fe
- " 2

1

O^ aJ u

O

S

^ .*-

CO

S

S

"5

10

ir>

J>*

oo

fc

i

A

c

». — f

03

136

FRANCIS MARSH BALDWIN.

as typical the data of the 0.07 vol. per cent, solution, an interest-
ing curve is obtained (Fig. 2) which is fairly comparable with
the one shown for i-amyl alcohol. There is a gradual rise in
resistance up to the period of first cleavage, with a sharp drop
during cell-division followed by a slow recoverry.

OCTYL ALCOHOL.

Normal octyl alcohol is apparently considerably more toxic
than its isomere capryl alcohol. In a series of five experiments
with octyl alcohol in concentrations ranging from o.oio to 0.030
vol. per cent., the best concentration for five minute times of
exposure was found to be in the neighborhood of 0.015 VOL Per
cent. On the other hand the outcome of fourteen experiments
with capryl alcohol showed the optimum concentration for the
same time of exposure to be between 0.035 and 0.045 v°l- Per
cent., which is between two and three times the favorable con-
centration of normal octyl alcohol. Table IV. summaries a

TABLE IV.

NORMAL OCTYL ALCOHOL.

Fertilized eggs were subjected for five minutes to 0.013 vol. per cent, of normal
octyl alcohol at intervals of ten minutes.

Intervals After
Fertilization.

Observed Condition on Removal from
Fluid.

Observed Condition the Following
Day.

(1) 15-20 m.

(2) 25-30 m.

(3) 35-40 m.

(4) 45-50 m.

(5) 55-60 m.

(6) 65-70 m.

(7) 75-8o m.

(8) 85-90 m.

Fertilization membrane well
formed. Slight loss of pigment.

No noticeable cytolysis although

very marked loss of pigment.
Decided loss of pigment. No

marked change in membrane

or cytoplasm.
About 2 per cent, show first

furrow. Slight loss pigment.

No cytolysis.

Over half in first cleavage.
About 90 per cent, in two-celled

stage.

Aside from loss of pigment no
noticeable change.

A few (i per cent.) just begin to
show second cleavage furrow
No marked change in appear-
ance.

Nearly 50 per cent, active blas-

tulse. Others badly cytolyzed,

few ruptured.
About 65 per cent, active blas-

tulse. Others cytolyzed.
Nearly 80 per cent, active.

Others intact.

Practically all active blastulae.

About 85 per cent, active.

Almost 60 per cent, active blas-
tulte, numbers of two-celled egg
present and mostly intact.
Some badly cytolyzed and
ruptured.

Between 65 and 70 per cent,
active blastulae. Most others
cytolyzed but intact.

Between 85 and 90 per cent,
active blastulse. Others cy-
tolyzed but intact.

PHASES OF DIVIDING SEA-URCHIN EGG.

137

typical experiment using normal octyl alcohol of 0.013 vol.
per cent, concentration. For exposures of five minutes duration,
this concentration gave the best results, and showed very clearly
the resistant and susceptible phases.

CAPRYL ALCOHOL.

As mentioned before, experiments with various concentrations
of capryl alcohol showed that for brief exposures, the most
favorable concentration was nearly three times that of normal
octyl alcohol. This may perhaps be accounted for in some
measure by the fact that not all samples of capryl alcohol are

ree wpRn. ALCOHOL.

CEffT TLOT OF CDHVES HHH1 US1HO 0 035 and 0.045 TOL. FER CEHT

OF CJ1PRT1. ALCOHOL FOR FIVE MIHUTB EXPOSURES AT SUCCESSIVE TEN

SOSVIVOBS. HHHTEE IHTEHFAlS.

10

83 TIME.

uniform in chemical composition and purity; a slight difference
in this respect is known to make a decided difference in its
chemical and physiological activity. In suitable concentrations,
this alcohol is without doubt one of the most satisfactory for

138

FRANCIS MARSH BALDWIN.

showing susceptible and resistant phases in dividing eggs.
Several experiments were tried with this alcohol in which the
concentration was kept constant and the time of exposure was
varied; and from the data thus gathered, it seems probable
that of the two factors, concentration is the more important.
In other words, if the concentration is such that it gives the best

TABLE V.

Fertilized eggs were subjected for five minutes at ten minute intervals to 0.035
and 0.045 v°l- Per cent, capryl alcohol.

Intervals.

(A) 0.035 Vol. Per Cent.

(B) 0.045 Vol. Per Cent.

Observed Condition.

Condition Follow-
ing Day.

Observed Condition.

Condition Follow-
ing Day.

(l) IO-I5IT1.

Membranes well

75-80 per cent.

No marked dif-

About 65 per

formed. No

active. Others

ference from

cent active.

great difference

cytolyzed but

normal eggs.

Others badly

from normal

mostly intact.

cytolyzed.

eggs.

(2)20-2501.

Membrane

About 83 per

No marked

Nearly 85 per

slightly swollen.

cent, active.

change.

cent, active.

No marked

Others badly

Others intact.

cytolysis.

cytolyzed.

(3)30-35m.

Few show loss

About 92 per

Marked fading

Between 85-90

of pigment, cy-

cent, active.

of pigment. No

per cent, active.

toplasm shrunk-

Others intact.

marked cytoly-

Some ruptured

en in some cases.

sis.

eggs motile.

(4) 40-45m.

No marked

About 95 per

No marked iAbout 70 per

change. None

cent, active

change, slight

cent active.

show cleavage

blastulae.

loss of pigment.

Others mostly

furrow.

Others intact.

intact.

(5)50-55m.

Nearly half show

80 per cent, ac-

No marked

About 50 per

first cleavage

tive blastulae.

change.

cent . active

furrow No

Few ruptured.

blastulae.

marked change.

Most others

badly cytolyzed,

many ruptured.

(6) 60-6501.

Nearly all show

About 50 per

Practically all in

About 40 per

first cleavage

cent, active

two cell stage.

cent, active. All

furrow. No

blastulae.

Some loss pig-

others badly

marked cytoly-

Most others

ment.

ruptured. Some

sis.

ruptured.

still in two cells.

(7)?o-75m.

No marked

About 30 per

No marked

Nearly 30 per

change. Few

cent, active

change, except

cent, active.

scattered cells

blastulae.

loss of pigment.

Others badly

show second

Others badly

cytolyzed. Few

furrow.

cytolyzed.

persist in two

cells.

(8)80-85111.

No marked

Nearly 80 per

No marked

Nearly 50 per

cytolysis.

cent, active

change.

cent . active

blastulae.

blastulae.

Others mostly

Others cytoly-

intact.

zed but mostly

intact.

When plotted the data gives interesting curves as shown in Fig. 3.

PHASES OF DIVIDING SEA-URCHIN EGG. 139

results with a five-minute exposure, when the exposure is pro-
longed to eight minutes, very little or no difference is detected.
This generalization, however, could probably be applied only
within narrow limits.

The data from two experiments using capryl alcohol in 0.035
and 0.045 v°l- Per cent, concentrations respectively are given in
Table V. These records are fairly typical of results of other

experiments.

SUMMARY.

1. The developing sea-urchin egg when subjected to suitable
concentrations of various lipoid-soluble substances — i-amyl,
hexyl, heptyl, octyl and capryl alcohols — shows unmistakable
rhythms of susceptible and resistant phases, which when taken in
connection with the earlier observations of Lyon, Herlant,
Mathews, Spaulding, Lillie and others, constitute additional
evidence that a very intimate relation exists between the
general physiological condition of the egg, and the physical state
of its plasma-membrane.

2. During the first ten or fifteen minutes after fertilization
the eggs are more susceptible than at any other time until the
period just preceding division. A comparatively resistant phase
gradually becomes more and more marked up to just before the
first cell-division (about 45 or 48 minutes after fertilization).
This is followed by a period of decidedly increased susceptibility
which lasts for about 15 or 20 minutes, during which time marked!
cytological effects are noted. Subsequently the resistant phase
is largely recovered, and maintained up to the time of the second
cleavage.

3. The most favorable concentrations of the various alcohols
for demonstrating the rhythm of susceptibility range as follows:
i-amyl, between 0.7 and 0.9 vol. per cent.; hexyl, between 0.13
and 0.17 vol. per cent.; heptyl, between 0.06 and 0.07 vol. per
cent.; normal octyl, about 0.015; while capryl was considerably
above its isomere (normal octyl) between 0.035 and 0.045 vol.
per cent. The best records were obtained in experiments using
i-amyl and capryl alcohols, possibly indicating a higher specific
toxicity of these when compared to the others.

4. When suitable concentrations were used, no marked

140 FRANCIS MARSH BALDWIN.

differences could be detected by varying slightly the durations ot
exposure. Eggs exposed for five, eight or even ten minutes to the
same concentration gave similar results. This, however,
would probably apply only within narrow limits.

LIGHT PRODUCTION IN CEPHALOPODS, I.

AN INTRODUCTORY SURVEY.
S. STILLMAN BERRY,

REDLANDS, CALIFORNIA.

CONTENTS.

PAGE

1. Introduction 141

2. Classification of Cephalopods 144

Tetrabranchiata 144

Dibranchiata 144

Octopoda 145

Decapoda 145

Synoptic Tables 146

3. Distribution of the Photogenic Function Among Cephalopods 147

Preliminary Remarks 147

Synopsis of the Genera and Species 149

Summary of Recent Cephalopoda 157

Summary of Preceding Tables 158

Recapitulation of Data 158

4. Actual Observance of the Phenomenon 159

i. INTRODUCTION.

Recent interest in the subject of biophotogenesis has been
so great, and bids so fair to continue at high ebb, until at least the
problem of the economical artificial production of chemical light
has been solved, that for the use of the many classes of investi-
gators, most of whom are not zoologists and can scarcely be
expected to possess accurate taxonomic knowledge of the group
with which they may chance to wish to work, it would be exceed-
ingly desirable if there could be placed on record in compact form
a summary of all the species of each principal division of plants
and animals which are known or thought to possess photogenic
properties. The writer's desire to see this service performed
on behalf of the Cephalopods, animals which must always stand
well up with the highest in the estimation of the student of or-
ganic light, furnished the initial stimulus which has finally
broadened into the production of the present paper. As a
taxonomist, however, and, in so far as this particular group of

141

142 S. STILLMAN BERRY.

animals is concerned, one of that despised species, the "closet
naturalist" —he can only go a certain way with his subject, and
by the same token, his remarks must perforce have only a very
limited value. Yet the effort seems worth while spending, and
he can fairly plead in extenuation of his temerity, if not of his
own limitations, that cephalopods are such active, delicately
balanced creatures, and so exquisitely adjusted to an environment
in which it is next to impossible to observe them accurately, and
which it is even more vain to attempt to establish, even partially,
under artificial conditions, that the difficulties of subjecting the
details of their life history and ecology to that searching exami-
nation required by the standards of modern biological investi-
gation have proven practically insurmountable. Therefore the
unfortunate circumstance that we have no specialist in this
branch, — no authoritative student of the bionomics of cephal-
opods, and that such halting summarization as can be done must
be handled by the systematist or general student of the group,
if at all. Admitting then the largely pragmatic and temporary
rather than permanently intrinsic value of the present dissertation
we may proceed with it, for even so small a contribution as this
can pretend to be should prove helpful.

Amid the wealth of remarkable features, structural and physi-
ological, with which the entire group of the Cephalopoda entices
the student, the variety and multiplication of those which in
an earlier day would have been as unquestioningly as delightedly
hailed as adaptive are supremely conspicuous. These are special
for the most part to the conditions and vicissitudes brought
about by an exceptionally active manner of life in an environ-
ment full of actual or potential diversity. Not even the fishes
are better swimmers, nor, with all their aristocratic vertebrate
organization, lead a more complicated struggle for existence.
It is perhaps concomitantly both a result and cause of all this
that the group so fairly teems with bizarre cells, tissues, organs,
complexes of organs, which, as we may as well admit without
further parley, can scarcely be interpreted otherwise than as
marvelously exquisite adaptations, each to its own definite end.
Such knowledge of most of these as we possess has been amassed
almost wholly since the time of Darwin, else the pages of the

LIGHT PRODUCTION IN CEPHALOPODS. 143

"Origin" might have been enriched by many examples as start-
ling in their way as any of the classical ones. The complex
cephalopod chromatophore, the inter-playing system of exactly
balanced musculature with scarcely any hard skeletal parts to
give it support and leverage which goes to make up the arms and
each single sucker, the delicate adjustment between eye, sucker
and chromatophore through the mediation of the nervous
system to result in one of our most perfect demonstrations of
concealing coloration, the innumerable types of hectocotylus
often involving the most astonishing modifications in sexual
behavior, the amazing and still insufficiently understood mech-
anism of the spermatophore, the eyes, and, without attempting
to prolong the list further, the photogenic organs, — each is in
its own way a triumph of adaptive development, how much so
we may perhaps infer to some extent from the widespread
occurrence of these structures in one form or another among
nearly all the now-surviving cephalopods. Continuing with the
structures last named, for instance, I think it can be truly said
that no other class of animals can compare with the cephalopods
in the complexity, diversity, beauty, brilliancy, — in brief, the
high specialization of organs devoted to the production and utili-
zation of that form of energy which to our human faculties finds
expression as light.

It has been said with considerable show of truth that the
generation of light by the plasm of animals and plants is really
far less to be marveled at than the transformation of their
energy into motion. But motion is practically a general prop-
erty of protoplasm in all its forms, without which it could scarcely
exist as living substance at all. The reason why the production
of organic light appears so remarkable to most of us is more
special: it is partly of course because of the apparent economy
of this light so far as the dissipation of heat energy is concerned,
but mainly to the average observer because of its evident highly
specialized adaptation to certain particular ends.

Dubois wrote in I8Q5:1 'The most resplendent of all animals
are insects, of which class the glowworm, beloved of the poets, is
one of the most brilliant examples." Cephalopods were scarcely

Smithsonian Report, 1895, p. 418.

144 s- STILLMAN BERRY.

noticed as being luminous at that period, but now we know that
the firefly and glow-worm pale in comparison, and that probably
not even the brilliant display of the tropical elaters can vie with
the gorgeous pyrotechnics of certain squids. It is indeed quite
possible that the latter exhibit the highest development of the
photogenic function known in the entire animal kingdom.

2. CLASSIFICATION OF CEPHALOPODS.

Living members of the Molluscan Class Cephalopoda cleave
simply and naturally into two well-defined, easily separable
groups. The first of these, and that universally regarded as
the most primitive, is the Order Tetrabranchiata, comprising
only the few species contained in the single genus Nautilus.
The animals of this group are characterized especially by the
possession of a massive, chambered, external shell; of a "funnel"
formed by the appression of two lateral folds which remain
unfused in the median line below; of a system of suckerless
lobes around the mouth, bearing retractile, annulated tentacles;
of such traces of metamerism as the presence of two pairs each
of ctenidia, "branchial hearts," auricles, renal organs, and
osphradia; and of a simple "pin-hole" eye, open to the exterior.
Ink-sac and chromatophores are absent.

The second group, Order Dibranchiata, comprising all living
cephalopods except Nautilus, is characterized by either the
complete atrophy of the shell or its reduction in the adult to a
concealed loose coil (Spirulidae), a calcareous plate (Sepiidse), or
a horny pen; by an entire, tubular funnel; by the development of
the anterior portion of the primitive foot into a series of eight or
ten muscular, sucker-bearing, tentacle-like arms about the
mouth; by but a single pair each of ctenidia, branchial hearts,
auricles and renal organs; by the presence, at least typically, of
highly developed eyes; by the development of complex, special-
ized pigment cells in the skin; and by the presence of the peculiar
"ink-sac." Osphradia are absent.

The Dibranchiata in their turn are sharply divisible into
two subgroups, the Decapoda and Octopoda. The former are
mainly pelagic; have finned, generally elongate bodies; have not
only the eight "primary" arms of the octopods, but a pair of

LIGHT PRODUCTION IN CEPHALOPODS. 145

specially modified "tentacles" as well; have the suckers pedun-
culate, reinforced with a chitinous ring, and often very curiously
modified; and have a so-called buccal membrane surrounding
the mouth.

The Octopoda are principally shore or bottom-loving forms;
short bodied; lack fins, or have them only secondarily developed;
have eight arms only, with their suckers sessile and lacking
chitinous rings; and lack the buccal membrane. There are also
important internal characters which need not concern us here,
but it may be said that few living groups are more sharply de-
limited. As a whole the group of the Decapoda has seemed to
most students more archaic than the Octopoda, at any rate is
less uniformly divergent from what must have been the ancestral
stock, but it includes many highly specialized types, and prob-
ably neither group as we now know it can be taken a<s especially
"primitive."

Of the two groups the Decapoda are much the less uniform,
and are therefore still further to be divided. The classical
bifurcation, which has found general acceptance until quite
recently, is that of d'Orbigny, into CEgopsida, or those forms in
which there is a free eyelid, and the Myopsida, in which there is a
simple fold-like pseudo-lid, or the skin passes uninterruptedly
over the eyeball. There seems to be no doubt but that the
CEgopsida, at least, form a monophyletic, natural group, for on
this point there is reasonable agreement. This cannot be said
of the Myopsida.

The principal alternative system is that of Naef (1916). l
He sets apart the Sepias and Sepiolids together as one of the two
main subgroups, the Sepioidea. The other group, which not alto-
gether satisfactorily he denominates Teuthoidea, is subdivided
into Myopsida (here restricted by the elimination of the Sepioidea
and hence comprising only the single family Loliginida?) and
CEgopsida, the latter as outlined above. Naef's classification
avoids several very serious difficulties involved in the standard
arrangement, but encounters, perhaps, certain others, which
need not be dwelt upon here.

For convenience the two systems are summarized in the

1 Naef, A. J., Pnbblicazione Slazione Zoologica Napoli , V. i, pp. 14-17, 1916.

146 S. STILLMAN BERRY.

accompanying table, that of Naef being slightly modified from
his printed synopsis in order better to serve the purposes of the
moment.

STANDARD SYNOPSIS OF THE CLASS CEPHALOPODA.
Phylum MOLLUSCA.
Class CEPHALOPODA.

Order I. Tetrabranchiata.

Suborder i. Nautiloidea.
Order II. Dibranchiata.
Suborder i. Decapoda.

Division A. CEgopsida.
Division B. Myopsida.
Suborder 2. Octopoda.
Division A. Pteroti.
Division B. Apteri.

SYNOPSIS OF THE CLASS CEPHALOPODA ACCORDING TO NAEF.

Phyltim MOLLUSCA.
Class CEPHALOPODA.

Sub-class I. Tetrabranchiata.

Order i. Nautiloidea.
Sub-class II. Dibranchiata.
Order I. Decapoda.

Suborder A. Teuthoidea.

(a) Teuthoidea myopsida.

(b) Teuthoidea cegopsida.
Suborder B. Sepioidea.

Order 2. Octopoda.

Suborder A. Pteroti.
Suborder B. Apteri.

In number of families and genera now living, the CEgopsida
easily preponderate, but the tremendous modern development
of the genera Polypus, Sepia and LoUgo throws the preponder-
ance in species over to the side of the Myopsids (+ Sepioids)
and Octopods. For instance, among recent Cephalopoda are
to be recognized some 32 families, and in round figures about
1 20 genera and 600 species.1 Of these the (Egopsid Decapods
claim 16 families (one half the total), 66 genera (slightly over
one half), and around 175 species (nearly one third). The
Myopsid (+ Sepioid) Decapods account for but 7 families
(nearly one quarter), and 27 genera (nearly one quarter), but

1 More critical figures compiled from the author's card register will be found
in Tables I., II. and III. to follow.

LIGHT PRODUCTION IN CEPHALOPODS. 147

include around 225 species (over one third of the total). The
Octopoda, with only 8 families (one quarter) and 25 genera
(slightly over one fifth), yet develop nearly 200 species (one
third of the total). In comparison to this the single family and
genus of Tetrabranchiata (3 generally recognized species) hold
a minor place in the fauna.

The extraordinary development of the CEgopsida in families
and genera is indicative, as the reader may anticipate, of the
more than usual modification to which the different branches of
this group have been subject, and we are to see that this is
particularly true of the photogenic organs. The number of
species on the other hand seems to have been held down by the
circumstance that a very considerable proportion of the genera
are pelagic types of widespread distribution, and, like so many
other animals showing somewhat similar ecologic relations, do
not break up well into species with our present degree of refine-
ment in perception.

This brief survey of the classification may seem a digression,
but without it as a guide, no consideration of the distribution
of the photogenic function within the group as a whole or within
its components could be entirely intelligible.

3. DISTRIBUTION OF THE PHOTOGENIC FUNCTION AMONG

CEPHALOPODS.

By no means all cephalopods are luminous. Among the entire
major division of octopods but two species, Melanoteuthis
lucens Joubin and Eledonella alberti Joubin, have been described
as possessing photogenic organs, while even here the fact that
the structures so described are actually designed for the produc-
tion of light still remains to be demonstrated. The various
instances where octopi have been observed to emit light are
almost always poorly authenticated, though it is not impossible
that in some cases, such as the observation by Darwin during
the Voyage of the "Beagle," which will be noted later, are ex-
plicable on the assumption of infection by photogenic bacteria
or protozoa. However that may be, and perhaps the point is
not yet definitely settled, in the morphological evidence offered
the two species mentioned stand quite alone.

148 S. STILLMAN BERRY.

On the other hand among the decapods photogenic properties
are so widespread that taking the class as a whole, and even
including the Octopoda, I am aware of no other major division
of metazoan animals which shows such a proportional develop-
ment of luminous species as the Cephalopoda.

It is therefore no wonder that the scattered literature, as
well as the fragmentary character of the information to be
gleaned therefrom, offers almost insurmountable difficulties to
the inquiring student, while even reasonably complete informa-
tion is scarcely to be found in any text-book or work of reference.
There seems in fact no more recent effort on the part of teutho-
logists to meet this need than that of Chun (:io), which seems
to have been carried out simply as an incident to the preparation
of his monumental work on the "Valdivia" CEgopsids, an
expensive and in the United States a relatively inaccessible
volume. Furthermore he dealt with but one group of cephalo-
pods, while even for this group there has now accumulated a
considerable mass of additional information. In the English
language by all odds the best and most trustworthy summary
is the brief one of Hoyle (:o8).

Probably the best way to convey an accurate idea of the man-
ner of distribution of the photogenic function within the group
is to present summarily a systematic survey of the entire class,
carried down at least as far as genera, and including at the same
time such appropriate supplementary data concerning the posses-
sion of this function as in each case is possible. An effort to do
this is constituted in the following synopsis. In all cases an
attempt is made to state as exactly as one writer can the number
of valid or recognized species in each genus, and, in the case of
photogenic forms, a full list of the species themselves, as well
as an indication of the situation of their photophores. In a
catalogue of this kind it must at the outset be admitted that
the figures given are only approximate, due to the fact that
entire elimination of the personal element is quite impossible,
even though the concrete numbers quoted are not mere estimates,
but represent in every instance an actual weighing of the validity
of each specific name in the light of all available information,
information which from the very nature of this paper cannot be

LIGHT PRODUCTION IN CEPHALOPODS. 149

gone into in any detail here. The attempt has been made to be
both conservative and judicial, but it of course goes without
saying that the result attained must be regarded as far from
final. Aside from matters of judgment even, it would be
presumptious to claim for this list either completeness or freedom
from error, but every effort has been made to reduce unneces-
sary mistakes to a minimum, and after all it is utility rather
than finality which must always be remembered as the end in
view. In certain special cases where the number of valid species
seems to be more than usually problematic, this circumstance
has been so indicated by the use of + or ± signs.

TABLE I.1

SYNOPTIC TABLE OF THE CLASS CEPHALOPODA, SHOWING THE
OCCURRENCE OF PHOTOGENETIC ORGANS.

Class CEPHALOPODA.
Order TETRABRANCHIATA.

Suborder NAUTILOIDEA.
Family NaulilidcB.

Genus Nautilus Linnaeus 1758. (No photogenic species known.)

3 or 4 species.

Order DIBRANCHIATA.

Suborder DECAPODA.

Division CEcopsiDA.
Superfamily Architeuthoidea.

Family Architeuthid.ee. (No photogenic species known.)
Genus Architeuthus Steenstrup, 1857.

14 species.

Superfamily Enoploteuthoidea.

Family Gonatids. (No photogenic species known.)
Genus Gonatus Gray, 1849.

3 species.
Family ONYCHOTEUTHID.-E.

Genus Onykia Lesueur, 1821. (No photogenic species known.)

12 + species.

Genus ONYCHOTEUTHIS Lichtenstein, 1818. (2 axial photogenic organs in
pallial chamber.)

1 established species, banksii; several doubtful.

Genus Tetronychoteuthis Pfeffer, 1900. (No photogenic species known.)

2 species.

Genus CHAUNOTEUTHIS Appellof, 1891. (Photogenic organs in ventral in-
tegument of mantle — Chun; no luminous organs — Pfeffer.)
i species: mollis.

1 Photogenic families and genera are printed in small capitals.

150 S. STILLMAN BERRY.

Genus Ancistroteuthis Gray, 1849. (No photogenic species known.)

1 species.

Genus Moroteuthis Verrill, 1881. (No photogenic species known.)

3 species.
Family LYCOTEUTHID^E.

Genus LYCOTEUTHIS Pfeffer, 1900. (Photogenic organs on eyes, in stalks of
tentacles, and in pallial chamber.)

2 species: diadema, jattai; possibly identical.

Genus NEMATOLAMPAS Berry, 1913. (Photogenic organs on eyes, in arms, in

stalks of tentacles, in pallial chamber, and at posterior tip of body.)
i species: regalis.
Family LAMPADIOTEUTHID^:.

Genus LAMPADIOTEUTHIS Berry, 1916. (Photogenic organs on eyes, in stalks

of tentacles, and in pallial chamber.)
I species: megaleia.
Family ENOPLOTEUTHID^.

Genus ENOPLOTEUTHIS d'Orbigny, 1844. (Photogenic organs on eyes and in
the integument of mantle, funnel, head, and arms, but almost entirely
confined to ventral aspect.)

3 species: leptura, chunii, galaxias.

Genus ABRALIA Gray, 1849. (Photogenic organs on eyes and in the integu-
ment of arms, head, funnel and mantle, but almost entirely confined to
ventral aspect.)
7 species: andamanica, armata, astrolineata, astrosticta, steindachneri,

trigonura, veranyi.

Genus ABRALIOPSIS Joubin, 1896. (Photogenic organs on eyes, at tips of
ventral arms, and in the integument of arms, head, funnel, and mantle,
but almost entirely confined to ventral aspect.)
5 + species: affinis, hoylei, lineata, morisii, owenii, pfefferi.
Genus WATASENIA Ishikawa, 1914. (Photogenic organs as in Abraliopsis.)

i species: scintillans.

Genus ENOPLOION Pfeffer, 1912. (Larval form; photogenic organs on ten-
tacle stalks and ventral integument of ventral arms, head, funnel, and
mantle.)

i species: eustictum.
Genus ASTHENOTEUTHION Pfeffer, 1912. (Larval form; photogenic organs on

eyes.)

I species: planctonicum.

Genus ANCISTROCHEIRUS Gray, 1849. (Photogenic organs in ventral integu-
ment of mantle.)
i species: lesueurii.
Genus THELIDIOTEUTHIS Pfeffer, 1900. (Photogenic organs on tentacle stalks

and ventral integument of head and mantle.)
i recognized species: alessandrinii.

Genus PTERYGIOTEUTHIS H. Fischer, 1896. (Photogenic organs on eyes, in
tentacle stalks, and in pallial chamber.)

4 ± species: gemmata, giardi, hoylei, microlampas.

Genus PYROTEUTHIS Hoyle, 1904. (Photogenic organs on eyes, in tentacle

stalks, and in pallial chamber.)
3 species: aurantiaca, margaritifera, oceanica, + several named larval forms.

LIGHT PRODUCTION IN CEPHALOPODS.

Family OCTOPODOTEUTHID^E.

Genus OCTOPODOTEUTHIS Rtippell, 1844. (Photogenic organs on ink sac?)

i species: sicula.
Genus Oclopodoteuthopsis Pfeffer, 1912. (No photogenic species known.)

i species.
Genus Cucioteulhis Steenstrup, 1882. (No photogenic species known.)

i species.
Family HISTIOTEUTHID^.

Genus CALLITEUTHIS Verrill, 1880. (Numerous photogenic organs in integu-
ment of arms, head, and mantle; best developed ventrally.)
12 ± species: asteroessa, chuni, dofleini, goodrichi, hctcropsis, hoylei,japonica,

meleagroteuthis, meneghini, miranda, ocellata, separata, verrilli.
Genus HISTIOTEUTHIS d'Orbigny, 1839. (Photogenic-organs as in Callileuthis.)

i species: bonnellii.
Genus HISTIOCHROMIUS Pfeffer, 1912. (Larval form; photogenic organs in

integument of mantle on ventral aspect?)
i species: chuni.
Family BENTHOTEUTHID.E.

Genus BENTHOTEUTHIS Verrill, 1885. (Photogenic o.gans on arms )

i species: megalops.

Genus CTENOPTERYX Appellof, 1889. (Photogenic organs on eyes.)
i species: siculus.

Superfamily Ommastrephoidea.
Family Brachioteuthidce. (No photogenic species known.)

Genus Brachioteuthis Verrill, 1881. (+ Tracheloteuthis Steenstrup 1881.)

4 species.
Genus Cirrobrachium Hoyle, 1904 (?)

1 species.
Family OMMASTREPHID.-E.

Genus Illex Steenstrup, 1880. (No photogenic species known.)

2 species.

Genus Todaropsis Girard, 1889. (No photogenic species known.)

1 species.

Genus Ommastrephes d'Orbigny, 1835. (No photogenic species known.)

6 =b species.
Genus Nolotodarus Pfeffer, 1912. (No photogenic species known.)

2 species.

Genus HYALOTEUTHIS Gray, 1849. (Photogenic organs in ventral integument

of mantle.)
i species: pelagicus.
Genus Sthenoteuthis Verrill, 1880. (No photogenic species known.)

4 species.
Genus Symplectoteuthis Pfeffer, 1900. (No photogenic species known.)

i species.

Genus EUCLEOTEUTHIS Berry, 1916. (Bands of photogenic tissue on ventral
aspect of head and mantle.)

1 species: luminosa.

Genus Dosidicus Steenstrup, 1857'. (No photogenic species known.)

2 species.

152 S. STILLMAN BERRY.

Family ThysanoteuthidtB. (No photogenic species known.)
Genus Thysanoleuthis Troschel, 1857.

2 species.

(Position uncertain.)

Family Lepidoteuthidce. (Naef refers this poorly known group to the Myopsida.)
Genus Lepidoteuthis Joubin, 1895. (No photogenic species known.)
i species.

Superfamily Chiroteuthoidea.
Family Chiroteuthidae.

Genus Doratopsis de Rochebrune, 1884. (+ Planctoteuthis Pfeffer, 1912, and

Leptoteuthis Verrill, 1884; no photogenic species known.)
7 species.
Genus CHIROTEUTHIS d'Orbigny, 1839. (Photogenic organs on eyes, ventral

arms, and in pallial chamber.)

7 species: imperalor, lacertosa, macrosoma, pellucida, picteti, regnardi (photo-
genic organs undescribed), veranyi.
Genus MASTIGOTEUTHIS Verrill, 1881. (Photogenic organs in integument of

the arms, funnel, mantle, or fins, or even absent.)

ii species: dentata, famelica, levimana, magna. (Not known to be photo-
genic.)
cordiformis (small tubercles, possibly photogenic, thickly distributed

in dorsal integument of body).
agassizii (photogenic organs numerous in integument of head, arms,

tentacle stalks and mantle, both dorsally and ventrally).
grimaldii (photogenic organs on dorsal surface of fins, and ventral

surfaces of head, arms, funnel and mantle).

flammea (photogenic organs comparatively few; in integument of dorsal
surface of fins, and on ventral surfaces of head, ventral arms, funnel
and mantle).

talismani (photogenic organs on ventral aspect of fins).
hjorti (photogenic organs on eyes).

glaukopis (a photogenic organ in ventral border of each eyelid sinus).
Genus Joubiniteuthis nov.1 (No photogenic species known.)

i species.
Genus Idioteuthis Sasaki, 1916. (No photogenic species known.)

i species.

Family Grimaldileulhidee. (No photogenic species known.)
Genus Grimalditeuthis Joubin, 1898.

i species.

Superfamily Cranchioidea.
Family CRANCHIID.E.

Subfamily Cranchiinae. (Series of small photogenic organs on eyes.)
Genus CRANCHIA Leach, 1817.

3 species or forms: hispida, scabra, tenuitentaculata.
Genus LIOCRANCHIA Pfeffer, 1884.

3 species or forms: globulus, reinhardtii, valdivicE.

1 Chiroteuthis Portieri Joubin, 1916, does not seem strictly referable to the
genus in which it is placed, as it is said to possess no luminous organs, while the
extreme attenuation and length of the three dorsal pairs of arms are at variance
with the well known state of affairs in Chiroteuthis. Hence I would propose the
new group Joubiniteuthis, with this species as type.

LIGHT PRODUCTION IN CEPHALOPODS. 153

Genus PYROGOPSIS de Rochebrune, 1884.

4 species: pacificus, rhyncophorus, schneehageni, zygaena.
Genus LEACHIA Lesueur, 1821.

3 species: cyclura, ellipse ptera, eschscholtzii.
Genus LIGURIELLA Issel, 1908.

I species: podophthalma.

Subfamily Taoniinae. (A large single or duplex photogenic organ on each eye-
ball.)

Genus PHASMATOPSIS de Rochebrune, 1884. (Photogenic organs not yet

described.)
i species: cymoctypus.

Genus TOXEUMA Chun, 1906.
i species: belone.

Genus TAONIUS Steenstrup, 1861. (Photogenic organs not yet described.)
i species: pavo.

Genus VERRILLITEUTHIS Berry, 1916. (Photogenic organs not yet described.)
i species: hyperborea.

Genus MEGALOCRANCHIA Pfeffer, 1884.

5 species: abyssicola, fisheri, maxima, pardus, pellucida.
Genus LEUCOCRANCHIA Joubin, 1912.

I species: pfefferi.
Genus TAONIDIUM Pfeffer, 1900. (Photogenic organs not yet described.)

4 species: chuni, incertum, pfefferi, suhmi.
Genus CRYSTALLOTEUTHIS Chun, 1906.

i species: glacialis.
Genus PHASMATOTEUTHIS Pfeffer, 1912.

1 species: richardi.

Genus GALITEUTHIS Joubin, 1898.

2 species: armata, phyllura; possibly identical.

Genus CORYNOMMA Chun, 1906. (A pair of photogenic organs embedded in

the liver in addition to the subocular photophores.)

i species: speculator.

Genus HENSENIOTEUTHIS Pfeffer, 1900 (-\-Sandalops Chun, 1906, Helico-
cranchia Massy, 1907, and Teuthoivenia Chun, 1919.)

5 species: antarctica, joubini, megalops, melancholicus, pfefferi.
Genus BATHOTHAUMA Chun, 1906.

3 species: bergeti, boureei, lyromma.

Division MYOPSIDA.
Superfamily Loliginoidea.

Family Loliginidce. (No photogenic species known.)
Genus Acroteuthis Berry, 1913.

3 species.
Genus Doryteuthis Naef, 1912.

5 species.
Genus Loligo Schneider, 1784.

32 ± species.
Genus Lolliguncula Steenstrup, 1881.

3 or 4 species.

154 s- STILLMAN BERRY.

Genus Loliolus Steenstrup, 1856.

4 species.
Genus Sepioteuthis de Blainville, 1824.

21 ± species.

Superfamily Spiruloidea.

Family SPIRULID/E. (A single organ thought to be photogenic at posterior end of

body.)
Genus SPIRULA Lamarck, 1799.

ii named forms (number true species uncertain): atlantica, australis, blakei,
fragilis, indopacifica, laevis, peronii, prototypus, reticulata, spirilla, vulgaris.

Superfamily Sepioidea.

Family Promachoteuthida. (No photogenic species known.)
Genus Promachoteuthis Hoyle, 1885.

1 species.

Family Idiosepiidcz. (No photogenic species known.)
Genus Idiosepius Steenstrup, 1881.

2 species.
Family SEPIOLID/E.

Subfamily Rossiinse. (No photogenic species known with certainty.1)
Genus ROSSIA Owen, 1834.

14 species.
Genus Semirossia Steenstrup, 1887.

2 species.

Subfamily Heteroteuthinae. (A fused pair of glandular photogenic organs on

ink sac in all known cases.)
Genus HETEROTEUTHIS Gray, 1849.

3 species: dispar, hawaiiensis, weberi.

Genus STOLOTEUTHIS Verrill, 1881. (Photogenic organs still undescribed.)

I species: lencoptera.
Genus IRIDOTEUTHIS Naef, 1912.

i species: iris.
Genus NECTOTEUTHIS Verrill, 1883. (Photogenic organs still undescribed.)

I species: pourtalesii.
Subfamily Sepiolinae.

Genus SEPIOLA Schneider, 1784. (Paired glandular photogenic organs on

ink sac.)
ii ± species: affinis, atlantica, aurantiaca, intermedia(l) , ligulata, pacifica

(?), penares (?), robusta, rossiceformis (?), sepiola, steenstrupiana.
Genus RONDELETIA Naef, 1916. (A fused pair of glandular photogenic organs

on ink sac.)
i species: minor.
Genus Sepietta Naef, 1912. (No photogenic species.)

4 or 5 species.

6 This lack is stated by Naef (:i2, p. 245) as one of the diagnostic characters of
the subfamily, but the possession of photogenic organs by Rossia macrosoma is
definitely affirmed by Meyer (:o6, pp. 390, 392). One is perforce still of unsettled
mind in the matter, especially as both observers worked at Naples, and Naef goes
into no details beyond the mere negation.

LIGHT PRODUCTION IN CEPHALOPODS. 155

Genus INIOTEUTHIS Verrill, 1881. (Paired glandular photogenic organs on

ink sac.)

3 species: japonica, maculosa, parva.
Genus Sepiolina Naef, 1912. (No photogenic species known.)

1 species.

Genus EUPRYMNA Steenstrup, 1887. (Paired glandular photogenic organs on

ink sac.)
8 ± species: bursa, morsel, pusilla, schneehageni, scolopes, similis. steno-

dactyla, tasmanica.

Subfamily Sepiadariinje. (No photogenic species known.)
Genus Sepiadarium Steenstrup, 1881.

2 species.

Genus Sepioloidea d'Orbigny, 1855.

1 species.

Family Sepiidce. (No photogenic species known. 'i
Genus Sepia Linnaeus, 1758.

80 ± species.
Genus Metasepla Hoyle, 1885.

2 species.

Genus Sepiella Gray, 1849.

1 1 ± species.
Genus Hemisepius Steenstrup, 1875.

i species.

Suborder OCTOPODA.

Superfamily Cirroteuthoidea.
Family CIRROTEUTHID.E.

Genus Cirroteuthis Eschricht, 1836. (No photogenic species known.)

8 species.
Genus Stauroteuthis Verrill, 1879. (No photogenic species known.)

3 species.

Genus MELANOTEUTHIS Joubin, 1912. (A pair of supposed photogenic organs
on dorsal aspect of mantle.)

i species: lucens.
Genus Opisthoteuthis Verrill, 1883. (No photogenic species known.)

6 species.
Genus Cirrothanma Chun, 1911. (No photogenic species known.1)

I species.
Genus Froekenia Hoyle, 1904. (No photogenic species known.)

1 species.

Genus Vampyroteuthis Chun, 1903. (No photogenic species known.)

2 species.3

Genus Lcetmoteuthis Berry, 1913. (No photogenic species known.)

i species.

Superfamily Argonautoidea.

Family Ocythoidce. (No photogenic species known.)
Genus Ocythoe Rafmesque, 1814.

i recognized species.

'A possibility seems to exist that this genus is photogenic, — cf. Chun, : 13, p.
23-25, 27.

- Cirroteuthis macrope Berry, 1911, appears to belong to this genus.

156 S. STILLMAN BERRY.

Family Argonautidcs. (No photogenic species known.)
Genus Argonauta Linnaeus, 1758.

12 ± species.

Family Tremoctopodida. (No photogenic species known.1)
Genus Tremoctopus delle Chiaje, 1829.
Only i certainly established species.
Family Alloposidce. (No photogenic species known.)
Genus Alloposus Verrill, 1880.
2 species.

Superfamily Amphitretoidea.

Family Amphitretidce. (No photogenic species known.)
Genus Amphitretus Hoyle, 1885.
i species.

Superfamily Polypodoidea.
Family BOLIT/ENID.E.

Genus Bolitcena Steenstrup, 1859. (No photogenic species known.)

i species.
Genus ELEDONELLA Verrill, 1884.

5 species: (i species, alberti, described as " probablement photogene")-
Genus Vilreledonella Joubin, 1918. (No photogenic species known.)

i species.

Family Polypodida. (No photogenic species known.)
Genus Polypus Schneider, 1784.

125 ± species.
Genus Tritaxeopus Owen, 1881.

i species.
Genus Pinnoctopus d'Orbigny, 1845.

1 species.

Genus Scaurgus Troschel, 1857.

4 species.
Genus Cistopus Gray, 1849.

2 species.

Genus Moschites Schneider, 1784.

3 species.

Genus Graneledone Joubin, 1918.

9 species.
Genus Eledonenta de Rochebrune, 1884.

2 species.
Genus Velodona Chun, 1915.

i species.

The increase afforded by the present list over the numbers
included in the earlier catalogs of photogenic cephalopods is
quite remarkable for the small number of years that has elapsed.
Hoyle's list (:o8, p. 14) records as photogenic 6 families, 26
genera, and 30 species of (Egopsida, i family, 3 genera and 3
species of Myopsida, and none of the other groups, or a total

1 But cf. Tryon, '79, p. 131.

LIGHT PRODUCTION IN CEPHALOPODS.

157

TABLE II.

RECAPITULATION OF RECENT CEPHALOPODA.

Class CEPHALOPODA.
Order TETRABRANCHIATA.

Number of
Genera.

Number of
Species.

Number of

Photogenic

Species.

Suborder NAUTILOIDEA.

Family Nautilidce I

I i
Order DIBRANCHIATA.

Suborder DECAPODA.
CEgopsida.

Family Archileuthidtf I

Family Gonalidce i

Family ONYCHOTEUTHID.E 6

Family LYCOTEUTHID^E 2

Family LAMPADIOTEUTHID.E i

Family ENOPLOTEUTHID^E 10

Family OCTOPODOTEUTHID^: 3

Family HISTIOTEUTHHXE 3

Family BENTHOTEUTHID.E 2

Family Brachioteuthidce 2

Family OMMASTREPHID.E 9

Family Thysanoteuthidce i

Family Lepidoteuthidce I

Family CHIROTEUTHID.E 5

Family Grimalditeuthidce i

Family CRANCHIID.-E 18

66

f j

Myopsida.

Family LolighiidcB 6

Family SPIRULID^E i

Family Promachoteuthidce i

Family Idiosepiida i

Family SEPIOLIDJE 12

Family Sepiadariince 2

Family Sepiidce 4

27

Suborder OCTOPODA.

Family CIRROTEUTHID.*:

Family Ocythoidcs I

Family Tremoctopodidce i

Family Argonantidce i

Family Alloposida i

Family Amphitretidas i

Family BOLIT/ENID.E 3

Family Polypodidce 9

14
2
20 ±

2

I

27±

3

I4±

2

5

20 ±
2

I
28

I
41

O
O
2
2

I
27

14
2
O
2
O
O

14
0

34 +

i?3±

99 +

i +

i

2
SI±

3

94 ±

o

o

26-

0

o

27 +

23

I

1 +

25

2

I

7
148:

195:

I
o
o
o
o
o
I
o

158

S. STILLMAN BERRY.

TABLE III.

SUMMARY.

Familes.

Genera.

Species.

"s

o

i-1

° 0

2'5

~ V
0- M

S 2 o

u o'c

J= OJ

£a, M
a.

"5

0

H

2 °"

0 C

J3 W
PH M

C 0 0

_ _ ._

r, o c

0_c „

S^ ""
PH

"5
H

2 «'

o'c

J= «j

a, be

S 2 °"

U o'c

, J3 11

fea, IM
0<

CEgopsida

16

7

IO
2

62.5
28.6

66
27

39 +
6 +

59-0

22.2

I73±

224±

99 +

27 +

57-2
n.6

Myopsida

Total Decapoda

23

8

i

12
2
O

52.2
25-0
o.o

93 45 +

25 2
I O

48.4

8.0

0.0

397 ±
I95±
3

126 +

2
0

31-6

I.O
0.0

Octopoda

Tetrabranchiata

Total.

T.?.

Trf

4T. 7 IIQ 474-

^?0.^

<;o=; ±

1284-

21.3

of 7 families, 29 genera, and 33 species. Chun (:io, p. 39),
treating only of the CEgopsida, increases these figures to 8
families, 26 genera and 39 species. The two sets of figures
should be compared with those given in the numerical summary
in Table III at the top of this page. Here it appears that of
the 32 families of recent cephalopods now recognized, 14 (or
more than two fifths) contain luminous species; out of 119
genera, 47 (or nearly two fifths) are light producing; and out of
595 species, 128 (or over one fifth) are now held on good ground
to be luminous. The rich development in species of the genera
Loligo, Sepia and Polypus, which has already been noted, is the
circumstance chiefly responsible for the cutting down of the
proportion which the luminous species bear to the whole to less
than one half that which is exhibited by the luminous families.
Similarly the slight proportional decline in the case of the lumi-
nous genera is due to the large number of ranking genera in
certain mainly non-luminous families such as the Ommastre-
phidae, Sepiolidae, Cirroteuthidae and Polypodidae.

The table also indicates very strikingly what is really the
outstanding feature of the taxonomic distribution of the photo-
genic forms, namely, the preponderance both of CEgopsida among
the species known to be light producing, and of light producing
species among the CEgopsida. In the former instance this
preponderance is enormous. 71.4 per cent, of the luminous
families, 83.0 per cent, of the luminous genera, 77.3 per cent, of

LIGHT PRODUCTION IN CEPHALOPODS. 159

the luminous species, are (Egopsid. Among the QEgopsida
themselves over one half of all the families, genera, and species
are described as possessing photogenic organs. Five entire
families — the Lycoteuthidee, Lampadioteuthidae, Enoploteu-
thidae, Histioteuthidae, and Benthoteuthidae, all of them of more
or less deep sea habit, — have all their species so equipped, and
this seems almost certainly true of the very aberrant but num-
erous Cranchiidae as well. Among other groups of cephalopods,
only the Spirulidae can aspire to inclusion in the same category,
and regarding them our information is still deficient. There
may be only one valid species in this family. In addition to
those named, one other cegopsid family (Chiroteuthidae) has
more than half its species light producing. On the other hand
luminous species for five cegopsid families (the Architeuthidae,
Gonatidae, Thysanoteuthidae, Lepidoteuthidae and Grimalditeu-
thidee), five of the seven myopsid families, six of the eight
octopod families, and the Nautilidae, are as yet unknown.

For multiplicity and variety of luminous forms the palm must
be awarded to the Enoploteuthidae and Cranchiidae, though, as
will subsequently appear, the maximum attainment and diversity
of structure of the photogenic organs themselves is reached not
in either of these families, but in the Lycoteuthidae.

4. ACTUAL OBSERVANCE OF THE PHENOMENON.
As compared with other Mollusca, or even with other general
groups of Invertebrata, Cephalopoda, and especially those of
the decapod section, are extremely difficult either to capture, to
maintain alive under artificial conditions, or even to observe
with any degree of satisfaction in their free condition. Among
the (Egopsida it is probable that a sheer majority of the genera
have never been seen at all in the living state, at any rate by any
human eyes but those of fishermen. It is therefore not to be
wondered at that actual observance of the phenomenon of light
production in this group of animals is an extremely rare event,
possible only occasionally or under very exceptional conditions.
The published records of such observations are consequently so
scattered that they have fallen into obscurity, or else, in the
case of some of the more spectacular ones become all the more

l6o S. STILLMAN BERRY.

conspicuous by very reason of their paucity and desultory
character.

A brief historical survey of this subject has been given by
Hoyle (:o8), but the most valuable contributions thereto have
been made since that time, while Hoyle himself omitted one or
two quite interesting accounts from his summary. It will
therefore be well to review briefly the entire field.

I have been no more successful than previous authors in the
discovery of any recorded observation of photogenic phenomena
in living Cephalopoda prior to that of Verany in the case of
Histioteuthis bonnellii (bot^elliana) , ('51, p. 119), a translation
of which is quoted in full by Hoyle in the paper cited and is well
worthy of repetition here.1

"As often as other engagements permitted, I watched the
fishing carried on by the dredge on the shingly beaches which
extend from the town of Nice to the mouth of the Var. On the
afternoon of September 7, 1834, I arrived at the beach when the
dredge had just been drawn in, and saw in the hands of a child a
cuttle-fish, unfortunately greatly damaged. I was so struck by
the singularity of its form and the brilliance of its color that I at
once secured it, and, showing it to the fishermen, asked whether
they were acquainted with it. Upon their replying in the
negative I called their special attention to it, and offered a
handsome reward for the next specimen secured, either alive or
in good condition, and then passed on to other fishermen and
repeated my promise. Shortly afterwards I was summoned and
shown a specimen clinging to the net, which I seized and placed
in a vessel of water. At that moment I enjoyed the astonishing
spectacle of the brilliant spots, which appeared upon the skin of
this animal, whose remarkable form had already impressed me:
sometimes it was a ray of sapphire blue which blinded me;
sometimes of opalescent topaz yellow, which rendered it still
more striking; at other times these two rich colors mingled their
magnificent rays. During the night these opalescent spots
emitted a phosphorescent brilliance which rendered this mollusc

1 Verany's monograph is a very rare one in the United States, especially in
the West, where I am not aware that any complete copy exists. I am accordingly
entirely dependent for the information quoted upon the translation given by Hoyle.

LIGHT PRODUCTION IN CEPHALOPODS. l6l

one of the most splendid of Nature's products. Its existence was,
however, of short duration, though I had placed it in a large
vessel of water. Probably it lives at great depths."

Although not mentioned by Hoyle, the next student whose
published observations concern us was none other than Charles
Darwin. Among the melange of odd notes in the course of his
account of the voyage of the "Beagle" ('60, pp. 7-8) appears
the following: "I was much interested, on several occasions, by
watching the habits of an Octopus, or cuttle-fish. . . . I observed
that one which I kept in the cabin was slightly phosphorescent
in the dark." As we have already seen, photogenic organs or
tissues are practically unknown among octopods, so that this
observation would be quite an anomalous and puzzling one,
were it not for the at least plausible explanation that the phe-
nomenon in this instance as so many others in the literature of
biophotogenesis was due not to the active functioning of any
tissues of the cephalopod itself, but to bacterial infection or even
to the presence of effulgent Protozoa in the slime surrounding
its skin. Another possibility which occurs to me is that the
animal may not have been examined in absolute darkness, but
that sufficient light penetrated into the chamber, though imper-
ceptible to the unadjusted human eye, to enable the iridocytes
in the skin of the octopus to yield a pseudo-luminous reflection,
analogous to that so notorious in the case of the eyes of many
mammals. The description by Giglioli of luminescent specimens
of a squid which he identified as " Loligo sagittaliis" and certain
Chilean octopods, referred to by Holder in the quotation given in
the next paragraph, may be susceptible of similar explanation.
To the original of this work with the description of his observa-
tions I have unfortunately not been able to gain access. For
my own part I have on several occasions attempted to discover
similar properties in captive specimens of the common southern
California devilfish, Polypus bimaculatus (Verrill), but so far
with only negative results. Final settlement of the question
can only be accomplished by careful experiment.1

1 Since this paragraph was put in type I am reminded that Tryon ('79, p. 131)
in his paraphrase of the description of Tremoctopus gracilis (Souleyet 1852) says
that this species is " phosphorescent and with metallic reflections when living."
I have been unable to check this observation by reference to the original work of
Souleyet.

1 62 S. STILLMAN BERRY.

In a popular volume by C. F. Holder ('87, p. 46), descriptive
of luminous organisms in general, but unfortunately none too
carefully compiled, occurs the following paragraph on the
Cephalopoda:

'The highest forms of the Mollusca, the Cephalopods, cuttle-
fishes, are probably at times luminous. I have noticed what I
presumed was a delicate, sensitive glow about an Octopus in a
semi-darkened tank, but I am not satisfied to make the state-
ment as a fact. These forms are so remarkable for the waves of
color that pass over them, and which seem to make them trans-
parent, that one could readily be deceived.

'The little Cranchia (Plate IV., Fig. 2) is a light-giver, its
phosphorescence having been distinctly observed. It is an
ally of the giant squids, which have been found fifty-five feet in
length, and which, if luminous like their pygmy relative, would
present a marvelous spectacle, darting veritable living arrows
through the depths of the sea.

"Giglioli refers to the phosphorescence of Loligo sagittatus,
and to that of several small Octopods observed by him at Callao
and Valparaiso. Their bodies gave out a pale whitish light,
uniformly distributed."

It happens that Cranchia is a genus which is now known to
possess definite photogenic organs, but these have been found to
occur only on the eyeball, whereas the rather poor figure given
by Holder represents the animal as brightly and evenly glowing
over the entire surface, — body, head, arms, tentacles, and all.
As to the supposed photogenic properties of Polypus and related
octopods, — both Darwin and Giglioli would seem to have been
too accurate observers for the explanation advanced by Holder
to be entirely satisfactory.

Chun, in his narrative of the cruise of the "Valdivia" (:O3,
pp. 569-570; also :03a, p. 81 ; :io, p. 50) gives up to this time the
fullest account of the actual display of photogenic propensity by
a cephalopod we have been able to find, and he followed this in
later publications by a very considerable contribution to our
morphological knowledge of the organs responsible for the
manifestation. The specimen observed proved to belong to a
wonderful undescribed species, the Lycoteutliis diadema (Chun).

LIGHT PRODUCTION IN CEPHALOPODS. 163

Although it was taken from a considerable depth, he was able to
keep it alive in ice water long enough to make a photograph of it
by dint of its own light. Again I must quote from a translation
by Hoyle : "Among all the marvels of coloration which the animals
of the deep sea exhibited to us, nothing can be even distantly
compared with the hues of these organs. One would think that
the body was adorned with a diadem of brilliant gems. The
middle organs of the eyes shone with ultramarine blue, the lateral
ones with a pearly sheen. Those towards the front of the lower
surface of the body gave out a ruby-red light, while those behind
were snow-white or pearly, except the median one, which was
sky-blue. It was indeed a glorious spectacle." It is altogether a
pity that similar observations have not been possible for the
doubtless even more spectacular Nematolampas regalis, which,
although very nearly related to Lycoteuthis is equipped with an
entire further battery of photophores.

More detailed from the standpoint of physiology is the account
given by Watase (105) of a little squid, the "hotaru-ika" of
Japanese writers,1 which is extremely abundant at the proper
season and locality on certain of the shores of Japan, and which
has since become the best known of all the luminous squids.
Watase's paper is an important one as the first dealing with
this species, but, being semi-popular in character and published
in Japanese, escaped notice for a considerable time and has only
lately received a little of the attention it deserves. Through the
kindness of Mr. Sotaro Matsushita, formerly of Redlands,
California, I have for some time been in possession of a translation,
and a very free transcription of some of its more interesting if
quaint passages should be neither inappropriate nor unwelcome
here. "Hotaru-ika, when seen externally, does not differ much
from other ika [squids]. Yet there are many interesting features
which we do not see in other ika. At each end of the two ' legs '
there are three oblong, black spots. These small spots were
first discovered by the French scientist Joubin. Yet even he did
not know their function. According to the results of my own
study of these in living Japanese specimens, the spots were found
to produce a considerable light, penetrating to the space of about

1 The "firefly squid" — Watasenia scintillans (Berry).

164 S. STILLMAN BERRY.

a foot. . . . While the animal is living these spots are trans-
parent.

"Again there are hundreds of other small spots all over the
body. . . . When seen in daylight they appear to be small
black spots, but in the night all these spots shine with a brilliant
light like that of the stars in heaven. . . . When these spots
(while the hotaru-ika is alive) are viewed under the microscope,
they are very interesting. When the animal is about to produce
the light, the membranes [chromatophores] covering the spots
will concentrate and remove themselves, thus opening a way for
the light. The light is so brilliant that it seems like a sunbeam
shot through a tiny hole in a window curtain. Again when the
hotaru-ika wishes to shut off the light, the membranes will
expand and cover the spots. . . ."

In the following year Meyer (:o6) described briefly the photo-
genic activity of the myopsid, Heteroteuthis dispar (Riippell),
similar observations having been made some time previously
by Lo Bianco, but never published. Meyer found that in the
case of the specimen observed by him at the Naples Zoological
Station he "could in the dark room easily locate the position of
the photophore through the transparent mantle, lying on the
ventral surface just behind the anus." He further found that
when the animal was irritated, "it shot rapidly through the
water, and spurted through its funnel a luminous secretion which
floated in the water as separate globules, these being drawn out
by the currents into shining threads, a pyrotechnic display
(Feuerwerk) which he was able to repeat many times. The light
of the secretion and of the light organ itself had the same pale
greenish hue which we observe with our glow-worms." Meyer
further reports the discovery by one of his colleagues, Marchand,
of somewhat similar photogenic properties in Sepiola, except
that in this genus "the luminous secretion is not discharged into
the water but remains on the surface of the gland. Furthermore,
Sepiola only shines if it be very powerfully stimulated, as when,
for instance, the mantle is cut open." From the foregoing it is at
once evident that in both these myopsid genera the mechanism
of light production is very different from that of any of the other
forms studied, and this conclusion is borne out by the anatomical

LIGHT PRODUCTION IN CEPHALOPODS. 165

features. The possible utility of this peculiar development of
the function, so far as Heteroteuthis is concerned, is the subject
of some interesting speculation in one of Meyer's subsequent
papers (:o8, pp. 507-508), which will receive more attention
later on.

We are now brought to a consideration of some important
recent work performed by various Japanese observers in con-
tinuation of Watases pioneer studies on Watasenia scintillans
(Sasaki, :i2, 113, :i4; Ishikawa, :i3). This constitutes probably
the chief work which has been done in this field, and therefore
merits consideration in considerable detail.

"In the region where the squids live, that is, in the waters
of Namerikawa on the coast of the Japan Sea," writes Ishikawa
(:i3, pp. 167-169), "this circumstance [the luminosity of the tips
of the ventral arms described by Watase] had already long been
known ; but none of our zoologists were aware of it until Watase
by chance made the discovery. At a time when he was engaged
in the study of fireflies, he was apprised by a schoolmaster that
there occurred a species of squid in the sea at Namerikawa which
lighted very strongly. Pursuant to this suggestion he sought
the village named, found in due course a species of small squid
with powerful light organs, and recognized the same as a species
of Abraliopsis. As he has orally told me, it was on the 28th
of May, 1905, on the memorable day of the battle of Tsusima,
hat he saw the light of this squid for the first tim2

"The large swellings at the tips of the ventral arms, as well
as two or three smaller dots, are, as he remarks, luminous organs
of the first order. They shine so brilliantly that when one
observes the animals in dark water, one sees only two effulgent
bodies moving in the dark water, like the glow of an electric
contact, and the lively oscillations of the invisible arms produce
a very wierd effect. Next to these in the intensity of their light are
the eye organs, and then come the remaining organs. The
three types of organs do not always shine simultaneously; often
only one or the other. But it can also happen that the animal sets
all the organs into action at the same time. When the mantle
organs light up, the form of the animal springs out spectre-like in
the dark water. These organs, arranged in rows, when one

1 66 S. STILLMAN BERRY.

examines them close at hand, shine like an electric illumination.
The color of the light is a beautiful clear blue.

"As Watase writes, the arm organs in dead animals are en-
tirely surrounded by pigment cloaks and only when alive can
the animal retract these. The retraction of these cloaks takes
place very qu:ckly, and when they are retracted, the organ ap-
pears in daylight as a delicate dull-green colored body."

In a paper which comprises a most notable contribution to
our knowledge of the ecology and habits of ten-armed cephalo-
pods, Sasaki (:i4, pp. 77-80) adds materially to the accounts of
his predecessors. Some of his observations are so pertinent to
some of the discussion which must follow later that they should
be quoted rather fully. Treating the three types of photophore to
be seen in Watasenia under separate headings, this author writes:

"Brachial Organ. This is the largest organ, and when I made
observations in the fishing season, it was much more active in
phosphorescence than other organs. It is situated at the end
of each ventral arm, composed of 3 globules arranged in a
series. The globules are ovoid in shape and nearly equal in
size, but the middle one in the series is generally a little larger
than the others, the dimensions being 1.4 mm. long and about
I mm. broad. In fresh specimens they show a greenish cobalt
colour, and there are 2 or 3 layers of large brownish chroma-
tophores covering a part of the preceding substance. These
chromatophores are constantly contracting and expanding.
When they were observed at night on the living animals, they
were seen to discharge light in all directions much brighter
than any of Japanese fireflies. The color of the light is Prussian-
blue or tinged a little with purple, and the luminosity is strong
enough to outshine the other luminous organs. When the living
animal was placed on a glass plate, which was put directly on
the case of the dry plate of the photographic camera, and then
exposed for four seconds with the Lion's dry plate of the special
rapid no. 230, the light of this brachial organ was distinctly
taken on the dry plate, although those of other organs made no
impression.

"Minute Organs Scattered on the Ventral Surface of the Whole
Body. There are numerous minute organs distributed on the

LIGHT PRODUCTION IN CEPHALOPODS. 1 67

ventral surface of the mantle, head and siphon, and they are
also on the third and fourth arms. . . .

"Each organ in the fresh specimen has a substance of purplish
hue in the centre; this substance seems to be that discharging
light when the animal is living. When the organ is exposed in
the air, the purplish hue of the substance changes to greenish
blue after a while, and finally resolves into a true green. The
substance is covered by a pigment layer of darkish brown or
deep purple which has a hole resembling the pupil of an eye,
through which the substance can easily be seen. The light of
the substance at night is whiter and less luminous than that of
the brachial organ.

"Ocular Organ. When the eyelid of the fresh specimen is
removed and the eyeball exposed, there are seen 5 luminous
organs arranged in a series along the ventral circumference of
the eyeball, the organ on either end of the series being a little
larger than the remaining 3. The colour of all these organs
is pearly white. When the organ is seen at night in the living
animal, the phosphorescence is not distinguishable from that of
the minute organ on the body.1

"Difference of Phosphorescence in the Sexes. On examining
the preserved specimens to discover the difference of the external
forms as well as the histological structures of their luminous
organs as occurring in the male and female, none could be
discerned. But in the female specimens there are one hundred
or so more of the minute organs of the mantle than in the male.
Whether there is any meaning as to sexual selection, it is difficult
to say, the data concerned being insufficient at present to an-
nounce any opinion.

"Next, as to the difference of phosphorescence between the
sexes in their living state, the means of investigation proved to
be very difficult. At first I repeatedly undertook to keep the
animal in an aquarium, but no success was attained. The
reason for the failure is that first of all, the animals are very
delicate, and next the aquarium was defective. The animals

1 A slight discrepancy is noteworthy between the account of Sasaki and that of
Ishikawa concerning the character of the light of the subocular organs. According
to the latter these photophores are more or less intermediate in brilliancy between
the large brachial organs and those of the general integument.

1 68 S. STILLMAN BERRY.

are so weak that in carrying them from the sea to the aquarium
they wasted and died. As they wasted, the luminosity in ques-
tion became very feeble, and naturally with their expiration,
the light of the luminous organs gradually vanished altogether.
This being so, I then tried to observe the animals directly while
they were swimming in the net. But no good means were found
easily to distinguish the sexes on such dark nights, even with the
feeble light of the moon or of a lantern.

"However in my examinations at night, no special variety of
the light could be found, the colour of the light being always the
same. And in one case, putting in a vessel and observing about
thirty specimens in a fishing boat while they were yet actively on
motion, I verified the fact that their luminosity is uniform. In
the morning, to my surprise, a male was found dead among those
30 specimens; this proves that it had the same colour of light
with the female on that night. The above data seem to prove
the fact that the colour of the light of the luminous organs is the
same in both sexes.

"Again, in late July of the same year, I made another obser-
vation on the phosphorescence under consideration and then it
was quite evident to me that the luminosity of the brachial organ
was at this season noticeably feebler than in the spring.

'The phosphorescence of the immature animal can never be
studied in Namerikawa, young ones thus far not being found
there."

In the same paper (pp. 98-99), Sasaki incidentally records the
fact that he observed the photogenic property in living specimens
of the myopsid, Inioteuthis japonica Verrill ( = inioteuthis
(Naef)). These he found to be "discharging a faint cobaltish
light from a great luminous organ which is situated in the mantle
cavity near the ink-bag." From anatomical observations we
know that the luminous organs of this genus are essentially
similar to those of the nearly related if not actually congeneric
Sepiola.

Lastly, Dahlgren (:i6, pp. 70-71) describes in a little greater
detail than before the photogenic behavior of Heteroteuthis
dispar, the myopsid species already observed by Meyer. He
writes: "When brought into the laboratory in good condition

LIGHT PRODUCTION IN CEPHALOPODS. 169

and allowed to rest quietly it may be taken into the dark-room
and gently struck, as it swims in the aquarium, with a glass rod.
Fig. 1 8 is a drawing to illustrate what may and usually does
happen under these circumstances. The animal throws out of
its siphon several little masses of mucus which show no light
at the moment of ejection, but almost instantly, as the oxygen of
the water begins to work on them, show a number of rod-shaped
particles of a brilliantly luminous matter embedded throughout
the very delicate mass. As the mass continues to expand this
light continues to glow brightly for as much as three to five
minutes, after which it rather suddenly dies out. In color the
light is the usual blue-green of luciferine when burning outside
the body. The animal can repeat this process for a number of
times, when it appears to have exhausted its supply of luciferine,
and it is not possible, apparently, to keep it in captivity for a long
enough period for the supply to be restored."

From this scanty, but for all practical purposes probably ex-
haustive summary, we find that except for the doubtful obser-
vations by Souleyet, Darwin and Giglioli, the actual process of
light production in cephalopods has been observed directly in but
seven species, of which three belong to the myopsid family
Sepiolidse and have photogenic organs of a peculiar discharging
type, while the other four belong to the (Egopsida. We are
fortunate, however, in that each of these latter species is repre-
sentative of a different family and thus ample support is given
to the inferences necessarily drawn from the outward appearance
and histological structure of the many types of photophore that
they are of a fact photogenic.

(To be Continued')

Vol. XXXVIII. April, 1920. No. 4.

BIOLOGICAL BULLETIN

LIGHT PRODUCTION IN CEPHALOPODS, II.

AN INTRODUCTORY SURVEY.
S. STILLMAN BERRY,

REDLANDS, CALIFORNIA.

CONTENTS.

PAGE

5. Color and Intensity of Light .............................. ••

6. Distribution of Photophores on Animal ................. • i?4

On Eyeball ............................................... *75

In General Integument of Body ............................. • • i?6

On Mantle ............................................ i?6

On Head ................................... • • * 76

On Arms ................................... z 7°

On Funnel ......................... :77

Around Eye-openings ................ • • i?T

On Fins ............................... Z77

In Arms ........................... :79

In Tentacles ........................ : 8o

Intrapallial ................... • • l8°

Anal ................ l81

Branchial ...................... . . 181

Gastric .................. . . 181

Axial ........................ l81

At Tip of Body ............... l82

7. Structure of Photogenic Organs .... •• l83

Diversity ..................... l84

Component Structures .............. • • • • l85

Duplex Organs ................. • • • l85

8. Polymorphic Nature of Photogenic Organs. . 186

Genera with Isomorphic Photophores. . I87

Genera with Dimorphic Photophores ............. • • l87

Genera with Trimorphic Photophores. . l87

Genera with Polymorphic Photophores. *87

9. Systematic Significance of Photogenic Organs. . . 189

10. Probable Polyphyletic Origin of Photogenic Organs. 190

11. Concluding Note ....... I9I

12. Summary ...................

13. Literature ............. .................... :93

171

172

S. STILLMAN BERRY.

5. COLOR AND INTENSITY OF LIGHT.

But little is known concerning the color, particularly what may
permissibly be termed the intrinsic color, of the light produced by
cephalopods, in fact next to nothing of any of its fundamental
physical qualities. This of course follows as a natural corollary
of the scanty nature of the recorded human observations of
these animals in the living state. Such as they are the appro-
priate data gleaned from the preceding section of this paper
are briefly tabulated.

Verany's observations previously quoted are a little ambiguous
and it is not just evident whether the "sapphire blue " and " topaz
yellow" rays which he describes with such naive enthusiasm for
the photophores of Histioteuthis apply to the result of their
functional activity at night, or merely to their ordinary brilliant
coloration in the daytime. The fact that he was "blinded"
would seem to indicate the former.

TABLE IV.

COLOR OF LIGHT IN CEPHALOPODS.

Species.

Date of Published
Observation.

Color of Light.

CEgopsida:
Family Lycoteuthidae,

Lycoteuthis diadema, (Chun)

Chun 1902

ultramine sky blue

Family Enoploteuthiclae,
Watasenia scintillans (Berry) . . .

Watase, 1905, 1912.
Ishikawa, 1913 . .
Sasaki 1914

ruby red, pearly
white.

clear blue.
Prussian blue to pur-

Family Histioteuthidae,

Histioteuthis bonellii (Ferussac)

Verany 1851

plish.

?

Family Ommastrephidse,
? " Loligo sagittatus Lamarck" . .

Giglioli

?

Family Cranchiidae,
Cranchia sp.

Holder, 1887 .

p

Myopsida:
Family Sepiolidae,
Sepiola sp. (rondeletii Leach ?) . .

Marchand, in
Meyer, 1906 . . .

9

Inioteuthis japonica Verrill

Sasaki, 1914 • . .

"cobaltish "

H eteroteuthis dispar (Guppell)

Meyer 1906

pale greenish

Octopoda:
' Polypus sp

Dahlgren, 1916. . . .
Darwin 185

blue green.

?

? small octopod

Giglioli

"pale whitish."

Tremictoous eracilis CSoiilevet) .

Soulevet.

?

LIGHT PRODUCTION IN CEPHALOPODS. 173

Of the subsequent observations, only a few trouble to specify
the apparent color of the light rays. I use the word apparent
advisedly, not alone because of the ever-present subjective
considerations by which one and the same ray may yield diverse
impressions to different persons at the same time and under the
same conditions, or to the same person at different times or
under different conditions, but also because there is evidence that
the original color values of the light rays may suffer modification,
either by reason of the physical features of some of the super^
vening tissues of the photophore itself, or by the interposition of
the chromatophoric color screens to which attention has already
been drawn.

The extent to which the brilliantly varied illumination which
was described by Chun for Lycoteuthis diadema is due to such
considerations as these, rather than to differences inherent in
the light rays produced by the respective organs is therefore a
matter for considerable speculation. In this species Chun
(:O3, pp. 569-570; :o3a, p. 81; :io, p. 50) described the light of
the central organ in each subocular series as "marvelous ultra-
marine blue," of the anterior axial organ as "sky blue," of the
two anal organs as "ruby red," of the remaining organs as
"snow-white" or "pearly." But it should be remembered that
no matter what other rays may have suffered absorption to
result in the described effect on the human eye, no sort of screen
or filter could manufacture those which evidenced themselves and
they must therefore have been produced within the photogenic
tissues. If, as in all other luminous organisms which have been
subjected to examination, this is still a relatively efficient and
therefore "cold" light, the question is yet before us whether
the "ruby red" rays of Lycoteuthis are none the less as "cold"
as the ultramarine and blue ones, or as the blue-green lumines-
cence of the firefly. The biochemist and biophysicist have here a
tempting field, once the technical biological difficulties of securing
and handling the animals can be fairly overcome.

The light of the luminous secretion of Heteroteuthis dispar is
described by Meyer (:o6, p. 389) as "pale greenish," and by
Dahlgren (:i6, p. 71) as "the usual blue-green of luciferine when
burning outside the body."

174 S- STILLMAN BERRY.

Sasaki and other observers of Watasenia scintillans describe
the light of the large organs at the tips of the ventral arms as
purplish or Prussian blue, the body organs appearing "whiter
and less luminous" than these. In spite of their absolutely
different histological structure, the rays emanating from the
integumentary and subocular organs do not here appear to be
respectively distinguishable and one wonders whether, in this
regard, the observations recorded convey the whole truth.

That such elaborate variety in the size, morphological detail,
and possession of accessory contrivances as will shortly be
described, must find at least partial expression in differences
in the physical qualities (intensity and color) of the resulting
light rays, seems as inescapable to the present writer as it did
to Chun (:O3a, p. 81). And on the whole the scanty evidence
just outlined is in accord, showing that the hues of the light are
different, often most strikingly so, not alone as between inde-
pendent species, but between the organs occupying different
situations on the body in one and the same species.

6. DISTRIBUTION OF PHOTOPHORES ON ANIMAL.

The photogenic function in cephalopods is, as has been seen,
not a general attribute of the body surface, but is always, so
far as is known, localized in the specialized tissue of definitely
circumscribed organs disposed in equally definite regions of the
body. It therefore becomes appropriate to examine what posi-
tion or positions on the body these structures have come to occupy.
Proceeding accordingly, one is at once struck with the fact that
although strong evidence of partiality for certain special situations
exists, yet no hard and fast rule may be laid down. The region
where the organs occur most commonly seems to be by all means
the surface of the ventral hemisphere of the eyeball. Photo-
phores are found in this position in most (probably all) of the
Cranchiidae, in Enoploleuthis , Abralia, Abraliopsis, Watasenia,
Asthenoteuthion, Pyroteuthis and Pterygioteuthis of the Enoplo-
teuthidae, in all the Lycoteuthidae, in Lampadioteuthis , in
Ctenopteryx, and in Chiroteuthis, — at least some 25 and more prob-
ably around 29 of the entire 44 photogenic genera in the sub-
order Decapoda. Most of the cranchiid genera, comprising, so far

LIGHT PRODUCTION IN CEPHALOPODS. 1 75

as known, the entire subfamily Taoniinae, are peculiar in that the
subocular photophores are reduced to one or at most two organs
which are frequently so large as to cover nearly the entire lower
surface of the eyeball. When two such are present they are semi-
circular or more or less crescentic in outline, the smaller or ante-
rior organ fitting into the concavity of the larger. The eye-organ
in Ctenopteryx is a single large falciform structure. In most
genera, however, the subocular photophores are smaller and more
diffuse in their arrangement, the commonest system being an
alignment in a simple, bead-like, longitudinal series on the
ventral periphery of the eyeball. Curiously enough, the series
usually includes organs belonging to two or more diverse
structural types. Such is the arrangement to be found in
Lycoteuthis, Nematolampas Abralia, Abraliopsis, Watasenia and
Enoploteuthis, the last-named genus having nine or ten organs
on each eye, all the other genera five. Liocranchia and
Pyrgopsis have four organs similarly located, but all of one
type. Chiroteuthis picteti and C. imperator are figured by Chun
as having three longitudinal chains of isomorphic organs,
22 to 29 in all, upon each eye. In the latter species he
found the number to be somewhat variable, which is an unusual
circumstance with the subocular organs. This is a particularly
striking fact when the remaining five genera having this type of
photophore are considered. In all of these, namely, Lampadio-
teuthis, Pyroteuthis, Pterygioteuthis, Cranchia and Leachia the
photophores of the eyes, varying in number from four in
Lampadioteuthis to fifteen in Pterygioteuthis giardi, have lost
their simple serial arrangement, and the individual organs are
scattered to a greater or less degree over the lateral as well as
the ventral region of the eyeball. Their distribution thus
becomes highly irregular, yet it is almost always absolutely
definite and practically invariable within the bounds of each
single species. Chiroteuthis veranyi, as described by Chun, is
unique in having two large bands of photogenic tissue on the
ventral convexity of each eye, accompanied by a few small
isolated photophores of the more ordinary form, by the coales-
cence of a number of which they perhaps originated. Since the
genera possessing subocular organs are all cegopsid, it follows that

176 S. STILLMAN BERRY.

these photophores are covered by the double fold of the integument
which forms the eyelid, and consequently in preserved specimens
are often invisible without partial dissection. But in the Cran-
chiidse the overlying membranes are thin, transparent, and very
insufficiently equipped with chromatophores, so that in good
specimens the organs may be clearly seen from the exterior.
And likewise in certain other groups such as Enoploteuihis and
the Abralioid genera, we find a delicate, transparent, elongate-
oval "window" in the integument, nearly or entirely free of
pigmented chromatophores, and overlying that portion of the
eyeball where are borne the photogenic organs. There can be
little doubt that this functions in aid of the latter by facilitating
the passage of their rays.

The next most frequent topographic type of photophore to be
met with comprises those occurring in the general integument of
the body, primarily on the mantle, head, and arms. A remark-
able peculiarity of the integumentary organs is that they, like
the subocular photophores, are generally confined to the ventral
aspect and this circumstance has given rise to some interesting
theories regarding the origin and ecologic significance of the
whole phenomenon of light production in this group of animals.
Some writers have gone so far as to state that the distribution of
these organs is entirely ventral, but this is not in strict accord
with the facts, there being a few scattered photophores on the
dorsal aspect of the mantle in such forms as Abralia astrolineata
and most of the Histioteuthidse, while Verrill's figures show them
to be quite as strongly developed in this region in his Masti-
goteuthis agassiziiu as they are below. Certain other species of
Mastigoteuthis have them in plenty on the dorsal surfaces of
the fins, even if not upon the body proper. Again in Professor
Joubin's anomalous Melanoteuthis the supposed photophores
are entirely dorsal. The possibly photogenic tubercles of Masti-
goteuthis cordiformis should likewise be recalled in this connection,
and finally the presence of photophores on the dorsal arms of
Nematolampas and Benthoteuthis. But even as many exceptions
as this serve principally to accentuate the prevalence of the rule.
In some genera the integumentary organs are developed on the

» Bull. Mus. Comp. Zoo/., V. 8, PI. I., 1881.

LIGHT PRODUCTION IN CEPHALOPODS. 177

ventral surface of the mantle only (Aneistrocheirus, Flyaloteuthis ,
and, according to Chun, though he is controverted by other
writers, Chaunoteuthis) . In others (Eucleoteuthis] they occur on
the ventral surface of the head as well. In Calliteuthis, His-
tioteufhis and some species of Mastigoteuthis they are found not
only on the mantle and head, but on the aboral surfaces of the
ventral and ventro-lateral arms. In Mastigoteuthis agassizii
they are figured as occurring even on the tentacle stalks, as they
do likewise in Thelidioteuthis , although this would appear to be
an unusual situation for organs of the integumentary type. In
the former of these two genera they are numerous in the inte-
gument of the head, arms, and mantle as well as the tentacles;
in the latter, they are less numerous and although found along
the outer side of the tentacles, occur elsewhere only on the ven-
tral aspect of the mantle and head, where they have a very
regular and characteristic arrangement. Finally in a number of
well-known genera (Enoploteuthis, Abralia, Abraliopsis, Wat-
asenia, Mastigoteuthis}, integumentary organs are plentifully
distributed in indefinite number over the entire ventral aspect
of mantle, head, arms, and funnel.

On the fins these organs appear less frequently, but they are
described as occurring dorsally in several species of Mastigo-
teuthis, and in one (M. talismani) on their ventral faces.

In a number of species there is a particular development of
the integumentary photophores in the neighborhood of the eyes,
usually in the form of a circlet around the margin of the lid
opening, and such a circlet may occur, as in Enoploteuthis,
Abralia, Abraliopsis and Watasenia, in addition to a well de-
veloped series of subocular organs. As a general rule, and cer-
tainly in the four genera named, the organs comprising this
circlet are not to be distinguished from those of the general
integumentary surface save by their peculiar arrangement and
position. In the Histioteuthidse, however, comprising the genera
Histioteuthis and Calliteuthis, a most singular modification of
this feature is encountered. A peculiar attribute of these
genera is that, probably without exception, all the species have
the left eye enormously more developed than the right, so much
so in fact that a strong lateral torsion or displacement of the entire,

178 S. STILLMAN BERRY.

in both genera relatively enormous, head is produced, which would
seem to render it a physical impossibility for the animal to propel
itself in a straight path without recourse to spiral movement or
some violent sort of counter twisting. This asymmetry extends
quite inexplicably to the photogenic organs, inasmuch as the
"normal" right eye has a well developed circlet of photophores
surrounding the lid opening as above described, while the Brobding
nagian left eye has the photophores of its circlet not only pulled
farther apart by the distention of the lid, but its every component
reduced almost to a rudiment, some of them quite atrophied,
or they may even be, as Sasaki has stated for Calliteuthis sep-
arata,15 absent entirely. It seems as though from the very
nature of the case there must be some correlation between such
pronounced asymmetry and the habits of the animals, but no
reasonable explanation of what might be necessary to bring
about or to render advantageous such an anomalous condition
seems ever to have been suggested. In Mastigoteuthis glaukopis
there is no circumocular circlet of photophores, but a single
photogenic organ is described as occurring in the ventral edge of
each lid sinus.

In a few species the integumentaty photophores are few and
consequently definite in number and position (Ancistrocheirus,
Thelidioteuthis, Hyaloteuthis , Eucleoteuthis) . This is probably
true also of the very young or larval stages of all the species
possessing photogenic organs, but in adults of most species,
though still continuing to retain more or less evidence of the
primal bilateral symmetry, they are apt to increase to such an
extent as to become practically or quite impossible of separate
identification and enumeration and thereupon show little con-
stancy in either number or position.

Eucleoteuthis is a genus which deserves discussion by itself.
It is unique among known cephalopods in that its photogenic
organs instead of forming small rounded or ovoid capsules as in
practically all the other genera, are developed as a pair of narrow,
more or less interrupted stripes or bands of photogenic tissue
extending along the ventral aspect of the mantle for nearly its
entire length. A small oval tract of similar tissue flanks the

16 Journal College Agriculture Tohoku Imperial University, V. 6, p. 137, 1915.

LIGHT PRODUCTION IN CEPHALOPODS. 179

outer side of each band at its anterior end, and, in the type
species at least, there is a pair of somewhat larger, transversely
ovoid photogenic areas on the head at the base of each ventral
arms. That these curious tracts should be classified with the
remaining organs here collectively referred to as integumentary
is by no means certain.

The arms are a favored situation for photogenic organs.
The extension along their outer surfaces of the ordinary inte-
gumentary photophores in the case of such genera as Enoploteu-
this, Abralia, Abraliopsis, Watasenia, Callileuthis, Histioteuthis ,
and certain forms of Mastigoteuthis, has already been noted.
In addition to this certain special types of organs are sometimes
developed. One of the generic characters of Chiroteuthis is the
presence of a series of conspicuous dark photophores along the
oral aspect of each of the greatly enlarged ventral arms. Nemato-
lampas has a small dark photophore embedded in the extreme
tip of each arm of the two dorsal pairs. Not only this but each
arm of the third pair bears immersed in its tissues along the outer
margin a series of plainly visible photogenic organs which con-
tinue as the principal component of a long, chain-like, filamentous
extension of the arm which in life must extend like a string of
fiery beads far in advance of the animal. There are in excess
of thirty individual organs in each chain, but the true number
may be much greater as no specimens of the species still retaining
these extraordinary structures entirely in their pristine state have
yet been captured. In Abraliopsis and Watasenia, genera so
closely allied to one another that one could with about equal
ease be regarded as but a subgenus of the other, there are three
large, black, bead-like photophores, with perhaps some smaller,
more rudimentary ones, in close juxtaposition at the tips of each
ventral arm. As previously related, these are known to give
forth a brilliant light. Rudiments of similar organs corres-
pondingly situated are known in at least one species of Abralia,
another nearly related genus. This is A. astrolineata Berry of
the Kermadec Islands. The curious deep-sea Benthoteuthis
has a single photophore on the outer periphery of each arm of the
three dorsal pairs near the base, and none elsewhere on the body,
an arrangement wholly unlike that.met with in any other cepha-
lopod.

l8o S. STILLMAN BERRY.

Until very recently photophores on the tentacles have been
supposed to be of rare occurrence, but it has lately been shown
that they do actually so exist in quite a number of diverse forms,
having tended to escape observation by reason of being embedded
so deeply in the fundamental substances of the tentacle stalk as
to be quite invisible in preserved material unless thoroughly
cleared or otherwise specially treated. In Pyroteuihis Vivanti
and Mortara have recently established the presence of a series of
four such organs in the stalk of each tentacle. I had not only
independently made the same discovery in material from both
the Atlantic and Pacific, but had likewise found that there is
yet a fifth tentacular organ present, and that the same condition
obtains as well in the nearly related genus Pterygioteuthis.
Lycoteuthis and Nematolampas have two such organs in each
tentacle stalk. Lampadioteuthis is unique in possessing not only
a series of four photophores embedded in the stalk proper, but
in addition tucked away at its very base, a single large spherical
organ of peculiar structure which is quite invisible without
extraction of the entire tentacle from its socket. Conspicuous
tentacular photophores are also shown in Verrill's figures of his
Mastigoteuthis agassizii,16 but the inference seems to be, as has
been indicated above, that these are simply of the ordinary
integumentary type, as seems to be true also of the tentacular
photophores in the genus Thelidioteuthis.

We now come to the class of photogenic organs which is per-
haps the most distinctive of the Cephalopoda as compared with
other luminous animals, and which, next to the subocular photo-
phores, exhibits the most general distribution within the group.
Included here are a large array of very diversely constructed
photophores found in quite various situations upon the visceral
mass within the pallial chamber. These one and all, however,
except in the case of those myopsids which eject their luminous
secretion through the funnel, must naturally depend in life upon
the more or less complete transparency of the mantle tissues to
permit the unobstructed emanation of their beams. In pre-
served specimens, as would be expected, they can rarely be seen
without laying open the pallial chamber, whereupon they are

16 Bull. Mus. Comp. Zoo/., V. 8, 1881, pl.i.

LIGHT PRODUCTION IN CEPHALOPODS. l8l

generally easy to distinguish, many of them being of unusual
size and often of conspicuous coloration, while the situations
which they occupy are peculiarly limited and, within a given
species, constant. By reason of this last fact the intrapallial
organs may readily be subclassified into four series, (i) anal,
(2) branchial, (3) gastric, and (4) axial. Such a classification,
too, in spite of its obviously superficial foundation, is a con-
venient one. That it is at the same time in all respects a natural
or phylogenetic arrangement is probably not true, and it will no
doubt be greatly improved upon by the first worker who takes
up the relationships of these organs in any sort of adequate
detail.

The term anal organs is misleading, but has become so well
established in the literature that I use it pending the invention
of a more appropriate term. The photophores so classified
appear usually as a pair of quite large, often very brightly
colored organs of rounded or ovoid outline, lying on the ink sac
on either side of the rectum, with which they would otherwise
appear to have no particular connection. Being often situated
just back of the funnel, or sometimes almost within it, they are
therefore sometimes termed the siphonal photophores, a . name
which in its turn is open to objection as inappropriate to the
actual morphological relationships involved. Anal organs occur
in a considerable number of little related genera, and the dis-
charging photophores of the luminous Sepiolidae are noteworthy
for occupying an analogous situation.

The branchial organs are always paired, being situated one
near the base of each gill. They are confined, so far as known,
to the Lycoteuthidse, Lampadioteuthidae, and the pterygiomorph
section of the Enoploteuthidse.

The gastric and axial organs are classed together by most
writers under the general term abdominal, but I prefer to separate
the mesially situated, unpaired organs, which are often extended
into a considerable series in the hinder portion of the mantle
cavity, from the paired organs which sometimes occur near the
middle of the body on either side of and often in close association
with the anteriormost of the axial organs. There is evidence
that in at least some genera the division here postulated into the

1 82 S. STILLMAN BERRY.

paired gastric and unpaired axial organs is founded upon a good
morphological as well as merely topographical basis, but at the
same time it is impossible to emphasize too strongly that we are
here dealing primarily with the mere somatic distribution of the
organs, and not with a true genetic classification based on the
embryology or finer anatomy, save where the latter becomes
incidentally involved. The need for this qualification has no
doubt already been patent to the reader from the foregoing dis-
cussion.

Of the dozen genera listed in the synopsis as possessing intra-
pallial photophores, only Heleroteulhis , Sepiola, Eiiprymna,
Chiroteuthis and Corynomma" are described as having anal organs
only, a single pair or organ formed by the fusion of a pair being
present in each instance. Lampadioteuthis has paired anal and
branchial organs (the latter very large) and a single posterior
axial organ. Pterygioteuthis has paired anal and branchial
organs, and four axial organs, the most anterior of which is vastly
the largest, the most posterior very minute and pushed far down
past the fins into the sharp-pointed tip of the body. Pyro-
teuthis has a quite similar illumination system, but the foremost
axial organ is more anterior in position, is only a little larger than
the others, and is flanked on each side by a small gastric organ.
Lycoteuthis and Nematolampas have a single pair each of anal,
branchial and gastric organs as above, a small anterior axial
and a very large posterior axial organ. Branchial, gastric and
anterior axial organs are placed at about the same transverse
plane so that they form a belt of fiery jewels near the middle of
the body. Onychoteuthis (banksii) is unique in having but two
large unpaired photophores, both of which are intrapallial and
lie upon the ink sac in the median line, one very large and en-
sconced in a specially constructed depression on the ink sac
proper, the smaller upon the narrow, neck like, anterior portion
of the sac.

The minute unpaired organs which have been mentioned as
occurring in the spine-like tip of the body in Pterygioteuthis
and Pyrotheuthis are probably correctly interpreted as but the
terminal members of an unusually developed axial series. Lo-

17 Chun rather doubtfully adds Octopodoteuthis to this list.

LIGHT PRODUCTION IN CEPHALOPODS. 183

cated in the same general region as these, and, by their appear-
ance, seeming to bear a closer relation to the intrapallial organs
than to the other systems outlined, yet scarcely to be regarded as
lying actually within the mantle cavity, are the conspicuous
paired photophores placed at the extreme posterior tip of the
body in Nematolampas. Lycoteuthis does not possess them.
They stand in a class quite by themselves at present, but if the
peculiar swellings to be noted in the same situation in certain
species of Abralia are susceptible of a photogenic interpretation,
or if Chun's identification of the posterior disk of Spirula as a
luminous organ be accepted, a further extension of this division
of the classification is afforded.

7. STRUCTURE OF PHOTOGENIC ORGANS.

Another most remarkable feature of the development of photo-
genic systems in Cephalopoda is, so far as I am aware, the quite
unparalleled variety of structural type manifested by their con-
stituent organs. It is entirely beyond the scope of this paper to
enter into any extended account of the histological detail, but
it will be useful to call attention to at least a few of the main
features. Suffice to say that since the first observations on the
finer morphology of cephalopod photogenic organs made by
Joubin in 1893, a most bewildering variety of structure within
the confines of this single, narrowly limited group of animals has
been brought to light, ranging all the way from the simple
discharging glands of the luminous myopsids, and the lump of
photogenic tissue which forms the proximal photophore in the
tentacle of Lycoteuthis, through almost innumerable intermediate
types, to the astonishingly complex bull's-eye lanterns of Abra-
liopsis and the mirrored searchlights of the Histioteuthidse.
Each species has in fact its own peculiar modifications and some-
times many of them. The histology of all affords a fruitful field
of investigation, which, with all due respect to the fine work of
Chun, Hoyle and Joubin, we can truly say has been hardly
skimmed. This is especially true of the embryology and he who
attempts to work out the origin and homologies of even the
simplest of these organs will have a virgin field.

Cephalopod photophores appear only rarely to be made up

184 S. STILLMAN BERRY.

of masses of photogenic tissue without accessory structures
(intrapallial organs of Sepiolidae; proximal tentacle organs of
Lycoteuthis and Nematolampas; eyelid organs of Mastigoteuthis
glaukopis}. As a general rule they are more or less complicated.

The principal division of the organs on morphological grounds
is that already noticed which places the discharging glands of
the Sepiolidae on the one hand, — the enclosed glands of the
remaining photogenic genera on the other. The latter it is again
possible to roughly separate into three types: the no doubt rela-
tively primitive invaginated epithelial organs of which the
subocular photophores of Cranchia, Liocranchia and Leachia
are interesting examples, band-like expanses of photogenic tissue
as in Eucleoteuthis, and the spherical, ovoid or discoid organs,
often provided with the most extensive array of accessory mech-
anisms, which are found in most of the other genera.

The organs of the last mentioned class in their highest de-
velopment attain to an almost unbelievable degree of complexity
To the primary photogenic tissue, with its invariably abundant
blood and nerve supply, are here added more or less efficiently
developed reflector mechanisms, pigment cups, lenses, diaphragms
directive muscles, mirrors, windows, color screens, — even in some
cases accessory photophores, giving rise to the puzzling "double
organs" which are met with now and then in the most dissimilar
situations, so that their purpose and manner of functioning is left
even more than it otherwise would be a complete enigma. In
some cases only certain ones of these accessory structures are
developed, in other cases nearly all, as in the miniature search-
lights which yield such beautiful microscopic preparations in
the integument of the Abralioid and Histioteuthicl forms.
Space will not permit a complete description, but the presentation
of these various accessories in outline form will give an idea
their wonderful variety and serve likewise as a convenient
summary. The student desiring more detailed information is
referred to the works cited in the bibliography, particularly
those of Joubin ('93, '930, '936, '93*;, '94, '95, 105, 1050), Hoyle
(:02, 104, :09), Meyer (:o6, :o8), Vivanti (114), and the beautiful
memoirs of Chun (1030, :io).

LIGHT PRODUCTION IN CEPHALOPODS. 185

TABLE V.

COMPONENT PARTS OF THE CEPHALOPOD PHOTOPHORE.

I. Primary (photogenic tissue).

1. Photogenic cells.

2. Veins and arteries.

3. Nerves.

4. Connective tissue.

II. Secondary (accessory structures).

1. Pigment cup (almost always present, but sometimes lacking where photo-

phore is surrounded by other pigmented tissue, as the ink sac or
eyeball).

(a) Chromatophores.

(b) Specially modified pigment cells (an adaptation of preceding ?).

2. Reflector, or Tapetum.

(a) Nucleated cells.

(b) Fibers.

3. Scale cells, or "Schuppenzellen" of Chun.

(a) as reflector.

(b) as lens or cornea.

(c) in photogenic tissue.

4. Lens.

(a) Fibrillar.

(b) Cellular.

(1) Connective tissue.

(2) Modified mantle musculature.

5. Diaphragm.

(a) Chromatophores.

(b) Muscles.

6. Windo'vv.

7. Mirror.

8. Accessory photophores ("double organs").

The duplex photophores deserve a further word. These
comprise two separate masses of photogenic tissue so closely
associated together that the conclusion seems unavoidable that
in some way they function in common. Organs of this type
seem to have been first discovered by Chun, who described them
in some detail for a number of species. There is small doubt
that histological examination will show the occurrence of similar
organs in many other instances also. The double crescentic
subocular photophores of certain Cranchiidse have been briefly
described on an earlier page. Lycoteuthis (and most probably
Nematolampas also)17 possesses a number of duplex organs, the

17 Nematolampas certainly agrees in having the terminal subocular photophores
equipped with an accessory photophore. The other organs mentioned have not
yet been investigated.

1 86 S. STILLMAN BERRY.

distal organs of the tentacles, the terminal members of the suboc-
ular series, and the gastric organs, all being of this category.
In the gastric organs, the respective masses of photogenic sub-
stance, though entirely distinct from one another, are contained
within the same capsule. In the case both of these and the
terminal subocular organs, which are separated, the accessory
photophore lies beneath the principal one and the rays which
emanate from it must accordingly pass through the latter if they
are to have egress at all. In Pterygioteuthis the branchial organs
are duplex, the accessory organ being contained a little to one
side of its principal, but still within the same pigment cup.

8. POLYMORPHIC NATURE OF PHOTOGENIC ORGANS.

The question is now very near, whether so many simple and
elaborate morphological types of light-producing organs have
any especially closer genetic relationship to one another where
they are found within one and the same species or genus. And
this leads easily to another, whether the photophores of any
given species exhibit such manifold structural diversity as to
render improbable their ultimate reduction to a single primor-
dial type. The affirmation of this latter question implies the
negation of the former, and I think we may certainly say that
this seems most truly to express the facts as we have them. The
accompanying table (Table VI.), which it has seemed worth
while to elaborate upon the basis of the interesting outline given
given by Chun, shows that whereas about a third of the genera
cited each possess photophores belonging to a single general
type, nearly as many have strongly dimorphic photophores,
and an even greater number have trimorphic or polymorphic
organs. It is nothing unusual therefore to find organs of extreme
simplicity functioning as components of the same photogenic
system which contains also organs exhibiting the most varying
degrees of complexity in structural plan. While this seems to
take place almost in hit or miss fashion, I think it may be taken as
a general statement of fact that those species having a relatively
abundant development of integumentary photophores distrib-
uted over the body generally fail to evolve a great variety of
other types, the Abralioid genera providing the nearest to an

LIGHT PRODUCTION IN CEPHALOPODS. 187

exception to this rule (see Table VII.). Those species showing
the richest development of structural type in general are the ones
which depend upon intrapallial rather than integumentary organs
to serve the light producing function. Here there is sufficient
divergence among the various organs as to discourage almost at a
glance any attempt to homologize them on the basis of reference
to a single primal type. Not only their diversity, but their
extremely sporadic appearance in connection with organs and
tissues of heterogeneous origin, is strongly inhibitive of any such
view.

TABLE VI.

POLYMORPHISM IN CEPHALOPOD PHOTOPHORES.

I. Genera with Isomorphic Photophores.

Thelidiotenthis Integumentary.

Histioteuthis

Callitenthis

Benthoteuthis On arms.

Mastigoteuthis Integumentary.

Cranchia Subocular.18

Liocranchia Subocular.

Pyrgopsis

Hensenioteuthis

Bathothanma

II. Genera with Dimorphic Photophores.

Enoploteuthis Integumentary: subocular.

Leachia Subocular.

Megalocranchia

Crystalloteuthis

Toxeuma

Taonidium

Corynomma Subocular (?); intrapallial.

III. Genera with Trimorphic Photophores.

Abralia (except .4. astrolineata') Integumentary; subocular (latter di-
morphic).
Chirotenthis On ventral arms; subocular; intrapallial.

IV . Genera with Polymorphic Photophores.

Lycoteuthis In tentacles; subocular; intrapallial; 10

types (13 if 3 types of accessory organs are
counted separately).

Nematolampas In arms; in tentacles; subocular; intra-
pallial; at tip of body; probably 12 or 13
types (15 or 16 if accessory organs are
counted separately).

Lampadioteuthis In tentacles; subocular; intrapallial; — prob-
ably 7 or 8 types.

Abralia astrolineata, \

Abraliopsis, .Integumentary; tips of ventral arms; sub-

Watasenia, ocular (dimorphic).

Pterygiotenthis Tentacular; subocular; intrapallial; — 8 types. ,

Pyroteuthis Tentacular; subocular; intrapallial; — prob-
ably 8 or more types.

18 Unequal in size, but showing clear structural evidence of homology.

188

S. STILLMAN BERRY.

u
z

w

O

g
o

<!

33
6,
fa)

u

fal

a!
O
B

CH
O

i— ; o

^H X

> ^

S s

1-4 a,

O
P
0

H

§

U

z
u

D
O

§

z
3

D

O

a
u

u

Q

Pterygio-
rcuthis
giardi.

1 1 1 1 1 1 II

1 1

HH M hH HH

O IN W ^

H

^s

Pterygio-
teuthis
gemmata.

1 1 1 1 1 1 1 II

1 1

i „ 33 i 3 i

CS (N ' *t '

rt.

ro

111

1 1 1 1 1 1 "5 II

1 1

i 3 3333 i

O tN (N CS ^" '

M

?§

11 "at as en ia
scintillans

3 3 1 i i

++++++ 2

O

II 1 1 1 1 1

8 ^_

+ 3

0 ^

M

Aln-aliopsis
sp.

3 Siii

++++++ 2

o

II 1 1 1 1 1

8

+ 3

•0

h-l

|-tp

3 3 1 i

++++++ 2

'§

11 1 1 1 1 ~-

8

+ 3

M

Lamfiadio-
teuthis
iiifgaleia.

i i i i i i S 11

•O N

1 1

N 00 'rT ^T ^M"

^g

~ <•>

r^.

"TO

M

"~ ^<:

0) M M l-t f^ M

+ u

|l§

1 1 1 1 1 X 1 ' 'V-^

O

hH 01 01

1

o

-^- (N N PI CS <N

0 0

1 ^
o.|

f — v

1 1 1 1 1 1 ^ 1 II

M

1 1

W M M M IN

\ • ^ ^^ ^^ ^^ — 1

0) O

N M

HI

a

^j

_

c, 3

*j c c ~

o t;

B * c S w

2 P 3 <y . -^
E .„ ^ ^ s

: : : : : >,

•§. ^

•*- o o o ^ ii

'O

pO

41 <U H Si o, JS

.cuouu J3 ^£~
K-^oJairt^b D. (SO

•^' ^ '-*« **- '-•- C n T1 Cfi K

I- C, i_ k, 1-, C 73 >t >-.!-,

fl"-i323t,3 0; OC

'c S K * K ^ o •• ^* *o *o

rn "T3 ^3rtnJOJCrt^o3 OO
-^^^"ti^'tifj^-^c^^MW

MoggSS.SoSia ».&.&
Ba»»u^> j SHH

« C/3 <

Ventro-laterals .
Tips of ventrals

; ; ! o
: - • • ''•''• 1

en OJ w •-« *p fj .

Ql W |^J »™^ .^5 . *^

•• ^C ^C o_> ^C CQ CJ ^C w

o o
E^ ii p

Total number
Probable nurn

LIGHT PRODUCTION IN CEPHALOPODS. 189

The photogenic systems of all the species of the eight genera
having polymorphic organs are outlined in further detail in
Table VII. Those species considered having the mere largest
number of photophores are the three Abralioids, occupying the
three central columns of the table, but those exhibiting the
highest degree of polymorphism are Lampadioteuthis megaleia,
which has not been investigated histologically but must have not
less than seven or eight types of photophores in all, Lycoteuthis
diadema, with ten types, or thirteen, if the accessory organs are
counted in, and Nematolampas regalis. Lycoteuthis diadema,
with the immeasurable advantage of having had its marvelous
photogenic properties observed in the living state, is usually
cited as the example par excellence of a luminous cephalopod.
However, it is evident from sheer morphological grounds that
even this wonderful creature must yield the palm to another, if
nearly related, genus and species, — the truly amazing Nemato-
lampas regalis of the Kermadec Islands. Whether this species
will ultimately be found to display all the varied brilliance of
the red, white, and blue lights of Lycoteuthis, the fact remains that
in addition to a complete series of exactly homologous organs,
it has an entire battery of pyrotechnic engines of its own, so
there is every reason to expect a more rather than a less elaborate
illumination. The total number of photophores in this species
is in excess of ninety, which are elaborated upon no less than
twelve or thirteen different structural principles of uncertain
homologies with one another. Counting in the three types
of accessory photophores which are to be found in the eight
"double" organs (proximal tentacular, terminal subocular, and
anal), the total number of types is increased to fifteen or sixteen.
Which of the alternative figures quoted is the correct one is still
to be established by histological work.

9. SYSTEMATIC SIGNIFICANCE OF PHOTOGENIC ORGANS.

It follows almost as a corollary from what has been said in the
foregoing sections of this paper that the photogenic system
evinces a complex of features of the utmost value to the taxono-
mist. Of late years ever increasing weight has been given to it,
and the presence of constant differences, even though minute, in

I9O S. STILLMAN BERRY.

its components, is now admitted practically without debate as
ample ground for taxonomic discrimination. Where such differ-
ences are shown to occur, further differences in the remaining
organization seem practically predestined for eventual discovery.
Good characters for specific discrimination are to be found, not
only in the presence or absence of photogenic organs, but also in
their distribution on or within the body, in their number, in
their size, and in the veriest details of their intrinsic structure.
The taxonomist has in fact few more convenient points of attack
in the pursuit of his primary objects of classification and relation-
ship than that afforded by the light organs. And this is exactly
what we find, if to somewhat less degree, among the fishes and
the few other groups where the photogenic organs have attained
some considerable complexity. One can construct a fairly
workable taxonomic key based on the photogenic organs alone,
for such species as possess them.

10. PROBABLE POLYPHYLETIC ORIGIN OF PHOTOGENIC ORGANS.

Before concluding this paper a somewhat general answer may
be attempted to a question which has no doubt occurred more
than once in the mind of the reader, and which indeed has been
touched upon very nearly on more than one occasion — Is photo-
genesis a primitive function among cephalopods? In other
words, are our present day species descended from an ancestral
photogenic stem, some branches of which have now yielded up
the function? Or has photogenesis arisen several times in this
class of animals, possibly to meet altogether diverse conditions or
associations in the environment, so that its presence therefore
becomes of secondary ratl er than primary significance?

At first glance the widespread distribution of the function in
the great and, comparatively, primitive oegopsid group of cephalo-
pods favors an affirmative answer to our first query. But in
reply to this it may be said that the varied pelagic environment
of these forms would almost per se favor the development of the
light-producing function after a manner which would be hardly
likely to hold true among the more littoral Myopsida and Octo-
poda, the former of which are mainly frequenters of much
shallower water than the (Egopsida, the latter hardly ever

LIGHT PRODUCTION IN CEPHALOPODS.

pelagic at all, and then generally surface forms or confined to
the shallower water like so many of the myopsids.

There are many other arguments which may militate against
any theory of monophyletism and as strongly support the con-
trary view as brought out by the last query above. These,
having already been largely elaborated elsewhere or to be dealt
with in another connection later on, need be merely summarized
here. Such considerations are:

1. The uneven distribution of photogenic organs throughout
the entire group, and, as a corollary of this, their appearance in
distantly related groups more or less sporadically.

2. The variety and sporadic character of the development
of photogenic organs in different regions of the body.

3. The large number of strongly diverse structural types.

4. The evidence from ecological considerations, the distribution
upon the body, and similar facts that these organs have arisen
in response to very diverse environmental requirements.

How then may one bespeak a photogenic sy 'stem? Exactly as
one speaks of a muscular system, or a skeletal system, or a
receptor system in almost any animal body. The term is used
in the sense not necessarily indicating an aggregation of homo-
logous structures, but an assemblage of organs within a single
organic body exhibiting more o'r less similar or coordinate
physiological reactions, if at times neither in fact phylogenetically
nor ontogenetically related.

ii. CONCLUDING NOTE.

This paper is mainly a compilation from the scattered work of
other authors. No doubt there are omissions, but the aim has
been to present simply a concise summary of the knowledge of
this subject which has been gained to the present time. It cannot
be too strongly emphasized that not only are many more species
of luminous cephalopods likely to be discovered in the future,
but some of those now known but not yet recognized as possessing
photogenic properties are likely to be revealed as having them.
Of the known luminous forms some will no doubt prove to
possess luminous organs or properties additional to those de-
scribed. Bearing all this in mind, if this little paper but fur-

192 S. STILLMAN BERRY.

nishes some delving student just a little better base of attack on
his problem than might otherwise have been afforded him, its
purpose will have been fulfilled.

12. SUMMARY.

1. Light production is an unusually widespread phenomenon
in the molluscan Class Cephalopoda.

2. Although unknown in the Order Tetrabranchiata, scarcely
developed in the octopod section of the Dibranchiata, and occur-
ring little more than sporadically among the Myopsida, over one
half of all described (Egopsida are known to possess photogenic
properties.

3. The actual production of light by living cephalopods has
been observed only rarely, but in species of sufficiently diverse
relationship to confirm the evidence drawn from the morphology
and histology of organs found in the remaining species.

4. The light of some species exhibits remarkable brilliance.

5. The color of the light emanating from the respective organs
within the same species or in different species may exhibit striking
differences in both intensity and quality, but it is not known to
what extent this is actually due to inherent diversity in the physi-
cal properties of the light rays themselves.

6. Photogenic organs may occur in almost any portion of the
body in this group of animals, but the outer integument, eyeball
and pallial chamber are the situations most favored. They are
often internal and able to function only by reason of the trans-
parency of the body tissues in the living state.

7. The organs are predominantly, but by no means exclusively,
ventral in distribution.

8. The organs are strongly polymorphic, even in the same
species, varying from comparatively simple bodies of photogenic
tissue to the highly complex "searchlight" types.

9. Numerous duplex organs, or organs with accessory photo-
phores, are known to occur.

10. Luminous organs in the Myopsida are usually of the type
known as discharging. Those of the other groups are entirely of
the enclosed or ductless type.

11. The maximum polymorphism in the photophores of any

LIGHT PRODUCTION IN CEPHALOPODS.

single species occurs in Nematolampas regalis Berry, from the
Kermadec Islands, where the 90 or more organs are elaborated
upon 12 or 13 more or less diverse structural principles.

12. The occurrence, distribution, arrangement, and morpho-
logical detail of photogenic organs in cephalopods are features
of considerable taxonomic importance and yield valuable clues
as to the relationship and classification of the genera and species
even where still unknown anatomically.

13. The best evidence seems to indicate that the photogenic
organs in this group of animals are polyphyletic and more or
less sporadic in origin, hence that light production in cephalopods
is not an essentially primitive or ancestral function to be regarded
as now lost in many members of the group.

13. LITERATURE.16

Berry, S. S.

'13 Nematolampas, a Remarkable New Cephalopod from the South Pacific.

BIOLOGICAL BULLETIN, v. 25, pp. 208-212, i fig., AuguSt, 1913.
'isa Teuthological Miscellany No. i. Zoologischer Anzeiger, Bd. 42, pp. 590-

592, October, 1913.

Chun, C.

'oo Aus den Tiefen des Weltmeeres. Jena, 1900.

'03 Aus den Tiefen des Weltmeeres. Zweite Auflage, Jena, 1903.

'033 Uber Leuchtorgane und Augen von Tiefsee-Cephalopoden. Verhandl. d.

deutschen Zoologischen Gesellschaft, 1903, pp. 67-91, text figs. 1-14, 1903.
'10 Die Cephalopoden. i. Teil: Oegopsida. Wissenschaftliche Ergebnisse d.

deutschen Tiefsee-Expedition Valdivia, Bd. 18, pp. 1-402, 2 pis. and 32

figs, in text, Atlas of 61 pis., 1910.
'13 Cephalopoda. Report Scientific Results " Michael Sars " North Deep Sea

Expedition, 1910, V. 3, pt. i, Zoology, pp. 1-28, text figs, i-n, pi. 1-2,
1013.

Dahlgren, U.

'16 The Production of Light by Animals. Journal Franklin Institute, Feb.-
April, 1916, pp. 1-75, text figs. 1-20.

Darwin, C.

'60 Journal of Researches into the Natural History and Geology of the Coun-
tries Visited during the Voyage round the World of H. M. S. " Beagle "
under the Command of Captain Fitz Roy, R.N. 8vo, pp. 1-551, illus-
trated, London, 1913 (first edition May, 1860).

Holder, C. F.

'87 Living Lights. A Popular Account of Phosphorescent Animals and
Vegetables. 8vo, 127 p., 28 pis., New York, 1887.

16 A more exhaustive bibliography of the subject is in preparation for a later
paper of this series.

194 s- STILLMAN BERRY.

Hoyle, W. E.

'02 The Luminous Organs of Pterygioteuthis margaritifera, a Mediterranean
Cephalopod. Memoirs and Proceedings Manchester Literary and Philo-
sophical Society, V. 46, No. 16, pp. 1-14, text fig. i— 6, June, 1902.
'04 Reports on the dredging operations off the west coast of Central Americ-
... by the ..." Albatross," etc. Reports on the Cephalopoda. Bula
letin Museum Comparative Zoology, V. 43, pp. 51-64, pi. 9-10, March, 1904.
'08 Address to the Zoloogical Section, British Association for the Advancement
of Science, Leicester, 1907. Report British Association Advancement
Science 1907, pp. 520-539 [1-20], text fig. A-D, 1908.

'09 The Luminous Organs of Some Cephalopoda from the Pacific Ocean.
Proceedings 7th International Zoological Congress, Boston Meeting (ad-
vance print), pp. 1-5, Cambridge, 1909.
Ishikawa, C.

"13 Einige Bemerkungen iiber den leuchtenden Tintenfisch, Watasea nov. gen.
(Abraliopsis der Autoren) scintillans Berry, aus Japan. Zoologischer
Anzeiger, Bd. 43, pp. 162-172, text fig. 1-6, December, 1913.
Joubin, L.

'93 Recherches sur 1'appareil lumineux d'un cephalopode, Histioteiithis ruppellii,
Verany. Bulletin Societe Scientifique Medicale de 1'Ouest, V. 2, pp.
49-80 [1-32], text fig. i-ro, 1893.

'933 Note complementaire sur 1'appareil lumineux d'un cephalopode: Histio-
teuthis Ruppellii, Verany. Bulletin Societe Scientifique Medicale de
1'Ouest, V. 2, pp. 161-169 [1-9]. 1893.

*93b Note sur une adaptation particuliere de certains chromatophores chez un
cephalopode (1'oeil thermoscopique de Chiroteuthis-Bomplandi Verany?).
Bulletin Societe Zoologique France, V. 18, pp. 146-151 [1-6], i text fig., 1893.
'93C Quelques organes colores de la peau chez deux cephalopodes du genre
Chiroteuthis. Memoires Societe Zoologique France, V. 6, pp. 331-343
[1-13], text fig. 1-12, 1893.

'94 Nouvelles recherches sur 1'appareil lumineux des cephalopodes du genre
Histioteuthis. Bulletin Societe Scientifique Medicale de 1'Ouest, V. 3, pp.
[1-15], text fig. 1-7, 1894.

'95 Note sur les appareils photogenes cutanes de deux cephalopodes: Histiopsis
atlantica Hoyle et Abralia oweni (Verany) Hoyle. Memoires Societe
Zoologique France, V. 8, pp. 212-228 [1-17], text fig. i-n, 1895.
'05 Note sur les organes lumineux de deux cephalopodes. Bulletin Societe

Zoologique France, V. 30, pp. 64-69, text fig. 1-2, 1905.

'053 Note sur les organes photogenes de 1'oeil de Leachia cyclura. Bulletin
Musee Oceanographique Monaco, No. 33, pp. 1-13, text fig. 1-7, April, 1905.
'12 Etudes preliminaires sur les cephalopodes recuellis au cours des croisieres
de S. A. S. le Prince de Monaco, ire Note: Melanoteuthis lucens nov.
gen. et sp. Bulletin Institut Oceanographique Monaco, No. 220, pp. 1-14,
text fig. 1-12, January, 1912.

'iza Sur les cephalopodes captures en 1911 par S. A. S. le Prince de Monaco.
Comptes Rendus Seances Academic Science, Paris, V. 154, pp. 395-397
[1-3], 1912.

Meyer, W. T.

'06 Uber das Leuchtorgan der Sepiolini. Zoologischer Anzeiger, Bd. 30, pp.
388-392, text fig. 1-3, July, 1906.

LIGHT PRODUCTION IN CEPHALOPODS. 1 95

'08 Uber das Leuchtorgan der Sepiolini: II., Das Leuchtorgan von Hetero-
teuthis. Zoologischer Anzeiger, Bd. 32, pp. 505-508, text fig. 1-4. January,
1908.

Naef, A.

'12 Teuthologische Notizen, 1-2. Zoologischer Anzeiger, Bd. 39, pp. 241-248,
March, 1912.

Sasaki, M.

'12 Hotaru-ika. Toyama-ken, Sui-san Kumiai Hokoku, pp. 1-29, 2 pis. and
chart, 1912. (In Japanese.)

'13 Hotaru-ika. Zoological Magazine, Tokyo, V. 25, pp. 581-590, pi., 1913-
(In Japanese.)

'14 Observations on Hotaru-ika Watasenia scintillans. Journal College Agri-
culture, Tohoku Imperial University, V. 6, pp. 75-107, i text pi., pi. 1-3,
November, 1913.

Tryon, G. W., Jr.

'79 Cephalopoda. Manual Conchology, V, i, pp. 1-316, pi. 1-112, Philadel-
phia, 1879.

Verany, J. B.

'51 Mollusques mediterraneans, observes, decrits, figures et chromolitho-
graphies d'apres le vivant, I., Cephalopodes de la Mediterranee. G£nes,
1851.

*

Vivanti, A.

'14 Contributo alia conoscenza dei Cefalopodi abissali del Mediterraneo.
Ricerche sulla Carybditeuthis maculata n. g. n. sp. dello Stretto di Messina.
Archivio Zoologico Italiano, V. 7, pp. 55~79. text fig. i, pi. 3~5. I9M-

Watase, S.

'05 Luminous Organs of Abraliopsis, a New Phosphorescent Cephalopod from
the Japan Sea. Zoological Magazine, Tokyo, V. 17, PP- 119-122. l text
fig., June, 1905. (In Japanese.)

RAPIDITY OF ACTIVATION IN THE FERTILIZATION

OF NEREIS.

H. B. GOODRICH.

The following experiments were designed as a test of the
rapidity of the action of the spermatozoon of Nereis in relation
to the initiation of the processes of maturation. The almost
instantaneous effect of the contact of the spermatozoon in stimu-
lating the formation of the fertilization membrane and, in this
case, of the jelly is well known. It is also usually observed that,
in the case of those eggs in which maturation follows insemination,
that polar-body formation will occur without fail even if later
observation shows that no cleavage follows. It might be con-
ceived that the continued action of the spermatozoon were
necessary to cause maturation. The results outlined below
indicate that maturation in Nereis follows almost, if not quite as
brief an application of the stimulus as is necessary to initiate the
jelly formation.

The experiments of Lillie ('n) on Nereis limbata showed
clearly that the removal of the spermatozoon later than twenty-
one minutes after insemination by a process of centrifuging
did not interfere with the maturation of the egg. It was, how-
ever, found impossible by this method to remove the sperm
earlier than twenty -one minutes after fertilization. Various
workers have suggested (I am indebted to Dr. Chambers for
first calling this to my attention) that the Barbour apparatus
for micro-dissection offered a means of removal during this earlier
period. The manipulation of the instrument for this p'urpose
has proved most successful. The spermatozoon may readily
be removed shortly after attachment and with more difficulty
later because of the increasing strength of adherence to the
egg and the viscidity of the head of the spermatozoon. The
viscidity is shown in attempts to remove the spermatozoon at
about 35 minutes after insemination. The head, remaining
attached in the region of the perforatorium may be extended by

196

FERTILIZATION OF NEREIS. 1 97

the needle some five or ten times its original length. In these
later stages the method of removal by centrifuging is more
practical.

The experiments were made during the summer of 1919 at the
Marine Biological Laboratory at Woods Hole, Mass. Sperm and
eggs were obtained from animals caught the evening before and
kept in the upper compartment of the refrigerator during the
night. Sperm dilutions of about 1/400 to 1/500 were used for
most of the experiments. This dilution is not such as will
insure that only one spermatozoon will come in contact with
each egg (Lillie, '15), but it is sufficiently dilute so that poly-
spermy rarely results. Polyspermy was noted in a few of the
eggs in experiments 6 and 7. The eggs were placed in a dilute
suspension of chinese ink, at once fertilized and transferred to
the coverslip which forms the roof of the moist chamber. The
coverslips had been previously carefully cleaned of all traces of
oil or grease in order to allow the drop of water to spread, thus
forming a thin film which compresses the eggs slightly as is neces-
sary in order to hold the egg firmly during the operation. The
apparatus was always in complete readiness for the operation
and it was usually possible to locate promptly an attached
spermatozoon and to carefully push or rub it from its point of
attachment to the egg. Spermatozoa remained immobile after
detachment. Two or three operations were usually carried out
on each coverslip. More were not attempted because the delay
would allow too great an evaporation of water in the necessarily
imperfect moist chamber, thus causing greater compression of the
egg and also greater concentration of the sea water. The posi-
tion of the operated eggs among others was plotted, the coverslip
removed and placed under a binocular microscope, the selected
eggs were isolated and removed to separate dishes for observation.
Frequently the remainder of the eggs on the cover slip was also
removed for cleavage counts (Control 2). These were the last
to be removed and in some cases had by that time undergone
considerable compression (in other cases the cover slip was slightly
flooded to prevent this). This may in part account for the low
percentage cleavage recorded frequently under Control 2.

The eggs were then kept under observation to note the forma-

198 H. B. GOODRICH.

tion of the polar bodies and in the second series (Table II.) to
note possible cleavage. In all but two of the sixty eggs the
polar bodies formed. I am inclined to think that as these two
were among the earlier ones observed in this case the negative
record may be due to faulty observation. Also in only three
cases and one of these from a clearly polyspermic lot of eggs was
cleavage observed. The time of removal varied from i%
minutes to 13^ minutes after insemination. It is impracticable
to remove the sperm in most cases earlier than two minutes after
insemination as it is difficult to discriminate between spermatozoa
lying against the egg and those that have effected an attachment.
The average of elapsed time from insemination to removal was
6.2 minutes and in sixteen cases the spermatozoon was removed
in less than 4 minutes after attachment, and in five cases in less
than 3 minutes. It seems therefore clear that in so far as
maturation is concerned, the full stimulating effect of the sperma-
tozoon is effective within a very few minutes after attachment
and quite possibly it is only a matter of seconds.

The results clearly support the concept that the first phase of
activation of the egg-membrane formation and maturation is
initiated by the spermatozoon which "activates a substance, or
ferment-like bodies, contained within the egg" (Lillie, '19, p.
159), rather than by the continuous introduction of some lysin-
like substance through the slender perforatorium. For it seems
improbable that the spermatozoon could introduce in less than
two minutes through the perforatorium which has a cross section
area of perhaps one ten-millionth of the surface area of the egg,
a sufficient amount of material to take part in reactions through-
out the egg. It should, however, be noted that under certain
conditions (heating, Just, '15) it is possible to initiate jelly forma-
tion without maturation following. This may be taken to indi-
cate that a lesser stimulus gives a lesser result and possibly if it
were practicable to remove the sperm more promptly that jelly
formation only would result. On the other hand in the cited
case we may be dealing with a stimulus differing in kind rather
than in degree. It should also be noted that these results are
not at variance with the concept that cortical changes form an
all important intermediate step in the activation of the egg.

FERTILIZATION OF NEREIS.

199

In some cases (Table II.) it was possible to note the time of
formation of the first polar body. The average of elapsed time
after insemination was 35.8 minutes and no correlation was
noted between the duration of attachment of the sperm and the
time of formation of the polar bodies.

TABLE I.

Date.

Experi-
ment

Num-
ber.

Time of

Fertiliza-
tion.

Egg-

Time of
Removal
of Sperm.

Minutes
Elapsed.

Polar
Body
Forma-
tion.

Cleav-
age.

Controls.

No. i.

No. 2.

July

17

I

10.48 '

A

10.56

8 +

—

+

60%

B

11.02

14 +

—

+

1 1

2

2.44

A

2-54

10 +

_l

+

90%

B

2-58

14

+

_i

+

"

18

3

5-29

A

5-32^

31A

—

_i

+

65%

B

5-37^

8^

+

_i

+

' '

iQ

4

9-59

A

10.03

4

—

_i

+

70%

B

io.o8H

91A

+

i

+

i i

C

10. IO

ii

+

_i

+

i i

20

5

2.29^

A

2.33J4

4

+

—

+

60%

6

2.40

B

2.51

ii

+

— 2

+

"

21

7

9-40lA

A

9.48

7^

+

+2

+

90%

8

io.i4lA

A

10. 18^

4

+

+

+

40%

B

10.21%

7M

+

+

+

**

9

11.26

A

n.27%

i%

+

1

+

50%

B

11.32

6

+

1

+

tt

10

2-54lA

A

3-00

51A

+

—

+

70%

B

3-03

&yz

+

—

+

"

ii

3-13

A

3-I6H

31A

+

_1

+

B

3-21 Yi

81A

+

_1

+

As noted in Table I. some eggs were fixed at about time of the
first cleavage with view to study of cytological changes. Too
few of these have survived the ordeal of embedding and sectioning
to make any detailed study valuable. I noted that in some cases
only chromosomal vesicles and in other cases well-formed chro-
mosomes were present at the time of the first cleavage spindle.
This may be compared with the observation of Lillie ('15) where
chromosomes were formed in eggs from which sperm were re-
moved by centrifuging but not in eggs which were caused to
maturate by centrifuging without insemination.

The appended tables outline the experiments. Full data are

1 Egg fixed at about time of expected first cleavage. No indication of cleavage
at that time.

2 Poly-spermy observed in same lot on cover-slip.

2OO

H. B. GOODRICH.

not present in all cases. The experiments were in part pre-
liminary and the desirability of various controls and observation
became apparent as they progressed. In as much as the work

TABLE II.

u
rt

Q

ll

ll
w

•ojd

•5 *•• tj!

be

em

W

« ° S
Jg£

Minutes
Elapsed.

Polar Body
Formation.

Time
Formed.

Minutes
Elapsed.

Cleavage.

Controls.

No.
i.

No.

2.

No.
3-

July

23

12

9-47

A

9-50

3

+

IO.I9

33

—

+

60%

B

9-5334

634

3

+

C

9-5434

734

+

10.20

34

—

+

"

13

IO.2O

A

10.2334

3M

+

IO.52

32

—

+

80%

B

10.2534

sH

+

—

+

"

14

11.0514

A

11.0834

3

+

—

-f-

15

H.3634

A

11-3834

2

+

—

+

B

11.4234

6

+

—

-|-

C

n-4334

7

+

—

-{-

16

2.40

A

2.4834

834

+

3-21

31

—

+

90%

B

2-5334 1334

+

3.17

37

—

+

' '

17

3.1634

A

3.2434

8

no. obs.

4

+

B

3-2434

9

+

3-49H

33

4

-f

C

3-26M

10

+

_4

+

D

3-28

n34

+

_4

+

24

18

IO.I5

A

10.1834

334

+

10.46

31

4

-j-

B

10.22

7

+

10.49

34

4

+

19

I0.5I34

A

10-55

334

+

11.2234

31

4

-f

B

10.5734

6

+

4

+

C

10.5834

7

no obs.

— 4

+

20

11.26

A

11.29

3

+

H.5634

3034

4

+

B

H-3334

634

+

11-5734

3i34

— 4

+

21

3-i8}4

A

3-21%

334

+

3-53

35

_4

+

B

3-24

534

+

3-53

35

— 4

+

C

3-25

634

+

3-52

34

4

+

25

22

9-33

A

9.36

3

+

10.0534

3234

—

-j-

60%

B

9.38

5

+

10.06

33

—

+

' '

C

9-39

6

+

10. 06

33

—

+

1 1

23

10.16

A

10.1834

23^

+

—

+

58%

B

10.2134

534

+

—

+

*'

24

10.4834

A

10.5134

3

+

—

+

57%

B

10.52

334

-(-

—

-(-

* '

C

10.55

634

+

—

+

11

D

10.57

834

+

—

+

' '

25

11-34

A

n-3734

334

+

12.15

41

—

+

45%

B

n-4334

934

+

—

+

' *

26

2.19^4

A

2.4234

4M

-(-

2-57

3734

—

-j-

25%

21

2.47

A

2-49

2

+

3-25

38

—

+

55'-

+

28

3.5434

A

3-57

234

+

3-37

4234

—

+ 80%

+

B

4.00

534

+

3.36

4i34

—

+ 80%

+

26

29

9-52

A

9-55

3

+

10.27

35

—

+ ! 66%

+

30

10.29

A

10.32%

3/4

+

ii. 02

33

—

+

' '

+

B

10.35

6

+

11.03

34

—

+

57%

31

II-0534

A

11.0734

234

+

H.3834

33

—

+

75%

+

B

II. IO^

5',

-f

11.42 y<>

37

—

+

* *

+

3 Egg disintegrated.

4 No specific record in original notes of the absence of cleavage.

FERTILIZATION OF NEREIS. 2OI

was stopped by the end of the Nereis " Run" and as the observa-
tions in regard to maturation were complete it seems to me to be
best to present the work in its present form.

TABLES.

The headings are mostly self-explanatory. Under Polar Body
Formation and Cleavage a plus sign ( + ) indicates that the
process was observed and a minus ( — ) the reverse. Control
No. i are eggs from the same lots as in the experiments fertilized
a few minutes before the experiment began from the sperm of
the same male used. Control 2 are eggs not operated on from
the coverslip and Control 3 are eggs of the same inseminated lot
mixed with chinese ink from which those placed on the cover
slip were taken. The plus sign ( + ) under Controls I and 3
indicates the practically 100 per cent, cleavage usually realized
with Nereis eggs.

WESLEYAN UNIVERSITY,
January, 1920.

LITERATURE CITED.
Just, E. E.

'15 Initiation of Development in Nereis. BIOL. BULL., Vol. 28.

Lillie, F. R.

'n Studies of Fertilization in Nereis, I and II. Jour. Morph., Vol. 22.

'15 Studies of Fertilization, VII. BIOL. BULL., Vol. 28.

'ig Problems of Fertilization. University of Chicago Science Series.

IMMUNITY AND THE POWER OF DIGESTION.1

VERA DANCHAKOFF,
COLUMBIA UNIVERSITY.

The proteins are the fundamental chemical constituents of the
living protoplasm in all our tissues and organs. But the cells of
the tissues do not take up the proteins as such, they are known to
live on the products of protein digestion, on their building stones,
the various amino-acids. And by a long process of synthetic
character the cells of a multicellular organism build up out of
these building stones their own individual proteins of a highly
specific structure.

The multicellular organism possesses special organs of digestion
which successfully perform the work of splitting the ingested
food to the stage of amino-acids, in which stage they become
capable of passing through the intestinal wall ; they are absorbed
here and transported to all our tissues, which in this manner are
provided with various substances for their synthetic work in
order to replace the wear and tear of the living substance during
adult life or to build up organs and tissues during developmental
stages.

Now are the cells of all our organs and tissues which, under
usual conditions, are supplied for their synthetic work with ready
made building stones, without one of the most characteristic and
important properties of the living matter, viz., the power of
splitting the chemically more complex proteins to the stage of
amino-acids? Or are these cells not manifesting a digestive
potency inherent in them, because they have no chance of
manifesting it, because the ingested food does not reach them
primarily and undergoes the hydrolytic changes before it is
capable of being absorbed? Is it a lack of power or a lack of
opportunity which prevents the cells of our tissues from mani-
festing a digestive capacity under usual conditions?

1 Given at the meeting of the Section of Biology, New York Academy of Sciences,
October 13. 1919.

202

IMMUNITY AND THE POWER OF DIGESTION. 2O3

A most wonderful property exhibited by the tissues themselves
is to inhibit growth of foreign tissue. This resistance of the
organism against proliferation of any heterogeneous cellular
elements, be it normal tissues, tumor cells or microorganisms, is
more generally known under the term of immunity. There is
little need to emphasize the importance of any new information
concerning this problem. Disregarding, however, the most earn-
est efforts and attempts to throw upon this property light suffi-
cient to master it, no definitive knowledge has yet been acquired
concerning the nature of this property.

In all organisms, plants and animals as well, the power of
resistance against proliferation of any foreign tissue does not
seem to depend upon the activity of a special organ in the
organism. A graft of heterogeneous tissue simply does not
take, a tumor graft does not grow, microorganisms do not pro-
liferate, die and disappear in an immune or immunized animal.
A living cell endowed with the faculty of proliferation, for ex-
ample, a tumor cell, well known for its high rate of proliferation
and on account of this property disastrous to the organism, on
which it has settled, if grafted on an immune organism, will
certainly find here all of the elements needed for the building
up of its own protoplasm, for we know that the tissues of the
organism are abundantly provided with various amino-acids,
and these not being specific, can be used equally well by any
tissue. And still a tumor cell will not live on an immune organ-
ism. It dies and disappears.

If now we inquire into the relation between the digestive power
of the cell and resistance in a unicellular organism, we will
see that both phenomena in this case are very closely connected.
A unicellular organism can ingest another living organism, bac-
teria as well and the latter, while within a cytoplasmic vacuole of
the former, are killed and consecutively digested by the enzymes
of the phagocyte. The unicellular organism in this case proves
to be immune against a possible intracytoplasmic proliferation
of the ingested living individual, primarily and solely because
of its digestive capacity.

The question whether in a multicellular organism any relation
exists between its resistance against proliferation of a foreign
tissue in it and a digestive power of its tissues is obscured by the

2O4 VERA DANCHAKOFF.

fact that the tissues of the organism do not exercise under normal
conditions any digestive power, and it is not definitely settled,
which of the various tissues, if any, besides those connected with
the digestive tract, are endowed with a digestive power. And
still there exist sufficient indications that the tissues of a multi-
cellular organism, besides those highly specialized of the digestive
tract, are equipped with enzymes which confer on them a diges-
tive power similar to that so characteristic of the special elements
in the digestive tract. If the tissues of a multicellular organism
be put under the stress of starvation, under which condition no
amino-acids will reach them from the absorbing surface of the
digestive tract, the tissues will in this case use up tissue proteins,
they themselves hydrolyze the proteins or digest them, which
work is made possible by the presence of enzymes in these cells.
And another example of digestive work performed by tissue cells
of a multicellular organism can be found in the growth of tissue
on artificial media. Here again the transplanted bits of tissue
do not find within the culture medium ready building stones for
the building up of their protoplasm in the form of amino-acids,
but only proteins of the blood plasma clot. Here again the cells
of the tissues themselves have a chance of performing the splitting
or digestion of the proteins, and they do it. In both these cases
we base our conclusions of the presence of enzymes in the tissue
cells on the results of the experiments: on the continuation of the
output of nitrogen by the organism during a diet free from
nitrogen in the starvation experiment, and on the further growth
of the tissue in the culture experiment. The fact only not the
mechanism of the digestive activity by tissue cells is determined
in both cases.

Some information concerning the mechanism, or rather con-
cerning one of the mechanisms, by which cells of our body may
digest, i.e., split and assimilate protein may be gained by micro-
scopical study. And this study informs us that in embryos
developing from mesoblastic eggs, i.e., from eggs containing
great quantities of yolk, as for example, from birds eggs, all of
the tissues contain at an early stage a great quantity of yolk
granules in their cytoplasm, which are digested intracellularly
and assimilated. Endodermal cells as well as ectoderm and
mesoderm cells, even primitive germ cells, all are endowed with

IMMUNITY AND THE POWER OF DIGESTION. 2O5

an intracellular digestive capacity. And this capacity is surely
not lost during the development of the embryo, at least the
mesenchymal cells continue to exercise this faculty, and wher-
ever cells in the developing organism have lost their normal
correlation with other tissues, as for example, red blood corpuscles
during the rearrangement of vessels, mesenchymal cells are seen
to ingest them and digest them within their cytoplasm. The
mechanism of this digestive activity is analogous to that observed
in unicellular organisms. Enzymes are not given off as in the
digestive tract, they work within the cytoplasm, they are endo-
enzymes. To a digestive activity of the mesenchymal cells is
due the disappearance of aseptic emboli, trombi and infarcts;
small sequesters may be entirely resorbed and catgut sutures
disappear in the organism in a short time.

An intracellular digestive activity is exhibited in an intense
way by the tissues of embryos developing from mesoblastic eggs
in the early stages of embryonic development and only occasion-
ally during adult condition. This power is dormant in the
tissues of mammals during embryonic development, the embryo
receiving all the materials needed for its development in the
form of amino-acids from the maternal placenta, but is awakened
at the first opportunity of being confronted with unsplit protein.

The digestive power of the tissues is well evident in respect
to cellular protein in the dead form. Only in a unicellular
organism did we see that their digestive activity confers on them
a power over living organisms and cells as well and therefore
becomes the source and the cause of their immunity against a
possible proliferation of the ingested organisms. May the
digestive capacity of the tissues of a multicellular organism, of
which we have seen a few examples, be also in some way con-
nected with the failure of a heterogeneous graft to take? Is
this power not the deleterious factor which cannot be overcome
by the grafted heterogeneous tissue and which inhibits its growth
and proliferation?

The results of a long series of experiments, which I am going to
illustrate, show without any doubt that at least in some cases
it is the digestive activity of an adult mesenchymal cell which
inhibits the growth of a heterogeneous tumor or rather destroys
the actively growing tumor. The experiments were made with

206 VERA DANCHAKOFF.

different tumors, but only the Ehrlich sarcoma, and the tumor
known at the Crocker Fund Laboratory, under the number 180,
gave demonstrable results. It is known that mammalian tumors
may be grown easily on chick embryos, but not on adult animals.
Also normal adult chick tissue grows well on embryos of the same
species. Therefore, embryonic tissues, more particularly the
chick allantois, have been used by me as a culture medium, to
bring together mammalian tumors and various adult chick
tissues. It was expected that the study of the interaction of the
tissues of the adult naturally immune fowl and of the mammalian
tumor cells grown in a culture medium, for both equally favor-
able, might show whether the digestive capacity of the tissues
more particularly that of the adult splenic mesenchymal cells may
in any way be connected with the resistance offered by the adult
animal to the grafting of the tumor on it. The experiments
have shown that the Ehrlich sarcoma gives invariably a good
growth if grafted alone, but if grafted in a mixture with the
spleen, disappears even after a short period of proliferation. The
short-lived fame of the small lymphocytes thought to be re-
sponsible for this disappearance is still in our memory. Only an
absolute disregard of microscopical findings can explain how
microscopical pictures similar to thor.e shown here (Figs. I, 2,
3 and 4), have been overlooked and how the small lymphocyte
could become the fetish of the immunity.

Microscopical preparations, as seen in retouched photographs,
which accompany this paper, illustrate in a striking way the
process of disappearance of tumor foci surrounded by the adult
splenic mesenchyme in the allantois. Two lines of activity are
observed in the mesenchyme, it splits off numerous mobile cells
of hemoblastic nature which differentiate further into granular
leucocytes. This developmental potency is exhibited in an even
more intensive way by the splenic adult mesenchyme, if the
splenic tissue is grafted alone on the chick allantois. But the
fact of grafting the splenic tissue together with tumor reveals
in it a new potency and this is its power of isolating and surround-
ing tumor cells, of enclosing them in vacuoles and of digesting
them within these vacuoles (Figs. I, 2, 3 and 4). The adult
splenic mesenchymal cell is apparently attracted toward the
mammalian tumor cell. Contrary to the embryonic mesen-

IMMUNITY AND THE POWER OF DIGESTION. 2O7

chyme, the adult mesenchymal cells, once close to the tumor
cell, will not indifferently pass by, but together with other
mesenchymal cells, will tightly surround it. The tumor cell, in
response to the approach of the adult mesenchymal cell, with-
draws its cytoplasmic processes, becomes immobile and assumes
soon a spherical shape. The Figure I illustrates a tumor focus
in which the cells still intensely proliferate in the center as may
be seen from numerous mitoses present. The tumor focus is
however surrounded by a zone of mesenchymal syncytium of
splenic origin. The cells of this tissue encircle the tumor cells
at first surrounding them very tightly (Figs. I and 2, x). The
adult splenic mesenchymal cells of the fowl treat a heterogeneous
mammalian tumor cell, which, if present alone in the embryonic
allantois, would live and proliferate in no other manner than if
it were a block of dead protein. They gather around the tumor
cells and enclose them into a capsule and then secrete a fluid
within the little cavity occupied by the tumor cell. The nature
of this fluid is such that a disintegration of the tumor cell takes
place and a complete splitting of its proteins, it must therefore
contain proteolytic enzymes. The tumor cell gradually loses its
structure (TV, Figs. 1,2,3 and 4) is transformed into a block of
structureless protein (TV, Figs, i, 2, 3 and 4) and finally disap-
pears completely. Tumor cell after tumor cell is digested in this
manner. In the case of the Ehrlich sarcoma, the rate of digestion
of the tumor cells by the splenic mesenchyme is higher than the
rate of proliferation of the tumor cells, and the grafted tumor, even
after a good start of development, disappears. The same process
is observed in relation to the sarcoma 180. Only in this case the
rate of proliferation of these tumor cells is higher than the rate
of their digestion by the splenic mesenchyme, and the tumor
still grows in spite of the existence of a peripheral zone around it,
in which the process of digestion of tumor cells by mesenchymal
cells is most evident.

The study of microscopical preparations allows us to follow
the gradual changes which the tumor cells undergo within the
vacuoles surrounded by mesenchymal cells, changes which lead
to the full disappearance of the tumor cells. Figures 2, 3 and 4
show very clearly how healthy tumor cells H Tc are cut off from
the tumor focus, the tumor cells withdrawing their processes

208 VERA DANCHAKOFF.

and rounding up, how the tumor cells gradually lose their
structure TV, their cytoplasm becoming vacuolized, their nuclei
pycnotic and how finally the tumor cells disappear entirely.

Now in respect to the small lymphocytes, may they not be
still connected in some indirect way with the disappearance of
the tumor cells? Round-cell infiltration has been described
around the disappearing tumor grafts within the allantois. But
though the small lymphocyte is a round cell, not every round cell
is a small lymphocyte. Round cell infiltration exists indeed
around such grafts. This infiltration, however, at a closer study
proves to consist not of small lymphocytes, but of round, mobile
cells of hemoblastic nature which, as mentioned above, differ-
entiate into granular leucocytes. As has been shown in one of
my previous papers, the small lymphocytes are not very viable
in the allantois and in most cases quickly disappear in the grafted
splenic tissue.

It is a positive tropism between the adult mesenchymal cell
and the tumor cell expressed in the phagocytic power of the
mesenchymal cell over the tumor cell and the digestive activity
of the adult mesenchymal cell which prove to be the direct
factors in the disappearance of the tumor cells in a double mixed
graft of tumor and spleen. One of the remarkable results of the
various experiments undertaken in this direction is the fact, that
only the adult splenic mesenchyme, not the embryonic, exhibits
the capacity of digesting tumor cells. This observation well
corresponds with the fact that the embryo fails to resist growth
of heterogeneous tissue.

I have already mentioned that only in respect to the Ehrlich
sarcoma and to the sarcoma 180 could I obtain results which
clearly demonstrate the phagocytic and digestive activity of the
adult splenic mesenchymal cells. The study of the growth of
other tumors and of the check of their growth is sufficiently
advanced to enable me to conclude that the inhibition of the car-
cinoma growth is effected by a mechanism not altogether iden-
tical to that described in relation to the Ehrlich sarcoma cells.
Must this fact astonish us? I do not think so. To expect a liter-
ally identical response of the mesenchymal cells to different agents
would be just as inconsistent as to expect the digestive activity
in all multicellular animals to proceed in a perfectly identical

IMMUNITY AND THE POWER OF DIGESTION. 2OQ

manner. And we know that this is not the case. We know that
in multicellular organisms there exists a digestive cavity, into
which free enzymes are poured, we know, however, that in sea-
anemones, though a digestive cavity does exist, no ferments had
ever been discovered in it and the digestion of the food proceeds
through the immediate apposition to it of definitecellular elements.
The fact that at least in some cases the phagocy tic and digestive
activity of the mesenchymal adult cells is found to be the dele-
terious factor, which cannot be overcome by some of the tumors
and which inhibits the growth of these tumors, or rather destroys
an actively growing tumor, throws a new light on the immunity
problem in general, since the power of resistance against pro-
liferation of foreign tissue, though manifested in different ways,
can certainly not be of a fundamentally different nature.

210 VERA DANCHAKOFF

EXPLANATION OF FIGURES.

The figures are retouched photographs from original preparations. The Figure
i at about 400, the Figure 3, 700 and the Figures 2 and 4, 1000 diameters.

ABBREVIATIONS.
Gr Lc. Granular leucocytes.
H Tc. Healthy tumor cells.

5 Ms. Splenic mesenchyme exercising a phagocytic and digestive activity.
Tc'. Tumor cells in various stages of disintegration.

Tc" . Tumor cells transformed into blocks of almost structureless protein.
V. Vessel.
X. Tumor cells tightly surrounded by mesenchymal cells.

The figures illustrate the result of growth of mixed grafts, consisting of mam-
malian tumor and adult chick spleen on the allantois.

PLATE I.

FIG. i. Actively growing tumor focus of 4 days growth in the allantois. In
the center groups of healthy tumor cells with numerous mitoses. The focus of
healthy tumor tissue is surrounded by a zone in which the splenic mesenchymal
cells encircle the tumor cells and digest them in closed vacuoles.

FIG. 2. Small area of the peripheral zone around a tumor focus of 4 days growth.
Only two tumor cells (H Tc) still exhibit a healthy structure. All the others (Tc'
and Tc"} show various changes dependent upon the digestive activity of the splenic
mesenchymal syncytium enclosing them in vacuoles.

BIOLOGICAL BULLETIN VOL. XXXVIII.

PLATE

FIG .1.

n *

• r '••- •

• ••
. _,

* «• * •/*

* * . «

V

• • V-

» *

• Vr

* *

/•<

"*"

• •

^

».v

•\v; •

•i> »'••*!

• r

; -V*

-V •

•• . -

V

"" ;.' .

.
"

,-j .t--. ^ m :

_,_' ' ' -;V.- v r>''"- "' '^ •-' ', •

S.Ms. •>.% ' " i'j '*• •

&. * ' • « . , t ." /-,. V v ,

«V» » • ^ • . * •' • * i' •'

^\;S^' ''v'^oV^,

*ni :fA>'v

**. c • ' v

^

,v* i *

«
*».* .

.**

H.Tc.

W ^

i '

^ ., -*t *,', ,

S.Ms. ^ ,

j

V. DANCHAKOFF.

FIG. 2.

212 VERA DANCHAKOFF.

PLATE II.

FIG. 3. Small region of the peripheral zone around a tumor focus of 4 days
growth, showing a group of healthy tumor cells (H Tc) and numerous other tumor
cells in various stages of disintegration.

F'iG. 4. Small area of the peripheral zone around a tumor focus of 4 days
growth, illustrating the disintegration of the tumor cells within the vacuoles,
formed by the splenic mesenchymal cells.

BIOLOGICAL BULLETIN, VOL. XXXVMI

FIG. 3.

'*.,

9

•

£k*

.

. - ' -

'

Tc.'

S.Ms.

-

' I.*,: V--

m -

H.Tc.

H.Tc.

•" V '

••

.**•

r
i

Tc.'

Gr.Lc

> '

Tc.'

H.Tc.

S.Ms. -

; Gr.Lc.

d *

•&.

- Tc.'

™

.<**.

V. OANCHAKOFF.

FIG. 4-

CHROMOSOME STUDIES IN TETTIGID^. II. CHRO-
MOSOMES OF PARATETTIX BB AND CC
AND THEIR HYBRID BC.1

MARY T. HARMAN.

Introduction 2I3

Observations • 2I4

1. Spermatogonial Divisions 2I4

a. Chromosomes of CC 2I5

b. Chromosomes of BB 215

c. Chromosomes of BC 2I°

2. Growth Period 216

3. Spermatocyte Divisions • 2I7

4. The X-chromosome • 2I&

Discussion . . 218

1. Synapsis and Heredity 2I9

2. Chromosomes in Hybrids • • • 221

Conclusions :

Literature Cited • 223

Explanation of Plates • • • 226

INTRODUCTION.

As was stated in a previous paper (Harman '15) the work on
the cytological constitution of the germ-cells of Paratettix has
been undertaken for the purpose of discovering whether or not
the microscope will reveal any differences in the germ-cells of
very closely related forms which may be correlated with the
differences in hereditary characteristics. Three forms have been
considered in this paper; two breeding true to type, BB and
CC and their hybrid BC (Nabours '14 and '17).

In his breeding work with Paratettix Nabours ('17) has con-
sidered only the color patterns of the pronota and femora of
the jumping legs. He has found fourteen distinct color patterns
which behave as a unit and cannot further be broken up. These
color patterns he designated as multiple allelomorphs. In most
crossings between different pure types the resultant hybrid is
readily discernible from either parent. In those hybrids which

1 Contribution from the Zoological Laboratory, Kansas State Agricultural Col-
lege, No. 27.

213

214 MARY T. HARM AN.

are not so readily discerned from the parents, a closer examina-
tion shows that the pigment of each parent is distributed in
about equal proportions. There is no evidence of dominance nor
crossing over, and the proportions of offspring are almost per-
fectly Mendelian even in small numbers.

It is from these grouse locusts that the material for the present
paper was obtained. The author wishes again to thank Dr.
R. K. Nabours for this material and express her appreciation to
him for furnishing her the material at a time when she could
use it and for giving her access to his records from which she has
been able to know the pedigrees of her animals.

OBSERVATIONS.

Only the male germ-cells have been studied. Observations
have been made upon the spermatogonial divisions, the growth
period, synapsis and the maturation divisions. Special con-
sideration has been given to the metaphase plate, polar view, of
the spermatogonial divisions, the behavior of the chromatin
material during the growth period and the manner of the forma-
tion of the first spermatocyte chromosomes.

i . Spermatogonial Divisions.

In all three forms the thirteen spermatogonial chromosomes
in the metaphase are arranged on the spindle, having the appear-
ance of the spokes on the hub of a wheel (Fig. i). As may be
seen in a metaphase plate, polar view (Figs. 3 to 12), these
chromosomes are in pairs and the members of each pair are
approximately of equal size and of similar shape. The accessory
chromosome is larger than the two smallest pairs and smaller
than the other chromosomes. However, many times the differ-
ence in size between the chromosomes of the second and third
pairs and the X-chromosome is so slight in the spermatogonia
that it is difficult to distinguish them. In Fig. 9, the homologous
chromosomes of the different pairs are designated by the same
number.

In the anaphase the chromosomes take a position more in the
same direction of the fibers of the spindle rather than at right
angles to it as in the metaphase (Fig. 2). The change from the

CHROMOSOME STUDIES IN TETTIGID/E. 215

perpendicular position in the metaphase to the parallel position
in the anaphase would indicate a telomitic fiber attachment.
In the late anaphase and early telophase the chromatin material
begins to take on a granular appearance before the chromosomes
have formed into a spireme (Fig. 13). In fact the cell has
completed its division before there has been much reconstruction
of the chromatin material.

(a) Chromosomes of CC. — Figures 3, 4 and 5 are metaphase
plates, polar view, of CC spermatogonial cells. Four of the
thirteen chromosomes are decidedly larger than the others. The
six smallest chromosomes are,,somewhat narrower at the proximal
end than at the distal end. The third pair, marked 3 in these
figures, are distinctly pointed at the proximal end, but there is no
indication of a bend or hook on this pointed end. The X-chromo-
some is more ovoid and when lying in certain positions shows a
slight constriction in the middle. Figure 3 shows one of the
larger chromosomes split at its distal end. This must be a
precocious division, as it was not observed in any of the hundreds
of other cells examined. Often one of the chromosomes is
observed in the center of the spindle (Fig. 4), or the proximal
end may be near the center (Fig. 5).

(&) Chromosomes of BB. — In general the arrangement of the

chromosomes of BB are like those of CC (Figs. 9-12). There

are the two large pairs, the three smaller pairs and an intermediate

pair. The X-chromosome is ovoid with a slight constriction

near the middle (Fig. 9). There also occurs the crowding of the

chromosomes toward the center of the spindle. The most evident

difference between BB and CC is in the third pair, marked 3 in

all the figures. Not only are these chromosomes pointed at

their proximal ends, but the point is considerably bent to form

a definite hook in BB. Sometimes this hook forms an acute

angle with the other part of the chromosome (Fig. 12), but more

often the bend is rounded (Fig. 10). The hooked chromosome

on the left in figure 10 is much smaller than its homologue on

the right, but this is due to the fact that a portion of its distal

end was cut off as was also the large chromosomes on either side

of it. These parts of chromosomes were found in the preceding

section.

2l6 MARY T. HARMAN.

(c) Chromosomes of BC. — The hybrid BC (Figs. 6-8) has only
one hooked chromosome, while its homologue is merely pointed
like in CC. This difference is constant in the many metaphase
plates examined. There is some variation in the shape of the
chromosomes and the size of the bend of the pointed end. In
Fig. 6 the hooked member of the third pair is almost crescent
shaped, while in the other cells the hooked chromosome is thick
at the distal end, tapers toward the proximal end, and then

bends abruptly.

2. Growth Period.

Figures 13 to 24 illustrate the qhanges which take place in
the growth period. At the end of the last spermatogonial divi-
sions the chromosomes have a crinkled appearance. Figure 13
shows all thirteen of the chromosomes in this condition. How-
ever the X-chromosome is more compact than the others. This
crinkled appearance is due to the chromatin becoming less
compact and more granular. There is no sign of a nuclear
membrane nor any indication of a separation of the nuclear
area from the cytoplasm. Soon the chromatin becomes very
finely granular and the chromosomes unite end to end (Fig. 14).
Shortly after this stage the chromatin has the appearance of a
finely granular continuous thread and a light area separates the
cytoplasm from the nuclear material (Figs. 15 and 16). The
cell begins to increase in size and the chromatin thread becomes
contracted into a ball (Fig. 17). There seems to be no definite
polarization to the loops of the spireme, but it has more the
appearance of a tangled mass. Sometimes it is with difficulty
that the individual fibres are distinguished. Now both the
nucleus and the entire cell increase greatly in size (Figs. 18-21)
and the chromatin material forms into a thicker thread (Fig. 18).
Then the chromatin thread becomes looser and the granules are
more concentrated, particularly in places throughout the strand
(Fig. 19). The portion of chromatin marked X never becomes
as granular as the remainder of it. This chromatin is the sex-
chromosome and can be identified in all the stages.

Following the opening up of the tangled knot the chromatin
thread breaks into pieces (Figs. 20 and 21). Figure 22 shows
thirteen pieces, twelve of which are uniting in pairs by an end

CHROMOSOME STUDIES IN TETTIGID/E.

to end union. Figures 23 and 24 show successive stages in the
formation of the primary spermatocyte chromosomes. As the
chromosomes become more compact in appearance the nuclear
membrane disappears and there is no definite boundary again
between the nucleus and cytoplasm (Fig. 24).

During all these changes there has not been the least appear-
ance of a side by side pairing of the chromosomes, neither has
there been any indication of a longitudinal splitting of the chro-
matin material either as chromosomes or as a chromatin thread.
With the breaking up of the chromatin thread always more
than the haploid number of chromatin elements may be seen,
and in many instances there is evidently the cliploid number.
This surely justifies the conclusion that at the end of the growth
period previous to the formation of the primary spermatocyte
chromosomes there appears the diploid number of chromosomes
and that the real pairing of the chromosomes does not take place
in the spireme stage. It seems, also, very evident that the union
of the chromosomes is by telosynapsis. As was previously stated
(Harman '15), the dumb-bell shape of the chromosomes is not
due to the manner of the division of the chromosomes, but is a
result of the way they are formed. This shape is noticed in
some of the chromosomes even before the entire chromatin
thread is broken up (Fig. 21), and certainly before the first
spermatocyte spindle is formed (Fig. 24).

3. Spermatocyte Divisions.

There are six dumb-bell shaped bivalent chromosomes and
an ovoid univalent chromosome in the primary spermatocyte.
These arrange themselves longitudinally on a large spindle (Figs.
25 and 26). The univalent, or X-chromosome, lies near the
periphery of the spindle. The other chromosomes are crowded
toward the center. Most often they are so closely crowded
together that it is difficult to see them all at once in a lateral
view of the spindle. Always one of the medium-sized chromo-
somes lies very near the second-largest chromosome (Figs. 26
and 30).

In the first division all the bivalent chromosomes divide at
the constricted part of the dumb-bell (Fig. 27). In other words,

2l8 MARY T. HARMAN.

they have divided at the point of union as we have previously
seen. As the result of this division the chromosomes of the
anaphase are ovoid. The X-chromosome does not divide in this
division but passes to one pole much in advance of the others.
The first division then is truly a reductional division.

The spindle of the secondary spermatocyte is nearly spherical.
Early in the metaphase the chromosomes show a longitudinal
split (Figs. 28 and 29). All of the chromosomes divide in this
division. It is an equational division in the sense of the dividing
of the chromatin material of original chromosomes into halves.

4. The X-Chromosome.

The X-chromosome is a persistent portion of chromatin ma-
terial which can easily be identified throughout the growth period.
It never becomes finely granular, nor does it have the woolly
appearance of the other chromatin elements. The other chromo-
somes lose their identity in the coiled thread, but the X-chromo-
some does not form a part of this coiled knot. With the formation
of the primary spermatocyte chromosomes it may be distin-
guished from the others both by its shape and by its more compact
appearance. On the primary spermatocyte spindle it is often
found half way to the pole before the other chromosomes have
completely divided. It is more difficult to recognize it in the
secondary spermatocyte because all the chromosomes here are
ovoid in shape, and the differences in size between some of the
chromospmes are so much less. It behaves similarly to the
other chromosomes here, neither being precocious in its division

nor lagging.

DISCUSSION.

Concerning many of the most obvious and essential points of
the behavior of the chromatin in the maturation of the sperm,
there is no longer any debate. It is agreed that during this
period the number of chromosomes is reduced to one half. That
this reduction in number is brought about by the union of chro-
mosomes by pairs is also generally accepted. As to how this
pairing has taken place has been the subject of much discussion,
and upon the answer to this question depends the acceptance or
rejection of other theories. It seems to the writer that when we

CHROMOSOME STUDIES IN TETTIGID.E. 2 19

have accepted the thought that this reduction in number takes
place by the union of chromosomes in pairs we have essentially
also accepted the theory of the individuality of chromosomes.
Then how do these individuals come together and finally separate
again?

i . Synapsis and Heredity.

One can scarcely follow out the discussion of Janssen on
Batracoseps attennatus, and more especially when close attention
is given to his drawings, without coming to the conclusion that
with this form, at least, the chromosomes unite by parasynapsis
and that it is perfectly possible in the twisting of certain chroma-
tin elements to have an exchange of parts of homologous chromo-
somes, thus forming a convenient mechanism for the "crossing
over" of hereditary characteristics if we accept the chromosome
hypothesis of heredity. The "chiasmatypie" theory of Janssen
('09) furnishes a very convenient means for explaining some
ratios in Mendelian inheritance.

An application of this chiasmatype theory, which should not
pass without mention, is the work on Drosophila, particularly
the work of Morgan and his students ('15). From the behavior
of the hereditary characteristics a chromosome map has been
constructed in which not only are the determiners of the char-
acteristics located on definite chromosomes, but also the relative
distances that these genes are from each other, is given. The
position of the genes on the chromosomes is calculated from the
percentages of cross-overs. In addition to the vast amount of
genetic evidence for this condition in Drosophila there is also
some cytological evidence.

Metz ('14) in dealing with five different types of chromosome
groups of Drosophila shows that the chromosomes exhibit a
close association in pairs at nearly all times and that before
each cell division the pairs become so intimately associated that
they may be said actually to conjugate. He further states that
the union of the chromosomes is "unquestionably a side-by-
side, or parasynapsis one." In a later paper (Metz '16) the
same author in considering chromosome pairing in about eighty
species of Diptera, many of which belong to the genus Drosophila,
states that the pairing "certainly involves the essential features
of a synaptic (parasynaptic) union."

22O MARY T. HARMAN.

Bridges ('17) confirmed the observations of Metz as to the
spermatogonial and oogonial pairing, but deduced the kind of
synapsis from his genetic data.

The genetic behavior of Paratettix is very different from that
of Drosopliila. In the hundreds of matings of Paratettix not
a single instance of crossing-over has occurred. In accounting
for the hereditary behavior of Parateltix without a knowledge
of the cytological behavior Morgan has suggested the theory of
identical loci. As has been previously shown, at no time during
the growth period is there any indication of a double thread,
and also there is evidence of an end to end union of the chromo-
somes to form the tangled thread of the contraction figure.
Moreover at the end of the growth period the chromatin material
forms the diploid number of chromatin elements. Also these
chromatin elements agree in relative sizes to the spermatogonial
chromosomes. Undoubtedly these parts unite end to end. In
this case there is little chance for crossing-over, as only the ends
of homologous chromosomes come in contact with one another.
In view then of the cytological behavior, it seems more likely
that the series of multiple allelomorphs may be accounted for
by the kind of synapsis rather than by the theory of identical loci.

The writer is well aware that there is not perfect agreement
as to how synapsis takes place in the Orthoptera, and even within
a single family of the Orthoptera. McClung ('14) has recently
reviewed the literature on the subject of synapsis in Orthoptera,
and Wenrich ('16) has summarized the results. Briefly those
who have described or assumed telosynapsis are: Montgomery
('05), Syrbula; Stevens ('05), Stenopalmatus, ('05) Blatta, ('10)
Forficula; Wassilieff ('07), Blatta; Zweiger ('06), Forficula;
Davis ('08), Acrididae and Locustidse; Buchner ('09), Gryllus,
CEdipoda; Brunelli ('09), Gryllus., and ('10) Tryxalis; Sutton
('02, ''03), Brachystola; Baumgartner ('04), Gryllus; McClung
('05, '08, '14), various Orthoptera; Nowlin ('08), Melanoplus;
Pinney ('08), Phrynotettix; Robertson ('08), Syrbula; Carothers
('13), Acrididae. Those who have assumed or described para-
synapsis are: Gerard ('09), Stenobothrus ; Morse ('09), Blattidse;
Stevens ('12), Ceuthophilus; Robertson ('15), Tettigidae; Ve-
jodvsky ('11-12), Locustidae; and Otte ('07), Locusta. In addi-

CHROMOSOME STUDIES IN TETTIGID^E. 221

tion to these Wenrich ('16) gives strong evidence for parasynapsis
in Phrynotettix and ('17) Chorthippus and Tr inter otro pis.

May it not be that within the order Orthoptera, or even within
the same family, that the kind of synapsis is different? The
behavior in heredity is certainly different. The behavior of
Paratettix in heredity has already been discussed. Bellamy ('17)
has described a similar condition in the genus Tettigidea. He
found five allelomorphic color patterns without any indication
of cross-overs. The cytological constitution of the germ-cells
of his material has not been examined. However, Robertson
('17) has described parasynapsis in Tettigidea parvipennis.
Nabours ('19) has not only found crossing-over in Apotettix, but
has also found parthenogenesis and crossing-over in those repro-
ducing parthenogenetically. Certainly if we accept the chromo-
some hypothesis of heredity there must be a difference in the
cytological behavior of the germ-cells of these genera. Certainly
enough has already been said in this paper to show that the chro-
mosomes pair by telosynapsis in Paratettix. With the many
cross-overs in Apotettix, and the similar behavior in inheritance
to Drosophila, one would expect the chromosome behavior to be
different than in Paratettix and more like Drosophila. The
writer has not yet studied the chromatin behavior in Apotettix.

2. Chromosomes in Hybrids.

In working with the chromosomes of hybrids Moenkhaus,
Morris, Federley, Harrison and Doncaster, Pinney and others
have found that despite the fact that the paternal chromosomes
are in a foreign cytoplasm they retain their characteristic form,
size and number. Moenkhaus ('04) pointed out that when
Fundulus is crossed with Menidia two kinds of chromosomes
are present in the fertilized egg and can readily be distinguished
in later divisions, furthermore, that these two kinds of chromo-
somes are like the chromosomes of each parent respectively.

Morris ('14) found two types of chromosomes in the early
cleavage stages of the hybrid of Fundulus heteroclitus 9 and
Ctenolabrus adspersus c? . These types of chromosomes she
identifies as the two types of parental chromosomes.

Pinney (' 1 8) made a number of crosses with teleosts and also

222 MARY T. HARMAN.

made the reciprocal crosses. She found that some cytoplasm
seemed to be more favorable to foreign sperm than others. In
some instances a few chromosomes were eliminated, but those
that did remain could be identified as to their paternal or maternal
origin. Similar results were obtained by Federley ('13) and
Harrison and Doncaster ('14) in hybrids of certain moths.

In the above cases cited there is no indication that these
hybrids may occur in nature, and these offspring were obtained
from animals which are accepted as distinct species. In the
case of Paratettix, P. BB and P. CC are not yet admitted to be
species by taxonomists, and they are known to cross in nature,
judging from hybrids obtained in the wild and segregated in the
laboratory. The chromosomes of each form are so very similar
to those of the other forms in so many ways that it is only with
close study that the differences are recognized. Yet, when once
recognized it is found to be constant. The chromosomes of the
hybrids of Paratettix are different from those of the fish mentioned
above in that the parental chromosomes are more alike and that
the cytoplasm of the egg seems to be perfectly compatible with
the chromatin of the foreign sperm. They are different from
those of the moths in that there are the same number of chromo-
somes in both forms and there is a complete synapsis with the
homologous pairs of chromosomes.

CONCLUSIONS.

1. The third pair of spermatogonial chromosomes of BB are
bent at the proximal end so as to form distinct hooks.

2. The third pair of spermatogonial chromosomes of CC are
pointed at the proximal end, but there is no bend.

3. The third pair of spermatogonial chromosomes of the hybrid,
BC, is composed of one hooked member and one pointed member.

4. During the growth period there is no indication of a parallel
condition of the chromatin either in the chromatin thread or in the
chromosomes.

5. At the end of the growth period there is evidently the
diploid number of chromosomes formed which correspond in
relative sizes to the respective spermatogonial chromosomes.

6. Synapsis does not take place in the thread but at the end

CHROMOSOME STUDIES IN TETTIGID^. 223

of the growh period by an end to end union of the homologous
pairs of chromosomes.

7. The chromatin of the sex-chromosomes does not lose its
identity during the growth period.

8. The first maturation division is a reductional division, and
the second maturation division is equational.

9. The formation of the diploid number of chromosomes at
the end of the growth period and the union of their homologous
pairs by telosynapsis may explain the absence of any crossing-
over in Paratettix.

LITERATURE CITED.
Baumgartner, W. J.

'04 Some New Evidence for the Individuality of Chromosomes. BIOL. BULL.,

Vol. 8.
Bellamy, A. W.

'17 Studies of Inheritance and Evolution in Orthoptera, IV. Multiple Allelo-
morphism and Inheritance of Color Patterns in Tettigidea. Jour. Gen.,
Vol. 7.
Bridges, Calvin B.

'17 Non-disjunction as a Proof of the Chromosome Theory of Heredity. Gen .,

Vol. i.
Brunelli, G.

'09 La spermatogenesi del Gryllus desertus Pall. (Division! spermatogoniali

e maturative.) Reale Academia dei Lincei, T. 2, serie sa.
'10 La spermatogenesi della Tryxalis. Division! spermatogoniali. Memarie

della Societa Italiana Scienze, T. 16, serie 33.
Buchner, P.

'09 Das accessorische Chromosom in Spermatogenese und Ovogenese der
Orthopteran, Zugleich ein Beitrag zur Kenntnis der Reduktion. Archiv.
f. Zellforsh., Bd. 3.
Carothers, E. Eleanor.

'13 The Mendelian Ratio in Relation to Certain Orthopteran Chromosomes.

Jour. Morph., Vol. 24.
Davis, H. S.

'08 Spermatogenesis in Acrididse and Locustidse. Bull. M. C. Z., Vol. 53.
Federley, H.

'13 Das Verhalten der Chromosomen bei der Spermatogenese der Schmettalinge,

etc. Zeit. Abst. Vererb., Bd. 9.
Gerard, P.

'09 Recherches la Spermatogenese Chez Stenobothrus bigutulus (Linn.).

Archives de Biologie, T. 24.
Harman, Mary T.

'15 Spermatogenesis in Paratettix. BIOL. BULL., Vol. 29.
Harrison, J. W. H., and Doncaster, L.

'14 On Hybrids between Moths of the Geometrid Subfamily Bistoninse, with
an Account of the Behavior of the Chromosomes in Gametogenesis in
Lycia (Biston) historic,, Ithysia (Myssia) Zonaria and in their Hybrids.
Jour. Gen., Vol. 3.

224 MARY T. HARMAN.

Janssens, F. A.

'09 La theorie de la chiasmatypie Nouvelle interpretation des cineses de

maturation. La Cellule, T. 25.
McClung, C. E.

'05 The Chromosome Complex of Orthopteran Spermatocytes. BIOL. BULL.,

Vol. 9.
'08 The Spermatogenesis of 'X.iphidium faciatum. Kansas Univ. Sci. Bull.,

Vol. 4.
'14 A Comparative Study of Chromosomes in Orthopteran Spermatogenesis.

Jour. Morph., Vol. 25.
Metz, Charles W.

'14 Chromosome Studies in Diptera I. A Preliminary Survey of Five Different
Types of Chromosome Groups in the Genus Drosophila. Jour. Exp. Zool.,
Vol. 17.

'16 Chromosome Studies on the Diptera II. The Paired Association of
Chromosomes in the Diptera and its Significance. Jour. Exp. Zool.,
Vol. 21.
Moenkhaus, Wm. J.

'04 The Development of the Hybrid between Fundulus heteroclUus and Menidia
nolata with Especial Reference to the Behavior of the Maternal and Paternal
Chromatin. Am. Jour. Anat., Vol. 3.
Montgomery, T. H., Jr.

'05 The Spermatogenesis of Syrbula and Sycosa, with General Considerations
upon Chromosome Reduction and the Heterochromosomes. Proc. Acad-
emy Nat. Sci. Philadelphia, February.
Morgan, T. H. and others.

'15 The Mechanism of Mendelian Heredity. Henry Holt and Co. New York.
Morris, Margaret.

'14 The Behavior of the Chromatin in Hybrids between Fundulus and Cteno-

labnis. Jour. Exp. Zool., Vol. 16.
Morse, Max.

'09 The Nuclear Components of the Sex Cells of Four Species of Cockroaches.

Archiv. f. Zellforsch., Bd. 3.
Nabours, Robert K.

'14 Studies of Inheritance and Evolution in Orthoptera I. Jour. Gen., Vol. 3.
'17 Studies of Inheritance and Evolution in Orthoptera II, III. Jour. Gen.,

Vol. 7.
'19 Parthenogenesis and Crossing-over in the Grouse Locust Apotettix. Amer.

Nat., Vol. 53.
Otte, H.

'07 Samenreifung und Samenbildung bei Locusta viridissima. Zool. Jahrb.

Abt. f. Anat., Bd. 24.
Pinney, Edith.

'08 Organization of the Chromosomes in Phrynoteltix magnus. Kansas Univ.

Sci. Bull., Vol. 4.
"18 A Study of the Relation of the Behavior of the Chromatin to Development

and Heredity in Teleost Hybrids. Jour. Morph., Vol. 31.
Robertson, W. Rees Brebner.

'08 The Chromosome Complex of Syrbula admirabilis. Kansas Univ. Sci.
Bull., Vol. 4.

CHROMOSOME STUDIES IN TETTIGID.E. 225

'15 Inequalities and Deficiencies in Homologous Chromosomes; their Bearing

upon Synapsis and the Loss of Unit Characters. Jour. Morph., Vol. 26.
'17 Chromosome Studies IV. A Deficient Supernumerary Accessory Chromo-
some in a Male of Tettigidea parvipennis. Kansas Univ. Sci. Bull., \ ol. 10.
Stevens, N. M.

'05 Studies in Spermatogenesis with Especial Reference to the "Accessory

Chromosome." Carnegie Inst. Publ. 36.
'10 An Unequal Pair of Heterochromosomes in Forficula. Jour. Exp. Zool.,

Vol. 8.
'12 Supernumerary Chromosomes and Synapsis in Ceuthophilus. BIOL. BULL.,

22.
Sutton, W. S.

'02 On the Morphology of the Chromosome-group in Brachyslola magna.

BIOL. BULL., Vol. 4.

'08 The Chromosomes in Heredity. BIOL. BULL., Vol. 4.
Vejdovsky, F.

'12 Zum Problem der Vererbungstrager. Bohm. Gesell. wiss., Prag.
Wassilieff, A.

'07 Die Spermatogenese von Blatta germanica. Archiv. f. Mikr. Anat., Bd. 70.
Wenrich, D. H.

'16 The Spermatogenesis in Phyrnotettix magnus, with Special Reference to
Synapsis and the Individuality of the Chromosomes. Bull. Mus. Comp.
Zool., Vol. 60.

'17 Synapsis and Chromosome Organization in Charthippus (Stenobothrus)
curtipennis and Trimerotropis suffusa (Orthoptera). Jour. Morph., Vol. 29.
Zweiger, H.

'06 Die Spermatogenese von Forficula auricularia L. Jena. Zeit., Bd. 42

226 MARY T. HARMAN.

EXPLANATION OF PLATES.

All figures were made at table level by means of a Zeiss compensating ocular
No. 6 and a 1.5 objective with the aid of a camera lucida. Figures 13, 14 and 15
were enlarged three and one half diameters. All others were enlarged two and
one half diameters. All were reduced one third.

PLATE I.

FIG. i. BC, metaphase plate, lateral view, of spermatogonial division.

Fig. 2. BC, late anaphase, spermatogonial division.

FIG. 3. CC, metaphase plate, polar view, spermatogonial division; X is the
odd-chromosome, 3, 3 is the third pair of chromosomes, and 2 is one of the chromo-
somes of the fifth pair showing a precocious split.

FIGS. 4 AND 5. CC, metaphase plates, polar view, spermatogonial divisions;
3, 3 third pair of chromosomes.

FIGS. 6, 7 AND 8. BC, metaphase plates, polar view, spermatogonial divisions;
3, 31 third pair of chromosomes, 31 is the hooked chromosome of the third pair.

FIG. 9. BB, metaphase plate, polar view, spermatogonial division; the numbers
i, 1-2, 2, etc., to 6, 6 are homologous pairs of chromosomes, X is the odd-chromo-
some. It will be noted that the members of the third pair are both hooked chromo-
somes in BB.

FIGS. 10, ii AND 12. BB, metaphase plates, polar view of spermatogonial
divisions, 3, 3 is the third pair of chromosomes.

BIOLOGICAL BULLETIN, VOL. XXXVIII.

PLATE I.

9

M. T. HARMAN

228 MARY T. HARMAN.

PLATE II.

FIG. 13. Early telophase of last spermatogonial division. X, the odd-chromo-
some.

FIG. 14. Beginning of growth period showing the uniting of the chromosomes
to form spireme; X is chromatin of the odd-chromosome in all the figures of this
plate.

FIG. 15. Early loose chromatin thread.

FIG. 16. Early contraction stage.

FIG. 17. Later contraction stage and the appearance of the nuclear membrane.

FIG. 18.' Large tangled thread.

FIG. 19. Beginning of the concentration of the chromatin in various places
throughout the thread.

FIG. 20. Breaking up of chromatin thread into the diploid number of chromatin
elements.

FIG. 21. Further breaking of the chromatin thread and the beginning of the
formation of the bivalent chromosomes.

BIOLOGICAL BULLETIN, VOL. XXXVII

16

W¥P

I T

fev'' ..;.«?&..**.

., ^ ...;-^M_

; ,

18

&•'***«-

""vX/^'V*.

20

M. T. HARMAN.

230 MARY T. HARMAN.

PLATE III.

FIG. 22. Union of univalent chromatin elements to form bivalent chromo-
somes. X is the odd-chromosome in all figures of this plate.

FIG. 23. Bivalent chromosomes with the nuclear membrane present.

FIG. 24. Bivalent chromosomes before the formation on the spindle. The two
elements of the third chromosome have not yet united.

FIG. 25. First spermatocyte spindle showing the seven chromosomes.

FIG. 26. Metaphase, lateral view of first spermatocyte division.

FIG. 27. Early anaphase of first spermatocyte division. The odd-chromosome
approaching one pole in advance of the other chromosomes.

FIG. 28. Metaphase plate, lateral view of second spermatocyte division. Four
of the chromosomes show the longitudinal split.

FIG. 29. Metaphase plate, polar view of second spermatocyte division.

FIG. 30. Metaphase plate, polar view of first spermatocyte division.

BIOLOGICAL BULLETIN, VOL. XXXVIII.

PIAIf III.

30

M. T. HARMAN.

WHITE-OCELLI— AN EXAMPLE OF A "SLIGHT' MU-
TANT CHARACTER WITH NORMAL VIABILITY.

CALVIN B. BRIDGES.

Most of the non-lethal mutant characters of Drosophila melano-
gaster may be described as "slight" in the amount of visible
change. This is fortunate, for, in general, the mutant genes
that cause great changes in an organ, or that cause several
distinct changes in different organs, produce races of poorer
viability than the wild type of the species. The amount of
disturbance to viability is roughly proportional to the extent of
the somatic change. In nature these "extreme" mutant types
would not be able to survive in competition with the wild type,
and in our experimental cultures their presence leads to aberrant
ratios that tend to obscure the simple genetic relations. It is
possible to reduce the shortage of these relatively inviable mutant
characters only by making the conditions of culture exceptionally
favorable. Thus, the number of eggs per culture must be re-
stricted by breeding from a single mother; and the amount of
food per culture should be carefully regulated. Great progress
has been made in improving the quality and methods of using
culture media. While it is possible by such improved methods
to make the relatively inviable mutants usable in many kinds
of experiments, there remain other experiments that require so
high a degree of exactness or the simultaneous use of so many
mutants that even one poorly viable mutant is inadmissible.
As the number of mutant types increases, we are able to drop the
use of more and more of the poorly viable mutants and to replace
them by new mutants of normal viability. As already implied,
these mutants of normal viability are nearly all characterized
by somatic changes of slight degree, or by effects, so far as observ-
able, upon a single organ only. Many of these "slight" char-
acters are perfectly definite in demarkation and offer as great
precision of classification as do more extreme mutant characters.
Another great advantage of slight mutant characters is that they

231

232 CALVIN B. BRIDGES.

do not preclude the simultaneous use of other mutants. On the
other hand, the use of such an extreme mutant character as
white-eyes prevents the effective use in the same experiment of
all other eye colors, and likewise the use of vestigial-wing hinders
the use of all other wing and venation characters. The most
valuable mutants are, then, those of slight but definite somatic
change, free from disturbance to viability or masking effects on
other mutant characters.

ORIGIN AND DESCRIPTION OF THE WHITE-OCELLI MUTANT.

One of the best examples of a slight mutant character that
fulfills these conditions is that of "white-ocelli," which was found
very early in the breeding work with Drosophila (June 21, 1912).
The ocelli of wild flies are three small simple eyes in a group on
the dorsal posterior part of the head. In color they are of a
dilute brownish-red, about that of a coffee infusion. Close
examination shows that the color of the ocelli themselves is
quite light, about that of weak tea, but that there is a crescent"
shaped deposit of dense brownish-red pigment about the median
side of the two posterior ocelli and against the posterior side of
the anterior ocellus. The apparent color of the ocellus is largely
due to this outside pigment seen through the transparent lens-like
ocellus, and consequently the color changes in intensity with the
angle at which the ocellus is viewed.

It was observed that the ocelli of white and also of vermilion-
eyed flies were without color or pigment deposit. These ocellar
changes are only other effects of the white and vermilion genes.
That the color of the ocelli could vary independently of that of
the eyes became apparent when it was found that about half
of the flies of the stock of the mutant black-body color had
white ocelli, while the remainder had the normal coffee-colored
ocelli. Some of the white-ocelli flies were bred together and
gave a pure-breeding stock of white-ocelli flies (June, 1912).
About a year after this, it was found (July, 1913) that the stock
of the third-chromosome recessive spineless was also pure for
the white-ocellar color. The gene for white-ocelli was thought
to be in the third chromosome, since, in crosses in which the
spineless was used, the white-ocellar character generally reap-
peared in association with the spineless, though sometimes not.

WHITE-OCELLI.

233

The usefulness of the mutant white-ocelli was not appreciated
for some years after its discovery. This neglect was due, in
large part, to the fact that the regular examination of flies during
this period was carried out by aid of a hand lens only, and the
separation of the white from the normal ocelli was difficult
because of the small size of the region affected. The later work
has been done with a binocular microscope, with special attention
to proper illumination and magnification; and under these condi-
tions the separation is complete and entirely accurate, though
still somewhat slow.

THE LINKAGE RELATIONS OF WHITE-OCELLI.

An accurate localization of the gene for white-ocelli was made
easy by the use of the two excellent dominant characters dichsete
and hairless. The locus of dichaete was near the left end of the
map of the third chromosome as then known, being about 13
units to the right of sepia, while the locus of hairless was some-
what to the right of the middle (about 42 units to the right of
sepia and 21 to the left of rough). Spineless white-ocelli males
were outcrossed to dichaete hairless females; and the FI dichsete
hairless females were back-crossed by spineless white-ocelli males,
with the results shown below:

o

.

I

K.

2

;

.

i,

2,

i,

3-

2,

3-

1918

D

D

D

D

D

D

D

Sept.

ss

ss

ss

ss

ss

ss

SS

Total

27

H

H

H

H

H

H

H

\vo

wo

wo

wo

wo

wo

wo

Total

660

528

88

80

116

87

So

S3

5

4

2

8

o

I

1,682

* Crossovers in the first region, that between dichsete and spineless, are headed
by " i," etc.

The results of this experiment showed that the locus of white-
ocelli is to the right of that of hairless by about 6.8 units (a total
of 114 crossovers involving region three). The locus of white-
ocelli, as thus established, is in what had been the longest un-
occupied region of the third chromosome. There had been no
workable mutant in the entire distance of about 20 units from
ebony (1.5 units to the right of hairless) to rough (21.2 units to

234

CALVIN B. BRIDGES.

the right of hairless). The ebony rough distance was so great
that in constructing a map a correction was required on account
of double crossing over. The presence of white-ocelli between
ebony and rough gave an opportunity to make a direct test
of the amount of double crossing over and consequently of the
amount of correction required. The results of the spineless
white-ocelli X hairless rough back-cross are given below:

o

2

•

I,

3-

2,

3-

1919

SS

SS

SS

SS

SS

SS

Feb.

H

H

H

H

H

H

Total

25

WO

wo

wo

wo

WO

wo

ro

ro

ro

ro

ro

ro

Total..

1,084

1.126

16?

TO=;

TTS

OT

?.?,f>

2^

TT

o

0

T

T.I7Q

There was 6.6 per cent, of crossing over between hairless and
white-ocelli, which agrees with the value 6.8 found in the previous
experiment. Likewise, the white-ocelli rough value of 15.8 is in
agreement with the expectation from the usual value of 22 for
hairless rough. There was only one double crossover in the
hairless white-ocelli rough section — a percentage relatively very
low. Comparisons show that in the third chromosome (as in
the second) the region near the end of the chromosome has a far
lower 'coincidence' than has the mid-region. The amount of
correction of the observed crossover value for the hairless ebony
interval is thus .063 per cent., or somewhat less than one tenth
of one unit. Other back-cross tests involving this region have
produced a total of 45,971 flies, of which 19.6 per cent, were
crossovers. This value is to be corrected to 19.7, which is the
map-distance between ebony and rough.

THE VIABILITY OF THE WHITE-OCELLI MUTANT.

As just seen, the linkage of white-ocelli was worked out
through use of the more convenient spineless white-ocelli stock,
while the black white-ocelli stock was discarded. The white-
ocelli character persisted in the original black stock. No effort
was made to eliminate it, nor, on the other hand, to aid in its
survival. In May, 1919, a census of the flies of this black stock
showed that approximately half were white-ocellars. That is,

WHITE-OCELLI. 235

the character had persisted in undiminished frequency from June,
1912, to May, 1919, a period that represented fully 175 genera-
tions of flies. During this period the black stock had been carried
on in mass-cultures. Every two weeks a new culture was started
by transferring, without examination or selection, enough flies
to insure breeding. In such mass-cultures overcrowding is ex-
treme, and, in spite of. the great numbers of parents, not many
more offspring succeed in hatching than hatch from successful
pair-cultures. The competition grows keener with the age of
the culture, since the number of larvae is continually increasing
from eggs laid each day, while the quantity of available food soon
begins to diminish and its quality becomes progressively poorer.
The mass-culture method of breeding thus exercises a strong and
continuous selection against the perpetuation of the weaker or
slower hatching individuals or types. In several instances mixed
stocks have been started with equal numbers of different muta-
tions, and this stock transferred without selection through several
generations. Watch was kept, and in these cases there has been
a progressive change in the composition of the stock, rapid at
first, until the numbers of one type were quite small, and there-
after slower but in the same direction. Recessive characters of
very low viability may persist for many generations as a small
proportion of the population. Their existence is maintained by
the inter-crossing of the heterozygotes, whereby the mutant gene
escapes the adverse selection that the mutant character suffers.
Certain of our mutations are so sensitive to larval overcrowding
that the ratios in mass-cultures and in pair-cultures seem to
belong to different systems of heredity. Thus, the character
strap approaches I in 4 in pair-cultures, but may approximate
I in 1 6 in sister mass-cultures.

The persistence of the white-ocelli character in undiminished
proportion through 175 generations of forced competition means
that the mutant is under no disadvantage. Such a mutant might
easily survive in nature, and one slightly advantageous might
ultimately supplant the original type.

236 CALVIN B. BRIDGES.

THE MODIFICATION OF EOSIN EYE COLOR BY WHITE-OCELLI.

An examination of the various stocks of eye color mutations
showed that there was a strong correlation between the eye color
and the ocellar color. The ocelli of white-eyed flies are entirely
colorless. The ocelli of vermilion-eyed flies show a slight trace
only of color. Indeed, in the case of vermilion, the vermilion
gene has a relatively greater effect upon ocellar color than upon
eye color. The ocellar color of pink is so faint that pink can
not be used in the same experiment with white-ocelli without
some confusion in classification. The ocellar color of the dark
eye 'sepia' is itself also darker. In the ten multiple allelomorphs
of the white series, the ocellar color is proportional to the eye
color. This direct effect of eye color genes on ocellar color
suggested that the reverse relation might also hold — namely,
that the white-ocelli gene might dilute the eye color. A careful
examination of the eye color of white-ocelli flies did not show
any certain effect. White-ocelli was crossed to vermilion, and
the p2 vermilion white-ocelli flies were not distinguishable from
the simple vermilion flies. In the ¥2 of the cross between white-
ocelli and eosin, a definite modification of the eosin by the white-
ocelli gene was observed. In the case of the males, the eye color
of the double form was lighter in intensity and less yellow in
tone than that of the eosin brothers. In the females, the
change was in the same direction but was less marked in degree.
Probably 95 per cent, of the diluted males were separable from
the simple eosin, while only about 60 per cent, of the females were
thus separable. Eosin is known to be especially subject to
specific modification,1 and the effects of the white-ocellar gene
give a color intensity and tone and a sexual difference practically
identical with those observed in the case of the modifier
'pinkish.' The gene for pinkish was, however, in the second
chromosome, and there are other differences between the two
cases.

Jour. Exp. Zoo/., July, 1919.

OBSERVATIONS ON THE SEXUAL CYCLE OF THE

GUINEA PIG.

O. ISHII,

PATHOLOGICAL LABORATORY OF THE WESTERN PENNSYLVANIA HOSPITAL,

PITTSBURGH, PA.

In the following work, which was undertaken at the suggestion
of Professor Leo Loeb, my aim has been to correlate the micro-
scopical and experimental analysis of the sexual cycle which this
author had previously made and further studies which he had
planned at that time with a careful study of the cyclic changes as
far as they are accessible to the naked eye. These observations
were made at the breeding establishment of Miss A. E. C.
Lathrop, in Granby, Mass., during the years 1913 and 1914. l

THE THREE STAGES OF MENSTRUATION AND HEAT.

i . Prcestrum. — This period continues for one to one and a half
days, during which time the external genital organs become
congested and swollen and a slight serous secretion is noticeable.
This secretion, while in some instances turbid or viscid, is in
the majority of cases more or less transparent; while, as before
said, this period averages one to one and a half days, in some
instances it may continue for two or three days or more. In the
last 4 or 5 hours toward the end of this period a maximum in
these changes is reached. This is followed by the period of heat.
The duration of the procestrus period is indicated on the following
table. Ninety-four (94) animals served for this observation.

Length of Procestrus Period.

1A Day.

i Day.

i% Day

2 Days.

3 Days.

4 Days.

No of guinea pigs

I

39

34

II

8

1

1 Accidental conditions greatly delayed the publication of this paper. In the
meantime there has appeared a study of Stockard and Papanicolaou on the oestrous
cycle of the guinea pig. Notwithstanding this long delay in publication, we believe
that our observations present new facts of interest and are worth recording.

237

238

O. ISHII.

On the basis of these observations it may be reasonably said
that the average length of this period in an animal of normal
physical condition is from one to one and a half day, but after
puerperium, or in case of weakness or of pathologic conditions,
a shortening or prolongation may be brought about.

2. (Estrus, or Heat. — The approach of this period is made
apparent by excessive menstrual secretion during the last 4 or 5
hours of proffistrum. This continues for 6 or 7 hours of the
oestrus period. This is followed by the" secretion of a small
quantity of moisture which may be just noticeable, or during the
last 3 or 4 hours of the oestrus period menstrual secretion may
be lacking altogether; still heat is present in this period.

I give the following data as to the length of period of heat:

Length of Heat Period.

5 to 6 Hrs.

7 to 8 Hrs.

1
9 to 10 Hrs. ii to 12 Hrs. 13 to

14 Hrs

No. of guinea pigs .

6

51

62 30

9

As can be seen from the above figures, from 8 to n hours is
the average length of the heat period in the healthy animal. I
found those guinea pigs in which the periods were longer (12 to
14 hours) or shorter (5 to 6 hours) to have been lacking in vigor
or to have shown abnormal physical conditions. In the normal
animal I feel justified in claiming, on the average, 9 to n hours
as the correct length of this period in the normal guinea pig.

3. Metcestrum. — This is the period following the oestrus period
and might be considered as the period of recovery from the
cestrous changes. The length of this period depends almost
wholly on the condition of the animal. In the normal animal
2^ to 3 days are required until the normal condition is re-
established. From the figures that follow it will be noticed
that the length of this period varies from i^ to 5 days, but my
observations have convinced me that where the recovery is
shorter or longer than the average period, the animal is not in a
normal healthy condition.

Length of Metoestrus Period.

*%

Days.

2 Days.

2^

Days.

3 Days.

31A
Days.

4 Days.

4^
Days.

5 Days.

No. of guinea pigs. . . .

4

9

30

2O

15

10

5

4

SEXUAL CYCLE OF THE GUINEA PIG.

239

I may state that at this penod a slightly transparent serous
or a turbid or viscid secretion can be noticed; but within 2 or 3
days this will disappear, and the swelling of the external organ
will cease and normal conditions will again prevail.

PHENOMENA OF FIRST MENSTRUATION OR HEAT PERIOD IN THE

GUINEA PIG.

The first heat or menstruation of the guinea pig differs in some
respects from the later periods of heat. The symptoms of
prooestrum are of much longer duration. While as stated above,
the normal prooestrum lasts about I to 1^2 days, during the
first menstruation a slight secretion and swelling continues for
3 to 8 days; only then appears the period of heat proper. In
some instances the secretion and swelling disappear temporarily.
Then a few days later it begins again and now develops into the
regular heat. As stated, generally the duration of the procestrum
is variable and usually longer in the first menstrual period.

DICESTRUS PERIOD OF THE GUINEA PIG.

The dioestrus period is the length of time between two consecu-
tive menstruations or heat periods. Through observation of the
condition of heat we determined the length of the sexual cycle in
the guinea pig; we used for this purpose a considerably larger
material than Loeb1 and Stockard and Papanicoloau2 used in
their observations. While our average agrees with that of the
aforementioned authors, we found a greater range of variation
than Stockard and Papanicolaou. From the observations of
Loeb it follows that in some exceptional cases the period may be
even shorter than that found by us.

•

Length of Time Between Heat.

1454
Days.

Days.

IS1/!

Days.

16
Days.

16^
Days.

*7
Days.

17^
Days.

18
Days.

No., of guinea pigs ....

I

40

72

62

36

12

8

I

From these figures we may conclude that the average duration
of this period was 15 to 16^2 days. I found that where the
length of time was decidely shorter (14.1/2 days) or longer (17 to

1 Leo Loeb, BIOLOGICAL BULLETIN, 1914, XXVII. , i.

2 C. R. Stockard and G. Papanicolaou, Am. Jour. Anat., 1917, XXII., 225.

24O O. ISHII.

1 8 days) than the average, the animals were either in a weak or
pathological condition, and that menstrual irregularities sub-
sequently followed. My continuous experiments made on this
subject covered both winter and summer of 1913, and the winter
of 1914 in Granby, Mass., and during all these seasons, through
all temperatures, the same condition prevailed.

GENERAL PHENOMENA OF MENSTRUATION AND HEAT AND

PREGNANCY.

Physical Phenomena of Menstrual Period.

One of the foremost indications of the approach of the period
of menstruation or heat is the physical change noticed in the
animal. The muscles and joints become tender and relaxed,
and by this change it can be determined, whether the animal is
normal and healthy. During this period nervousness and a
certain mental depression are noticeable; and when the period is
over the symptoms most apparent indicate fatigue or exhaustion.
However, by the end of the second or third day this fatigue has
entirely disappeared and within I ^ to 2 days after cessation of
heat a normal animal has regained its full vitality.

Relation between Menstruation and Falling Out of Hair.

I have found, particularly in the guinea pig, rat and rabbit an
additional visible sign of heat, viz., the falling out of the hair in
unusual quantities during the period of menstruation; as a
result of this happening, the external appearance of the animal
becomes somewhat glossy. We can determine this by comparing
the ease with which the hair can be made to fall off by rubbing the
skin in the menstruating animal, on the one hand, and in the
infantile, pregnant, amcestrous female or male, on the other hand.
In exceptional cases a slight menstrual secretion may develop
during pregnancy; in this case the falling out of the hair is slight
in accordance with the slight degree of menstrual activity.

Specific Odor of Females during the Period of Heat.

Many (but not all) males can evidently clearly distinguish
between females which are in the period of heat on the one hand
and those in which there is merely menstrual secretion and in

SEXUAL CYCLE OF THE GUINEA PIG. 24!

which both are lacking on the other hand. Therefore the male
attempts copulation only in the former case. When a number
of animals of both sexes are together and among them is a
female in the heat period, the males run around and search for
that particular female, ultimately finding it.

I am convinced from this searching of the male for the female
in heat that he is led to do so by a special odor characteristic of
the female, when it reaches the heat period. This peculiarity
is common to the mouse, the rat and the rabbit.

Influence of Food upon the Character of Heat and on the Time
when Sexual Maturity is Reached.

The vigor of the heat in guinea pigs depends upon the nourish-
ment which the animals have received. Guinea pigs which had
been poorly fed develop weak heat. This applies to the duration
of the heat as well as to the vigor of the symptoms. The length
of the dicestral period does on the other hand not seem to be
affected by the state of nourishment.

I made observations on the influence of food on the time when
sexual maturity is reached.

I. On especially prepared rich food the first heat was observed
45 to 60 days after birth.

II. On usual food 55 to 70 days after birth.

III. On food of poor quality 75 to 100 days after birth.

The usual food consisted of hay (or grass), carrots (or beets)
and oats. In the rich diet corn cake was added, while the poor
diet consisted merely of hay (or green grass) and oats. When
the quality of the food was above the ordinary, its effect on the
size of the guinea pig as well as their physical condition in other
respects could be seen. They were of the average condition,
when ordinary food was given at regular intervals, and the poorly
fed animals were poorly developed physically. The food is
therefore of the greatest importance as far as the time of sexual
maturity and the character of the heat are concerned.

Influence of Climate on the Character of the Heat and the Length

of Sexual Cycle in the Guinea Pig.

The extreme temperature of the summer months (a tempera-
ture of over 90° F. in Granby, Mass.) has a weakening effect on

242 O. ISHII.

the guinea pigs. And the heat is therefore not so vigorous as it
would be under other climatic conditions. The length of the
dioestrous period on the other hand seemed to be the same during
the warm and cold season of the year.

Does Contact with the Male Influence the Time of Appearance of

the First Heat?

If young female guinea pigs are kept in a cage with males and
other guinea pigs of the same descent are kept separated from
males, the first heat appears in both lots at the same age. This
shows that psychical factors such as the stimulation of the male
does not influence those conditions which lead to the first heat.

The Copulation-reflex of the Heat Period and the Diagnosis of the

Heat Energy.

This reflex can be elicited by the male guinea pig seeking
copulation, when it touches or gently massages with his forefeet
the lumbar portion of the spinal column of the female which is
in heat. The female will then lift up the tail and the vagina
will be opened in preparation of the copulation. The same
reflex can be called forth if an observer touches with his finger the
same part of the body of the female guinea pig. The readiness
and vigor with which the reflex occurs can serve as a measure of
the energy or vigor of the heat.

Menstruation during Pregnancy.

I very carefully watched twenty guinea pigs during their state
of pregnancy for a sign of menstruation and found in ten no
evidence of menstrual secretion, but in the remaining ten there
was some secretion present; not all of those however showed
signs of heat. The menstrual period is not definite during
pregnancy. Statistically my results were as follows:

Menstruation was noticed in 3 guinea pigs 27 days after copulation

" 3 " " 30 "

" 2 " " 40 "

2 " " had menstrual secretion twice, the first
time 15 days and the second time 30 days after copulation.

While in these cases the swelling of the external genitalia and the
secretions are manifest, they are slight. The secreted material

SEXUAL CYCLE OF THE GUINEA PIG.

243

is mostly transparent, but sometimes turbid. It continued for
3 to 8 days in my observation; then the external genitalia
returned to a normal condition.

Some authors suppose that the diagnosis of the menstrual
secretion is very difficult in the guinea pig; however careful
observation has shown me that the diagnosis can be readily
made and hardly misinterpreted.

Diagnosis of Pregnancy about 14 to 15 Days after Copulation.

About 14 or 15 days after copulation we can correctly determine
the success or failure of the impregnation. If at that time the
external genitalia are in a normal condition we may almost always
assume a successful pregnancy. If, on the other hand, there
should be within that time a noticeable swelling of the external
genitalia, accompanied by secretion, this may in all cases be
taken as an indication of failure. As we have stated above,
the same phenomenon may be present, perhaps in a milder form,
at this period of pregnancy, but this is a very rare occurrence;
and as a general rule in the case of failure (early abortion) the
secretion and swelling is much more marked.

The Sexual Cycle in Cases of Early Abortion.

In cases of early abortion I found the period of heat in some
cases somewhat accelerated and more irregular than in normal
animals.

DlOESTRUM.

Length of Heat Following Copulation.

1.2% Days.

13^ Days.

14% Days.

15^ Days.

16 Days.

No. of guinea pigs examined

I

2

2

-j

5

As shown in the table, the heat period may appear as early
as 12^2 or J3^ days after copulation. In control cases the
shortest period observed was 14^ days and this occurred only
once; and, as I stated above, I believe that in normal cases 15
days represents the minimum and that the shorter periods are
usually found in cases of abortion.

244

O. ISHII.

The Length of the Period of Gestation in the Guinea Pig.

On the basis of our observations we can give the following
statistical data concerning the length of the period of pregnancy
in the guinea pig.

Length of Gestation Period.

61
Days.

62
Days.

64
Days.

66
Days.

67y2

Days.

68
Days.

r>

Days.

Days.

No. of pregnant guinea
viss .

2

2

4

4

8

15

5

I

The shortest period was 61 days; but those young animals
which are born 61, 62 and 64 days following copulation show a
certain lack of development; they are really born prematurely
and in the majority of cases they soon die, although they may
be able to survive. Those born 66 days after copulation usually
lived and continued to grow, but those born from 67 to 69 days
following copulation show the maximal development after birth ;
these have the best chance to develop into absolutely healthy,
perfect animals. We may therefore conclude that 68 days is the
regular period of gestation in the guinea pig.

Separation of the Symphysis Pubis in the Guinea Pig at the End
of Pregnancy, at the Time of Birth and During the
Period Directly Following Birth.

Through examination with the finger we discern in the guinea
pig a gradual separation of the symphysis pubis which begins
about 6 1 to 63 days after copulation and increases as the time
of labor approaches to a width of 7 to 8 millimeters or more;
but within one or one and a half days after birth the symphysis
pubis assumes again its normal condition. The following table
shows the time when the separation becomes apparent :

Beginning Separation of Symphysis Pubis.

61 Days, Following
Copulation.

62 Days, Following
Copulation.

63 Days, Following
Copulation.

No. of pregnant guinea pigs.

5 5

4

The length of time from the beginning of the separation of the
symphysis to the onset of labor is shown on the following table:

SEXUAL CYCLE OF THE GUINEA PIG.

245

Duration of Period of Separation of Symphysis
Pubis Prior to Birth.

4% Days.

5 Days.

7 Days.

No. of pregnant guinea pigs

'

6

7

The following table shows the length of time necessary for the
symphysis pubis to return to its normal condition following labor :

Time when the Symphysis Pubis has again
Become Normal Following Labor.

X Day.

i Day.

iyz Days.

•zY2 Days.

No. of guinea pigs

8

c;

I

From these data we may conclude that the separation of the
symphysis pubis begins from 5 to -7 days previous to birth, or
from the 6ist to the 63d day after copulation. Under ordinary
conditions birth should not take place before the 5th day following
this separation. The length of the period of gestation should
therefore usually be 68 days.

Preparation for Labor.

In the guinea pig the separation of the symphysis pubis
serves as a preparation for parturition, inasmuch as the orifice
of the normal pelvis is too narrow to permit birth to take place.
Moreover we find at the same time the same remarkable softness
and elasticity of the muscles and joints as during the ordinary
menstrual period. This same effect (emanating from the ovary
or pituitary gland) I have also noticed in the rat and in the
rabbit and I presume it is the same in all mammals.

The Time of the First Heat in the Puerperium.

The time of the beginning of heat in the period directly follow-
ing labor is somewhat variable. In 20 guinea pigs in which
parturition set in at the normal time (67 to 69 days after copula-
tion) I found the following figures :

Beginning Puerperal Heat After Parturition.

Soon
After

3 Hours
After.

5 Hours
After.

7 Hours
After.

No Heat.

No. of examined guinea pigs

7

7

3

I

2

246

O. ISHII.

As our table shows, in 2 out of 20 guinea pigs no heat at all
was observed at this period. There was evidence pointing to the
conclusion that the heat is less vigorous at this stage than in
the, normal menstrual period. On the other hand the heat seems
perhaps to be of a somewhat longer duration at this period as
shown in the following table:

Duration of Puerperal Heat.

7 to g Hrs.

10 to ii Hrs.

12 to 13 Hrs.

No. of guinea nigs . ,

8

7

7

We find therefore that in about one third of the guinea pigs
at this period the average duration of heat was 12 to 13 hours,
while in normal animals the heat continues for that length of
time only in a small minority of cases. Whether or not heat
occurs in the period directly following parturition depends upon
the duration of the period of gestation as shown in the following
table :

Duration of Pregnancy.

Efl
M >,

*o n

a

en

N >*

^o rt

Q

t/j
•«• >*

>o a
Q

in
MD >*

^o a

P

x£

f~ !8

VOQ

tn

00 >*

\o ri
Q

Ifl
o >*

^o «

Q

n

M >*

^

No. of observed guinea pigs.
Puerperal heat .

2

— 2

2

— 2

4

— 4

4
-I/+3

o

-I/+7

15

— 2/+I3

5

+ S

r

+ T

+ coming in heat; — not in heat.

This table shows that heat does not occur in the period directly
following parturition in cases in which the duration of pregnancy
was less than 66 or 67 days. In cases in which the duration of
pregnancy was less, birth occurred prematurely and the phe-
nomena of the puerperium had not yet fully developed at the
time of the completion of labor. If in cases of parturition
occurring 61 or 62 days after copulation the first heat following
parturition would take place 6 or 8 days later, the first heat
would correspond to a typical puerperal heat, provided the
duration of pregnancy would have been 68 or 69 instead of
61 days.

SEXUAL CYCLE OF THE GUINEA PIG.

247

On the Duration of the Sexual Cycle following Parturition.
The time of the second menstruation in the period following
parturition is rather variable as shown in the following table:

Development of Second Menstruation Following Parturition.

^£

m rt

MQ

tfl

VO X

M B

p

NT 1 «
WX >>

*o rt

"P

E/>
t^X

M «

Q

:«£

1- B

HP

yi

00 >i

>- rt

P

(fl

H >*

M Ctf

P

ui

M ><

"p

ui
m >%

N rt

P

rt
O

H

No. of guinea pigs

2

I

4

5

2

3

I

I

I

20

Among 20 guinea pigs in the period following the puerperium
I found complete menstruation and heat in 15 cases, while in 5
animals I found only menstrual secretion without heat. The
majority of these animals suckled their young and among those
the deviation from normal was somewhat less; in these cases the
duration was mostly 151/2 to 18 days.

Effect of Lactation and Non-Lactation on the Periodicity of the
Sexual Cycle after Parturition.

In cases in which the young were suckled the next period of
heat took place approximately at the normal time and the condi-
tions continued from then on in a normal manner. It was dif-
ferent in cases in which bodily weakness of the mother prevented
the suckling of the young. In the latter cases the return of
menstruation was delayed so that it occurred as late as 21 to 23
days after parturition. In the normal suckling animal the sexual
periodicity following labor is therefore similar to that of the
normal animal, although certain minor variations depending on
the vitality of the animal occur even in such cases. The second
or third menstruations occur usually after the regular interval of
15 to 16 days even in the weak animal. In the latter, hbweven
we> may at these times instead of heat merely find menstrual
secretion.

The Effect of Suckling on the Mother and the Young.

Marshall1 expresses the opinion that in the young separated
prematurely from the mother growth will continue. My own
observations lead me to a different conclusion. I have found

1 F. H. A. Marshall, "The Physiology of Reproduction." London, 1910.

248 O. ISHII.

that if a mother suckled three young, the growth of the latter was
very slow, or death resulted; in addition the weakening effect
on the mother was quite noticeable; it sometimes caused her
death, and in most cases the death of the young occurred. If
the young died, the mother soon regained her health and strength.
The result will therefore be disappointing either as far as the
condition of the mother or of the young is concerned. It has
been my experience that when the young were taken away from
the mother within 15 or 20 days after birth and the ordinary
vegetable and oat diet was given them, development was very
slow and in most cases death occurred. In any case the young
to attain healthy growth should not be taken from the mother
until more than 25 to 30 days of age. Under natural conditions
they cease suckling after 35 to 40 days, when they seek their own
food. I am therefore of the opinion that a litter of three very
seldom shows normal development when suckled by one mother.
When I combined one mother with two and another with
three young, each mother suckled on the average 2^/2 young.
In such cases I obtained good results. While two or three young
is the average size of a litter, I have known the number to vary
from one to six; the latter number, however, only occurred once
in my observation. When but one young is born, it is usually
well developed and vigorous; but to be successfully raised it
must be suckled for the usual period or at least for a time ap-
proaching it. There is another symptom in addition to the
bodily weakness, which the mother shows as the result of too
intensive suckling: it consists in the changing of the color of
the pupil of the eye which becomes pale or white. It may be
associated with general exhaustion.

Under what Condition is Copulation followed by Pregnancy?

As we stated above the normal duration of heat is about 10
hours. If copulation took place within 3 or 4 hours after the
beginning of heat, pregnancy followed only rarely. If the copula-
tion took place at a later period of the heat, better results were
obtained. According to my observations pregnancy occurred
in about 73 per cent, to 80 per cent, of the cases. In such cases
in which copulation was not followed by pregnancy a second or

SEXUAL CYCLE OF THE GUINEA PIG. 249

third attempt during the next two periods of heat were mostly
successful.

Sobotta1 states that the guinea pig in captivity can become
pregnant more frequently in summer than in winter. My
observations do not bear out this statement. In my experience
the pregnancies resulted almost as often in the cold as in the
hot season. The -winter climate is better adapted for the guinea
pig in captivity than the hot summer weather when the tempera-
ture is liable to exceed 90° F. Under those conditions the animal
suffers and becomes poorly developed as the result of under-
nourishment. On the other hand, it is almost always in good
condition in winter. Therefore the cold season agrees better
with guinea pigs than the hot season under conditions of cap-
tivity. In exceptional cases the guinea pig may become in
winter time too fat to be favorable for impregnation; and this
may even happen in summer time.

Relation between Growth, the First Period of Heat and the Period
of Sexual Maturity in the Guinea Pig.

In my observations the weight of the guinea pig is ordinarily
three ounces soon after birth. However, the weight of the
young is in accordance with the age of the mother and also the
number of the litter. As a rule the young born from a young
mother are small in size and in litter. On the other hand, when
the mother is older than 5 months, the weight of the young is
mostly larger. For instance, in one litter the young weighed
6 or 7 ounces soon after birth.

Minot2 has concluded that the guinea pig is slow in growing
and that it attains its full size one year after birth. My observa-
tions are as follows: The young guinea pig at the time of the
first heat was about two months old; about 7 to 10 days before
the beginning of heat, . the growth almost stopped or at least
progressed very slowly. At the time of the first heat period the
average weight was n to 13 ounces. After the first heat period
had passed it began to grow very rapidly. About two or three

1 Sobotta, Anal. Hefte, 1906, XXXII.

2 Minot, C. S., "Growth and Senescence," Journ. Physiol., 1891, XII., 97.
"Problems of Age, Growth and Death," Popular Science Monthly, 1907, LXX..
481, and LXXI., 97 and succeeding numbers.

250 O. ISHII.

months later, at the age of 4 to 5 months, it attained its full
weight, about 20 to 28 ounces, but one exceptional animal
weighed 34 ounces in a non-pregnant condition. However, if
we feed a special diet of nutrient material and the animal is
otherwise well taken care of, then the growth progresses rapidly
and the heat begins earlier than usual. The guinea pig which
weighed 34 ounces was fed on the usual farm or laboratory food.
Therefore we must remember that the growth of the young
depends considerably upon the character of the food.

Vol. XXXVIII. May, 1920. No. 5.

BIOLOGICAL BULLETIN

SELECTIVE FERTILIZATION IN POLLEN MIXTURES.

D. F. JONES,
CONNECTICUT AGRICULTURAL EXPERIMENT STATION, NEW HAVEN.

Equality in fertilizing power of gametes dissimilar in the
hereditary factors they carry is a corollary of Mendelism. It
has been thought that certain exceptions to this general rule
may exist. Differential fecundating ability has been suggested
from time to time as a possible interpretation of otherwise un-
accountable results. The earlier uses of this hypothesis of
inequality in sperm efficiency, such as Castle's (1903) theory of
sex determination, Cuenot's (1908) conception of the reason for
the non-appearance of homozygous yellow mice, have since been
found to be unnecessary. The interpretation of linkage phe-
nomena according to the reduplication hypothesis, while essen-
tially different in application, holds somewhat the same implica-
tion as selective fertilization, that is, differential operative power
of cells of unlike germinal construction.

As the yellow mouse problem was finally resolved to a selective
elimination of zygotes so many puzzles have had considerable
light thrown on them by a more complete understanding of the
factor relations. Many instances of the non-appearance of cer-
tain types are known to be the result of the action of lethal factors.
Good illustrations of/ this are found in Drosophila (Muller, 1918).
Other cases of elimination immediately after fertilization or early
in development are known in maize.

Sometimes abortion of a part of the gametes before fertilization
takes place and it is assumed that the elimination is selective.
Belling (1914) has shown that in Stizolobium 50 per cent, of the
pollen grains and ovules regularly abort in certain types. The
aberrant results from the (Enotheras are now generally considered
largely, if not wholly, to be due to differential destruction of both

251

252 D. F. JONES.

gametes and zygotes. Davis (1915-' 17) has called attention to
the great amount of pollen and ovule abortion in these plants
and to the low germination of the seeds which are produced
and to the results which are obtained when a more complete
germination is secured.

It is not always easy to distinguish between differential destruc-
tion of zygotes immediately after fertilization and selective
fertilization. It is still more difficult in the case of abortion of
germ cells. In fact it is only a matter of degree between selective
elimination of gametes and selective fertilizing ability- From
germ cells which are unable to complete development on account
of the particular inheritances they carry, to gametes which
appear normal but are unable to function perfectly under any
circumstances, it is only a step. However, no instance of the
latter condition is positively known and the cases of gametic
abortion clearly due to the factors which the gametophytes
carry are rare.

Although results still remain to be cleared up which indicate a
selective action of some kind (Kempton, 1919) the general con-
clusion holds, which is, that the pollen, although gametophytic
itself, has the function of the sporophyte which produces it.
This is supported by the well-known cases where pollen color
and shape are all of the type of the maternal parent irrespective
of the factors for pollen characters which they carry. Many
instances are known where gametes, containing lethal factors
which stop development immediately after fertilization, are able
to function fully as well as others permitting normal growth.
East and Park (1917) have shown that compatibility in Nicotiana
is alike for all the gametes of any one individual. Different
pollen grains may carry factors which will determine sterility
or fertility of the future zygotes but all function alike in fertiliza-
tion according to the sporophyte from which they come. Follow-
ing up these results East (1919) shows that in the frequency dis-
tributions of pollen tube lengths there is no significant difference
between pollen from plants greatly unlike in the degree of
heterozogosity. In other words gametes carrying markedly di-
verse germinal complexes are no more variable than those all of
like constitution in ability to send pollen tubes down the styles

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 253

and to accomplish fertilization. East considers that chance
segregation in the germ cells and random mating of these germ
cells is a fundamental genetic hypothesis applicable to plants
and animals alike. With this conclusion the writer is in accord,
with the reservation that the evidence from gametic abortion
should not be put aside as belonging to a different category.

The data to be presented here bear upon another phase of the
problem. The above conclusion, it should be clearly kept in
mind, applies only to gametes produced by one or more indi-
viduals of the same type. That is, the gametes may be unlike
in the factors they carry, but if they come from the same or
similar individuals they are potentially equal in ability to
fertilize. But what is the result when germ cells from two
individuals of different type are presented at the same time in
excess so that not all can fulfill their function? Will fertilization
take place at random or not? This is the problem to be con-
sidered here. In the one case doses of different kinds of medicine
come in the same capsules. In the other the capsules as well
as their contents may differ. Anyone accustomed to swallowing
a particular kind of capsule made of a familiar substance and all
of the same size and shape can take any kind of medicine with
equal readiness irrespective of the result which will ensue when
the materials within the containers begin to operate. Applying
this crude metaphor to plants and animals, in the one case, the
germ cells, however they differ in factorial composition, come
in the same cytoplasmic envelope; in the other, the cytoplasm
as well as the genes may differ.

It has already been shown that, in particular cases, there is a
selective action when pollen from different plants is applied to a
stigma at the same time. East (1919) has demonstrated that
in a mixture of compatible and -incompatible pollen placed on
the stigmas of self-sterile Nicotianas the compatible pollen alone
functions. Two experiments were devised to test this. Small
numbers of pollen grains were counted out under a microscope
and placed on stigmas known to be receptive to that kind of
pollen. The stigmas were then covered with a large amount of
incompatible pollen. The former only was able to fertilize as
shown by the results from eight such mixed pollinations which

254 D. F. JONES.

in no case produced a greater number of seeds than the number
of compatible pollen grains applied. In the other experiment
a constant white-flowered self-sterile plant was pollinated with
its own pollen. Several hours later pollen of a red-flowered
plant was applied. Abundant seed was secured and when grown
red-flowered cross-fertilized plants only were found. In both
these cases there was complete selective fertilization in favor
of the pollen from dissimilar plants.

According to the results to be reported here a directly reversed
effect is obtained from mixed pollinations in maize. In many
trials the plant's own pollen has been more efficient in accom-
plishing fertilization than that from other individuals which
differ only in minor features. This same pollen which is less
effective when in competition with the plant's own kind of
pollen is fully able to function when not applied in mixtures.
The results are remarkable in view of the notable advantages
which cross-fertilization gives to the immediately resulting seeds
and the plants grown from them. The material used consisted
largely of self-fertilized strains which had been brought to uni-
formity and constancy and were considerably reduced in size
and vigor. In this material crossing is known to increase the
weight of seed within the same inflorescences as much as 50
per cent, in some cases. This greater amount of material is
laid down both in the embryo and endosperm, and is apparent
in the greater size of the seeds which show a higher specific
gravity together with more rapid maturation indicated by their
lower water content at the close of the growing period. A large
series of mixed pollinations show the ability of the cross-fertilized
seeds to germinate better by an average of 16 per cent. The
resulting plants start to grow sooner, develop faster, mature in
a shorter time and at the end far surpass their self-pollinated sibs.
The amount of heterosis shown by maize is possibly greater than
that displayed by any other plants in intra-specific crosses.
Production of grain has been advanced on an average of 180
per cent., height of plant 27, length of pistillate spike 29, number
of rows of spikelets 5 and number of nodes 6 per cent, in a study
of a large number of crosses between inbred strains (Jones, 1918).

In spite of these great immediate advantages to be secured the

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 255

plants manifest a decided preference for their own kind of pollen.
This is a result which would have surprised students of flower
pollination in Darwin's time but which, as I shall attempt to
point out later, is in agreement with other results from biological •
investigation.

METHODS OF CARRYING ON THE EXPERIMENTS.

In an investigation in which it was desired to compare the
chemical composition of seeds of maize having different genetic
constitution but produced under as nearly identical conditions
as possible advantage was taken of endosperm characters to
enable proper classification of seeds produced in the same inflores-
cences. For example, two kinds of pollen carrying yellow endo-
sperm color and white endosperm color were mixed together
and applied to a plant which normally produces uncolored seeds.
The resulting yellow and white seeds were distributed at random
on the pistillate spikes and were as nearly comparable in respect
to external and nutritional factors as it is possible to obtain.
It was found that such pollen mixtures could be applied to both
strains furnishing the pollen and the two kinds of seeds easily
distinguished. On the white-seeded strain the seeds resulting
from the plant's own pollen were white and the cross-fertilized
seeds yellow. On the yellow-seeded plants the self-pollinated
endosperms were dark yellow while the crossed seeds were dis-
tinctly lighter in color, in most cases they had a white cap, and
were as a rule readily separable.

Attention is directed to the fact that the material used for
these mixed pollinations consisted largely of inbred strains which
had been reduced to uniformity and constancy so that the
genetical differences between the two kinds of seeds sharply
differentiated them, much more than in ordinary cross-pollinated
varieties of this plant in which the yellow color is usually variable
due to more than one factor for this color and various modifying
conditions such as the consistency of the endosperm in respect to
corneous and floury starch.

A number of pairs of plants were treated in this way by mixing
their pollen and applying to both types. After harvesting it was
realized that here was an excellent method of determining

256 D. F. JONES.

whether or not any selective action was shown by the plant's
own pollen as compared to that from a plant of somewhat
different type. If one member of the pair of plants which
furnished the pollen for the mixture is designated A and the
other B, the two kinds of seeds grown on A plants are A X A,
self- fertilized, and A X B, cross-fertilized; on the B plants
B X A cross-fertilized and B X B self- fertilized. Since the same
pollen mixture is applied to both, the ratio of the seeds resulting
from A pollen to the seeds resulting from B pollen, on A plants,
should be the same as the ratio of the seeds resulting from A and
from B pollen on B plants. In other words, the numbers form
a proportion which, irrespective of the relative amounts of A
and B pollen in the mixture, should be a perfect proportion
within the limits of random sampling if fertilization takes place
at random. The end terms of the proportion comprise the
self-pollinated seeds and the middle terms the reciprocally cross-
pollinated seeds. If a true proportion is obtained the products
of the end terms should naturally equal the products of the middle
terms. If they do not the deviation is either in an excess of
cross-fertilized or of self-fertilized seeds indicating a selective
action in one or the other direction.

The advantages of this method of attacking the problem are
readily apparent. It is practically impossible to make up a
mixture of large amounts of pollen in which the proportion of
each is known. Either measuring or weighing the pollen is
out of the question because maize pollen takes up moisture from
the air rapidly and when any quantity is brought together it
becomes aggregated into a flocculent mass. Moreover this pollen
soon loses its viability so that even in case equal numbers of
pollen grains could be had there would be no proof that there
were equal numbers of functional pollen grains in the mixture.

The method of reciprocal application and arranging the results
in the form of a proportion automatically overcomes all these
difficulties and the experiment is as simple as could be devised.
An attempt was made to have as nearly equal quantities of pollen
as possible by measuring out the two kinds roughly. But in
many cases the results showed that one kind of pollen was far
more effective than the other. This, however, does not destroy

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 257

the value of the figures as it is the relative efficiency of each type
of pollen, when it is applied to its own and to foreign stigmas,
that is being investigated.

In the preliminary experiments there were 18 pairs of plants
which produced seed as the result of the application of mixed
pollen. When these were counted and the deviations of the
proportions found from the closest perfect proportions were
obtained there were 12 mixtures which showed a deviation in
favor of the self-fertilized seeds and 6 in the opposite direction.
The results as a whole showed a tendency to favor the plant's
own pollen. This was somewhat unexpected so that it was con-
sidered worth while to investigate the matter more fully.

The inbred material was so reduced in growth that the number
of seeds produced on one plant was not large enough to give the
results much weight. It was therefore decided to pollinate a
number of plants of two different self-fertilized strains with the
same mixture. Most of the strains used had been self-pollinated
for six generations or more, some as many as ten, so that the
plants within one strain were practically identical in hereditary
constitution. Pollen was collected from about the same number
of plants as the pollen was applied to. The two lots of pollen
were put together in a paper sack and thoroughly mixed by
shaking. This mixture was then applied to plants of the two
strains which supplied the pollen. It was desired to have from
ten to fifteen plants in each of the paired strains so as to give
from 1,000 to 2,000 seeds in each of the two parts of the propor-
tions but the flowers were not always ready at the right time
and some pollinations were failures for a variety of reasons so
that not as large numbers as desired were secured in every case.

Every effort was made to prevent the entrance of undesired
pollen. The technique has been described and the amount of
experimental error due to contamination to be expected in
artificial pollination of maize has been considered previously
(East and Hayes, 1911). In the course of these experiments the
effects of extraneous pollen were seen in very small numbers
compared to the total number. This source of error could be
detected in the seeds when colors or other characters differing
from either of the strains used were brought in by the undesired

258 D. F. JONES.

pollen. In a total of 63,000 seeds produced in these experiments
only about 30 illegitimate seeds were observed. A larger number
of contaminations, however, would have the same appearance as
the legitimate seeds and so could not be detected but giving a
reasonable allowance to this source of unreliability the results
could not be appreciably altered. Moreover the error of this
kind is never all in one direction so that in the main it can be
disregarded.

The accuracy with which the seeds have been classified de-
serves particular attention. Strains were selected to be used
which gave sharp differences between self-fertilized and cross-
fertilized seeds and in most cases separation was made very
satisfactorily. In a few mixtures there was some doubt and in
two experiments the seeds on the yellow endosperm plants could
not be distinguished. In these two cases the seeds were planted
and classification was made with the mature plants. Also in all
the other mixtures involving yellow and white endosperms a
sample was taken, after all the seeds of one class were mixed
together, and grown to determine the per cent, of error in sepa-
rating the seeds. Since the self-fertilized seeds give small inbred
plants, pure for yellow or white color, while the cross-fertilized
seeds produce large vigorous hybrids segregating into yellow
and white seeds, classification of the mature plants can be made
without the least doubt. However, it should be noted that the
better germination and greater vigor give the advantage to the
cross-fertilized classes in every case if there is any difference.
About 1 20 plants in each lot were grown and the per cent, of
error obtained was used to calculate the total amount of error
if all the seeds had been grown. Since the numbers of seed ran
up into the thousands it was impossible to grow all of them. The
figures showing the per cent, of errors found are arranged in
Table III. Only in one case is the number of wrongly classified
individuals above 3 per cent. In 25 out of 44 lots no faulty
separations were discovered. In the remaining cases the mis-
placed seeds tend to balance each other so that this source of
doubt can be largely removed. Calculating the figures without
regard to the error of classification gives practically the same
results as when this is taken into consideration.

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 259

Different characters were used in other mixed pollinations
which permitted even more positive classification than the single
character difference of yellow and white endosperms. Plants
with yellow sweet seeds were paired with white starchy plants.
Each strain brought in a dominant character. White starchy
plants of the pop or Zea mays everta type were used because the
clear corneous endosperm differentiates very clearly between
yellow and white. In the reciprocal cross the smooth opaque
seeds show up plainly contrasted with the wrinkled translucent
seeds. Similarly purple sweet and white starchy types were
paired. In a few cases the purple crossed seeds were not as
distinct as could be desired, but the error is certainly small.
These latter mixtures were made the past season and no plants
have been grown to test the reliability of their separation but the
writer is confident that the number of wrongly classified seeds,
if any, is not sufficient to alter the results, appreciably.

In many inflorescences a few seeds were found which had
failed to reach a stage of development so that they could be
classified. Seeds at the tips of the spikes and where the seeds
were closely crowded were abortive. This introduces another
source of unreliability, that of selective elimination of zygotes.
Since crossing gives to seeds of maize an enormous advantage in
development, it can confidently be expected that as a rule more
of the self-pollinated seeds will be found among the abortions
than cross-pollinated. What we are attempting to study is the
•relative fertilizing efficiency of different types of pollen grains.
But one can only arrive at this by counting the zygotes sometime
after fertilization has taken place. In the meantime a differential
destruction of zygotes may have taken place. This effect must
be considered in any organisms employed but because of the
short time which elapses between fertilization and the maturation
of the seeds, and from the fact that they develop in an exceedingly
favorable and uniform environment, maize is the very best
material the writer can think of in which this problem can be
attacked, especially when the numbers which can be obtained
are taken into consideration. All animals and those plants
which d3 not show zenia in the seeds have the objection that a
comparatively long time elapses between fertilization and suffi-

26o

D. F. JONES.

cient maturity to permit classification. In plants many cases
are known in which there is a selective elimination of certain
classes of individuals due to a lesser germination and unequal
ability to grow. In Drosophila (Hyde, 1914) crossing does not
influence the number of eggs laid but markedly regulates the
per cent, that hatch. Therefore the error from this source
always tends to show an apparent deviation in favor of cross-
fertilization. In maize where the seed progenies can be used this
differentiating effect is at a minimum and probably is not suffi-
cient to affect the numbers appreciably but since the tendency is
in the opposite direction to the results which have been obtained
the data are even more convincing.

TABLE I.

CHARACTERS OF THE SEEDS AND PEDIGREE NUMBERS OF THE INBRED STRAINS AND FIRST

HYBRIDS USED IN THE MIXED POLLINATIONS.

GENERATION

Pollen

Mixture
Number.

Characters of the Seeds of

Pedigree Number of

A Strain.

B. Strain.

A Strain.

E. Strain.

I

Yellow Starchy

ti ( t

a «

4 ( I (

It ( (
ft tt

Yellow Sweet
Yellow and White Sweet
Purple Sweet

t < * (

n n
1 1 it

White Starchy

ti u
it it

It It

14 tl

tl It
tl It

I ( tt

tt 41

41 tt

tl I t

White Starchy
tt it

tt it
tt tt

White Starchy

tt it

14 tt

"

1-9-1-2-4-6-7-5-6-2-1
1-9-1-2-4-6-7-5-3-2-1

14-4-6-16-2-12-22
1 4-4—6—16—2—1 2-2 2
I4-4-6-I6—2—I2-22
I4-30-4-3-7-II-IO

I4-30-4-3-7-II-IO
I4-4-6-I6-2-I2-22

I-9-I-2-4-6-7-5-6-2-I-I

(1-6-1-3 X 1-9-1-2) Fi
(1-6-1-3 X 1-7-1-1) Fi
(1-7-1-2 X 1-6-1-3) Fi

146-1-1
146-1-1

(126-1-1-1 X 77) Fi
(126-1-1-1 X 77) Fi

76-2-2-1-1
76-2-1-2-1

76-2-1-2-1
76-2-2-1-1

10-4-8-3-5-3-4-8-2-1
I 0-4-8-3-5-3-4-5-2-1

21-13-9-7-57-11
21-13-9-7-57-11
21-13-9-7-57-11
2I-I3-9-7-57-I I

20-4-25-47-4
20-8-5-35-20

20-4-25-47-4

(20-8 X 21-13) Fi
(20-8 X 21-13) Fi
(21-13 X 20-8) Fi

65-8-2-2-6-5-3-2-1-1-1
65-8 -2-2-6-5-3-2-1-1-1

117-3-1-1

117-1-1-1

117—1—1—1

117-1-1-1

20-4-25-47-24-1
20-4-25-47-24-1

2

3

4

?

6

7

8

Q ...

IO

II

12

I"?

14

15

16

17

18

IO. .

20. . .

i

C Strain

C Strain

Yellow Starchy
White Sweet
White Starchy

1-6-1-3-4-4-4-2-4-4-2-5
77-2-1-1-1
117-1-1-1

19

20

SELECTIVE FERTILIZATION IN POLLEN MIXTURES.

26l

MATERIALS USED AND PRESENTATION OF THE RESULTS.

The figures obtained from the preliminary experiments in
which pairs of single plants only were used are not given here
because the numbers are too low to give the results much value
and for fear of making this report too bulky. It is sufficient to
remember that the data taken together indicated a slight selec-
tive action favoring the plant's own pollen.

TABLE II.

SUMMARY OF ALL THE POLLEN MIXTURES GIV.ING THE TOTAL NUMBER OF SEEDS,
THE NUMBERS IN EACH CLASS FORMING PROPORTIONS, THE DEVIATIONS OF
WHICH, EXPRESSED AS PER CENT., FROM THE CLOSEST TRUE PROPORTIONS, ARE
ALMOST WHOLLY IN FAVOR OF THE PLANT'S OWN POLLEN. P is THE PROB-
ABILITY THAT THESE DEVIATIONS ARE DUE TO THE DIFFERENCES OF RANDOM

SAMPLING.

Pollen Total
Mixture Number
Number, j of Seeds.

AX A.

AXB.

BXA.

BXB.

Deviation
from True
Proportion,
Per Cent.

X2-

P.

I
2

3,430
5,636

1,738
2,133

46
145

1,602
3.080

44
278

+ -045
+ -955

.027

7-572

-994
.063

3

4
5
6

1,362

3,344
1,956
424

229

710

589
40

14
126

6
7i

770
1,856
1,290

187

349
652

7i
126

+ 12.715

+ 5-465
+ 2.105
-11.850

146.196
61.203
28.718
23-858

.000
.000
.000
.000

7
8

3-459

7,783

23

2,185

89
956

1,507
2,619

1,840
2,023

-12.245
+ 6.570

235-357
144.108

.000
.000

9

8,729

2,550

1,288

2,922

1,969

+ 3-350

42.061 ; .000

10 5,408
ii 3-314

12 3-561

1,084
448
1,724

I-I54
264

719

997
1.505
749

2,173
1,097

369

+ 8.495
+ 2.540
+ 1-790

162.687
8-937
5.313

.000
-030
•053

13

14

736
792

185

424

39i

150

95

156

65
62

-13-625
+ I-I55

55-051
•533

.000
.894

15

16

3,i68
2,224

2,609
723

47
8

14

74

498
1,419

+47-750
+46.975

2,889.561
1,965.981

.000
.000

17 1,410 1,303
18 1,599 4

3

21

4

i

100

1,573

+47-960
+ 7-970

1,298.993
I37-532

.000
.000

19

20

2,606
2,753

528
897

392

77

343
1,174

1.343
605

+ 18.525
+ 13-050

376.394
283.002

.000

.000

63,694

Of those experiments in which pollen was applied to several
plants of the same type 20 pollen mixtures in all have been made.

262 D. F. JONES.

Each mixture is given a number and the seed characters of the
materials used and their pedigree numbers are given in Table I.
The first number in the pedigree designates the variety from
which the inbred strains were derived. The following numbers
indicate the progenitors in the successive self-fertilized genera-
tions. The total number of units in the pedigree number less
one, show the number of generations the material had been
self-fertilized at the time the pollinations were made. The inbred
strains used in these experiments are as follows :

i. Several distinct strains from a yellow dent variety originally
obtained in Illinois and known as Chester's Learning, self-
fertilized ten or more years.

10. A strain with white floury seeds with no traces of corneous
starch, self-fertilized nine generations.

14. Two distinct strains from a yellow dent variety from Con-
necticut known as Stadtmueller's Learning, selected for high
protein content during six generations of self-fertilization.

20. Two distinct strains from a white dent variety originally
selected for high protein at the Illinois Exper. Station and further
selected during four generations of self-fertilization.

21. One strain from same source as above but selected for low
protein in field pollinated cultures and during five generations of
self-fertilization.

65. A small, white, round-seeded strain from a variety of pop-
corn, Zea mays everta, self-fertilized ten times and characterized
by clear corneous starch.

76. Two similar strains from a sweet variety of latent flint
type having purple aleurone and known as Black Mexican,
self-fertilized four years.

77. From a sweet variety of latent dent type with deeply
wrinkled white seeds known as Evergreen and self-fertilized four
generations.

126. From a small, early maturing, yellow, sweet variety of
latent flint type with dark yellow kernels, known as Golden
Bantam, and self-fertilized three generations.

117. From a variety of popcorn with sharp pointed seeds
having clear corneous endosperm, self-fertilized three times.

146. From a yellow, sweet variety, Golden Bantam, of different

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 263

source than 126, and somewhat different in type, self-fertilized
two years.

Mixtures number i to 9 inclusive comprise various inbred
strains with yellow starchy and white starchy endosperm. Some
of these strains have been described previously (Jones, 1918) and
all show marked heterosis in the crossed seeds and in the resulting
first generation hybrid plants. A sample of all the different lots
of seed secured in these mixtures have been grown to test the
accuracy of classification. Pollen mixtures number 10 to 12
are not from inbred strains but from first generation hybrids, one
having all yellow seeds the other all white. They were of such
a constitution that the second crossing gave still more increase
in vigor although not as great as the stimulus following the
first cross. The plants being vigorous a large amount of seed
was obtained from a few plants. It was desired to know whether
the same selective action would be shown by vigorous plants
with segregating gametes as contrasted with non-vigorous plants
whose gametes were all alike. Classification of the seed was
easily carried out and the per cent, of error when tested was
found to be quite low. Mixtures 13 and 14 involved yellow
sweet in one strain and white starchy endosperm in the other.
One dominant factor was carried by each so that differentiation
was perfectly distinct in the reciprocal applications. In mixtures
15 and 16 it was intended to make use of the same characters
as in the two preceding numbers. The plants which were sup-
posed to be inbred individuals of a yellow sweet strain, and so
labeled, when grown in the field were seen to be too vigorous for
this material as it had behaved in previous years. The plants
were all alike, however, in this respect. It was suspected that
this was a lot of first generation hybrids instead of plants from
self-fertilized seed. As some crosses had been made with plants
of this line the previous season it is now certain that in this
instance the seed was not properly labeled at harvest and so was
planted for self-fertilized seed when in reality it was all cross-
fertilized. As no other plants were available at the time they
were needed these were used. Several self-pollinations were
made at the same time to show what the seed characters were of
this undoubted hybrid. At maturity it definitely showed itself

264 D. F. JONES.

to be a hybrid as all the ears were segregating for yellow and
white sweet seeds. It therefore had certainly been crossed with
a white sweet inbred strain the year before as other pollinations
of this sort were made at that time. The yellow color of the
endosperm was not sufficiently diluted to cause the seed to be
suspicioned before planting.

Since the plants were segregating for yellow and white it was
not material which would have been used ordinarily for mixing
with a white starchy strain. The effects of the starchy-carrying
pollen showed up all right among the all sweet seeds but the
reciprocal cross-pollination showed only the yellow cross-fertilized
seeds. The white cross-fertilized seeds of course could not be
distinguished from the self-fertilized seeds. But since half of the
pollen grains carried yellow and half of them white the number
of yellow seeds can be doubled to give the total number of cross-
pollinated seeds and the assumed number of white cross-fertilized
seeds subtracted from the white seeds. This increases the error
from random sampling somewhat but since the number of yellow
seeds is very low in comparison with the white in these mixtures,
the data are reliable in view of the great selective action shown in
these two mixtures. The fact that the yellow color in this
material was a unit factor difference and that there were equal
number of pollen grains carrying yellow and white is proven by
the self-fertilized ears produced by the hybrid which gave very
good 3 to i ratios (one ear counted gave 318 yellow and 97 white
seeds). Furthermore the starchy crossed seeds produced on the
hybrid plants were of two kinds, yellow and white, and were
produced in equal numbers (actual numbers: 28 yellow and 27
white. Since the ovules were segregating equally and the self-
fertilized seeds gave the mono-hybrid ratio the pollen grains
must have carried the two colors in equal amount.

In the last four mixtures the characters, purple sweet and white
starchy, were used. In numbers 17 and 18 the plants were not
producitve and the numbers of seeds are low. Also the classifica-
tion of purple starchy cross-fertilized seeds and white starchy
self-fertilized was not as sure as in the other mixtures. In the
last two mixtures satisfactory numbers were obtained and the
differentiation was clear-cut on both sides.

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 265

The pollen mixtures I, 19 and 20 were also applied to a third
strain distinct from either of the two used in supplying the
pollen. The resulting two lots of seed in each case were both
cross-fertilized and probably showed hybrid vigor in the seed in
about equal amount. Heterosis is clearly apparent in the plants
of similar crosses involving the same or closely related material
and is approximately equal.

The data from each pollen mixture are presented in the form
of an appendix. Since the arrangement is the same in all the
tables these are given in the simplest form possible and a descrip-
tion of one applies to all. In the tables the headings for the
four different classes of seed are A X A, A X B, B X A, B X B
and in the three mixtures I, 19 and 20 there are in addition the
out-crossed seeds C X A, and C X B. In every case the pistil-
late parent is given first. These headings when expanded in
detail are as follows :

A X A = Self-fertilized seeds from color-carrying pollen.

A X B == Cross " non-color "

B X A == Cross " color

B X B = Self " non-color "

C X A == Out-crossed " color

C X A == Out " non-color"

On every plant there are two kinds of seeds indicating the
relative fertilizing efficiency of the two kinds of pollen in the
mixture used. The number of seeds from individual plants are
given in the tables under their respective headings, A X A and
A X B seed from the same inflorescence and B X A and B X B
seeds from another and C X A and C X B from a third. The
total number of seeds is summed up below the line. Next to that
is given the per cent, of error found upon testing the accuracy of
the classification. This figure is used to calculate the amount of
error based upon all the seeds obtained. Not more than about
125 seeds were grown in each lot except in two cases as shown in
Table III. The calculated amount of wrongly classified seeds is
given on the line below the per cent, of error, then follows the
corrected numbers after the proper additions and subtractions
have been made. After that is placed the observed proportion

266

D. F. JONES.

of the two kinds of seed in each member of the pair stated as
per cent. The closest perfect proportion is calculated from this,
based on the results from the A and B plants but not from the
C plants, and this subtracted from the actual proportion found
gives the deviation in per cent, which appears in the last line.
This description applies to mixtures I to 12 inclusive. Of the
remainder no correction for misplaced seeds is made.

TABLE III.

NUMBER OF PLANTS GROWN AND THE PER CENT. OF WRONGLY CLASSIFIED SEEDS
IN THE POLLEN MIXTURES INVOLVING YELLOW AND WHITE, STARCHY ENDO-
SPERM.

Pollen

A.xA.

A. >

T>

B.)

<A.

B. )

<B.

Number.
Mixture

Number. ] Per Cent.

Number.

Per Cent.

Number.

Per Cent.

Number.

Per Cent.

I

41 J. 121

26

7 60

IO7

O

27

22 22

2

481 2 60

^O

8 oo

11^

O

128

2 7.1

7

105 o

12

o

I IQ

O

T 14

I 7 r

4

116 o

IOI

99

I2O

8?

I Id

I 72

t; .

no o

6

O

120

o

S6

o

6.

40 —

71

112

o

76

2 67

1 .

2T.

80

I IO

o

118

o

8

119 o

118

o

I2O

o

118

I 69

9

118 2.^4

118

2.S4

I2O

1.67

117

o

10

117 I.7I

118

O

118

I 60

117

I 71

ii

114 .88

119

o

12^?

o

I2O

o

12. .

116 I .86

117

o

116

o

116

0

The total number of seeds, the numbers in each class and the
deviations in per cent, from the closest perfect proportion for
the 20 pollen mixtures are summarized in Table II. The devia-
tion, if it is an excess of self-fertilized seeds, is given as plus and,
if the opposite, as minus. The deviations can range from + 50.0
to - 50.0 per cent. The extremes indicate complete non-func-
tioning of each kind of pollen on one set of plants and exclusive
functioning on the other.

Altogether the number of seeds amounts to 63,694. Large
numbers are, of course, necessary to be convincing in any experi-
ment on selective fertilization as such investigations are largely
studies in sampling. Mixtures number 6, 7, 13, 17 and 18 are
less reliable than the others, because they have rather small
numbers from either the A or B plants, less than 200 in each case.
Like the chain with the weak link the value of each set of figures

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 267

is dependent upon the numbers in the less populous half of the
proportion.

Some of the mixtures have over one thousand individuals in
each of the four classes. Since the experimental error is low
and not all in one direction as has been shown and the selective
elimination of zygotes tends to obscure the result which has
been obtained, such figures as these can not be gainsayed. Of
the 20 mixed pollinations 17 show a deviation indicating a
selective action in favor of the plants' own pollen. While three
of the mixtures show the opposite effect. These three are all
low in number of individuals on one or the other side of the pro-
portion. Mixtures 6 and 7 could not be classified by the seeds on
the yellow seeded plants so consequently the progenies were
grown and classified at maturity. This brings in other sources
of error — differential germination and competition between plants
which are weak with those that are vigorous — which certainly
tend to result to the apparent advantage of cross-fertilization.
Two of the deviations showing an excess of self-fertilized indi-
viduals are not significant when compared to the allowable
differences from random sampling but all the others are. There-
fore the conclusion is inevasible that in maize the plant's own
pollen is more effective in consummating fertilization than pollen
from plants of only slightly different construction. This selec-
tive action is shown even though the foreign pollen is perfectly
capable of fertilizing the plants when not acting in competition
with the plant's own kind of pollen as has been definitely proven.
Mixtures number 15, 16 and 17 show 47 out of a possible 50
per cent, deviation, almost complete non-functioning of the
unfamiliar pollen. Numbers 15 and 16 include the first genera-
tion hybrids in which some calculations had to be made to allow
for segregation but these are perfectly justifiable adjustments
and the results can be discounted but very little. Mixture 17
is low in numbers having only 104 seeds on the B plants of which
4 are cross-fertilized. But observe the result when this same
mixture is applied to the A plants. Here only 3 cross-fertilized
seeds are to be found among 1,303 self-fertilized seeds. Surely
there is some powerful action working to hold back the unfamiliar
pollen. Mixtures number 3, 4, 8, 10, 19 and 20 are by them-

268 D. F. JONES.

selves convincing as in these the numbers are large, the differen-
tiation of the seeds in both groups is precise and the deviations
clearly show the superiority of self-pollination.

The data from pollen mixture number I have been published
previously (Jones, 1918) although at that time plants had not
been grown to test the error in separating the seeds. The greatest
number of mistakes of classification of any of the mixtures were
made in this lot and the deviation is now well within the limits
of random sampling. The data are included here to make this
report complete. In the previous publication the probable error
used was the familiar formula used for Mendelian ratios. The
determinations applied to each half of the proportion alone.
Since the ratio on each of the paired plants is dependent upon
the ratio on the other, it seems to the writer now that the use
of this method of calculating the probable error in connection
with this particular problem is wrong.

The method of calculating the significance of the figures as
used here is that proposed by Elderton (1901) and is in general
use in presenting genetic data. It is obtained in the following
manner. The deviations of the terms in the actual proportion
found from the closest perfect proportion as calculated are
squared and divided by the terms of the perfect proportion. Their
sum gives a value %2> which by use of convenient tables calculated
by Elderton, gives a probability value varying from o to I
proportional to the goodness of fit.

The calculations must be based on the actual numbers of seeds
obtained and not on the percentages. To obtain a perfect pro-
portion from which the deviations of the numbers found will be
the smallest in the four terms it is necessary to balance the
figures so that the same number of individuals are represented on
the A and B plants. This is done by reducing the number of
seeds of the greater and increasing the lesser keeping the ratios
the same, of course. The deviations of the proportions, balanced
in this way, from the closest perfect proportions are then used to
obtain the probability value in the way described above. The
same result is obtained more quickly by calculating the value of
X2 from the percentages and multiplying this figure by one half
the total number of seeds.

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 269

With four terms the values of x2 greater than 18 have no
probability out to three decimal places. Since very much
greater numbers were secured in most cases, as shown in Table
II., the deviations are clearly not due to the differences of random
sampling alone if the application of this method of calculation
is justifiable. The writer is not perfectly sure that it is because
it should be noted that the theoretical proportion is calculated
directly from the results found, that is, there is no possible way
of knowing the real amounts of the two kinds of functional pollen
contained in the mixtures. Moreover the probable error does
not take into consideration the corrections which are made for
the mistakes of classification found in a sample drawn from each
lot.

There may be a selective action when the pollen is applied to
one plant but not to the other or the action may be reversed.
All that is measured is the combined effect if both are in the
same direction or the excess of one over the other if in opposite
directions. It seems reasonable to suppose, however, that the
selective fertilization is approximately the same on both members
of the pair as a large number of mixed pollinations are avai able
made with many different types of plants and the majority
give the same result. If this were always true, however, the
ratio obtained by out-crossing the mixture onto a distinct strain
should not deviate from the ratio of the closest perfect proportion
calculated from the figures of the reciprocal crosses beyond allow-
able limits. That is, the ratio obtained from the out-crossed
seeds is supposed to represent very nearly the actual ratio of
effective pollen in the mixture since both kinds are more nearly
on the same footing. In the three experiments, Nos. I, 19 and
20 in which such out-crossed seeds were obtained, the deviation
is even greater in two cases than that from the reciprocal applica-
tions. In pollen mixture number I all the deviations are small
and probably without significance.

Since the results are convincing when considered without a
probable error, it is not necessary to lay much stress on the
method of its calculation at this time. Considering the data
altogether, magnifying the actual experimental error to its fullest
extent, and taking a common-sense view of the allowances to be

270 D. F. JONES.

made for variations inherent in a problem of this kind the con-
clusion can be no other than that these plants manifest a definite
receptiveness to their own pollen, discriminating against foreign
pollen even though it comes from plants only slightly differen-
tiated from them, both of which might easily be descended from
the same individual at no very distant period back. This selec-
tive action is shown by plants of weak growth or full vigor,
whether each strain descended from a line of similar ancestors
or whether its immediate parents were diverse and, finally,
irrespective of the gametes being alike or unlike in germinal
contents. The one significant feature in common in all these
experiments is the fact that the cytoplasm which surrounds the
male nuclei and which makes up the vehicle that carries them to
the egg cells is alike for the gametes of any one type of plant
whether this plant is homozygous or heterozygous and in self-
fertilization this cytoplasm is the same as the medium in which
the pollen fulfills its function. This points very strongly to the
probability that the differential effect is due to the rate of
pollen-tube growth although it may be determined after the
male nuclei are brought to the egg.

The average weight of the seeds in the different classes of all the
pollen mixtures is given in Table IV. with the increase in weight
of the cross-fertilized over the self-fertilized seeds. Expressed
as per cent, these figures permit an estimation of the comparative
amount of heterosis shown in the crosses. In fact this is one
of the best means of measuring the stimulation of heterozygosis
as the environmental differences are reduced to a minimum.
It has not been definitely proven that there exists a correlation
between the amount of heterosis in the seeds and that shown
by the resulting plants grown from those seeds but the indications
are that there is a close relation between the two. Since hybrid
vigor is roughly proportional to the germinal differences in the
two forms united it can be determined whether or not there is a
relation between the diversity of the plants used in the several
pollen mixtures and the degree of preference shown by those
plants to their own kind of pollen. Table V. shows that the
coefficient of correlation between the average increase in weight
of seeds and the deviation in favor of self-fertilization, both stated

SELECTIVE FERTILIZATION IN POLLEN MIXTURES.

271

TABLE IV.

THE AMOUNT OF HETEROSIS SHOWN BY EACH CROSS-POLLINATION IN THE INCREASE
IN AVERAGE WEIGHT OF SEEDS COMPARED TO THE SELF-POLLINATED SEEDS
GROWN IN THE SAME INFLORESCENCES.

Pollen
Mixture
Number.

Average Weight of Seeds in Centigrams.

Per Cent. Increase.

AX A.

AXB.

Increase.

BxA.

BXB.

Increase.

A.

B.

Ave.

I
2

13-8
12.2

18.1

17.1

4-3
4-9

18.0
17.6

15-7
14-3

2-3
3-3

30-4
40.0

14.6
23-0

22.5

31-5

3

4
5
6

29-3
26.9
30.0

35-5
31-4
47-61

6.2

4-5
17.6

39-8
34-2
38-3
36.7

35.7
29-3

34-8

34-7

4.1
4.9

3-5
2.O

21.2
I6.7
58-7

ii. 5
16.7

IO.I

5-8

16.4
I6.7
34-4
5-8

7
8

25-8

30.6

4.8

19.6
23-8

16.3
22.5

3-3
1-3

18.6

2O.2

5-8

20. 2
12.2

9

12. 0

15-8

3-8

15-7

13.2

2-5

31.7 18.9

25-3

10

ii

32.2
27-5

33-4
28.6

1.2
I.I

32.5
34-5

30-9
32.8

1.6

1.7

3-7 5-2
4.0 5.2

4-5
4.6

12

32-4

32.8

•4

37-4

34-3

3.1

1.2 9.0

5-i

13

14

25-7
20.2

29.3

22.1

3-6
1.9

16.4
13-9

14.8
12.9

1.6

I.O

14.0

9.4

10.8

7.8

12.4
8.6

IS
16

24.2
32.0

35-1
40.0'

10.9
8.0

12.9
11.4

ii. 7
9.6

1.2

1.8

45.0
25.0

10.3
18.8

27.7
21.9

17
18

14-3
25.0

20.0

23.81

5-7

— 1.2

15.0

10. 01

10.2
8.2

4.8
1.8

39-9 47-1

— 4.8 22.0

43-5
8.6

19
20

15- 71
14-3

22.2
I7.I

6-5

2.8

I6.61
15-9

15-6
15-3

I.O

.6

41-4 6.4
19.6 3.9

23-9
ii. 8

in per cent., is +.496 ±.093. Although the numbers are scanty
there is a significant relation between the two. This means that
the more unlike the plants are the greater the distinction that is
made between the two kinds of pollen. In proportion as the
cross-fertilization benefits the immediate progeny in its development
the less effective is that pollen in accomplishing the union.

The same method of experimentation was applied to another
plant, the garden tomato, Lycopersicum esculentum Mill. Ad-
vantage was taken of plant characters such that the seedlings
could be distinguished in both reciprocal applications. Pollen
from a variety with entire leaves with a tall habit of growth was
mixed with pollen from a dwarf variety with normal, serrate
leaves. Tall stature and normal leaves are dominant so that the
cross-fertilized and self-fertilized seedlings from one variety were

1 Number of seeds too few to make averages reliable.

272

D. F. JONES.

visible because of differences in leaf formation and in the other
variety by habit of growth. Dwarf plants are characteristically
shorter and more compact in stems and leaves which gives them
a distinct appearance. The plants used were grown from un-
pedigreed seed but the tomato is usually self-fertilized and the
varieties employed were tested and found to come true to type.
Two experiments were made and the results from these are given
as pollen mixtures number 21 and 22 in the appendix. The
plants from individual fruits are grown separately. The total
numbers in the two mixtures are 340 and 272. Differentiation
was sharp in the seedling stage in the A lots with either serrate
or entire leaves. In the B lots the presence of tall and dwarf
seedlings could be easily seen but not all of them could be
separated as positively as could be desired. The plants were
therefore set in the field and classified after they had grown
about two months. They were set out too late to make a
satisfactory growth and even at the end of the season classifica-
tion was not made with certainty in every case.

TABLE V.

CORRELATION BETWEEN THE AMOUNT OF HETEROSIS, SHOWN BY THE PER CENT.
INCREASE IN WEIGHT OF CROSSED SEEDS, AND THE SELECTIVE ACTION IN
FAVOR OF THE PLANTS' OWN POLLEN, r = + .496 ± .093.

Per Cent. Increase in Weight.
5 7 9 ii 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43

c

13-5

i i

2

io.5

i

I

7-5

O

4-5

0

— 1.5

0

+ 1-5

ii i i ii

6

4-5

i i

2

7-5

ill

3

io-5

o

13-5

i i

2

i6.5

0

19-5

i

I

22.5

0

25-5

0

28.5

0

31-5

0

34-5

o

37-5

o

40-5

o

43-5

0

46-5

ii i

3

O 0

I 20

SELECTIVE FERTILIZATION IN POLLEN MIXTURES.

273

POLLEN MIXTURE No. i.

AX A.

AXB.

BXA.

BXB.

CX A.

CXB.

317

i

319

5

361

10

189

2

224

9

429

IS

270

7

369

10

421

13

389

8

348

26

445

i

332

8

330

6

260

i

1,757

27

1,590

56

1,656

39

1. 21

7.69

o

22.22

•39

6.66

21

2

o

12

6

3

1,738

46

1, 602

44

1.653

42

97.420

2.580

97-330

2.670

97.520

2.480

97-375

2.625

97-375

2.625

97-375

2.625

+•045

-.045

-•045

+.045

+•145

-.145

POLLEN MIXTURE No. 2.

AX A.

AXB.

BXA.

BXB.

236

ii

359

42

206

9

353

20

188

IS

380

42

141

3

230

15

194

8

372

39

237

ii

214*

12

232

ii

286

36

167

7

265

19

I5i

3

273

31

188

8

341

29

245

7

2,185

93

3,073

285

2.69

8.00

o

2-34

59

7

0

7

2,133

US

3,080

278

93-630

6.370

91.720

8.280

92.675

7.325

92.675

7.325

+ •955

--955

--955

+-955

The results, taken as they stand, are the same as obtained
from maize. There is a deviation favoring the plants' own
pollen of 2.06 and 6.84 per cent., in the two cases with the prob-
ability values .907 and .082 respectively. In the first instance
the difference can easily be due to random sampling, in the other
the odds are strongly against such an explanation. In view of
the fact that differences in germination of the seeds and viability
of plants most certainly tend to decrease the proportion of self-
fertilized individuals, the data have some value. They should
be corroborated by larger numbers using other characters which
can be more surely identified before they are as convincing as the
results with maize.

274

D. F. JONES.

POLLEN MIXTURE No. 3.

AX A.

A XB.

BX A.

BXB.

23

3

90

41

54

7

114

28

27

0

27

26

34

i

15

20

18

0

97

56

43

2

73

19

19

O

76

52

4

O

27

8

7

I

50

3

4i

14

66

55

33

5

22

ii

33

17

229

14

764

355

0

0

O

1-75

0

0

0

6

229

14

770

349

94.240

5.760

68.810

31.190

81-525

18.475

81.525

18.475

+ 12.715

-12.715

-12.715

+ 12.715

POLLEN MIXTURE No. 4.

AX A.

AXB.

BX A.

BXB.

189

38

301

67

173

42

208

IO2

165

ii

248

70

64

14

164

I67

41

7

194

38

77

15

254

II/

213

38

278

49

709

127

1, 860

648

0

-99

-83

1.72

0

i

15

ii

710

126

1,856

652

84.930

15.070

74.000

26.000

79.465

20.535

79-465

20.535

+ 5-465

-5-465

-5-465

+ 5.465

POLLEN MIXTURE No. 5.

Ax A.

AXB.

BXA.

BXB.

23
123
56
143
244

2

3

i

0
0

219
246
313
352

160

8

14
24
19

6

589
98.990
96.885
+2.105

6

I.OIO

3.115
-2.105

1,290
94.780
96.885
-2.105

7i

5-220

3-II5
+ 2.105

SELECTIVE FERTILIZATION IN POLLEN MIXTURES.

275

POLLEN MIXTURE No, 6.

AX A.

AXB.

BXA.

E/B.

40

71

184

129

—

—

0

2.63

—

—

o

3

40

71

187

126

36.04

. 63.96

59-74

40.26

47.89

52.11

47.89

52.11

-11.85

+ 11.85

+ 11.85

-11.85

POLLEN MIXTURE No. 7.

AX A.

A X B.

B X A.

BXB.

23

89

253

188

104

189

185

212

84

117

1 66

174

H5

139

118

156

123

190

116

169

109

158

134

148

23

89

1.507

1,840

20.540

79.460

45.030

54.970

32.785

67.215

32.785

67.215

-12.245

+ 12.245

+ 12.245

-12.245

POLLEN MIXTURE No. 8.

AX A.

AXB.

BXA.

BXB.

96

53

173

108

158

112

98

69

154

60

74

83

136

47

102

102

1 06

38

175

131

103

29

199

147

135

65

109

79

89

53

196

89

130

47

142

153

60

ii

203

131

128

49

117

93

63

38

52

52

69

29

1 66

idf)

87

34

109

7')

186

67

83

72

59

3i

138

86

103

54

133

138

77

34

76

75

124

5i

45

4i

58

18

119

116

64

36

75

108

2,185

956

2,584

2,058

0

o

o

1.69

o

o

0

35

2,185

956

2,619

2,023

69.56

30.44

56.42

43.58

62.99

37-01

62.99

37-01

+6-57

-6-57

-6-57

+6.57

276

D. F. JONES.

POLLEN MIXTURE No. 9.

A X A.

AX B.

BXA.

B XB.

277

149

209

US

116

69

173

99

168

69

264

156

135

78

188

114

151

52

160

IOI

78

30

2IO

142

128

71

172

134

254

IO2

191

160

286

147

166

112

208

97

195

134

185

88

196

107

H3

55

175

108

208

117

207

105

277

130

128

97

92

73

123

76

123

86

2,584

1-254

2,972

1,919

2-54

2-54

1.67

0

66

32

50

o

2,550

1,288

2,922

1,969

66.44

33-56

59-74

40.26

63.09

36.91

63.09

36.91

+3-35

-3-35

-3-35

+3-35

POLLEN MIXTURE No. 10.

AX A.

AX B.

BXA.

BXB.

344

331

189

397

200

272

195

447

284

250

189

522

275

282

233

399

169

430

1,103

I.I35

975

2,195

1.71

o

1.69

1.71

19

0

16

38

1,084

1,154

997

2,173

48.440

51-560

31.450

68.550

39-945

60.055

39-945

60.055

+8.495

-8.495

-8-495

+8.495

POLLEN MIXTURE No. n.

AXA.

AXB.

BXA.

BXB.

67

57

209

130

385

203

290

199

279

265

386

191

341

312

452

260

1,505

1,097

.88

0

0

o

4

0

0

o

448

264

1.505

1,097

62.92

37.08

57.84

42.16

60.38

39.62

60.38

39.62

+2-54

-2.54

-2-54

+2.54

SELECTIVE FERTILIZATION IN POLLEN MIXTURES.

277

POLLEN MIXTURE No. 12.

AX A.

AXB.

BXA.

BXB.

552

196

218

1 06

86

67

197

92

265

in

163

69

253

114

105

46

182

61

66

56

401

155

L739

704

749

369

.86

0

0

0

15

0

o

o

1,724

719

749

369

70.57

29-43

66.99

33-01

68.78

31.22

68.78

31.22

+ 1-79

-1.79

-1.79

+ 1.79

POLLEN MIXTURE No. 13.

A x A.

AxB.

BXA.

BxB.

37

52

14

II

17

87

13

18

II

16

17

II

28

86

6

4

50

9i

3i

8

42

59

14

13

185

391

95

65

32.120

67.880

59-370

40.630

45-745

54-255

45-745

54-255

-13-625

+ 13-625

+ 13-625

-13-625

POLLEN MIXTURE No. 14.

AXA.

A x B.

Bx A.

BxB.

41
91
83
70

139

13
35
39

27
36

58
35
46

17

31
15
IO

6

424
73-S70

72.715
+ I-I55

ISO
26.130
27.285
-I-I55

156

71.560

72.715
-I- 155

62
28.440
27-285
+ I-I55

POLLEN MIXTURE No. 15.

Ax A.

AxB.

BxA.

BxB.

309

II

3

Ill

415

II

O

53

393

9

I

7i

326

7

I

H3

276

2

O

67

135

I

2

90

400

4

355

2

2,609

47

(7)

(505)

14

498

98.23

1.77

2-73

97.27

50.48

49-52

50.48

49-52

+47-75

-47-75

-47-75

+47-75

278

D. F. JONES.

POLLEN MIXTURE No. 16.

Ax A.

AxB.

BxA.

BxB.

285

5

12

355

187

i

8

350

251

2

8

408

4

226

5

117

723

8

(37)

(1.456)

74

1,419

98.910

1.090

4.960

95.040

51-935

48.065

Si-9.15

48.065

+46.975

-46.975

-46.975

+46.975

POLLEN MIXTURE No. 17.

A x A.

A x B.

Bx A.

BxB.

255

i

4

49

1 66

o

0

5i

185

i

222

0

22O

I

255

o

1.303

3

4

IOO

99-77

-23

3-85

96.15

51.81

48.19

51.81

48.19

+47-96

-47.96

-47-96

+47.96

POLLEN MIXTURE No. 18.

Ax A.

A x B.

BxA.

BxB.

3

5

I

368

i

16

O

432

0

320

0

336

O

39

o

78

4

21

I

1-573

16.00

84.00

0.06

99-94

8.03

91.97

8.03

91.97

+ 7-97

-7-97

-7-97

+ 7-97

POLLEN MIXTURE No. 19.

A XA.

AxB.

BxA. BxB.

C x A

C x B.

2O

16

41

206

75

213

89

53

36

147

7i

124

68

86

42

200

61

315

55

47

45

178

69

39

29

150

45

20

63

156

52

37

30

88

65

8

57

218

65

86

528

392

343

1.343

207

652

57-390

42.610

20.340

79.660

24.100

75.900

38.865

6I.I35

38.865

6i.i35

38.865

6LI35

+ 18.525

-18.525

-18.525

+ 18.525

-14-765

+ 14-765

SELECTIVE FERTILIZATION IN POLLEN MIXTURES.

279

POLLEN MIXTURE No. 20.

A x A.

Ax B.

Bx A.

Bx B.

C x A. C x B.

172

II

193

86

40

52

43

6

104

58

46

56

165

IS

II?

86

Si

55

38

2

80

45

2

5

89

4

47

17

103

8

128

52

82

12

81

47

19

O

94

60

53

4

92

30

33

o

99

43

69

13

75

46

3i

2

64

35

89?

77

I-I74

605

139

168

92.09

7.91

65.99

34.01

45.28

54-72

79.04

20.96

79.04

20.96

79-04

20.96

+ I3-OS

-13-05

-13-05

+ I3-05

-33-76

+33.76

POLLEN MIXTURE No. 21.

A x A.

Ax B.

Bx A.

B x B.

39

16

17

25

12

14

I

5

32

25

36

18

14

18

97

73

54

48

57.06

42.94

52.94

47.06

55-00

45.00

55-00

45.00

+2.06

— 2.06

— 2.06

+2.06

POLLEN MIXTURE No. 22.

A x A.

A x B.

Bx A.

Bx B.

7

IO

107

22

27
103

13

7

17
27

124

44-93
38.09

+6.84

152
55-07
6I.9I
-6.84

20

31-25
38.09
-6.84

44
68.75
61.91
+6.84

PREVIOUS INVESTIGATIONS ON SELECTIVE FERTILIZATION.

From the work of Kolreuter, Herbert, Gartner, Darwin,
M tiller, Knuth and others we are familiar with the phenomenon
of self-sterility in plants in which the individual's own pollen is
wholly incapable of functioning on the plant by which it is pro-
duced although perfectly developed and able to fulfill its duties
when brought to other plants of the same species. Numerous
investigators have been giving attention to this problem in recent
times. East and Park (1917) have made a noteworthy contribu-

28O D. F. JONES.

tion to its solution and give a complete resume of the work
which has been done on this subject. They have been able to
demonstrate that groups exist within which the individuals are
all both self-sterile and cross-sterile, but any member of one
group is perfectly fertile with any member of any other group.
These investigators find that there are about 100 well-endorsed
instances of self-sterility in plants scattered over some 35 families.
Undoubtedly this discrimination is a means to promote cross-
fertilization of approximately the same significance as floral
contrivances, dichogamy and dioecism. Even though widespread
in its occurrence self-sterility is a special adaptive process ful-
filling a particular function. It is directly opposite in its effect
to the results found in maize which shows no self-sterility of the
type found in Nicotiana and other genera. At least no clear cases
are known of maize pollen, which is unable to fertilize the plants
which produced it, being able to fertilize other plants.

Darwin has furnished many instances of self-sterility. In
addition he reports some experiments which led him to believe
that even when a plant was normally self-fertile that pollen
from unrelated plants of the same species was prepotent over
the plant's own pollen. In discussing means which insure flowers
being fertilized with pollen from distinct plants, he says: "We
now come to a far more general and therefore more important
means by which the mutual fertilization of distinct plants is
effected, namely, the fertilizing power of pollen from another
variety or individual being greater than that of a plant's own
pollen. The simplest and best knowTn case of prepotent action
in pollen, though it does not bear directly on our present subject(
is that of a plant's own pollen over that from a distinct species.
If pollen from a distinct species be placed on the stigma of a
castrated flower, and then after the interval of several hours,
pollen from the same species be placed on the stigma, the effects
of the former are wholly obliterated, excepting in some rare cases.
If twro varieties are treated in the same manner, the result is
analogous, though of directly opposite nature; for pollen from
any other variety is often or generally prepotent over that from
the same flower" ("Cross- and Self-Fertilization, " pp. 391-392).

These statements were based on observations and experiments

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 28l

with various cultivated plants. Different types of crucifers—
kohl-rabi, borccoli, Brussels sprouts, cabbage — were grown near
each other and the seed resulting from pollination at will, when
grown, showed a large amount of intercrossing. The observation
was also made with different varieties of the radish, Raphanus
sativus. These plants are all partially self-sterile so that cross-
fertilization is expected in somewhat greater degree than would
result from random pollination. Mixing was also shown by
plants which are generally self-fertile such as tulip, hyacinth,
anemone, ranunculus, strawberry, orange, rhododendron, rub-
barb. The fact that vicinism occurs when varieties of these
plants are grown together is established by such observations
but this does not prove that one type of pollen is prepotent over
the other. Somewhat more significant results were obtained
from two other species. Mimulus luteus was found to be highly
fertile when insects were excluded. Uncastrated flowers of a
constant whitish variety were artificially pollinated by a yellowish
variety and of the 28 resulting plants all had yellowish flowers
so that the "pollen of the yellow variety completely over-
whelmed that from the mother plant." A crimson variety of
Iberis umbellate, which was self-fertile, was crossed with a pink
variety, the pollen being applied to uncastrated flowers as before
upon the stigmas of which he saw abundant pollen presumably
from the same flowers. Out of 30 plants raised 24 showed
themselves to be crossed by the altered color of their flowers.
Obviously experiments such as these are not sufficient to
establish the prepotency of foreign pollen in self-fertile plants.
A number of conclusions might be drawn from such results.
The cross-fertilized seeds may have germinated better and the
plants grown from them survived in greater numbers. The
types may not have been as constant for their flower color as
Darwin supposed or the ovules may not have been receptive
at the time the plant's own pollen was available but were when
the foreign pollen was applied. Taken as they stand the results
do indicate a prepotency of pollen from dissimilar plants and
it would be desirable to investigate this effect with these species
using mixed pollen in reciprocal applications as employed with
maize and the tomato.

282 D. F. JONES.

Darwin knew of many cases of total self-sterility and was so
convinced of the necessity for cross-fertilization that he was easily
persuaded from these observations that a prepotency of pollen
from unrelated plants did exist since he supposed this enabled
a plant to choose between its own and unrelated pollen when
both were brought at the same time to the stigmas by insects or
other agencies. So plausible have been the arguments in favor
of such an assumption that the prepotency of germ cells from
individuals of somewhat different constitution, even where com-
plete self-fertility exists, has been accepted as an established fact
and incorporated in textbooks on biology.

Similarly inconclusive experiments have been performed with
animals. Marshall (1910) artificially impregnated a pure bred
dog with a mixture of equal quantities of seminal fluids from the
same breed and from a mongrel of unknown ancestry. Of the
four young which resulted one died early, and three resembled
somewhat the mongrel sire. Marshall cites another instance
in which a dog copulated with a member of the same breed and
two days later with a sire of different type. Out of three puppies
one was pure bred and two half-breeds. These cases, according
to this writer, indicated a selective action favoring dissimilar
rather than related spermatozoa. King (1918) mentions some
preliminary experiments with albino and wild gray rats in which
advantage was given to the former, yet the results tended to
show a prepotency of the latter, so that there was apparently a
selective action favoring the out-cross. The details of these
experiments are not given.

In attempting to determine whether or not a selective action
exists small numbers can never be more than suggestive and
unless the mixture is applied at the same time to both types
furnishing the sperm cells there is no way of estimating the rela-
tive proportions of the two kinds of fertilizing elements present
in the mixture which are capable of functioning. Furthermore
a constant excess of cross-fertilized individuals over the others
may be due to the greater viability of the hybrids and hence
there will be a selective elimination of zygotes but not necessarily
slective fertilization. Hyde (1914) compared the matings of
different types of Drosophila within the strains and in reciprocal

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 283

crosses. The dissimilar unions gave greatly increased numbers
in both reciprocal combinations. However, the type of mating
did not influence the number of eggs laid and there is no proof
that cross-fertilization occurred more readily than self-fertiliza-
tion. The results show that the cross-fertilized eggs hatched
better and the offspring survived in greater numbers, a result
which is easily understood since there were lethal factors involved
in the material worked with.

The only evidence from the animal side of a definite selective
action comparable to the many instances of self-sterility in
plants is the well-known case of self-impotency in dona intes-
tinalis (Castle, 1896). Morgan (1905, '07, '10) has experimented
with this organism and has found that the self-sterility is not
always complete. Material gathered on the Pacific coast showed
somewhat greater receptiveness to the individual's own sperm
than eggs of the same species at Woods Hole which were almost
entirely unresponsive to sperm from the same individual which
produced the eggs. In another ascidian, Cynthia partida, he
found that self-fertilization takes place frequently but the sperm
of unrelated individuals is more effective. A third species,
Molgula manhattensis is self-fertilized as readily as cross-fertilized.
From this it seems that incompatibility of uniting gametes as a
means of insuring cross-fertilization exists in various grades of
effectiveness. Even in extreme cases the degree of self-sterility
may be modified by internal and external conditions. In Nico-
tiana East and Park find that self-fertilization sometimes takes
place towards the end of the growing period when the vigor of the
plants is reduced.

DISCUSSION.

As far as the writer knows the results obtained from maize
stand alone among plants in showing a selective action unfavor-
able to fertilization by sperm from individuals of different
hereditary constitution. The handicap placed upon the foreign
pollen is proportional to the germinal unlikeness. If the unequal
effect is due to a slower growth of the pollen tube through the
tissues of style the selective action may be restricted to plants,1

1 E. C. Miller (Jour. Agric. Research, Vol. 18. pp. 255-266, Dec. 1919), has
recently made a detailed study of fertilization in maize and finds that from many

284 D. F. JONES.

and would also not be surprising to find that the phenomenon is
greater in maize than in any other species for the reason that in
this plant the pollen tubes have a larger distance to traverse to
accomplish fertilization than in any other form known to the
writer.

The stigmatic hairs of maize are scattered along a filamentous
style which continues to grow until fertilization takes place.
The structure withers and dries shortly after poll'nation takes
place. Pollen may adhere at any point along the filament.
The total length of style through which the pollen tube grows is
normally from 10 to 20 centimeters but in extreme cases may
be as great as 50 or more. It will be worth while to see whether
or not the selective effect is more pronounced when the styles
are long than when short. Such an experiment can be easily
carried out and would give some indications as to whether the
handicap is placed during pollen tube growth or after the sperm
nuclei are brought to the egg.

The lessened ability of moderately different types to fertilize is
in line with the impossibility of effecting unions between widely
separated forms. In such cases the prevention of fertilization
is sometimes due to mechanical difficulties in the way of bringing
sperm cells to the eggs but even when this is accomplished there
still exists a firm barrier which prevents the passage through the
egg membrane. The differential effect demonstrated in maize
may be simply a reduced manifestation of this phenomenon.

It is possible that the experiments on anaphyllaxis may throw
some light on this problem. It is known that foreign proteins
when injected in animal tissues may have a toxic effect and
excite an extreme irritability so that in repeated doses they may
cause markedly injurious results. By this means it is possible
to distinguish between proteins of very slight differences in
composition. Since the differences in protoplasmic substances
between the types in which a selective action is shown seem to
be small there may be some relation between the two phenomena.
However, the male gametophyte growing upon the stigmas is, in

pollen tubes which start to grow down the style only one tube in every case in
nearly 100 observations was seen to reach the ovary cavity. This indicates that
the differential fertilizing power is determined by the rate of pollen tube growth
and not after the sperm nuclei have been brought to the egg.

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 285

a way, merely a parasitic organism. It would be difficult to
find evidence from true parasites that they are restrained by their
hosts in proportion as they are genetically dissimilar although
in most cases the differences are so great that probably there is
no basis for comparison.

The only evidence which has any direct bearing on the problem
comes from grafting experiments. It is well known with plants
that the affinity of stock and cion is directly proportional to their
phylogenetic similarity. In animals the same rule holds, and
very fine distinctions are manifested. Morgan (1910) cites the
results of Schoene in which the skin of the mouse is readily
grafted back upon the same individual or member of the same
litter but not upon unrelated mice. Such results as these are
quite similar to the greater receptiveness of plants to their own
pollen.

For some time there has been current in biological literature
the hypothesis that heterogeneity in the structure of organisms
favors growth and reproduction and conversely that homogeneity ,
is unfavorable. This is a heritage from Darwinism and has
appeared again and again in theories of rejuvenation, vigor
derived from crossing, and selective fertilization. It has been
stated in many different ways but in general terms it amounts to
the supposition that similarity in protoplasmic structure brought
about by a line of similar ancestors is not conducive to physio-
logical efficiency and that the differences brought about by the
union of unlike elements and the consequent lack of balance
stimulates growth. The only basis for this hypothesis is found
in the necessity, in most cases, for the union of two differentiated
sex cells to start the development in the egg. The attempt to
argue by analogy than the union of dissimilar protoplasms is,
in itself, an immediate physiological benefit is not supported by
the facts and is founded upon fallacious reasoning.

Rejuvenation in vegetatively reproduced organisms by sexual
union is no longer looked upon as beneficial in destroying simi-
larity in structure. That the process of forming gametes and
their reunion may bring about a reorganization of the proto-
plasmic substances and an elimination of waste products so as
to result in greater growth seems quite plausible. But the

286 D. F. JONES.

significance of such a process is not to be looked for in the bringing
together of differentiated germplasms.

The advantage derived from cross-fertilization is now under-
stood as a phenomenon of inheritance and the older hypothesis
of the stimulation of heterozygosis is no longer needed. Accord-
ing to present theory homozygous factor combinations are more
efficient than heterozygous combinations of the same factors.
In the lowest organisms which are illustrative of a primitive
sexuality there is direct proof that the union of different indi-
viduals does not result in an increased developmental efficiency.
Jennings (1913) finds that in Paramecium, in the generations
immediately following conjugation, there is a slowing down in the
rate of division. The advantage derived from the pairing of
individuals is the greater elasticity in adaptiveness resulting
from the mixing of different germ plasms giving to some of the
descendants a greater chance for survival. Pearl (1907) has
found that in this same organism there is a tendency for like
forms to conjugate due to mechanical hindrances to the pairing
of individuals dissimilar in form. This has been substantiated
by Jennings (1911). Also in gastropods Crozier (1918) has
demonstrated that assortative mating takes place between indi-
viduals of the same size and this, he considers, results in a greater
number of offspring than there would be if random pairing was
the rule. It has been proven by Pearson and Lee (1903) that
assortative mating occurs in man. This conclusion is reached
after extensive investigation in the inheritance of physical char-
acters in which they have found that there are positive correla-
tions between husband and wife with respect to stature, span of
arms, and length of forearm. Moreover they have shown that
homogamy is a factor favorable to fertility. Parents with like
characters are more productive of offspring. This is an impor-
tant observation and supports the main thesis of this paper.

The occurrence of homogamy in such widely diverse forms of
life as the higher plants, protozoa, mollusca and man cannot fail
to have significance. The importance of discriminate isolation
has long been recognized since it supplies a necessary factor in
divergent evolution. The existence of such an assortative action
can now be looked for in all forms of life since evidence has been

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 287

produced from the corners of the phylogenetical triangle. The
evolutionary significance of this phenomenon has been ably
reviewed by Peal in the above citation so that it is not useful to
go into that phase of the problem in detail here. It is interesting
to note that the results from maize fulfill his expectation as in
conclusion he says, "the fact that we find such a high degree of
homogamy in a protozoan form like Paramecium strongly sug-
gests the possibility that in higher organisms there may be
assortative mating of the gametes in the process of fertilization.
Should such a homogamy occur it would probably be of far
greater importance than any assorted mating of the somas."
While the selective agency in plants does not differentiate be-
tween gametes produced by one type in the end the result is the
same. Individuals with like characters tend to be brought to-
gether and virtually to be set apart from the general population.

Biological investigations unite to show that the importance of
sex is to make organisms more plastic in adaptability. The
advantages have been so great that sexual reproduction is now
established as the dominant method for the renewal of organism
in both kingdoms at the expense of economy and speed of multi-
plication. It is not strange then that accessory devices have
been developed to insure the fulfillment of the function for which
so much has been expended. Self-sterility or self-impotency is
one of the means developed to -serve this purpose.

The reverse phenomenon, that of self-prepotency, so far is
shown among plants only by maize. It would be surprising to
find it limited to this one species. Is it not more likely to be a
general manifestation? As a fundamental principle it may apply
even to those organisms which show self-sterility, this latter
being a special adaptation entirely overcoming the handicap
placed upon unfamiliar gametes in order to make certain the
advantages which exogamy holds out. One cannot insist upon
such an assumption with evidence from only one or possibly
two species. But the evidence, limited though it may be for
the present, is one more indication that homogeneity, similarity,
likeness, familiarity, or however it may be described, in proto-
plasmic structure is consistent with and favorable to the highest
physiological efficiency.

288 D. F. JONES.

In maintaining two opposing tendencies Nature is not neces-
sarily working at cross-purposes. Biparental inheritance with
the inclination towards exogamy serves to bring about plasticity.
The preferential mating of similar kind operates to make this
mixing discriminative. Probably it is not yet time to reconcile
completely these two contradictory forces. It may at least be
held that assortative mating which favors the pairing of like
with like has some importance in evolution since it is an agency
in orthogenetic changes. Perceptive reproduction of this kind
tends to hold organisms in certain paths once a break from the
common type has been made. Having been demonstrated in
three of the farthest separated branches of the organic world-
angiosperms, protozoa, and man — homogamy may take a some-
what more authoritative part in evolutionary theory.

LITERATURE CITED.
Belling, J.

'14 The Mode of Inheritance of Semi-sterility in the Offspring of Certain

Hybrid Plants. Zeitsch. f. ind. Abst. u. Vererb., Vol. 12, pp. 303-342.
Castle, W. E.

'96 The Early Embryology of dona inteslinalis Flemming (L.). Bull. Mus.

Comp. Zool., Harvard University, Vol. 27, pp. 201-280.
'03 The Heredity of Sex. Bull. Mus. Comp. Zool., Harvard University, Vol.

40, pp. 189-218.
Crozier, W. J.

'18 Assertive mating in a Nudibranch, Chromodoris Zebra Heilprin. Journ.

Exper. Zool., Vol. 27, pp. 247-292.
Cuenot, L.

'08 Sur quelques anomalies apparentes des proportions Mendeliennes (6e note).

Arch. Zool. Exp. et Gen., Vol. 9, pp. 7-15-
Davis, B. M.

'15 The Test of a Pure Species of CEnothera. Proc. Am. Phil. Soc., Vol. 54,

pp. 226-245.

'17 A Criticism of the Evidence for the Mutation Theory of DeVries from the
Behavior of Species of CEnothera in Crosses and in Selfed Lines. Proc.
Nat. Acad. Sciences, Vol. 3, pp. 704-710.
East, E. M.

'19 Studies on Self-Sterility. IV. Selective Fertilization. Genetics, Vol. 4,

pp. 346-355-
East, E. M., and Hayes, H. K.

'u Inheritance in Maize. Connecticut Agr. Exp. Sta. Bull. 167, pp. 1-142.
East, E. M., and Park, J. B.

'17 Studies on Self-sterility. I. The Behavior of Self-sterile Plants. Genetics,
Vol. 2, pp. 505-609.

SELECTIVE FERTILIZATION IN POLLEN MIXTURES. 289

Elderton, W. P.

'01 Tables for Testing the Goodness of Fit of Theory to Observation. Bio-

metrika, Vol. i, pp. 155-163.
Hyde, R. H.

'14 Fertility and Sterility in Drosophila ampelophila. Journ. Exper. Zool.,

Vol. 17, Nos. 1-4.
Jennings, H. S.

'n Assertive Mating, Variability and Inheritance of Size, in the Conjugation

of Paramecium. Journ. Exper. Zool., Vol. n, pp. 1-134.
'13 The Effect of Conjugation in Paramecium. Journ. Exper. Zool., Vol. I4>

PP- 279-391.
Jones, D. F.

'18 The Effects of Inbreeding and Crossbreeding upon Development. Con-
necticut Agr. Exp. Sta. Bull. 207, pp. i-ioo, pis. 1-12.
Kempton, J. H.

'19 Inheritance of Waxy Endosperm in Maize. U. S. Dept. of Agr., Bur. P. I.

Bull. 754, PP- i-99-
King, H. D.

'18 Studies in Inbreeding. III. The Effects of Inbreeding, with Selection,
on the Sex Ratio of the Albino Rat. Journ. Exper. Zool., Vol. 27, pp. 1-35.
Marshall, F. H. A.

'10 The Physiology of Reproduction. London.
Morgan, T. H.

'05 Some Further Experiments in Self-fertilization in Ciona. BIOL. BULL.,

Vol. 8, pp. 313-330.

'07 Experimental Zoology. New York.
'10 Cross and Self-fertilization in Ciona intestinalis. Arch. f. Entwicklungs-

mech. d. Org., Vol. 30, pp. 206-234.
Muller, H. J.

'18 Genetic Variability, Twin Hybrids, and Constant Hybrids, in a Case of

Balanced Lethal Factors. Genetics, Vol. 3, pp. 422-499.
Pearl, R.
'07 A Biometrical Study of Conjugation in Paramecium. Biometrika, Vol. 5,

pp. 213-297.

Pearson, K., and Lee, A.
'03 On the Laws of Inheritance in Man. I. The Inheritance of Physical

Characters. Biometrika, Vol. 2, pp. 357-462.
'03 Assortative Mating in Man. Biometrika, Vol. 2, pp. 481-498.

STUDIES ON THE CELLS OF CATTLE WITH SPECIAL

REFERENCE TO SPERMATOGENESIS, OOGONIA,

AND SEX-DETERMINATION.

J. E. WODSEDALEK,
ZOOLOGY DEPARTMENT, UNIVERSITY OF IDAHO. l

CONTENTS.

I. Introduction 290

II. Material and Methods 293

III. General Arrangement of the Male Germinal Cells 294

IV. Spermatogenesis 294

1. Spermatogonia 295

2. Primary Spermatocytes 296

3. Secondary Spermatocytes 297

4. Spermatids 297

V. Dimorphism in the Spermatozoa 297

VI. Oogonia 298

VII. Chromosomes in Somatic Cells 300

VIII. Sex- Chromosomes in Relation to Sex-Determination 302

IX. Sex Ratio in Cattle 303

X. Sex-Limited Inheritance in Cattle 304

XI. The Free-Martin 305

XII. Summary 305

I. INTRODUCTION.

Sex-determination with its many attendant problems has
always been a subject of great interest to practical animal
breeders; and the art of breeding has always been replete with
rules by which the sex ratio might be shifted in various ways
to the advantage of the breeder. These rules, however, have
been founded upon inadequate evidence and unsound reasoning.

The most common beliefs in regard to sex control have from
time to time been founded on heat relations, some maintaining
that the products of conception in early heat were more often
males, others that they were more often females. Pearl and
Parshley ('13) have published data on sex-determination in
cattle with the following conclusions: (i) That as the time of
coitus approaches the end of the cestrous period there is a pro-

1 This research was supported in part by the Adams Fund of the Idaho Experi-
ment Station.

290

STUDIES ON THE CELLS OF CATTLE. 2QI

gressive increase in the proportion of male young born. (2)
That in the extreme case this increase in the proportion of male
births is probably statistically significant and not to be attributed
to errors of random sampling. (3) That these modifications of
the sex ratio cannot be attributed to age differences or to any
other factor yet suggested.

More recent and extensive data, however, according to Pearl
('17) make the relation of time of service to sex extremely
doubtful. He says: "The apparent relation between these two
factors, which is believed by many breeders to exist and which
our earlier statistics appeared to indicate, seems now to be
purely accidental, and to have arisen only because of the com-
parative meagerness of the statistics on which the matter was
discussed." Sex in cattle is a matter of heredity, as is shown
by the results of this investigation, and remains a matter beyond
the control of the breeder; its ultimate control is problematical.
However, the suggestion of experimentally separating the two
types of spermatozoa, or destroying the one type without impair-
ing the nature of the other, does not appear to be entirely hope-
less. If this can be done, the problem of sex control will be
comparatively simple, since artificial insemination can be resorted
to.

It appears to be a well-established fact that sex is determined
at the time of fertilization. Sex, like other characters of the
individual, has a definite factorial basis; and the factorial con-
stitution of the individual with respect to sex as well as to other
characters is fixed by the constitution of the two gametes which
unite to form the zygote.

The dimorphic condition among the spermatozoa in many
of the lower animal forms is well known. Guyer pointed it out
in the rat ('10), and man ('10) ; other investigators have reported
it in several other mammals. Among the domestic mammals the
same condition has been clearly demonstrated in the pig (Wod-
sedalek, '13) and in the horse (Wodsedalek, '14). This study
on the sex cells of cattle shows that two types of spermatozoa are
also produced in the bull ; the one type at the time of fertilization
determining maleness and the other type determining femaleness.
The spermatozoan which determines femaleness is somewhat

2Q2 J. E. WODSEDALEK.

larger owing to the presence of a large sex-chromosome or acces-
sory element which is lacking in the type producing maleness.
Since it has been found in this study that the oogonia of cattle
possess two sex-chromosomes it is safe to conclude, in view of
our knowledge of oogenesis in general, that all of the mature ova
carry the reduced number or one sex-chromosome. If the ovum
is fertilized by a sperm which lacks the sex-chromosome the
resulting zygote naturally possess only one sex-chromosome and,
therefore,. develops into a male. This is evidenced by the fact
that the somatic as well as the germ cells of the male possess
a single accessory element. On the other hand, if the ovum is
fertilized by a sperm possessing the sex-chromosome the resulting
zygote possesses two sex-chromosomes and, therefore, develops
into a female. The somatic as well as the germ cells of the female
possess the two accessory elements which is in exact accord with
expectations.

The results presented in this paper were obtained through a
most careful and critical study extending over a period of more
than four years. The problem was started in the spring of 1915
and every summer since and a great deal of the spare time in the
intervening school years, with the exception of the year 1918, has
been devoted to it. In the aggregate this means a total of about
fifteen months of continuous work. Practically all of the results
presented in this paper were at hand at the end of the first two
and a half years of study. Over four hundred pencil sketches
were made of the various cells, especially those in mitosis, with
the aid of the camera lucida and the location of each, as indicated
by the mechanical stage, was carefully recorded in a separate
booklet. A brief summary of the results of the study was
written up and all of the material was put away for a whole year
(1918) during which the summer months and other spare time
was devoted to the study of the sex cells in another mammal.
No effort was made to remember any of the results of the study
on the sex cells in cattle.

In the meantime, considerable new cattle material was obtained
and hundreds of new slides were made by the departmental
technician. At the end of the year the new cattle material was
studied and the results were carefully recorded. The original

STUDIES ON THE CELLS OF CATTLE. 2Q3

material was then reinvestigated and new sketches were made
of many of the cells, the location of which was recorded before.
The results obtained through the study of new slides prepared
by the technician, as well as those obtained through the second
study of my original material were then compared with the
original results. Several hundred different germinal and somatic
cells from both sexes were carefully checked up. Especial atten-
tion was paid to the number of the ordinary chromosomes and
to the sex-chromosomes. Only in a comparatively few cases was
there any discrepancy in the interpretation of the nature and
number of the ordinary chromosomes. Conditions pertaining
to the sex-chromosomes were corroborated in every instance.
I was further checked up on my chromosome counts by several
of my assistants and senior pre-medical students who had con-
siderable training in microscopic anatomy. In general, these
men who knew nothing of my own interpretations, corroborated
my counts in a surprisingly large number of instances.

II. MATERIAL AND METHODS.

All of the material used in this investigation with the excep-
tion of some ovaries was obtained through the courtesy of the
management of the Hagan and Gushing Packing Plant which
adjoins the University farm. Some excellent ovarian tissue was
given to me by my colleague Dr. A. R. Hahner, formerly professor
of veterinary science of the University of Idaho. The ovaries
were removed from two five-months-old heifers of the university
herd. Testes were obtained from seven mature bulls and one
male foetus of five months and from six smaller foetuses varying
from two to eight and one half inches in length. The ovaries
were obtained from four heifers and as many cows, and from
five small foetuses, varying from two and one half to seven inches
in length. In addition to this six small embryos were sectioned.
Many slides were also made of various somatic structures from
the small foetuses of both sexes.

Several fixing fluids were tried on the testicular material, in-
cluding Hermann's, Gilson's, Flemming's, and Bouin's. Bouin's
fluid used straight or with slight modifications, at 38° C. was the
most universally successful fixing agent. When used at some-

294 J- E. WODSEDALEK.

what lower temperatures, the material showed little or no modi-
fications. When Bouin's fluid was modified, the alteration took
the form of the addition of a small amount of chromic acid, or
urea, or the reduction of the percentage of acetic acid. All
of these slight modifications gave very good results in the
testicular material, and the ovarian, embryological and foetal
tissues appeared to be best when fixed in the fluid modified with
chromic acid. The cold method (Hance '17) was also tried in
two instances with the testicular tissue but with less success.
And while I have not tried this out on the cattle tissue myself,
I have every assurance that the laboratory technician carried
out the process with great care.

In the study of the male germ cells, smears as well as sections
were used. Many stains and counter stains were tried. Iron-
haematoxylin when used alone was found to be the most satis-
factory. All of the figures represented in this paper were made
from material stained in this manner.

III. GENERAL ARRANGEMENT OF THE MALE GERMINAL CELLS.

The structure of the testes of the bull is similar to that of the
other well-known mammals and bears a great resemblance to the
conditions found in the testes of the horse (Wodsedalek '14).
The interstitial cells as in the horse are small and fewer in number
in comparison with their large size and great abundance in the
testes of the pig. The size of the seminiferous tubules, as well
as the general size of the various germinal cells, however, corre-
sponds to the condition found in the pig. The usual types of
cells, (i) spermatogonia, (2) primary spermatocytes, (3) second-
ary spermatocytes, (4) spermatids, and (5) spermatozoa in
various stages of development, are present in great abundance.

IV. SPERMATOGENESIS.

In general the spermatogenesis of the bull corresponds to that
of the pig and the horse. Since many of the finer cytological
points are given in detail in the papers on the pig and the horse
(Wodsedalek '13 and '14) they are omitted here to avoid un-
necessary duplication. And while all of the finer details involved
in a thorough piece of work in spermatogenesis were carefully

STUDIES ON THE CELLS OF CATTLE. 2Q5

studied in this animal, only the phases pretaining to the chromo-
some numbers and their behavior are emphasized in this paper.

i. Spermatogonia.

The spermatogonia usually lie in a single layer next to the wall
of the tubule, though occasionally some of the cells are crowded
out, thus forming a second layer which is always very irregular.
The cells which undergo the last spermatogonial division (Figs.
5-12) are usually beyond the first layer, though occasionally
they may be found next to the tubule wall along the entire
section of the tubule. At times the cells are far apart, in which
case they are flattened out on the tubule wall. The cells also
differ considerably in size and appearance, depending on the
stage of development they are in.

During the resting stage a large nucleolus is invariably present.
As a rule it assumes a somewhat heart-shaped appearance;
especially is this true in the larger cells and in those in which the
chromosomes begin to form. At the conclusion of the resting
stage numerous large chromatin granules appear and arrange
themselves along fine threads in an entangled mass. The chro-
mosomes soon become distinct and mitotic figures are fairly
numerous. And while, as a rule, there is considerable over-
lapping and massing of the chromosomes in the early spermato-
gonial divisions, hundreds of cells were found in which there was
little or no overlapping, making accurate counts possible.

Thirty-seven chromosomes appear in the late prophase of the
spermatogonial division (Fig. i). Thirty-six of these are vari-
ously shaped, mainly oblong or slightly curved, and differ
somewhat in size. One which is much larger is triangular in
form or heart-shaped. This is the accessory or sex-chromosome,
and is the same thing as the large nucleolus which appears in the
resting stages. This point is certain, as. the body can be easily
traced through the various stages of the cells. A similar condi-
tion was reported by Guyer ('10) in man, Wodsedalek in the pig
('13) and in the horse ('14). Several other investigators have
reported it in other forms since. During division each chromo-
some, including the sex-chromosome, divides in two (Figs. 3
and 4).

296 J. E. WODSEDALEK.

In the last spermatogonial cells the chromosomes appear in a
dense mass. The cells gradually increase in size and great expan-
sion takes place in the nuclei (Figs. 5-9). The chromosomes
become decidedly distinct and surprisingly well segregated
throughout the spherical nucleus (Figs. 10 and n). The cells
in these stages are numerous and beautiful. The chromosomes
appear to be dense in structure and are thicker than those of the
early spermatogonial cells. Several of the chromosomes are
almost spherical. They are so evenly distributed that hundreds
of accurate counts can be made within a short time. This
condition prevailed in all of the mature testes studied and the
cells were especially numerous in three two-year-old bulls. Even
in somewhat stale tissue (Fig. 12) the chromosomes in these
cells appear to remain well segregated. Altogether over one
thousand accurate counts were made in these cells alone.

2. Primary Spermatocyles.

The sex-chromosome can invariably be seen in the spermato-
cytes where it retains its individuality (Fig. 13). Just how
pairing takes place in these cells cannot be stated with certainty.
When the chromosomes appear for division they are of the
reduced number and bivalent in nature. The thirty-six ordinary
chromosomes pair while the sex-chromosome remains unpaired
and can easily be distinguished from the others (Figs. 14-18).
It occasionally shows its double nature in the late prophase
(Fig. 17), and more frequently in the later stages of the primary
spermatocyte division (Figs. 22, 24, 25, 27, 28 and 29).

During the primary spermatocyte division the sex-chromosome
usually passes to one pole in advance of the other chromosomes
(Figs. 19-25). This unequal division of the chromosomes in these
cells (Figs. 19-29) gives rise to two different types of secondary
spermatocytes. The one type containing the eighteen ordinary
chromosomes plus the sex-chromosome. Just before division is
complete the chromosomes become loosely paired (Figs. 28 and
29). This peculiar behavior of the chromosomes is apparently
quite common in mammalian tissue.

STUDIES ON THE CELLS OF CATTLE.

3. Secondary Spermatocytes.

No resting stage occurs in the secondary spermatocyte. This
condition also occurs in the horse and is frequently found in the
pig, according to my former studies. The secondary spermato-
cytes divide soon after they are formed and not infrequently the
spindles are formed in the two cells resulting from the first
spermatocyte division while they are still in close contact. Nine
chromosomes arrange themselves in the equatorial plate for
division in the one type of secondary spermatocyte (Fig. 31),
and nine plus the sex-chromosome in the other (Fig. 30). All
of the chromosomes, including the sex-element when it is present,
divide in these cells (Figs. 32-38).

4. Spermatids.

The division of the secondary spermatocyte gives rise in the
one case to spermatids containing nine chromosomes (Figs. 36-
38), and in the other case nine plus the sex-chromosome or ten
(Figs. 33 and 34). All of the chromosomes except the sex-
chromosome are bivalent in nature so that in reality we have the
equivalent of eighteen chromosomes in the one kind of spermatid
and eighteen plus the sex-chromosome in the other. The bivalent
chromosomes frequently begin to separate before the division of
the cell is complete. Occasionally the eighteen chromosomes
can be distinguished as independent elements after the cell
divides (Fig. 38), although the chromosomes usually disintegrate
before complete separation can be identified. All of the foregoing
evidence indicates that eighteen is the reduced number of chro-
mosomes.

V. DIMORPHISM IN THE SPERMATOZOA.

The spermatozoa of the bull vary considerably in size, and
careful measurements show that they may be arranged in two
separate classes, one type being much larger than the other.
Mature specimens, which were free in the lumen of the tubule
and parallel to the objective, were selected at random from a single
slide and outline sketches of six hundred heads were made with
the aid of a camera lucida. The lengths of the sketches were
then carefully measured and recorded in quarter millimeters.
Figure I in the text shows the variation in size of the six hundred

298 J. E. WODSEDALEK.

heads measured. It shows a distinct bimodal curve with modes
at 12.50 mm., and 14.75 mm. The intermodal depression is

. 50 IV. 75

FIG. i. Diagram showing the variation in size among six hundred mature
spermatozoa of the bull.

deep and wide and the two elements of the curve are approxi-
mately equal as regards number of individuals. The spermato-
zoa of the larger type undoubtedly possess the sex-chromosome,
while those of the smaller type are without it. Figures 39 and
40 show the comparative size of the two distinct types of mature
spermatozoa. They can be distinguished with ease under a
high-power microscope.

The general scheme of the development of the spermatozoon
from the spermatid of the bull is similar to that in the pig and the
horse (Wodsedalek '13 and '14) and, therefore, will not be
described here. Bimodal curves were also shown by the writer
in connection with the spermatozoa of the pig and the horse, and
in a number of other species by Zeleny and Faust ('14 and '15)
and by Zeleny and Senay ('15).

VI. OOGONIA.

The best ovarian material studied in connection with the
number of chromosomes in the oogonia was obtained from some
of the small foetuses. The results obtained from this material

STUDIES ON THE CELLS OF CATTLE.

were corroborated in the tissue obtained from cows and heifers,
although the adult tissue was not nearly as satisfactory. Most
excellent material was found in the ovaries of a four and a half
inch fcetus and one six-inch fcetus. These were fixed in the
Benin's plus chromic acid fixing fluid. The oogonia were ap-
parently in extreme activity at the time of fixation. Mitotic
figures are abundant and the chromosomes are very distinct,
especially in the late prophase and early metaphase stages which
are numerous (Figs. 44-46). In general these cells resemble the
last spermatogonial cells (compare Figs. 44 and 10), except that
the oogonia are somewhat larger.

In well fixed and favorably stained material the oogonia in
the resting stage invariably show two large nucleoli each corre-
sponding to the single nucleolus of the resting stage of spermato-
gonial cells (Fig. 41). These retain their individuality during the
spireme stage though at times they are somewhat distended
(Fig. 42). When the spireme breaks up the chromosomes are
long and narrow and variously curved (Fig. 43). In the late
prophase they become shorter and thicker and appear evenly
distributed throughout the large nuclei (Figs. 44 and 45). In
this stage as well as in the early metaphase stages of division
(Fig. 46) hundreds of accurate counts were possible. The two
sex-chromosomes can be easily distinguished in all of the prophase
stages (Figs. 41-45).

Thirty-eight chromosomes are present in the oogonia (Figs.
43-46). Thirty-six are the ordinary chromosomes corresponding
to the thirty-six ordinary chromosomes in the spermatogonia of
the male. The two other elements are the sex-chromosomes.
When the chromosomes arrange themselves in the equatorial
plate for division the sex-chromosomes are always at the
periphery (Figs. 46 and 47). During division all of the chromo-
somes, including the sex-elements, divide in two (Fig. 48). All
of the figures of the oogonial cells were made from the same
section (Figs. 41-48).

On account of the great significance of the two sex-chromo-
somes in the femal tissue a tremenduous amount of time was
devoted to this particular phase of the problem; this was also
true of the studies in connection with the chromosomes of the

3OO J. E. WODSEDALEK.

somatic tissue in both sexes. The final preparation of the paper
for publication was postponed on several occasions, not at all
because of any uncertainties, since the results were convincing
from the start, but rather because each additional survey of the
entire problem from new material proved more fascinating and
gratifying than those preceding.

VII. CHROMOSOMES IN SOMATIC CELLS.

Numerous slides were made of various somatic structures from
foetuses of both sexes. The organs most frequently used were
the brain, lung, liver, Wolffian body, kidney, and intestine.
The larger embryos were cut into pieces all of which were then
sectioned. Three small embryos, ranging from ten to fourteen
millimeter neck-lengths were sectioned in toto. While splendid
mitotic figures and late prophase stages were found in many
parts of the embryos the very best cells were found in the brain.
In the larger specimens the liver and kidneys showed the most
favorable cells.

The male somatic cells, like the spermatogonia, contain thirty-
seven chromosomes, of which thirty-six are the ordinary chromo-
somes and one is the accessory element or sex-chromosome (Figs.
51 and 52). The female somatic cells, like the oogonia, contain
thirty-eight chromosomes, of which thirty-six are the ordinary
chromosomes and the other two are the sex-elements (Figs. 49
and 50). The sex-chromosomes in each case were as distinguish-
able as they are in the germinal cells. Literally thousands of
somatic cells were carefully studied in each sex.

After the chromosomes of the first one hundred of the most
favorable male somatic cells were studied and carefully sketched
it was found that in ninety-three cases there were thirty-six
of the ordinary chromosomes present plus the one accessory.
In the other seven cases there were slight discrepancies, usually
one or two less. This was in all probability due to unnoticeable
overlapping. In two cases there were two extra chromosomes
present. This was in all probability due to the fact that two
of the chromosomes had divided, since these cells were in the
early metaphase stage. This interpretation appears to be correct,
since it was found later that occasionally some of the chromo-

STUDIES ON THE CELLS OF CATTLE. 3OI

somes divide considerably in advance of the others. This point
was observed in several of the early metaphase stages; it was
fairly common in polar views of late metaphase stages, and in a
few instances a chromosome was found almost completely split
in two even in the late prophase. For this reason, in selecting
cells for accurate chromosome counts great care must be exercised
not to select polar views of late metaphase stages. The increase
in number may also be due to fragmentation, although one would
expect this only in poorly fixed material or general poor technique.
After the counts in the first one hundred camera-lucida sketches
showed that over ninety per cent, of the cells contained thirty-
seven chromosomes, no more sketches were made. However,
hundreds of other counts were made from favorable cells off-hand
with about the same results. This shows that in at least ninety
per cent, of the male somatic cells thirty-six ordinary chromo-
somes, and a single sex-chromosome were present. The ten
per cent, of discrepancies is undoubtedly due to unnatural states
and probable sources of error. The studies of the chromosomes
in the somatic cells of the female were conducted in the same
manner. Among the first one hundred cells, ninety-one showed
thirty-six ordinary chromosomes and two accessories. In several
hundred further counts, the percentage of discrepancies was
about the same.

In the case of the six small embryos in which sex could not be
determined morphologically, or was uncertain, one was poorly
fixed, although it was cut into small pieces, and, therefore, could
not be used in this study. Of the other five two were unquestion-
ably female and three were male, according to evidences from the
cytological standpoint. In the case of the two specimens, cells
in various parts of the body repeatedly showed thirty-six ordinary
chromosomes and two accessories. In many instances where the
ordinary chromosomes could not be counted, the two sex chromo-
somes were recognizable. In the other three specimens the cells
invariably showed only a single sex-chromosome, and in many
cases the thirty-six ordinary chromosomes were counted.

These extensive studies indicate quite conclusively that thirty-
six is the number of ordinary chromosomes in the somatic cells
of both sexes and that the male cells contain one sex-chromosome

302 J. E. WODSEDALEK.

while the female cells contain two, making a total of thirty-seven
in the male and thirty-eight in the female. These numbers
correspond exactly with those of the spermatogonia in the male
and the oogonia in the female. This is very significant in relation
to our chromosome theory of sex-determination.

VIII. SEX-CHROMOSOMES IN RELATION TO SEX-DETERMINATION.

It was shown that in the process of spermatogenesis two distinct
types of spermatozoa are produced of exactly the same number.
The one type contains eighteen ordinary chromosomes plus one
Sex-chromosome, and the other type contains only the eighteen
ordinary chromosomes; this being the result of the unequal
primary spermatocyte division, where the eighteen bivalent
chromosomes divide and the unpaired sex-chromosome passes
over to one pole undivided. In the oogonia there are thirty-six
ordinary chromosomes plus two sex-chromosomes. Before the
reduction division of the primary oocytes, in all probability
(though this was not actually determined in this animal), all of
the chromosomes, including the two sex-chromosomes, pair.
This eventually gives rise to ova all of which contain the reduced
number of chromosomes or eighteen ordinary chromosomes plus
one sex-chromosome.

Since the two types of spermatozoa are produced in equal
numbers, fertilization by the one kind or the other is equally
possible, and the number of male and female calves born is about
equal if a fairly large number of offspring is considered. Sex
in the offspring, as determined at the time of fertilization of the
ovum by the one or the other type of spermatozoon may be
illustrated as follows:

Spermatozoa. Ova. Offspring.

(18 + i) + (18 + i) = (36 + 2) = female,
(18 + o) + (18 + i) = (36 + i) = male.

The results of the above combinations are in exact accord
with the number of chromosomes found in the germinal and
somatic cells of the two sexes in cattle; and the relation of the
sex-chromosomes to sex-determination can not be doubted. All
of the five hundred or more so-called theories or rules for con-

STUDIES ON THE CELLS OF CATTLE.

303

trolling sex, including the one in relation to time of service which
is commonly practiced by animal breeders, must be abandoned.
There is considerable literature on the metabolic theories of
sex-determination which will not be discussed here. However,
reference may be made to the brief though able discussion of this
subject by Babcock and Clausen ('18).

IX. SEX RATIO IN CATTLE.

It might be well to quote here Pearl's ('17) more recent results
of extensive studies on the control of the sex ratio in cattle.
He says: "Some earlier statistics appeared to indicate that there
was a possibility of influencing the sex ratio by paying attention
to this point. It was believed to be of such extreme importance
as to justify the careful study of the matter on the basis of much
more extended statistics. These statistics we have now collected
and analyzed and shall publish as soon as possible. In the mean-
time it may be reported that, with the more extended statistics
in hand, it appears to be conclusively established that there is no
definite or permanent relation between the time in the heat period
at which the cow is served and the sex of the offspring. The
apparent relation between these two factors, which is believed
by many breeders to exist and which our earlier statistics ap-
peared to indicate, seems now to be purely accidental, and to
have arisen only because of the comparative meagerness of the
statistics on which the matter was discussed.

TABLE II.

SHOWING THE SEX OF THE CALVES FOLLOWING SERVICE AT DIFFERENT PARTS OF

THE HEAT PERIOD.

Heat Period.

Lapsed Time in Hours from Ap-
pearance of Heat to Service.

Sex of Offspring.

Per Cent, of
Males.

Males.

Females.

Early .

Under 3 hours
Over 3 and under 8 hours
Over 8 hours

2OO
270
I87

IQ2

252
212

51.0

51-7
46.9

Middle

Late

Totals

657

656

50.0

'The summarized results of 1 ,313 separate and distinct matings
given in Table II. will demonstrate this point. In each one of
these 1,313 cases the following facts were accurately known, and

304 J. E. WODSEDALEK.

reported in such a way that any bias, conscious, or unconscious,
of the observer could not have influenced the result: (a) the
time in hours from the first appearance of heat (oestrum), as
noticed by the breeder, to the time the cow was successfully
served; (&) the sex of the calf resulting from this service.

"It is evident from this table that there is no significant
preponderance of females when service is early in heat. There
is not now known any method by which the sex ratio or propor-
tion of the sexes in cattle may be effectively controlled by the
breeder. A more detailed account of the results, together with
further statistics will be published elsewhere."

X. SEX-LIMITED INHERITANCE IN CATTLE.

The discovery of the remarkable behavior of certain characters
in heredity which can only be plausibly explained by supposing
that they are linked with a sex-chromosome or a sex-determining
factor still further strengthens our belief in the existence of such
a definite factor. Wentworth ('16) reported a case which seems
to fall under this general sex-limited group in the inheritance of
black-and-white in Ayrshire cattle. While the general breed
color is red-and-white, black-and-white animals have been known
for some time, as shown by Kuhlman ('15). It is difficult to
state whether the black is due to a true black pigment or whether
it is simply a very dense red, since chemical solutions have not
yet been attempted.

In summarizing the results of the different crosses, Wentworth
says, "If the factor of the black-and-white color is represented
by B, the hereditary constitutions are as follows: BB is always
black-and-white; bb is always red-and-white; Bb is always
black-and-white in the male and red-and-white in the female.
All of the nine possible matings were discovered, as shown in
Table I.

"The expectations here presented are based on the most
probable result of each of the matings, considered on an indi-
vidual basis with reference to the number of animals produced
by each type of mating, but without figuring the proportions of
the sexes as equal. From these data it would appear that the
black-and-white color of Ayrshire cattle behaves in an ordinary

STUDIES ON THE CELLS OF CATTLE.

305

sex-limited manner similar to the horns in sheep as discussed
by Wood ('05) and the rudimentary mammae in swine as reported
by Wentworth ('16)."

TABLE I.

RESULTS OF NINE POSSIBLE MATINGS OF AYRSHIRE CATTLE.

Sires.

Dams.

Male Offspring.

Female Offspring.

Black-and-
white.

Red-and-white.

B'ack-and-

white.

Red-and-wWite.

BB

BB

I
O
IO

3

I

4
o

2
0

0
O
O

0

o

5

0

I
7

3
o
o

2

I
O

o

0
0

0
I
IO
I
O
4

3

2

9

BB

Bb. . .

BB.

bb.

Bb. . . .

BB

Bb

Bb

Bb

bb

bb

BB

bb

Bb. . .

bb

bb. .

Total .

21
20.7^

13

I^.2S

6
=;.2<;

30

^0.7^

Expected. .

The simple Mendelian scheme of inheritance is quite common
in cattle but, to my knowledge, this is the only case of sex-limited
inheritance reported in this animal ; nevertheless, it is significant,
especially in view of our knowledge of the relation between this
scheme of inheritance and the sex-chromosomes in many other

species.

XI. THE FREE-MARTIN.

The case of the free-martin, the female of two-sexed twins in
cattle, is well known to animal breeders to be perfectly sterile
although rarely such females are perfectly normal. I do not wish
to enter upon a discussion of this subject here for it is a big
problem in itself. However, I wish to call the attention of
animal husbandmen, who are not in position to keep in touch
with all of the zoological literature, to the extensive and most
painstaking piece of research on the free-martin by Dr. Frank R.
Lillie ('17). In this connection I might also call attention to his
article on sex-determination and sex-differentiation in mammals
(Lillie, '17).

XII. SUMMARY.

i. Thirty-seven chromosomes occur in the spermatogonia.
One, the sex-chromosome, is distinctly larger than the others.

306 J. E. WODSEDALEK.

2. Nineteen chromosomes appear in the primary spermatocyte
division, of which eighteen are bivalent and the other is the
unpaired sex-chromosome.

3. In the primary spermatocyte division the heart-shaped
sex-chromosome passes undivided to one pole in advance of the
other chromosomes.

4. The primary spermatocyte division is evidently the reduc-
tion division, giving rise to two different types of secondary
spermatocytes, one with the sex-chromosome and the other
lacking it.

5. The one type of secondary spermatocyte, which contains
the sex-chromosome, gives rise to two spermatids, each containing
the sex-chromosome and eighteen ordinary chromosomes.

6. The other type of secondary spermatocyte, which lacks the
sex-chromosome, gives rise to two spermatids, each containing
only the eighteen ordinary chromosomes.

7. The mature spermatozoa are of two types, equal in numbers.
The one type is larger and contains the sex-chromosome. The
smaller type is without the sex-chromosome. The larger type
is female producing, while the smaller is male producing.

8. Thirty-eight chromosomes occur in the oogonia; two of
these are the sex-chromosomes.

9. The reduced number of chromosomes in the female is, in
all probability, eighteen ordinary chromosomes and one sex-
chromosome which apparently occurs in all of the mature ova.

10. The somatic cells of the male contain thirty-six ordinary
chromosomes and one sex-chromosome.

1 1 . The somatic cells of the female contain thirty-six ordinary
chromosomes and two sex-chromosomes.

12. The number of chromosomes in the somatic cells of the
two sexes is in exact accord with expectations.

13. There is no relation between the time in the heat period
at which the cow is served and the sex of the offspring.

14. Sex in cattle, for the present, remains a matter beyond the
control of the breeder. It is determined by the sex-chromosomes ;
it is a matter of inheritance.

15. Sex-limited inheritance strengthens the belief in the chro-
mosome theory of sex-determination.

STUDIES ON THE CELLS OF CATTLE. 307

LITERATURE CITED.
Allen, Ezra.

'18 Studies on Cell Division in the Albino Rat (Mus norivegicus albinus).

Jour. Morph., Vol. 31, No. i.
Babcock and Clausen.

'18 Genetics in Relation to Agriculture. New York.
Guyer, M. F.

'10 Accessory Chromosomes in Man. BIOL. BULL., Vol. XIX., No. 4.
Hance, R. T.

'17 The Diploid Chromosome Complexes of the Pig (Siis scrofa) and their

Variations. Jour. Morph., Vol. 30, No. i.
Kingery, H. M.

'17 Oogenesis in the White Mouse. Jour. Morph., Vol. 30, No. i.
Kuhlman, A. H.

'15 Black and White Ayrshires. Jour. Heredity, Vol. 6, No. 7.
Lillie, F. R.

'17 The Free-Martin; a Study of the Action of Sex Hormones in the Foetal

Life of Cattle. Jour. Exp. Zool., Vol. 23, No. 2.
'17 Sex-Determination and Sex-Differentiation in Mammals. Proc. Nat. Acad.

Sci., Vol. 3, pp. 464-370.
Pearl, R. and Parshley, H. M.

'13 Data on Sex Determination in Cattle. BIOL. BULL., Vol. 24, No. 4.
Pearl, R.

'17 Maine Agricultural Experiment Station Bulletin, Bulletin No. 261, Part 3,

The Control of the Sex Ratio, pp. 130-131.
Pratt, B. H. and Long, J. A.

'17 The Period of Synapsis in the Egg of the White Rat, Mus Norwegicus

Albinus. Jour. Morph., Vol. 29, No. 2.
Wentworth, E. N.

'16 Rudimentary Mammae in Swine a Sex-Limited Character. Science, N. S.,

Vol. XLIIL, No. 1114.
'16 A Sex-Limited Color in Ayrshire Cattle. Jour. Agr. Research, Vol. VI.,

No. 4.
Wodsedalek, J. E

'13 Spermatogenesis of the Pig, with Special Reference to the Accessory

Chromosomes. BIOL. BULL., Vol. XXV., No. i .

'14 Spermatogenesis of the Horse with Special Reference to the Accessory
Chromosome and the Chromatoid Body. BIOL. BULL., Vol. XXVII.,
No. 6.

'16 Causes of Sterility in the Mule. BIOL. BULL., Vol. XXX, No. i.
Wood, T. B.

'05 Note on the Inheritance of Horns and Face Color in Sheep. Jour. Agr.

Sci., V. i, pt. 3.
Zeleny, C. and Faust, E. C.

'14 Size Differences in Spermatozoa from Single Testes. Science, N . S., Vol.

39, No. 1003.
'153 Size Dimorphism in the Spermatozoa from Single Testes. Jour. Exp.

Zool., Vol. 18.

'150 Dimorphism in Size of Spermatozoa and its Relation to the Chrom osomes
Proc. Nat. Acad. Sc., Vol. i.

308 J. E. WODSEDALEK.

Zeleny, C. and Senay, C. T.

'15 Variation in Head Length of Spermatozoa in Seven Additional Species of
Insects. Jour. Exp. Zool., Vol. 19, No. 4.

EXPLANATION OF PLATES.
PLATE I.

(All of the drawings were made with the aid of a camera lucida, X 2,400, except
Figs. 39 and 40 which are X 2,200.)

FIG. i. Polar view of the metaphase of division in a spermatogonial cell show-
ing thirty-six ordinary chromosomes and the single heart-shaped sex-chromosome.

FIG. 2. Side view of the metaphase of division in a spermatogonial cell showing
the thirty-six ordinary chromosomes, and the sex-chromosome at the left.

FIGS. 3 AND 4. Late anaphases of division of spermatogonial cells showing the
division of all of the chromosomes, including the large sex-chromosome.

FIGS. 5-9. Late prophase stages of the last spermatogonial cells.

BIOLOGICAL BULLETIN, VOL. XXXVIII.

J. F. WODSEDALEK.

31 0 J. E. WODSEDALEK.

PLATE II.

FIGS. 10 AND ii. Last spermatogonial cells showing thirty-six ordinary chromo-
somes and the large sex-chromosome.

FIG. 12. Last spermatogonial cell taken from stale tissue showing the per-
sistence of the segregation of the chromosomes in this stage.

FIG. 13. Spireme stage of the primary spermatocyte showing the large sex-
chromosome.

FIGS. 14-18. Polar views of metaphase stages of division of the primary
spermatocytes showing eighteen bivalent chromosomes and the sex-chromosome
which is usually at the periphery of the plate.

BIOLOGICAL BULLETIN, VOL.. XXXVlll.

Hffitet

10

II

12

16

J. E. WODSEOALEK.

\.

18'

;':": 'v

:*jm

.-/ v^,

i -' ---:^

312 J. E. WODSEDALEK.

PLATE III.

FIGS. 19-25. Metaphase stages of division of primary spermatocytes showing
the passing of the sex-chromosome to one pole in advance of the other chromosomes.

FIGS. 26 AND 27. Late anaphase stages of division of primary spermatocytes
showing eighteen chromosomes at one pole and eighteen plus the sex-chromosome
at the other.

BIOLOGICAL BULLETIN. VOL. XXXVIII.

PLATE III.

20

21

22

24

26

£r?»ix

ll

27

J. E. WODSEDALEK.

314 J. E. WODSEDALEK.

PLATE IV.

FIGS. 28 AXD 29. Cells showing the formation of the secondary spermatocytes
from the primary spermatocyte division. The eighteen chromosomes pair loosely
to form nine bivalents which are present in the one type of secondary spermatocyte
and in the other type are shown the nine bivalents and the sex-chromosome.

FIG. 30. Metaphase stage of division of the one type of secondary spermatocyte
showing the nine bivalent chromosomes and the sex-chromosome.

FIG. 31. Metaphase stage of division of the other type of secondary spermato-
cyte showing only the nine bivalent chromosomes.

FIGS. 32-34. Division stages of the secondary spermatocyte with the sex-
chromosome showing the division of all of the chromosomes, including the sex-
element, giving rise to spermatids both of which contain the sex-chromosome.
When the division of the cell is complete (Fig. 34) each of the nine ordinary chromo-
some splits into two so that in reality there are eighteen ordinary chromosomes plus
the sex-chromosome in this type of spermatid.

FIGS. 35-37. Division stages of the secondary spermatocyte without the sex-
chromosome. Figure 37 shows the splitting up of the nine chromosomes at the
poles so that there are in reality eighteen chromosomes passed on to the other type
of spermatid.

FIG. 38. A newly formed spermatid showing the eighteen chromosomes.

FIG. 39. A mature spermatozoan of the smaller type undoubtedly without the
sex-chromosome.

FIG. 40. A mature spermatozoan of the larger type which undoubtedly contains
the sex-chromosome.

BIOLOGICAL BULLETIN, VOL XXXVIII.

PLATE IV.

2S

3°

32

34

36

J t. AODSEDALEK

m

33

......

••*....

37

/:' • 'v^";\ /°' ,;'s>.

/v /I ^:." Si

39

*,."•>. i

31 6 J. E. WODSEDALEK.

PLATE V.

FIG. 41. Resting stage of an oogonial cell showing two large nucleoli which are
undoubtedly the sex-chromosomes.

FIG. 42. Spireme stage of an oogonial cell showing the two large sex-chromo-
somes.

FIG. 43. An oogonial cell showing thirty-six newly formed chromosomes and
the two large sex-chromosomes.

FIGS. 44 AND 45. Late prophase stages of oogonial cells showing thirty-six
ordinary chromosomes and the two sex-chromosomes.

FIGS. 46 AND 47. Metaphase stages of division of the oogonia showing the two
sex-chromosomes at the periphery of the plate.

FIG. 49. Liver cell in late prophase stage taken from a female foetus showing
thirty-six ordinary chromosomes and the two sex-hromosomes.

FIG. 50. Brain cell in metaphase stage of division taken from a female embryo
showing thirty-six ordinary chromosomes and the two sex-chromosomes.

FIG. 51. Brain cell in metaphase stage of division taken from a male embryo
showing thirty-six ordinary chromosomes and only one sex-chromosome.

FIG. 52. Liver cell in late prophase stage taken from a male foetus showing
thirty-six ordinary chromosomes and only one sex-chromosome.

BIOLOGICAL BULLETIN, VOL. XXXVIII.

PLATE V.

41 ''•'•• .•"•/•"•'

PS&Jlt W; r--

!->;•;,

•;c •:":•>; -^sf*;. •••• •

42 ";" • - .. i-c""

- ,

45

49

-

%

48

>,*

'

J. E WODSEDALEK.

STUDIES IN ARTIFICIAL PARTHENOGENESIS. III.

CORTICAL CHANGE AND THE INITIATION OF

MATURATION IN THE EGG OF CUMINGIA.1

L. V. HEILBRUNN

This study is a record of experiments performed during the
summer of 1916 at the Woods Hole Marine Biological Laboratory.

As Morgan pointed out in 1910, the egg of the lamellibranch
Cumingia is very suitable for study. Like most other eggs it is
still immature when shed into the sea-water. Although the first
maturation spindle has formed, no polar bodies are thrown off
unless the egg is fertilized or treated with the proper reagents.
Doubtless some change is necessary before the egg can throw off
polar bodies and begin its development.

An effort has been made to determine the nature of this
change. Many diverse reagents cause the egg to ma-ture. Al-
though all of these reagents do not occasion the same morpho-
logical transformations nevertheless all of them agree in having
one specific physical effect on the egg. All release the egg cyto-
plasm from the restraint of a rigid enveloping membrane. The
immature unfertilized egg is surrounded by a stiff vitelline
membrane which presses tightly in on it and effectively prevents
the throwing off of polar bodies. It is only when the egg is
released from this restraint that maturation can proceed.

PHYSICAL MAKE-UP OF THE EGG.

As in Arbacia, the Cumingia egg is a mass of fluid protoplasm,
surrounded by a rigid membrane. Only a few turns of the
centrifuge are sufficient to throw to opposite poles of the egg the
substances suspended in the cytoplasm. To one pole pass the
presumably lighter oil globules, to the opposite pole the heavier
pigment. But these suspended particles can go no farther than
the poles, for there they are stopped by the vitelline membrane
which surrounds the egg. This is a stiff structure and is easily

1 Contribution from the Zoological Laboratory, University of Michigan, New
Series, no. 3.

317

31 8 L. V. HEILBRUNN.

visible under higher power. It is about one micron in thickness.
The vitelline membrane must be thought of as the plasma mem-
brane of the egg. As I have pointed out before, students of
cellular mechanics have been blind to the fact that egg cells
are provided with visible membranes which are plasma-mem-
branes. They have often insisted that no one has ever seen a
plasma-membrane. It is easy to prove that the vitelline mem-
membrane of the Cumingia egg governs osmotic intercourse and
is therefore a plasma-membrane.

When Cumingia eggs are put into hypertonic solutions they
shrink only very slightly. The stiffness of the vitelline plasma-
membrane prevents a marked shrinkage. Moreover, a weakly
hypertonic solution produces just as much shrinkage as a strong
one provided that it does not alter the membrane. On the other
hand if the hypertonic solution makes the membrane less rigid
by causing it to swell, then the egg shrinks to a great extent.
Facts such as these can be interpreted only on the assumption of
a stiff plasma-membrane. That this is the vitelline membrane
is certain, for there is no other membrane in the vicinity. Cer-
tainly there is none inside the vitelline membrane, for when the
egg has been left for some time in a hypertonic solution of con-
siderable strength then the coagulated cytoplasm shrinks away
from the vitelline membrane and presents a rough uneven sur-
face. Under these conditions it is obviously not surrounded by
any membrane.

As to the chemical nature of the plasma-membrane it is
essentially protein. It swells in dilute acids or alkalies, and in
sodium chloride or sodium iodide solutions. Moreover it does
not contain any large admixture of lipoid as can be shown by test-
ing it with Scharlach R solution.

Like many other marine eggs the Cumingia egg is surrounded
by a diffuse jelly of the same refractive index as sea-water and
therefore invisible. It can easily be demonstrated by various
vital stains (e.g., Nile blue sulphate) or by India ink suspensions.
This jelly has no apparent effect on the early developmental
phenomena. If the eggs are shaken a few times in a test-tube,
they are deprived of their jelly. Such eggs react in the same
way as those with jelly intact.

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 319

As in Arbacia the Cumin gia vitelline membrane is capable of
two kinds of cortical change. Both membrane elevation and
membrane swelling can occur. If the egg is placed in solutions
of low surface tension the tension of the membrane is lowered
and it rapidly lifts away from the egg. No doubt the explanation
of this process is the same as that I have offered for the similar
process in the sea-urchin egg. But in Cumingia membrane
elevation is not the normal cortical change. When eggs are
inseminated the membrane does not become elevated, it becomes
swollen.

Treatment of eggs with a reagent which causes either mem-
brane elevation or membrane swelling will result in a throwing
off of polar bodies. Thus just as in the sea-urchin egg either
swelling or elevation provides the necessary cortical change,
although in this egg it is swelling and not elevation which is
normal. Finally there is a third way in which maturation can
be initiated. This consists in the removal of the membrane.

I shall now proceed to the experimental data considering first
membrane elevation, second membrane swelling, and third the
removal of the membrane.

MEMBRANE ELEVATION.

Any substance which markedly lowers the surface tension of
sea-water is effective in producing membrane elevation. But if
elevation is to be followed by maturation only certain concentra-
tions and certain lengths of exposure can be used in each case.
Too high a concentration or too long an exposure generally leads
to coagulation. Frequently an over-exposure produces a rupture
of the egg membrane, the protoplasm then flows out to form an
exovate. Such eggs soon disintegrate.

Numerous reagents were used to produce a lowering of surface
tension. Of course the number of successful reagents could have
been increased many fold. Thus many of the higher alcohols
no doubt act in the same way as the amyl alcohol which I used.

In the experiments the same general proceedure was always
employed. In every case, the eggs of only a single female were
used. Owing to the fact that the number of eggs obtainable
from a single female is comparatively small, I found it best to

32O L. V. HEILBRUNN.

perform all of the experiments in low Stender dishes (containing
about 50 cc.) instead of in larger fingerbowls. These dishes were
very convenient, for they could without any trouble be put on the
stage of the microscope. One Stender dish was made to contain
the desired reagent and then the eggs were pipetted into it-
After varying lengths of exposure the eggs were transferred
from the reagent to fresh sea-water also contained in Stender
dishes. Then after twenty minutes or more had elapsed the
treated eggs were examined.

As the polar bodies of Cumingia are unusually large it is possible
to count the percentage of matured eggs under low power and
without removing the eggs to a slide. Counts were usually made
in this way. Obviously it is not possible to see the polar bodies
if the egg is lying with the animal pole down. Hence the per
cent, recorded is always too low. Even if all the eggs had polar
bodies not many more than fifty per cent, would show them.
If the eggs are turned about this difficulty can of course be ob-
viated, but turning is a tedious process and was only occasionally
resorted to. Thus the counts recorded represent minima and are
really only about half as high as they should be. Counts are
given in the form of fractions, in which the denominator repre-
sents the total number of eggs counted the numerator the number
of eggs with polar bodies. For example the fraction 10/50 would
indicate that out of 50 eggs counted 10 showed a polar body.

In making up per cent, solutions of volatile liquids, it was
found convenient to use 100 c.c. measuring flasks. Thus if a
2 per cent, ether solution was desired 2 c.c. of ether was placed
in a 100 c.c. measuring flask and sea-water was added until the
solution reached the 100 c.c. mark. Owing to the diminution
in volume on mixing the two liquids such a solution is not exactly
a 2 per cent. one. However, after the above method of procedure
it is easier to calculate the molecular concentration.

The accompanying table gives the results of experiments with
eleven substances which lower surface tension. For each sub-
stance used it was necessary first to determine the proper con-
centration. Sometimes this was a simple matter. Thus for
saponin almost any concentration is successful. But usually
only a narrow range of concentrations will produce the desired

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 321

effect. If the reagent is just a little too strong it can not be
effectively employed. In many cases only a certain length of
exposure is suitable. Too long exposures generally produce
exovates. Formation of exovates is represented by the symbol
"e" in the table.

With ether and chloroform it is particularly difficult to obtain
maturation. In both instances I at first despaired of success.
A 3 per cent, solution of ether in sea-water did not produce
maturation. With such a solution exovates generally appeared.
On the other hand 2.5 per cent, ether had no observable effect on
the eggs. With a concentration intermediate between 2.5 per
cent, and 3 per cent, success was attained. In the table are given
the figures for a representative experiment with ether.

In the case of chloroform the range of effective solutions is
even narrower. In making up chloroform solutions very small
quantities of the liquid had to be measured out. This was done
by counting the drops from a small pipette which was calibrated
for the purpose. About 60 drops from this pipette constituted
i c.c. By placing 8 such drops into a 100 c.c. measuring flask
and diluting to the mark a 0.13 per cent, solution was obtained.
This solution produced but little effect upon the eggs. Exposures
of 3-5 minutes showed only 1-2 per cent, of polar body formation.
Thus a 0.13 per cent, solution was apparently too weak. On the
other hand, a 0.17 per cent, solution proved too strong. Such a
solution was prepared by diluting 15 drops of chloroform (from
the pipette mentioned above) up to 100 c.c. with sea-water.
A 0.25 per cent, solution was thus obtained and this was then
diluted to 0.17 per cent, by adding 5 c.c. of sea-water to 10 c.c.
of the solution. The resultant 0.17 per cent, solution was found
to be too strong, for it produced exovates. Moreover viscosity
tests with the centrifuge showed that it coagulated the egg cyto-
plasm. Although neither 0.13 per cent, nor 0.17 per cent,
chloroform produced maturation, a concentration slightly under
0.17 per cent. did. This solution was made by diluting 10 drops
from the pipette up to 100 c.c. Such a solution would ordinarily
give 0.17 per cent., but one particular solution was made on an
extremely hot day (when the room temperature was 27°). No
doubt the drops from the pipette were smaller, owing to the

322

L. V. HEILBRUNN.

lower surface tension. Hence the concentration which resulted
was slightly under 0.17 per cent. In the table I have referred
to this solution as "0.16 per cent." As the figures show, it
proved highly effective.

TABLE I.

POLAR BODY FORMATION AFTER TREATMENT WITH SUBSTANCES WHICH LOWER

SURFACE TENSION.

Exposure
Minutes.

i

Per
Cent.

Amyl
Alco-
hol.

2.5

Per

Cent.
Ethyl
Ace-
tate.

0.25
Per
Cent.
Ethyl
Buty-
rate.

Pe5r
Cent.
Ethyl

Ni-
trate.

2.8

Per

Cent.
Ether.

'0.16
Per
Cent.'

Chloro-
form.

5
Per
Cent.
Aceto-
nitrile.

Satu-
rated
Solution
Phenyl
Ure-
thane.

O.2

Per
Cent.
Chlore-
tone.

Emul-
sion1
Toluol.

O.2

Per
Cent.
Sa-

ponin.

I

0

O

0

0

O

29

12

4

SO

SO

50

IOO

IOO

IOO

50

i

o

O

O

0

O

£3

2

0

18

3

2O

2

SO

50

SO

SO

IOO

So

50

IOO

IOO

IOO

50

I

i

0

o

2

I

29

12

o

o

10

33

50

50

50

50

' IOO

50

50

IOO

IOO

IOO

50

12

4

22

14

6

o

o

2

25

50

IOO

IOO

IOO

e

e

IOO

IOO

e

12

IO

15

o

0

33

3

1 *

e

e

e

e

e

25

50

IOO

IOO

IOO

50

3£

13

22

o

18

4

e

e

e

e

e

50

5°9

IOO

IOO

SO

19

18

I

20

5

e

e

e

e

—

50

IOO

IOO

50

£4

12 232

2

6

,

e

25

SO 5°

IOO

15

19

7

e

e

e

e

e

IOO

50

8

23

5

50

IOO

3

16

5

9

10

—

e

e

e

e

25

50

IOO

IOO

15

43

»

200

All of the substances mentioned in the table produced a similar
morphological change in the egg. All of them caused the vitelline
membrane to become lifted away from the cytoplasm. The
details of this membrane elevation were watched in a great many
different experiments. The membrane first detached itself from
the egg at a number of points around the periphery. As a result
it appeared slightly thicker and for a minute or two it often

1 This emulsion was made by stirring i c.c. ol toluol with 3 c.c. or sea water.

2 Eggs were shaken while counting. As mentioned above this gives a higher
count.

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 323

showed small outbulgings which gave it a more or less crenate
appearance. Very soon, the membrane lifted away with con-
siderable rapidity. If success was to be obtained the egg had to
be removed from the solution at just about the time the crena-
tions appeared.

As in Arbacia when the vitelline membrane is lifted from the
egg surface, a new membrane, no doubt a precipitation membrane,
is immediately formed about the egg cytoplasm. Sometimes it is
possible to cause this to become elevated also. Saponin produces
the best results. When eggs are exposed for 40 minutes to 0.2
per cent, saponin they are found to be surrounded by two
membranes. The outer one of these is quite far from the egg
and it was thought at first that it might represent the outer edge
of the jelly which had become visible. But this was shown not
to be true, for two membranes could be produced about eggs
from which the jelly had previously been shaken off.

I have used the eggs of Cumingia for a number of years and in
some years I have found a tendency for a small per cent, of the
eggs to mature without apparently any treatment. In the experi-
ments recorded in this paper careful controls were always kept.
In every case at least two hundred eggs of the untreated- control
were examined for polar body formation. Of the experiments
cited in the table, in only one instance did the control show any
maturation. In the control for the ethyl nitrate experiment one
egg showed a polar body out of considerably over two hundred
examined.

The table shows that eleven substances can produce polar
body formation and this is in every case preceeded by membrane
elevation. These eleven substances differ from each other very
widely in chemical constitution. It is almost inconceivable that
they should have any one chemical effect in common. Their
action must be primarily physical. It is believed to involve a
lowering of surface tension. The explanation which I have
offered for membrane elevation in Arbacia applies equally well
for Cumingia. For details of this explanation the reader is
referred to my earlier papers.

Of course by choosing substances similar to those listed in the
table numerous other successful reagents could no doubt be

324 L. V. HEILBRUNN.

discovered. Thus it is probable that benzol or xylol would
behave as toluol, and that many other alcohols and esters could
have been added to those given in the table. Such experiments
would scarcely add material for the general argument.

On the other hand it might be thought that the eleven sub-
stances in the table represent only a few out of the many that
I have tried. It perhaps not infrequently happens that experi-
menters suppress the record of their failures. But in these
experiments practically every substance selected because of its
effect on surface tension gave the expected result. There were
only three exceptions and in two of these cases I performed only
a single experiment with a single concentration of the reagent.
Moreover in each of these three cases the reagent employed had
some secondary effect on the egg. I shall consider each case
in detail.

Ethyl Urethane. — 3 per cent, ethyl urethane had practically no
effect, although in longer exposures (5-7 minutes) a few eggs
with polar bodies were observed. The reagent has some action
on the jelly that I have not analyzed. Possibly it is a shrinkage
effect. Two minutes after the egg entered the solution it was
surrounded by queer looking bubbles. Fifteen minutes later
the bubbles had disappeared and their place had apparently
been taken by a zone of radiating lines.

Nitromethane. — I used a 5 per cent, solution of nitromethane
and exposed the eggs to it for intervals of from ^-10 minutes.
No polar bodies were produced as a result of the treatment.
Exposure for 4 minutes followed by transfer to normal sea-water
resulted in rupture of the vitelline membrane and disintegration
of the egg. The eggs left in the nitromethane solution showed
a queer transformation. They lost their spherical shape and
flattened out into discs resembling huge red blood corpuscles.
In addition to this queer effect the reagent also appeared to have
some action on the jelly, for some morphological changes were
visible around the egg. The nature of these changes I did not
stop to investigate.

Acetone. — Two concentrations of acetone were tried. 25 per
cent, acetone did not produce membrane elevation when eggs
were exposed i-n minutes. 50 per cent, acetone produced

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 325

membrane elevation, but not polar body formation. Of the
eggs exposed \ minute, 72 per cent, had widely elevated mem-
branes. These eggs, however, could not produce polar bodies
for they were thoroughly coagulated. This was shown by a
viscosity test with the centrifuge.

MEMBRANE SWELLING.

If instead of being lifted off, the vitelline membrane is made
to swell, much the same effect is produced on the egg. The
increased fluidity of the vitelline membrane results in a lower
surface tension.1 Consequently it no longer exerts as great a
pressure upon the egg contents. Thus maturation follows mem-
brane swelling just as it follows membrane elevation.

In order to produce a swelling of the membrane the same
reagents were used that were previously found to have been
effective for the sea-urchin egg. Evidently the vitelline mem-
branes of both Cumingia and Arbacia are similar, for they swell
under the same conditions.

Sodium Iodide. — Eggs were exposed to 0.6 M sodium iodide.
After 15 and 22^ minutes they were removed to sea-water in
Stender dishes A and B respectively. Of the eggs in A, 14/100
showed polar bodies. The eggs in B formed no polar bodies.
The sodium iodide solution caused membrane swelling.

Hydrochloric Acid. — Eggs were placed in 25 c.c. of sea-water
plus 0.7 c.c. n/io HC1. After exposures of |, i, 2, 3, 4, 5, 7, 10
minutes the eggs were removed from the acid solution and placed
in ordinary sea-water in Stender dishes A— H respectively. In
the acidified sea-water the egg membrane swelled and the surface
of the egg became sticky. Often the eggs adhered to the bottom
of the dish. Counts of eggs with polar bodies gave the following
results:

A }<2 minute exposure 6/50

B i 5/50

C i 10/50

D 3 4/50 (This count was made too early)

E 4 13/50

F 5 IS/50

G 7 10/50

H 10 8/50

1 Many biologists apparently do not understand that solids and pseudo-solids
(i.e., gels) exhibit surface tension. This tension is greater for a gel than for the
corresponding sol. For references to literature on this subject consult Heilbrunn
'15, footnote, p. 166.

326 L. V. HEILBRUNN.

The eggs which were allowed to remain in the acid sea-water
also formed a few polar-bodies.

Potassium Hydroxide. — Eggs were placed in 40 c.c. of sea-water
plus i c.c. n/io KOH. In this solution membrane swelling
occurred. The eggs formed polar bodies while in the alkaline
medium. The first polar body observed was noted after 16
minutes exposure. After 33 minutes, a count gave 14/100 with
polar bodies.

Potassium Cyanide. — Eggs were placed in a 0.04 per cent.
KCN solution, made by diluting 5 c.c. of 2 per cent. KCN up
to 250 c.c. with sea-water. During the experiment the cyanide
was not allowed to evaporate, for it was kept in a tightly
stoppered weighing-tube. After a 36 minute exposure a count
showed 28/100 of the eggs with polar bodies. The potassium
cyanide caused membrane swelling probably because of its alka-
line reaction.

Hypertonic Sodium Chloride Solution. — In the experiments with
acids and alkalies and with sodium iodide the solutions used
were approximately isotonic with sea-water. In the course of
some other work it was noticed that solutions made by adding
2\ M NaCl to sea-water caused membrane swelling. Hence it
was expected that these solutions would also cause polar body
formation. Eggs were exposed for 7 minutes to a solution made
by adding 5 c.c. of 2| M NaCl to 25 c.c. of sea-water. As a
result 21/100 of the eggs formed polar bodies.

To sum up, in all of these experiments the swelling of the
vitelline membrane was in every case followed by the throwing-off
of polar bodies in a large percentage of the eggs. The controls of
untreated eggs did not form polar bodies. In these experiments
with reagents which cause membrane swelling there were no
failures. No reagent could be discovered which would produce
swelling of the vitelline membrane without at the same time
causing the eggs to mature.

RUPTURE OR REMOVAL OF THE MEMBRANE.

There is a third way in which the Cumingia egg may be freed
from the binding pressure of its vitelline membrane. If the eggs
are shaken vigorously, oftentimes a certain percentage of the

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 327

membranes will be shaken off. Apparently the membrane rup-
tures at the animal pole of the egg and owing to its elasticity it
shrinks away toward the vegetal pole. In many cases its
wrinkled remains can be found at this pole. No doubt after
the membrane is shaken off a new precipitation membrane forms
about the cytoplasm but this is very much less rigid than the
original vitelline membrane. The results obtained from shaking
Cumingia eggs are somewhat variable. The eggs must be shaken
sufficiently to rupture or remove the membrane, but they must
not be shaken too vigorously or too long. Too much shaking
interferes with the mitotic processes underlying maturation and
a smaller percentage of polar bodies results. This is in accord-
ance with the observation of Wilson ('01) that shaking prevents
cell division.

A number of shaking experiments were performed. In one
of the best of these, the eggs were placed in a small 10 c.c. test
tube and shaken vigorously by swinging the forearm from a
vertical to a horizontal position. They were given 40 such
swings in 10 seconds. When these shaken eggs were examined
an hour and a half later it was found that 27/56 had polar bodies.
Thus practically all of them had matured, for as pointed out in
the first part of this paper, the counts represent minima. That
the shaking process had actually resulted in a removal of the
membrane could be demonstrated in three ways. In the first
place the remains of the membrane could often be seen at the
vegetal pole. Secondly, when the polar bodies formed they did
not appear to be within a stiff membrane as in polar body
formation after fertilization. Lastly some shaken eggs were
placed in a drop of acetone, of these only 3 out of 50 showed
membrane elevation. When normal unshaken eggs were simi-
larly treated with acetone all of them showed an elevated
membrane.

Finally there is still another method of freeing the egg from
the restraint of its membrane. Although this method produces
practically the same results as shaking, the procedure is very
different. It was found that in diluted sea-water rupture of the
vitelline membrane occurred and as was to be expected, matura-
tion followed. Obviously in diluted sea-water the osmotic pres-

328 L. V. HEILBRUNN.

sure forces water to enter the eggs and the membrane bursts-
In many instances the ruptured membrane could be seen at one
side of the egg. Sometimes exovates were produced. Various
dilutions were employed, from pure distilled water to a mixture
of I part of distilled water to 2 parts of sea-water. In one
experiment eggs were exposed to distilled water for 40, 60 and
80 seconds and then returned to sea-water. All three exposures
were successful. The 40 second exposure was not counted, the
60 second exposure showed 51/100 polar bodies and the 80 second
exposure 52/100 polar bodies. The controls were normal and
over 200 eggs which were counted showed no signs of maturation.
All the exposures showed some signs of cleavage. One egg
reached a stage with about 16 cells. In another experiment
eggs were placed into 30 c.c. of sea-water plus 15 c.c. of distilled
water. Of these eggs 13/50 showed polar bodies.

THE SIGNIFICANCE OF CORTICAL CHANGE.

It has been shown that three types of cortical change can be
produced in the Cumingia egg. All of these free the egg from
the restraint of a stiff vitelline membrane. This release from
restraint may then be thought of as the essential feature of
cortical change and as the direct cause of maturation. When it
occurs maturation follows; without it no maturation takes place.

It might be argued that all of the types of cortical change
have some other common effect besides the one mentioned.
Perhaps they all directly produce an increase of oxidations,
which then causes maturation to follow. This would be a diffi-
cult relation to conceive of chemically. Moreover there is experi-
mental evidence that maturation does not depend on an increase
of oxidations.

The supporters of the oxidation theory of initiation of develop-
ment have always held that dilute cyanide solutions check
oxidations. Thus Loeb, '13, states on p. 26: "It has long been
known that the oxidations in the cell can be prevented by the
addition of a little potassium cyanide, even when oxygen is
present. I have found that the addition of 0.5 c.c. of a 1/20
per cent. KCN solution to 50 c.c. of sea-water is sufficient to stop
almost immediately the effect of the spermatozoon in the fertilized
sea-urchin egg." Such a solution is 0.0005 Per cent.

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 329

Some Cumingia eggs were placed in a 0.04 per cent, solution of
potassium cyanide, five minutes after they had been fertilized.
The solution was prepared by diluting 5 c.c. of 2 per cent. KCN
up to 250 c.c. with sea-water. Solution and eggs were kept in a
glass-stoppered weighing-tube to guard against evaporation of
the cyanide. Under these conditions practically all of the
fertilized eggs formed the first polar body. Actual count with-
out turning over the eggs, showed 24/50.

On the other hand it might be thought that all types of cortical
change produce an increase of permeability. In recent years
various observers have claimed that the sea-urchin egg under-
goes an increase in permeability either after fertilization or
after artificial membrane elevation. These observers have en-
deavored to show: (i) An increased penetration of dyes, (2) a
drop in electrical resistance, (3) a more rapid passage of water
into or out of the cell.

When fertilized sea-urchin eggs are placed in dilute solutions
of methylene blue, they stain more rapidly than do unfertilized
eggs, according to Lyon and Shackell, '10. Runnstrom, 'n,
obtained similar results with methylene blue although not with
neutral red. The experiments of Lyon and Shackell are fre-
quently cited and are always taken to indicate an increased
permeability after fertilization.1 When unfertilized and fertilized
Cumingia eggs are placed in dilute solutions of methylene blue
or neutral red, the unfertilized eggs take up the dye just as
rapidly as do the fertilized eggs.

Fertilized and unfertilized eggs were put into Syracuse dishes
containing methylene blue solutions of various strengths. From
time to time the eggs were examined over a light and over a dark
background, and under the microscope. The color of faintly
stained eggs can be much better observed with the naked eye
than with the microscope. A more accurate method of deter-
mination would involve the use of a colorimeter but the method
used was sufficient to show that no marked increase of permeabil-
ity occurred. The dilutions of the dye were made up from a 0.5
per cent, solution of Griibler's "Methylenblau rectif. nach

1 They might however indicate nothing more than an increased affinity for
dyes on the part of the cytoplasm. Such changes in staining properties are
common enough, especially after changes in the colloidal state.

330 L. V. HEILBRUNN.

Ehrlich." Five minutes after fertilization some fertilized eggs
were placed in dishes A2 to £2 containing o.i per cent., 0.05
per cent., 0.025 per cent., 0.0125 per cent., 0.00625 per cent,
methylene blue respectively. At the same time some unfertilized
eggs were placed in dishes Ai to El containing similar concentra-
tions of the dye. After ten minutes had elapsed, the eggs in
Ai and A2 were navy blue, those in Bi and 62 pale blue, those
in Ci and C2 scarcely colored and those in Di and D2, El and
£2 not colored at all. Obviously there was no difference between
the two sets of eggs. After fifteen minutes eggs in Ai and A2
were navy blue, those in Bi and 62 light navy blue, those in Ci
and C2 light blue, the unfertilized eggs in Di were a very pale
blue, the fertilized eggs in D2 uncolored, in El and £2 eggs were
still uncolored. The eggs were observed at various times, but
no marked changes could be observed. After an hour I thought
I might be able to detect a slightly deeper color in the fertilized
eggs in A2, 62, C2 than in the unfertilized eggs subjected to the
same concentrations of dye in Ai, Bi, Ci, but this was probably
due to the fact that the fertilized eggs were slightly more numer-
ous. If entrance of stain is a test of permeability, then certainly
there is no sharp difference in the permeability of fertilized and
unfertilized Cumingia eggs. The concentrations of stain used
in the above experiment were not injurious. Even in the most
concentrated of the solutions used the eggs proceeded in their
development and became motile larvse. A similar experiment
was tried with neutral red. A saturated solution of the stain
and 1/2, 1/4, 1/8, 1/16 saturated solutions were used. In no
case could any difference in permeability between fertilized and
unfertilized eggs be noted.

McClendon, '10, and Gray, '16, have maintained that following
fertilization there is a drop in the electrical resistance of the sea-
urchin egg. They have interpreted this as indicating an increase
of permeability, although of course various other explanations
might be given. In the sea-urchin egg the normal process of
cortical change is membrane elevation, which would interfere
with the experiment. McClendon and Gray therefore were
obliged either to \vait until the eggs lost their power of under-
going membrane elevation or to so treat the eggs that they lost

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 33!

this power, before they could make their measurements. In
the Cumingia egg electrical measurements would be easier inas-
much as the normal cortical change is not membrane elevation.

But the following point should be noted. In any measure-
ments of the electrical resistance of masses of egg cells, it is rather
doubtful if one is measuring the resistance of the cells at all. If
one conceives of a piled-up mass of spheres resting in a liquid,
it is obvious after a moment's consideration that the liquid is
broadly continuous from one side of the mass to the other. Thus
water flows readily through a pile of shot. If now the spheres
are poor conductors as compared to the liquid, and an electric
current is sent through the mass, it will flow almost exclusively
through the liquid. The resistance then depends on the size
and shape of the interspaces between the spheres. In the case of
egg cells this may vary in several ways. In the sea-urchin egg
it is certain that after fertilization the interspaces between
individual eggs in a mass are greater than those before fertiliza-
tion, for it has been shown (cf. Heilbrunn, '15) that after fertiliza-
tion the eggs offer much more resistance to compression and
hence tend to preserve their spherical shape. Thus after fertiliza-
tion one might expect the resistance of a mass of eggs to be lower
even though the electric current did not pass through the eggs
at all.

In the last few years R. S. Lillie ('16, '17) has shown that when
fertilized or activated eggs are placed in hypotonic solutions,
water enters them more rapidly than it does unfertilized eggs.
Similarly in hypertonic solutions (R. S. Lillie, '18) water leaves
the fertilized eggs more readily. This is due according to Lillie
to an increased permeability of the plasma membrane to water.
Lillie's reasoning is a bit difficult to follow. Originally he be-
lieved in an increased permeability of the membrane to salts
and dissolved substances. This would of course decrease the
speed of entrance of water from hypotonic solutions, or the speed
of exit to hypertonic solutions, for it would decrease the osmotic
pressure upon which the exit or entrance of the water depends.
Osmotic pressure is, as everyone knows, dependent upon the
impermeability of a membrane to dissolved substances. Increase
in permeability to salts would therefore produce the opposite

332 L. V. HEILBRUNN.

effect from increased permeability to water. Apparently there
is a dilemma.

As a matter of fact it appears to be rather far-fetched to
assume a change in permeability to water, since we know the
plasma membrane to be at all times permeable to it. R. Lillie's
results can be much more simply explained on the basis of my
conception of the plasma membrane (see Heilbrunn, '15). Nor-
mally before fertilization it is a more or less rigid structure and
as such resists the entrance or exit of water from the cell. I
showed by measurement ('15, pp. 155-158) that when the
membrane was made less rigid as a result of membrane swelling
then water left the cell more readily. After fertilization the
plasma membrane either itself becomes less rigid, as when
membrane swelling occurs, or it is replaced by a less rigid mem-
brane as a result of membrane elevation. Hence water enters
and leaves the cell more rapidly.

It is easy enough to decide between R. Lillie's interpretation
and mine. If the difference is simply one of relative permeability
to water, then in hypotonic or hypertonic solutions the water
should enter or leave the fertilized eggs more rapidly, but the
final equilibrium point should be the same for both fertilized and
unfertilized eggs. However on the basis of my view, not only
should the water enter and leave the eggs more rapidly, but the
actual equilibrium state should be altered. In hypertonic solu-
tions, more water should leave the fertilized eggs and in hypotonic
solutions more water should enter them. In the case of hypo-
tonic solutions Lillie's own figures seem to show that at equilib-
rium more water has entered the fertilized eggs than the
unfertilized.1

1 Lillie's measurements were made on the egg of the sea-urchin Arbacia. In
this egg the presence of the elevated vitelline membrane or fertilization membrane
introduces a complication. As is well known this membrane is a stiff structure.
When fertilized sea-urchin eggs are subjected to hypotonic solutions they increase
in size rapidly until they reach the elevated membrane. Then further increase in
diameter is dependent on the power of the eggs to stretch or rupture the membrane.
R. S. Lillie does not state which occurs. As a matter of fact these experiments
on endosmosis were done in September when the Arbacia season is practically
over. At this time the normal membrane elevation is difficult to obtain, and
usually, unless the sperm concentration falls within certain very narrow limits,
the membrane swells at fertilization. For eggs with swollen membranes the
experiment is uncomplicated, as in Cumingia.

STUDIES IN ARTIFICIAL PARTHENOGENESIS.

333

My experiments with Cumingia eggs show clearly that after
fertilization not only do the eggs swell more rapidly in hypotonic
solutions but their total imbitition of water is greater.

Five minutes after fertilization some eggs were placed in 25
c.c. sea-water plus 15 c.c. distilled water. At the same time some
unfertilized eggs were placed in a similar solution. Both sets of
eggs were measured after about 45 minutes. By this time the
eggs had reached an osmotic equilibrium for a second set of
measurements 45 minutes later showed no further change. The
results are given in Table II.

TABLE II.

DIAMETERS OF FERTILIZED AND UNFERTILIZED EGGS IN 25 c.c. SEA-WATER + 15

c.c. DISTILLED WATER.

Average Diameter of 12 Unfertilized Eggs in Sea-water 62.65 p..

Fertilized Eggs After 40-45 Min.

Fertilized Eggs
After 85-90 Min.

Unfertilized Eggs
After 47-55 Min.

Unfertilized Eggs
After 92-100 Min.

M

M

M

M

70.4

69.2

66.9

67-5

69.2

69.8

66.9

67-5

73.9 X 66.9

68.7

66.9

66.9

71.0 X 68.1

69.2

68.1

66.4

69.8 X 66.9

69.2

66.9

68.1

69.8

71.5 X 63.6

68.1

67-5

69.2 X 68.7

71.5

67-5

66.9

68.1

70.4

66.9

67-5

71.0 X 67.5

69.2

66.4

68.7

70.4

70.4

68.1

68.1

72.7

69.8

68.7

67-5

70.4

70.4 X 68.1

68.1

66.9

69.2

Average, 69.76

69.53

67.46

67.46

Measurements were made with a Spencer movable scale
micrometer at a magnification of about 650 diameters. Not all
the fertilized eggs remained spherical. For such eggs the longest
and shortest diameters are recorded on the table. In computing
the averages the mean of these two measurements was taken.
It is obvious from the table that a greater amount of water
entered the fertilized eggs. This can not be explained by assum-
ing an increased permeability to water.

It should be noted that the vitelline membranes of the eggs in
this experiment did not rupture as a result of being placed in the
hypotonic solution. Such a rupture sometimes occurs in solu-
tions of this strength.

334

L. V. HEILBRUNN.

In hypertonic solutions fertilized Cumingia eggs lose more
water than do the unfertilized. The water does not merely
leave the eggs more rapidly, more of it passes out and the osmotic
equilibrium is different in the two cases. This is shown by an
experiment in which fertilized and unfertilized eggs were placed
in solutions prepared by adding i part of 2 M MgSO4 to 2 parts
of sea-water. Measurements were made as in the previous
experiment. The results are given in Table III.

TABLE III.

DIAMETERS OF FERTILIZED AND UNFERTILIZED EGGS IN 30 c.c.

2M McSOj + 60 c.c. SEA-WATER.
Average Diameter of 10 Unfertilized Eggs in Sea-water 6i.~4 n-

Fertilized Eggs After 13-23 Min.

Fertilized Eggs
After 163-172 Min.

Unfertilized Eggs
After 27-33 Min.

Unfertilized Eggs
After 143-161 Min.

M

M

M

M

57-7

51.9 x 58.3

59-4

60.6

53-7 X 58.9

57-7

6O.O

6O.O

56.5

43-3

59.4 (58.3)

6l.2 (56.0)

57-7

42.7

61.7

58.9 (53-7)

57-1 X 58.3

57-7

60.0 (53.7)

60.6 (56.1)

57-1 (51.9)

56.5

59-4

58.9

56.5

57-1

59-4

59-4

59-5 X 54-2

53-7

60.6

57-7

56.0

53-1

61.2

58.9

55-4

49.0

59-4

59-4

54-8

58.9 X 61.2

59-4

Average, 56.78

5-2-79

60.05

59-55

The above table requires a little explanation. Some of the
eggs, usually the unfertilized ones, became slightly flattened at
one pole while in the hypertonic solution. These eggs therefore
had a slightly smaller volume than their diameter would indicate.
In order to show the extent of the flattening a second measure-
ment was taken from the flattened pole to the opposite pole of the
egg. This second measurement is in every case shown in
parentheses. Another point also needs explanation. The ferti-
lized eggs measured in the second column had lost so much water
in the hypertonic solution that their cytoplasm was coagulated
and had in most cases begun to shrink away from the vitelline
membrane. This shrinkage was most pronounced in the third
and fourth eggs measured and this accounts for the very small
size of these eggs.

STUDIES IN. ARTIFICIAL PARTHENOGENESIS. 335

The table shows clearly that more water leaves the fertilized
than the unfertilized eggs. This can not be due to an increased
permeability to water for an increase of this sort could produce
no such effect. It must be due to a loss in the rigidity of the
plasma membrane. In order to make this relation clear, I shall
quote from p. 1 54 of the second paper of this series : ' ' The plasma-
membrane of the Arbacia egg is a protein gel. As such it pos-
sesses a certain degree of rigidity. Suppose a hypothetical
system completely surrounded by an extremely rigid semi-
permeable membrane. If such a system were placed in a con-
centrated solution no exosmosis could take place, for if the
membrane were perfectly rigid, there could be no removal of
solvent from the system without the production of a vacuum.
But the membrane would be subjected to a considerable pressure
which would tend to make it rearrange its particles in such a
fashion that the volume enclosed within it might be lessened.
Whereas an extremely rigid membrane would resist such forces
one with only a certain degree of rigidity would yield (in the
case of sufficient pressure) and exosmosis would be possible.
Thus osmosis in an enclosed system depends to some extent at
least on the rigidity of the confining membrane. These con-
clusions apply in some measure to the sea-urchin egg, for the
vitelline membrane possesses a slight degree of rigidity." They
apply even more directly to the Cumingia egg, for its vitelline
membrane, which is also its plasma .membrane, is stifter than
that of Arbacia. Any loss in the rigidity of this membrane favors
either endosmosis or exosmosis. That is why the fertilized eggs
of Cumingia take up more water from hypotonic solutions and
lose more water to hypertonic sloutions than do the unfertilized
eggs.

Osmotic change in unfertilized eggs is fairly rapid and no one
can deny that the plasma membrane is permeable to water. If
then R. S. Lillie's measurements are correct and water enters
fertilized eggs more rapidly than unfertilized eggs in hypotonic
solutions and leaves them more rapidly in hypertonic solutions
than it seems certain that the egg plasma membrane has not
markedly increased its permeability to dissolved substances as
a result of fertilization. For such an increase in permeability,

336 L. V. HEILBRUXX.

by diminishing the osmotic pressure, would slow osmotic inter-
change and if sufficiently great, would prevent it altogether.

All these points show clearly that the permeability theory of
fertilization and artificial parthenogenesis rests on rather doubtful
evidence. In Cumingia certainly, there are no facts which

support it.

DISCUSSION.

Bataillon ('12) in discussing the relations between artificial
parthenogenesis in amphibia and sea-urchins, states, "II n'y a
pas une parthenogenese experimental des Oursins et une des
Amphibiens. Ce sont des materiaux differents chez lesquels le
rythme des cineses est suspendu et pent etre retabli. S'il y a
une Biologic generale, les conditions de 1'arret ont quelque chose
de commun, et les conditions de la mise en branle doivent etre
comparable." Presumably this is true. The essential factors
underlying stimulation to development are very probably alike
for every sort of artificial parthenogenesis. It is certain, how-
ever, that the subsidiary features of the process are different in
each case. It is necessary therefore to study each egg indi-
vidually, to determine exactly its physical make-up, and to
attempt to discover what changes are significant in producing an
initiation of development. This is what I tried to do in the
case of Arbacia.

F. R. Lillie ('19) in referring to my explanation of the process
of cortical change in Arbacia points out that this explanation
"can hardly apply to other cases where the cortical changes
present a different morphological form." This is true and I
never intended that it should. As a matter of fact, I clearly
recognized that even in the one egg there were two distinct
types of cortical change which might be produced by spermato-
zoa. The explanation which F. R. Lillie cites was only advanced
to cover one of them. At the same time I offered a different
explanation for the other.

F. R. Lillie also objects to my considering cortical change as
"a mere epiphenomenon . . . the phenomenon of the primary
cortical change is too general to be treated in this fashion and its
character in different animal groups is too varied for it to be a
mere phenomenon of decrease of surface tension." I must point

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 337

out that I have always realized the importance of cortical change.
On page 183 of my 1915 paper I stated in a section devoted to
the significance of cortical change, that "cortical change, whether
it be membrane swelling or elevation, always results in the re-
moval of this obstacle (i.e., a stiff membrane]. The vitelline
membrane is either rendered soft by swelling or it is lifted away
from the egg surface and its place taken by the no doubt less
rigid hyaline layer." I showed moreover that "at least two
processes which play a part in normal development would be
greatly hindered if some kind of cortical change did not occur."

My results with Cumingia fully bear out this point of view.
Cortical change in Cumingia may take the form of a membrane
elevation dependent on a sharp decrease in surface tension. It
may more simply be just a membrane swelling. In both cases
the result is the same. The egg is freed from an obstacle which
impedes development. This in Cumingia as in Arbacia is the
restraining influence of a stiff vitelline membrane. In both eggs
the" same forces are involved in cortical change. The essential
features of the process and the effect on further development are
as closely alike as they could possibly be in the two cases.

It should be noted that cortical change in Cumingia is not
ordinarily followed by segmentation. As in Arbacia cell-division
in Cumingia is preceded by a sharp increase in the viscosity of the
cytoplasm. This can be demonstrated by tests with the centri-
fuge. A number of such tests were made and the relation
established beyond a doubt.

SUMMARY.

1. The Cumingia egg is surrounded by a stiff vitelline mem-
brane which tightly encloses the fluid cytoplasm.

2. A release from the restraint of this membrane is followed
by maturation.

3. Such a release from restraint can be accomplished in three
ways; by membrane elevation, by membrane swelling, or by the
removal or rupture of the membrane.

4. Substances which themselves have low surface tension pro-
duce a lowered surface tension of the membrane and this results
in its elevation from the egg surface.

338 L. V. HEILBRUNN.

5. Acids, alkalies, and certain salt solutions cause the vitelline
membrane to swell.

6. The membrane may be removed from the eggs by shaking,
or it may be caused to rupture by immersion in dilute sea-water.

7. All of the above mentioned treatments produce polar-body
formation. All of them free the egg from restraint.

8. Maturation in Cumingia is not dependent on an increase in
oxidations.

9. Cortical change in Cumingia produces no increase in permea-
bility either to dissolved substances' or to water.

10. The essential features of cortical change in Cumingia are
the same as those previously shown for Arbacia.

BIBLIOGRAPHY.
Bataillon, E.

'12 La parthenogenese des amphibians et la "fecondation chimique" de Loeb
(etude analytique). Ann. des Sci. Nat. Zool., Ser. 9, T. 16, pp. 249-307.
Gray, J.

'16 The Electrical Conductivity of Echinoderm Eggs, and its Bearing on the
Problems of Fertilization and Artificial Parthenogenesis. Phil. Trans.
Roy. Soc., Series B, Vol. 207, pp. 481-529.
Heilbrunn, L. V.

'13 Studies in Artificial Parthenogenesis. I. Membrane Elevation in the

Arbacia Egg. BIOL. BULL., Vol. 24, pp. 343-361.
'15 Studies in Artificial Parthenogenesis. II. Physical Changes in the Egg of

Arbacia. BIOL. BULL., Vol. 29, pp. 149-203.
'17 Maturation and Initiation of Development in Cumingia. Proc. Amer.

Assn. Anat. Anat., Rec., Vol. n, pp. 362-363.
Lillie, F. R.

'19 Problems of Fertilization. Chicago: The University of Chicago Press.
Lillie, R. S.

'16 Increase of Permeability to Water Following Normal and Artificial Activa-
tion in Sea-Urchin Eggs. Am. Journ. of Physiol., Vol. 40, pp. 249-66.
'17 The Conditions Determining the Rate of Entrance of Water into Fertilized
and Unfertilized Arbacia Eggs, and the General Relation of Permeability
to Activation. Ibid., Vol. 43, pp. 43-57.

'18 The Increase of Permeability to Water in Fertilized Sea-Urchin Eggs and
the Influence of Cyanide and Anesthetics upon this Change. Ibid., Vol.
45, pp. 406-30.
Loeb, J.

'13 Artificial Parthenogenesis and Fertilization. Chicago: The University of

Chicago Press.
Lyon, E. P. and Shackell, L. F.

'10 On the Increased Permeability of Sea-Urchin Eggs Following Fertilization.
Science, N. S., Vol. 32, pp. 249-251.

STUDIES IN ARTIFICIAL PARTHENOGENESIS. 339

McClendon, J. G.

'10 On the Dynamics of Cell Division. II. Changes in Permeability of
Developing Eggs to Electrolytes. Am. Journ. of Physiol., Vol. 27, pp.

240-75-
Morgan, T. H.

'10 Cytological Studies of Centrifuged Eggs. Journ. Exp. Zool., Vol. 9, pp.

593-655.
Runnstrom, J.

'n Untersuchungen iiber die Permeabilitat des Seeigeleies fur Farbstoffe. I.

Arkiv for Zoologi, Bd. 7, No. 13, pp. 1-17.
Wilson, E. B.
'01 Experimental Studies in Cytology. II. Some Phenomena of Fertilization

and Cell-division in Etherized Eggs. III. The Effect on Cleavage of

Artificial Obliteration of the ist Cleavage Furrow. Arch. f. Entwick-

lungsmech., Bd. 13, pp. 353~95-

A CONTRIBUTION TO THE LIFE HISTORY OF
AMCEBA PROTEUS LEIDY.1

LEON AUGUSTUS HAUSMAN, PH.D.,
ZOOLOGICAL LABORATORY, CORNELL UNIVERSITY.

It is only within the last few years that a connected life history
of Amceba proteus has begun to be formulated. This does not
seem strange, even in view of its common use in laboratories,
since the difficulties attendant upon the study of the life cycle
of such an elusive form, are numerous.

The isolated observations upon the various activities of Amceba,
the accounts of suppositions appendages, and the descriptions of
supposedly new species or varieties, are numerous. The fact
that so much observational material upon so many apparently
nearly related forms has accumulated, and that so many diverse
accounts of the habits of one and the same form have been
given, has led some investigators to suspect that perhaps much
of this material might be combined to throw light upon the life
history of possibly some one or two valid species in whose life
cycles some of the forms heretofore described might prove to be
merely developmental stages. This suspicion has been strength-
ened recently by the fact that some of the Rhizopoda closely
allied to the Amceba proteus, whose life cycles are now known,
show both widely varying forms and habits during different
stages in their growth and reproduction.

Investigations of the life history of Amceba proteus which have
hitherto been made make it possible to generalize, to some
extent, upon the course of some of the changes which take place
during its development. There seem to be several modes of
reproduction in this form, and several intermediate stages be-
tween young and adult before the life cycle is completed. In this
communication an attempt is made; first, to review briefly
what we know of the life history of Amoeba proteus; second, to

1 Grateful acknowledgment is here made of the aid of Professor H. D. Reed,
of the department of zoology, Cornell University.

340

LIFE HISTORY OF AMCEBA PROTEUS LEIDY. 34!

describe an apparently hitherto unknown method of reproduc-
tion; and third to record some observations upon the develop-
ment and growth of the forms produced by this new method.

The first observer to suggest that another method of reproduc-
tion other than simple fission might obtain among the Amceba
was Varter ('56) who recorded the presence of numerous fine
granules which he observed filling the bodies of some Amceba
radiosa. These granules he looked upon as fragments of the
nucleus, and since, when they escaped from the body, they
moved with a curious jerky motion, he inferred that they must
be provided with flagella, and termed them spermatozoids. The
observations extended no farther. Again in 1863 he observed
the granulation of the nucleus in Amceba princeps, noting that
the original nucleus divided several times and gave rise to many
smaller nuclei, which he termed, though apparently with meager
foundation, reproductive cells.

Wallich ('63) noted that when Amceba villosa "died," the
nucleus, which had divided several times, escaped from the body
surrounded by bits of protoplasm, of globular form. He observes
that the fate of these globules is unknown to him, but believing
that they might function in some reproductive process, termed
them scaroblasts. Referring to the many different species of
Amceba which appeared in his cultures, he writes (with reference
particularly to Amceba princeps, diffluens, and radiosa} : " It will,
I think, eventually be found that these are mere transitory
phases of one and the same species." Later, he observes with
greater conviction: "... though not prepared to affirm that
the whole of the varieties of Amcebce are reducible to a single,
primary, specific type, I candidly confess that the balance of the
evidence appears to me to point to such a conclusion, and to
indicate that the divergence in the form and outward characters
may be wholly dependent on the local and even temporary
conditions of the medium in which the young animal happens
to make its appearance in the world."

It is to Scheel ('99) that we owe our first complete knowledge
of a type of reproduction different from the familiar method of
binary fission. In 1899 he described a process of reproduction
in Amceba proteus which he termed schizogony, after a somewhat

342 LEON AUGUSTUS HAUSMAN.

similar mode of multiplication occurring among the Sporozoa.
As it occurs in Amoeba protects its course is as follows: The
individual encysts; the nucleus divides into several smaller
nuclei; these migrate to the periphery of the cell; the protoplasm
of the cell divides into as many equal portions as there are
nuclei; the cyst wall ruptures; and the nucleus-containing bits
of protoplasm emerge, each a complete Amoeba and an epitome
of its parent. In analogy with the spores of the Sporozoa these
were given the name, pseudopodiospores. More recently Calkins
('05) has noted a method of reproduction in which the nucleus
produced by repeated division, gametic nuclei, whose fusion
resulted in other nuclei which became the nuclei of the smaller
individuals after the manner of the pseudopodiospores just men-
tioned. The young Amcebce soon assumed the form hitherto
called Amoeba radiosa; passed through this stage and became
the common Amoeba proteus.

Another apparent mode of reproduction was observed by
Metcalf ('10). In this method the observer is of the opinion
that the sequence of events is: Mature individuals produce
globular masses of protoplasm, which are termed gemmules.
These develop flagella and assume a cercomonadoid form; later
they fuse by two and two, lose the flagella, and develop into adult
Amceboz. It is stated that possibly the life cycle of Amoeba
proteus may require a year or more for its completion, and may
exhibit during its course three, or even more, modes of repro-
duction.

Schepotieff ('08, '09) states: "All of these examples suggest
that in the case of the Amoeba [proteus] the developmental cycle
may be completed in very different ways."

It has been shown, as we have said, that Amoeba radiosa,
formerly regarded as a distinct species, is merely a stage in the
development of Amoeba proteus. It may be that this also is true
of such recognized species of Amoeba as villosa, princeps, diffluens,
etc. It is suggested elsewhere in this paper that the species
known as Amoeba guttula is but a developmental stage in the
cycle of proteus. Vahlkampf ('05) has shown that Amoeba Umax,
at least, is to be regarded as a distinct species.

LIFE HISTORY OF AMCEBA PROTEUS LEIDY. 343

PREPARATION OF CULTURES.

The material from which the Amcebce herein described were
reared was obtained from a pool in a cattail marsh, in about
two feet of water among decaying lily pads and ceratophyHum.
The thin glutinous deposit upon the bottom and especially on
the lily pads which had fallen to the bottom was found to be rich
in Amoeba proteus. In the laboratory the material was dis-
tributed into several battery jar aquaria after having been
filtered through cheese cloth to remove the larger creatures.

An immediate examination of the material showed that beside
the active proteus, radiosa, gultula, and Umax, there were many
encysted protozoa. One encysted form which appeared in large
numbers, and which confusingly resembles some of the smaller
encysted Amcebce was Vorticella microstoma (Fig. 4). Under
ordinary magnification its distinguishing feature, the crescentric
nucleus, is not visible. With the 1.8 mm. objective it becomes
faintly discernible, but it is best seen, and the cyst is indubitably
identified apparently only when stained. Methyl green and
iodine gave the best results.1

After some weeks had elapsed proteus and radiosa appeared in
large numbers. The material was now transferred to a dozen
small petri dishes and kept in a constant temperature of about 75
degrees Fahr. After a space of a fortnight there was begun the
transfer of inoculation of Amoeba proteus to 4 cm. stender dishes,
furnished with straw infusion or oak leaf infusion, and free from
all protozoan forms. The infusions were prepared by boiling the
straw or leaves for several hours, and decanting off the dark
brown liquor, to be diluted to the optimum strength. A slimy
scum formed upon the surface of the infusions after a few days
time, which when stirred up and caused to sink to the bottom
furnished a nutritive substance upon which the Amcebce throve.
From the stender dishes individual Amcebce were removed from
time to time and kept in shallow cells sunk in slides ot unusual
thickness. The slides employed were furnished with a device

1 Methyl green stain: Saturated alcoholic solution methyl green, 3 parts;
2 per cent, aqueous solution acetic acid, i part; water 3 parts. Iodine stain:
saturated alcoholic solution iodine, i part; water, 2 parts. For various protozoan
stains, see Hausman, "Fresh Water and Marine Gymnostominan Infusoria" (in
press).

344 LEON AUGUSTUS HAUSMAN.

for supplying water to take the place of that carried off by
evaporation (Fig. 3).

The method of removing individual Amoeba from stender
dishes was as follows, and can be used with success for the isola-
tion of any of the larger protozoan forms : A drop of water con-
taining the Amozbtz was placed upon an ordinary slide, and,
uncovered, searched with the 16 mm. objective and 8 X eye-
piece. When Amoeba were located they were removed by means
of what is termed an isolation pipette (Fig. 2). A long rubber
tube attached at one end to a glass tube drawn out to a very
fine tip, and at the other to a small compression bulb, enables
one to select and withdraw very minute objects with considerable
precision. Both the stender dishes and growing slides were kept
at a temperature of 80 degrees Fahr. in a large aquarium jar
heated by a carbon filament lamp placed in the bottom, the
current being controlled by a rheostat. By means of this simple
device unvarying temperatures could be maintained for any
desired length of time. The advantage of such a culture oven is
that light is admitted freely on all sides (Fig. i).

FORMATION OF APSEUDOPODIOSPORES.

One of the stender dishes which had been inoculated with
adult Amoeba protens proved very productive, the individuals
increasing rapidly, apparently by means of binary fission, since
many were observed in process of division. The bodies of the
largest individuals became filled with minute bodies, which upon
staining seemed to be nuclei. The individuals bearing these were
extremely sluggish. The pseudopodia were short ; exhibited very
little movement, changed their shape but slowly, and upon the
functional posterior of the body absent altogether (Fig. 5).

After an interval of four days had elapsed a reexamination of
some of the sediment from the same stender dish was made and
it was found that the numbers of the large proteus had appre-
ciably decreased and that their places were taken by a multitude
of very small amoeboid forms, of an average diameter of 4 or 5
microns. The majority of these exhibited feeble movements of
an amoeboid kind. Some were globular, some possessed an
irregular body outline, though definite pseudopodia were lacking

LIFE HISTORY OF AMCEBA PROTEUS LEIDY. 345

(Fig. 10). These, I was later led to believe, originated from the
bodies of the larger multinucleated proteus. Several of these
latter individuals while under observation were observed, to give
rise to smaller individuals of the form already described. The
process was as follows: the animal, which had been moving
slowly and apparently without much vigor, gradually came to
rest with the pseudopodia upon the then functional anterior
portion of the body lobate with slightly pointed tips (Fig. 6).
The posterior portion became semi-globular and towards this the
greater number of the minute suppositive nuclei within the
body plasm migrated. After an interval of about seven to ten
minutes the ectoplasm surrounding this globular posterior ex-
tremity appeared to disintegrate and from the interior there
floated forth several hyaline globules about 4 or 5 microns in
diameter. These were followed, after a few minutes by several
others, and then a constant outflow began that continued until
upwards of thirty of the hyaline spheroids had been extruded
(Fig. 7). These were apparently identical with the forms which
had made their appearance in the culture. Since these are judged
not to be fundamentally different from the pseudopodiospores,
of Scheel, but since they exhibited no lobose pseudopodia, they
are here called for convenience, apseudopodios pores. Some of
these, immediately after extrusion began to move slowly, bulging
the very thin ectoplasm at several points, yet without forming
any definite pseudopodia. Others, synchronous with their
emergence, disintegrated. Still others floated away without
exhibiting any signs of motion (Fig. 8). It may be that these
were gametic forms, and fused before further development.

Towards the time of the completion of apseudopodiospore
ejection, the parent Amcebce usually gave signs of renewed ac-
tivity, elongated perceptibly, and then began apparently to make
efforts to move away (Fig. 9). However, after an interval of
from twenty minutes to half an hour, they disintegrated. . Some
disintegrated at once, leaving the apseudopodiospores behind,
but usually they were extruded from the globular posterior
portion of the animal, the anterior part retaining its integrity.
It appeared as though not all of the minute nuclei were used up
at the time of the production of one "litter" of apseudopodio-

346 LEON AUGUSTUS HAUSMAN.

spores, and that some of them (those in the anterior portion) were
lost.

M. Popoff ('u) has recorded a type of reproduction in Amoeba
minuta, a marine species, similar to the one described above,
with the difference, however, that the resulting spores were
gametes, which later fused. Schmidt ('13) likewise, described
this same sort of reproductive activity as occurring in another
marine species, Amasba aquitalis.

The apseudopodiospores however, at least the majority whose
development was watched, were not gametes.

The smallest of the young Amcebce (as we shall call them) those
which have just been separated from the parent body are about
3 to 5 microns in diameter and are extremely sluggish. The
body is sub-globular and changes its outline but little during
the very slow movement. No pseudopodia are developed. The
protoplasm is clear, and contains a few small, angular, trans-
parent granules. No contractile vacuole could be seen, nor
where the creatures observed to feed (Fig. 10).

With growth comes an increase in activity and a progressively
greater irregularity of the body outline (Fig. n), until at length
true lobate pseudopodia make their appearance (Fig. 12). The
number of granules within the body increases, food is taken by
engulfing, and the protoplasm assumes a grayish hue. This color
may be due both to the number of particles within the endoplasm
and to the augmentation of its volume.

At the time of the appearance of the true pseudopodia the
body is unsymmetrical, but as growth proceeds are more or less
radiate arrangement of the pseudopodia takes place, at first not
well defined, but becoming more and more pronounced with the
increase in size (Fig. 13).

The pseudopodia now become more extended, and tend to
develop more acuminate tips. With increasing length and sharp-
ness the pseudopods seem to become more rigid, and spine like,
and the granules migrate from them into the more globular
central mass of the body, leaving them clear (Fig. 14).

During the time when the young Amcebce are passing from
the apseudopodia stage to the radiosa stage they confusingly
resemble, if indeed they are not exactly similar to the species

LIFE HISTORY OF AMCEBA PROTEUS LEIDY. 347

known as Amoeba guttula (Fig. n), and it is suggested that
possibly this creature hitherto accorded specific rank, may be
merely a developmental stage in the cycle of proteus. That
Amoeba radiosa was named from the radiosa stage of Amoeba
proteus has been indicated. The radiosa stage which we have
observed in this developmental series may be similar.

A new type of modification now takes place, as has been said,
when the pseudopodia become longer and more spinous. During
this stage, in respect to size, configuration, and characteristic
spineous, immobile, hyaline, ray-like pseudopodia, the creatures
are apparently indistinguishable from the species known as
Dactylosph&rium radiosum. Hence we shall term this stage the
Dactylosphcerium stage (Fig. 15).

The genus Dactylosphcerium was established by Hertwig, and
Lesser ('74) to receive the organism which they described as
Dactylosphcerium vitreum. The species now known as Dactylo-
sphcerium radiosum, however, was not referred originally to that
genus, but to the genus Amceba, as Amoeba radiosa, by Ehrenberg
('30). It was transferred to the genus Dactylosphcerium by
Biitschli ('80), who however erroneously called it Dactylosphceria.
Cash ('05) says of it: "The body consists of granular protoplasm
and when all the pseudopodia are withdrawn it may become
spherical or bluntly lobed ; or it may assume an active amoeboid
phase, when it is hardly, if at all,1 to be distinguished from the
smaller forms of Amceba proteus." It occurs in marshes and
pools, "less common than Amoeba proteus, with which it is usually
associated." 2

It was observed that not all of the young Amcebce acquired this
Dactylosphcerium-\ike form. Some became small radiosa, passed
on to large radiosa, and thence on to the proteus form. It seemed
to be the usual thing, however, for the majority to assume the
form of Dactylosphcerium, and the suggestion is made that perhaps
Dactylosphcerium radiosum, like other former species, may be
forced to relinquish its specific distinction.

From the bodies of those individuals which had assumed the
Dactylosphcerium form, lobate pseudopodia were occasionally pro-
truded, and at length the creature became almost globular, and

- The italics are the author's.

348 LEON AUGUSTUS HAUSMAN.

then proceeded to take on a form similar to the adult proteus.
Such individuals were isolated, and soon grew in size to adult
form, and were, apparently, Amceba proteus. Fission was ob-
served to take place among some of these, but no further repro-
ductive activities were noted. The entire sequence of events
which have been enumerated took place within three months,
the cultures being kept in the glass oven already described, at an
average temperature of 80 degrees Fahr.

Some of the individuals of nearly adult proportions developed
long, whip-like, and flexible pseudopodia, often more like long
threads (Fig. 16). These were clear, even the stouter ones being
devoid of granules. Gruber ('n) reported a similar type of
pseudopodium (there was but one in this case) which was sent
out from the body and moved about as if an organ of exploration.

Some of the most bizarre forms occur during the transition
period between the smallest form and the Dactylosphcerium stage.
Fig. 14 represents forms, all but one of which were taken from
the same slide.

In Fig. 1 8 an attempt has been made to group the various
forms observed in a tentative cycle of development. The nine
arbitrary stages appear to occur in the following order:

1. Adult stage.

2. Division of the nucleus, and migration of the nuclei to the

posterior extremity of the body.

3. Escape of apseudopodiospores.

4. Amceba guttula stage.

5. Small Amceba proteus stage.

6. Amceba radiosa stage.

7. Dactylosphcerium radiosum stage.

8. Resumption of amoeboid form.

9. Growth to adult size.

BIBLIOGRAPHY.
Biitschli, O.

'80 Protozoa, in Bronn, H. G. Klassen und Ordnungen des Thier Reichs,

parts 8 and 9.
Calkins, G. N.

'05 Evidences of a Sexual Cycle in the Life History of Amoeba proteus. Archiv

fur Protistenkunde, Vol. 5, p. i.
'07 Fertilization of Amtzba proteus. BIOL. BULL., Woods Hole, Mass., Vol.

13. P- 219.

LIFE HISTORY OF AMOEBA PROTEUS LEIDY. 349

Carter, H. J.

'56 Notes on the Freshwater Infusoria of the Island of Bombay. Ann. &

Mag. of Nat. Hist., Vol. 18, p. 115.
'63 On Amoeba princeps and Its Reproductive Cells. Ann. & Mag. of Nat

Hist., Vol. 12, p. 30.
Cash, J.

'05 British Rhizopoda and Hdiozoa. London.
Ehrenberg, H.

'30 Beitrage zur Kenntniss der Organization der Infusorien, etc. Abhl. der

Konigl. Akad. der Wiss. Berlin, p. 30.
Griiber.

'n Ueber Einartige Korperformen von Amoeba proteus. Archiv. fur Protisten-

kunde, Vol. 23, p. 253.
Hertwig & Lesser.

'74 Ueber Rhizopoden Denselben Nahstehende Organismen. Archiv. fur

Mikr. Anat., Vol. 10, suppl., p. 35.
Metcalf, M.

'10 Studies upon Amoeba. Jour. Ex. Zool., Vol. 9, p. 311.
Penard.

'02 Faune Rhizopodique du Bassin du Leman, Geneva.
Popoff, M.

'n Ueber der Entwicklungscyclus von Amoeba minuta. Archiv fur Protisten-

kunde, Vol. 22, p. 179.
Schaudinn, F.

'03 Untersuchungen ueber die Fortpflanzen einiger Rhizopoden. Arbeit aus

dem Kaiserl. Gesundheitsamte, Vol. 19, p. 547.
Scheel, C.

'99 Beitrage zur Fortpflanzung der Amceben. Festschrift zur Siebensigsten

Geburtstag von Carl von Kupffer, Jena, p. 569.
Schepotieff, A.

'og-'io Amcebenstudien. Zool. Jahrbucher, Abt. fur Anat., Vol. 29, p. 485.
Schmidt, H.

'13 Faunistische und Entwicklungsgeschichtliche^Studien an Sarcodinen der

Umgegend von Bonn. Archiv fur Protistenkunde, Vol. 29, p. 203.
Wallich, C. G.

'63 (a) On an Undescribed Indigenous Form of Amoeba. Ann. & Mag. of
Nat. Hist., Vol. n, p. 287.

(b) Further Observations on A. villosa and other Indigenous Rhizopods.
Ann. & Mag. of Nat. Hist., Vol. n, p. 434.
Vahlkampf, E.

'05 Beitrage zur biologic und Entwicklungsgeschichte von Amoeba Umax.

Archiv fur Protistenkunde, Vol. 5, p. 167.
Vejdovsky, F.

'80 Ueber die Rhizopoden der Brunnewasser Prags. Sitsber. der bohm. Gesell.
der Wiss. Prag., p. 136.

35° LEON AUGUSTUS HAUSMAN.

PLATE I.

FIG. i. Aquarium jar culture oven, a, thermometer; b, cardboard cover;
c, aquarium jar; d, cultures; e, iron tripod; /, lamp; g, copper wire rack for lamp.

FIG. 2. Isolation pipette.

FIG. 3. Growing slide, a, sponge, for absorbing excess water; b, cover glass;
c, glass tube carrying thread, as siphon; d, lower half of glass vial, cemented to
the slide with balsam, as reservoir.

FIG. 4. Vorticella microstoma, encycted.

FIG. 5. Unusually large Amoeba proteus, with body filled with minute nuclei.

FIG. 6. Proteus, resting before extrusion of the apseudopodiospores.

FIG. 7. Extrusion of the apseudopodiospores.

FIG. 8. Immobile, globular apseudopodiospores.

FIG. 9. Proteus after extrusion of the apseudopodiospores.

FIG. 10. Active apseudopodiospores.

FIG. ii. Apseudopodiospores during the stage when they resemble Amceba
gultula.

FIG. 12. Small proteus stage.

FIG. 13. Amoeba radiosa stage.

FIG. 14. Apparent transitional forms between the radiosa and the Dactylo-
spha:rium stages.

BIOLOGICAL BULLETIN, VOL. XXXVlll.

PLATE I.

L. A. HAU8MAN.

352 LEON AUGUSTUS HAUSMAN.

PLATE II.

FIG. 15. Dactylosphcerium radiosum stage.

FIG. 16. Amoeba proleus in the adult stage, exhibiting long attenuated pseudo-
podia.

FIG. 17. Adult normal Amoeba proleus (stained).

FIG. 18. Tentative Cycle of Development, i. Adult stage. 2. Division of
nucleus; migration of nuclei to posterior. 3. Extrusion of the apseudopodiospores.
4. Amceba guttula stage. 5. Small A mxba proteus stage. 6. Amoeba radiosa stage.
7. Dactylosph&rium radiosum stage. 8. Resumption of amoeboid form; retraction
of pseudopodia. 9. Growth to adult.

BIOLOGICAL BULLETIN, VOL. XXXVIII.

PLATE

17

16

18

L. A. HAUSMAN.

Vol. XXXVIII. June, 1920. No. 6

BIOLOGICAL BULLETIN

THE AXIAL GRADIENTS IN HYDROZOA. III. EX-
PERIMENTS ON THE GRADIENT
OF TUBULARIA.

LIBBIE H. HYMAN,
HULL ZOOLOGICAL LABORATORY, UNIVERSITY OF CHICAGO.

A. INTRODUCTION.

In a recent paper Banus ('18) states that there is no difference
between the time of regeneration of oral hydranths on apical and
basal pieces of the stem of Tnbularia. The data presented by
Banus apparently support this conclusion. Such data are, how-
ever, of no significance unless complete information is given as
to the manner in which they are obtained. Perusal of Banus's
paper reveals the fact that absolutely no information is imparted
concerning the conditions under which the experiments were
performed or the manner of handling the material. Those who
have worked on the physiology of the lower forms are well aware
that experimental results can be readily controlled and modified
by conditions. It is, therefore, impossible for the impartial
mind to accept the validity of Banus's conclusions, until further
information concerning his experiments shall be forthcoming.
Grave doubt is cast upon the correctness of Banus's statements
by the fact, completely ignored by him, that experiments of this
kind had already been performed several times, with results
contrary to his. In addition to these omissions, Banus has
made a number of exaggerated and misleading statements.

Banus begins by saying that Child "assumes the existence of
metabolic gradients in a great number of species of animals and
plants and on this assumption he builds a theory of individual-
ity." In view of the great mass of data which has been presented
concerning metabolic gradients and the extent to which these

353

354 LIB3IE H. HYMAN.

data have been checked by several different methods it seems
hardly scientific to dismiss the matter under the word "assumes."
Gradients are not assumptions; they are facts. It is legitimate,
of course, for any one to question and criticize the interpretation
of these facts, and desirable that other possible explanations of
them should be suggested ; but to ignore such facts by designating
them as assumptions is not the way to arrive at scientific truth.
The nature of the axial gradients has been so thoroughly and
frequently discussed in numerous papers from this laboratory
that presentation of the subject here seems to me superfluous.

Banus next remarks that the hydroid Tubularia is extensively
used to support Child's conceptions. It is scarcely necessary to
point out to the zoological world the exaggeration conveyed by
this statement, as it is well known that other ccelenterates have
been used as extensively and other lower forms much more
extensively in accumulating the experimental evidence upon
which those conceptions rest.

We are next informed that Child "has made no measurements
of the rate of metabolism of different regions of the stem of
Tubularia." We would be pleased to carry out such experiments
if Banus would kindly suggest a suitable method. The matter
would be relatively simple were it not for the fact that the
relative proportions of perisarc and cosnosarc vary in different
regions of the stem of Tubularia. It is therefore difficult or
impossible to determine the amount of living material in portions
of the stem and impossible to establish any basis for comparison
of the metabolism of different regions. Naked hydroids, such
as Corymorpha, would be required for experiments of this kind.
The metabolic rate of different regions of the first zooid of
Planaria dorotocephala has been determined and has been found
to accord with the metabolic gradient conception.

Banus then proceeds to discuss the regional differences in
rate of regeneration of Tubularia. He says that Child "as-
sumes" the existence of such differences. In his summary he
states that "the rate of regeneration of the oral hydranth of an
apical piece is on the average identical with the rate of re-
generation of the oral hydranth of the basal piece"; and further
that "there is no evidence of the existence of level or regional

THE AXIAL GRADIENTS IN HYDROZOA.

355

differences in the stem of Tubularia." Such statements as these
display an unpardonable ignorance of the literature dealing
with the regeneration of Tubularia, This form has been in-
vestigated by a number of well-known zoologists, and twenty
years ago the regional differences which Banus denies were
demonstrated to exist.

The first experiments on the regeneration of Tubularia were
those of Loeb ('91 and '92), performed at Naples. In the first
of these publications (on p. 15) he states that he took very long
stems, cut off the roots and polyps, and then divided the remain-
ing portion in half. No difference wras observed in the rate of
regeneration of the oral hydranths on the apical and basal pieces.

FIG. i. Diagram of Tubularia to illustrate method of cutting apical and basa
pieces of the stem of equal length; employed in experiments recorded in Tables
III., IV., V., VII., IX., and X. a, apical; b, basal.

The following year he repeated the experiment and this time he
observed that in one experiment (he does not state how many
experiments were performed) the oral hydranths emerged about
twenty-four hours earlier on the apical than on the basal pieces.
Owing to the meager details which are furnished regarding these
experiments, it is impossible to determine why the same investi-
gator working on the same material should at one time obtain

356 LIBBIE H. HYMAN.

one result and on another occasion the contrary result. Loeb
has chosen to disregard the one contrary experiment since later
("Organism as a Whole," p. 171), he again asserts, in spite of all
the evidence at that time available to the contrary, that apical
and basal pieces of the stem of Tubularia regenerate hydranths
simultaneously. This statement has never been confirmed, ex-
cept by Banus; every other zoologist who has worked upon the
matter has found the contrary to be true. In 1899, Driesch
working also at Naples expressed himself as fully convinced
that Loeb was mistaken. He found many evidences of regional
differences in Tubularia. He observed that the length of the
primordium is greater the nearer the piece is to the original distal
end of the stem; that in very small pieces, the more apical
pieces produce larger hydranths and tend to give rise to distal
structures only, while the basal pieces produce smaller hydranths
and proximal structures; and that when long pieces of the stem
are cut in half, the apical halves give rise to oral hydranths
earlier than the basal halves. Driesch, therefore, as he em-
phatically stated in this paper, disagreed with Loeb on this
point. In Table X, p. 131, Driesch gives a record of thirty
pieces in which the apical halves regenerated oral hydranths
one to twenty-three hours earlier than the basal halves in twenty-
five cases and simultaneously with them in but five cases.
Table XI, p. 132, presents similar data. These statements of
Driesch were verified by Morgan ('01, '05, 'o6a, '08), and by
Morgan and Stevens ('04). Thus Morgan says ('05, p. 496),—
"the rate of both oral and aboral development is determined
by the level at which the end lies." Again in 1906, p. 497, he
states: "It has been shown in Tubularia that the time required
for the formation of a new hydranth depends on the distance of
the cut surface from the old hydranth. The nearer the cut
surface to the oral end the quicker the regeneration. The same
law also holds for the development of the aboral hydranth from
the aboral end of a piece." These statements were reiterated in
1908, p. 157 — physiologically polarity is "shown in the more
rapid regeneration of the cut surfaces the nearer they are to the
distal end."

Child ('07) again performed the experiment in question and

THE AXIAL GRADIENTS IN HYDROZOA.

357

agreed with the results obtained by Driesch, Morgan, and Stevens.
These experiments of Child's are the only ones considered by
Banus and the false impression is thereby conveyed that no one
but Child had ever experimented upon the matter. Banus has
criticized these experiments of Child's on the grounds that they
are not extensive and that the differences recorded are in some
cases so slight that they may be due to experimental error. It

A

B

FIG. 2. Diagram to show method of cutting unequal pieces for the experiments
given in Table VI. 2A, apical pieces half as long as basal pieces; 2B, apical
pieces twice as long as basal pieces.

is true that no very extensive series of experiments were carried
out because Child felt that the point had already been settled
by the work of Driesch and Morgan and that any further experi-
mentation was superfluous. Such observations as he made were
therefore incidental to other matters. It is also true, as Banus
says, that the differences obtained were not very large, but since
they always vary in the same direction, the results cannot be
due, as claimed by Banus, to experimental variation. In Table

358 LIBBIE H. HYMAN.

I., p. 2, of Child's paper we find that of 24 pairs of pieces, the
distal oral hydranths emerged first in 14 cases, at the same time
as the proximal oral hydranths in 5 cases, and later than the
latter in 5 cases. A result which is due to experimental error
should vary equally in both directions. It should also be stated
that part of the pieces given in this table were regenerating in
modified sea-water and not under normal conditions. In regard
to Tables II. and III., p. 5, Banus has misrepresented the facts.
In these tables the differences between the times of emergence
of distal and proximal oral hydranths are less than in the pre-
ceding table, although they still vary in the same direction;
but it is distinctly stated by Child in the text that the difference
is decreased owing to the smaller size of the pieces, and he further
shows that with still greater reduction in the length of the pieces,
the proximal oral hydranths will emerge first. In short, these
experiments were directed towards demonstrating the effect of
reduced length on the time of emergence of the oral hydranths,
a fact which Banus in quoting them omits to mention.

After Banus's paper appeared the experiment was repeated at
Woods Hole in the summer of 1918 by Dr. W. C. Allee. Dr.
Alice was entirely unable to agree with Banus's statements, but
found, on the contrary, that the oral hydranths arise earlier on
apical than on basal pieces of the stem of Tubularia. He com-
municated this result to Professor Child and other members of
this laboratory and also showed his experiments to a number of
people at Woods Hole. In the summer of 1919, Dr. Allee
assigned the experiment to his class in Invertebrate Zoology at
Woods Hole. Twenty-seven sets of experiments were performed
by the students, each set consisting of from two to eight pairs of
pieces. Twenty-four hours after cutting, the apical halves (as
indicated by the red color of the regenerating ends) were in
advance in twenty-five of the sets, the basal halves in advance in
one set, and in the other, there was no difference. Of the 112
pairs of pieces cut, 95 of the apical pieces survived, and 71 of the
basal pieces. After forty-eight hours, 52 or 55 per cent, of the
apical pieces had produced hydranths, while this had occurred
in only 15, or 20 per cent, of the basal pieces. It should be
stated that the material was not in first-class condition at the

THE AXIAL GRADIENTS IN HYDROZOA.

359

time, and, as I shall show, the difference between the time of
regeneration of apical and basal pieces is reduced under such
circumstances. Professor Child and I are greatly indebted to
Dr. Allee for his interest in the matter, and for his kindness in
putting through the experiments.

In view, therefore, of the overwhelming preponderance of the
evidence already at hand in support of the existence of the

FIGS. 3 AND 4. Diagrams to show method of cutting pieces for experiments
given in Table VIII. Figure 3, method for all experiments in Table VIII., except
number 36; figure 4, method used for experiment 36.

regional differences along the axis of Tubularia which are denied
by Banus, further experimentation seems superfluous. Under
the circumstances, however, a repetition of the experiment has
been deemed necessary by various members of this laboratory
as an answer to Banus's paper. I therefore undertook to repeat
his work and for this purpose made trips to Woods Hole in June,
and in December, 1919. The results were identical at both

360 LIBBIE H. HYMAN.

seasons of the year and were completely at variance with Banus's
statements. I was unable to verify any of his results. In
numerous experiments conducted for this purpose, the apical
halves of Tubularia stems regenerated markedly faster than the
basal halves of the same stems. I further believe that I dis-
covered the cause of Banus's peculiar results. My results are
presented in detail in the present paper.

Not only are the researches just enumerated opposed to
Banus's statements but a large number of other facts concerning
the regeneration of Tubidaria clearly point to the existence of
axial differences in metabolic rate in this form. Thus all of the
facts collected by Driesch, Morgan, and Child concerning the
phenomena of "polarity" in Tubularia are entirely irreconcilable
with the view point of Banus and Loeb. If there is no axial
difference along the stem of Tubularia, why should the apical
end of a piece produce a hydranth and the basal end a stolon,
or if a hydranth, only later than the apical end? Why do
heteromorphic hydranths arise simultaneously on the two ends
of very short pieces while on long pieces the aboral hydranth is
delayed? This question has received no adequate answer except
that based on the axial gradient conception; in short pieces,
there is practically no gradient, and hence each end of the piece
begins to produce a hydranth at the same or nearly the same time;
while, in long pieces, the apical end by virtue of its higher meta-
bolic rate gets the start in hydranth formation and hence gains
control of the stem for a certain distance, thus inhibiting the
formation of the aboral hydranth. Why, as shown inde-
pendently by Driesch, Morgan, and Child, in an axial series of
very short pieces, do the apical pieces produce larger distal
structures with much reduced or absent proximal structures
while the more basal pieces give rise to smaller distal structures
and larger proximal parts? Longer pieces from the distal region
thus resemble shorter pieces from the proximal region in the
structures which they produce, owing to the pronounced tendency
of the distal pieces to use up their substance in the formation of
distal structures only. W'hy is the primordium of the oral
hydranth larger in apical than in basal pieces and the emerged
hydranth likewise larger? These and numerous similar facts

THE AXIAL GRADIENTS IN HYDROZOA.

have been repeatedly ascertained by several investigators and
are totally inexplicable on the point of view maintained by Loeb
and Banus that there are no regional differences in the stem of
Tubularia.

There is one further statement made by Banus to which we
must take exception. This is the assertion on p. 266 and again
on p. 273 that regional differences in rate of regeneration con-
stitute the "actual basis" for the axial gradient conception as

FIG. 5. Method of cutting pieces usually employed by Banus and used for the

experiments recorded in Table XI.

applied to Tubularia. They are part of the basis, but not the
entire basis, as Banus implies. Axial differences in metabolic
rate along the stems of Tubularia had been demonstrated at the
time when Banus published his paper by two other methods—
the differences in susceptibility to potassium cyanide and other
substances, and the differences in electric potential. The results
yielded by both of these methods are again considered in this
paper.

In view of the fact that Banus's results are in total disagree-
ment with a considerable number of researches, a critical examina-

362 LIBBIE H. HYMAX.

tion of his method of procedure is necessitated. But, as has
already been pointed out, this is impossible because no description
is given by Banus concerning his material, his methods, or his
experimental conditions. We need to know the season of the
year, the temperature, the vegetative condition of the material,
the length of time material was kept in the laboratory before
being used, the presence or absence of lateral buds on the stems
employed (since buds mark the limits of the individual), and
particularly the level from which the pieces were taken, with
reference to the original hydranth. We have repeatedly pointed
out that metabolic gradients are not fixed and static things, but
markedly dynamic and labile, and particularly in the lower forms,
they may result from external conditions and may be readily
modified and altered by conditions. The failure of Banus to
describe or consider the various factors mentioned, any one of
which might alter the experimental result, is evidence that he
really does not understand the metabolic gradient conception
and has not interested himself in understanding it. This is
further shown by certain remarks made in his paper such as for
example the naive statement on p. 268 that the pieces were
"long enough to show a marked difference according to Child's
opinion," whereas in fact "according to Child's opinion," and
as even a hasty perusal of the work put out from this laboratory
would show, the axial differences are most clearly marked in
most respects in relatively short pieces.

In order, therefore, to obtain any information concerning the
experiments of Banus, it has been necessary for me to com-
municate with him. Mr. Banus replied to the first letter which
I wrote to him, but did not reply to two others requesting further
details. It is therefore not possible for me to furnish all of the
details necessary for a correct evaluation of these experiments
but the information I was able to obtain is sufficiently astonishing.
Banus states that his experiments were performed in New York
City in November and in Woods Hole in December. Since
most of the other researches on Tubularia were carried out in
the summer season, it was at first thought that seasonal differ-
ences in the vegetative condition of Tubularia might account for
Banus's results. I found, however, that Tubularia is in prac-

THE AXIAL GRADIENTS IN HYDROZOA. 363

tically the same condition at Woods Hole in June and in
December and yields identical results at these two seasons of the
year. Banus further states that the "temperature during re-
generation was in New York 22° C., and in Woods Hole it was
about 15° C." Upon examination of Banus's tables, one finds
that the time interval between section and emergence of hy-
dranths is much too great for these temperatures. At a tem-
perature of 22° C., apical pieces of Tubularia produce hydranths
as early as 36 hours after section and the great majority of such
pieces will have completed regeneration in 48-60 hours, yet in no
case do any of the pieces recorded in Banus's tables regenerate in
less than 53 hours and the majority of them require more than
60 hours. Even at a temperature of 15° C., the times given by
Banus are surprisingly long, since I found that at 12° C., the
majority of the apical pieces will regenerate within 70-80 hours
(see Table IV.). It is therefore evident that in Banus's experi-
ments some factor is acting to delay the time of regeneration
of the apical pieces, whose regeneration precedes by an average
of i o to 12 hours, according to my findings, the regeneration of
the basal pieces.

The cause of the delayed regeneration in Banus's experiments
lies in all probability in his method of cutting the apical pieces.
In reply to my inquiry concerning the level of the stem from which
he cut the pieces, Banus made the following statement: "the
most distal cut was usually made as near as possible to the
hydranth without including any part of it." Such a method of
procedure explains the aberrant results obtained by Banus. It
has long been known that the short stalk below the hydranth of
Tubularia is incapable of regeneration. When the apical pieces
are cut in the manner described by Banus, this stalk forms the
distal end of such pieces. It dies and disintegrates, thereby
markedly delaying the regeneration and time of emergence of
oral hydranths on these pieces. I shall present evidence in
this paper (see Table XI.) that when the apical pieces are
cut in such a way that their distal ends are just below the
base of the original hydranth, the time of emergence of the
oral hydranth is greatly delayed and falls behind that of the
basal pieces. Hence when such a procedure is followed, all

364 LIBBIE H. HYMAN.

sorts of irregular results are obtained and it is not surprising
that under such circumstances, the experimenter would be lead
to question the existence of axial differences along the stem of
Tubularia. When the correct method of cutting the apical
pieces is observed, namely, when the hydranth, its stalk, and
the first millimeter or two of the stem are discarded, then there is
not the slightest question that the apical halves of stems prepared
in this manner give rise to oral hydranths very much in advance
of the basal halves.

The experiments presented in this paper were performed at
Woods Hole in June and in December, 1919. I am greatly
indebted to Professor F. R. Lillie for a research room at the
Marine Biological Laboratory on both of these occasions and
for a grant from the departmental funds covering the expenses
of the December trip. I am also indebted to Professor C. M.
Child for advice and suggestions throughout the course of the
work.

B. SUSCEPTIBILITY GRADIENTS IN TUBULARIA.

The death gradients in lethal concentrations of various sub-
stances have already been described by Child ('19^) for a number
of hydroids including Tubularia. I have repeated and confirmed
these observations for a number of forms. In Tubularia, the
disintegration (i/ioo to 1/400 mol.KNC) begins at the tips of
the proximal tentacles and proceeds down these tentacles to
their bases; soon after the proximal tentacles have begun to
disintegrate, the process is initiated in the distal tentacles and
proceeds to their bases. At about the same time as the tips of
distal tentacles begin to die, the mouth region disintegrates and
the disintegration extends along the body of the hydranth to its
base. In many cases, it was observed that the outer surface of
the proximal tentacles (i.e., the surface that contains the most
nematocysts) preceded in disintegration the inner surface. In
the stalk of the hydranth there is a specialized region bearing a
ring of nematocysts; in many cases it was noted that this region
disintegrated early and without any definite relation to the
progress of disintegration in other parts of the hydranths, as was
also noted by Child. Such specialized regions, owing to func-
tional activity, are commonly highly susceptible to toxic agents,

THE AXIAL GRADIENTS IN HYDROZOA. 365

a fact that has been noticed by us on many different forms.
It is to be understood that the disintegration of the ectoderm
precedes that of the entoderm by a considerable time interval
except in the case of the rim of the mouth where both germ layers
seem to disintegrate almost simultaneously forming a great
bulging mass of disintegrated particles.

The remarks in the foregoing paragraph refer to fully developed
hydranths. In young hydranths, it was commonly observed that
the distal tentacles disintegrated first and the disintegration then
extended basipetally along the body of the hydranth, the. proximal
tentacles disintegrating later than the body of the hydranth.
In medusa-buds, the disintegration proceeds from the free to the
attached end. Young hydranths are more susceptible than fully
developed ones only after they have reached a certain stage.

The hydranths are much more susceptible than the stems.
The disintegration of the stems is obscured by the presence of the
perisarc. Nevertheless the progress of disintegration was ob-
served in many cases in the stems but was never followed for
more than a short distance. The disintegration proceeds from
the base of the hydranth down the stem. It might be said in
criticism that the killing agent can gain access to the stem only
from the open top of the perisarc and that the death gradient in
the stem may therefore be simply a consequence of the diffusion
path followed by the agent. To answer this objection short
apical pieces of stems bearing hydranths were cut off and their
disintegration followed in cyanide. In such cases the open cut
surface at the proximal end of the stem affords a readier point
of entrance for the reagent than does the top of the perisarc;
nevertheless except for a small area of disintegration around the
cut surface, the disintegration proceeds in the same manner as
before, from the distal end of the stem proximally.

Direct observation of the course of disintegration in the stems
is not, however, entirely satisfying, as the process is admittedly
obscured by the presence of the perisarc and could be observed
for only the most distal region of the stems. Another method
was therefore employed, namely, the differential survival of distal
and proximal pieces of the stem in nearly lethal concentrations
of toxic substances. The substance employed for this purpose

366 LIBBIE H. HYMAN.

was ether. It was found that distal pieces are more susceptible
to ether than proximal pieces, that more of them die when both
kinds of pieces are exposed for a certain length of time to a given
concentration, and that of those that survive, the regeneration is
more delayed in the apical than in the-drotal halves. An experi-
ment of this kind is recorded in Table I. In this experiment the

TABLE I.

DIFFERENTIAL SUSCEPTIBILITY OF APICAL AND BASAL PIECES OF EQUAL LENGTH
(10-12 MM.) TO 2 PER CENT. ETHER.

Exposed to ether for twenty hours. Table records condition of the pieces seven
days after cutting. Temp. 12 ± 2° C.

Condition. Apical. Basal.

No. of hydranths emerged 5 14

No. of pieces with primordia of hydranth1 5 7

No. of pieces living but without primordia1 2 15

No. of pieces dead1 38 14

hydranths and first millimeter or two of the stems were dis-
carded, the stems were then cut into two equal pieces, each
10-12 mm. long, and all of the distal pieces placed in one finger
bowl and the proximal in another. A solution of 2 per cent,
ether in sea-water was then poured on both sets of pieces as soon
as possible after they were prepared and they were left in this for
twenty hours. They were then thoroughly washed and left in
normal sea-water. The temperature throughout was 12° C. ± 2.
The condition of the fifty pairs of pieces seven days after cutting
is shown in the table. The greater susceptibility of the apical
pieces is perfectly evident. More of them have died than in the
case of the basal pieces and of those that survived, the basal
pieces are much in advance in the process of regeneration.

C. GRADIENTS IN REDUCTION OF POTASSIUM PERMANGANATE.
Child (190) has called attention to a new method of demon-
strating the metabolic gradients. This method consists in
exposing the organisms to appropriate concentrations of a readily
reducible substance like potassium permanganate. This sub-
stance is reduced by protoplasm, a brown precipitate of manga-

1 Pieces were examined with the compound microscope. Pieces showing smooth
munded ends and circulation were counted as living, while stems containing only
masses of granules were regarded as dead.

THE AXIAL GRADIENTS IN HYDROZOA. 367

nese dioxide being formed. It was found by Child that this
capacity of organisms to reduce permanganate exhibits the same
kind of a gradation in relation to the body axes as does their
time of death in toxic solutions. The apical ends reduce the most
permanganate and take on a very deep brown or almost black
color and the depth of this color decreases basipetally.

I have observed the staining of Tubidaria by potassium per-
manganate. The picture thus presented is identical with the
course of disintegration already described. The tips of the
proximal tentacles stain first, the tips of the distal tentacles next;
the stain progresses rapidly down the tentacles (on the outer
surfaces of the proximal tentacles before their inner surfaces) to
their bases. Meantime the stain appears on the distal end of
the body of the hydranth and progresses basipetally along its
ectoderm. After the staining is completed, it is found that the
tips of both sets of tentacles and the mouth region of the hydranth
are very deeply stained and the stain shades off to the bases of
tentacles and hydranth. This was best observed by turning the
hydranth with a needle so that it faced upward.

The staining of the stem was naturally difficult to observe.
The short stalk below the hydranth became stained soon after
the hydranth but the staining of the stem proper was very slow.
As far as could be observed the stain proceeded from the distal
end of the stem proximally. The observations were, however,
unsatisfactory.

The younger hydranths stain much more rapidly than the
larger ones, but exhibit less distinct graded differences between
different regions of the hydranth.

D. ELECTRICAL GRADIENTS IN TUBULARIA.
Differences in electrical potential along the axes of organisms
form still another manifestation of the metabolic gradients and
are at present being investigated by members of this laboratory
for a number of the lower forms. These electrical gradients
correspond in all respects to the death and staining gradients.
The regions of highest susceptibility and reducing power are
electronegative (galvanometrically) to regions of lower suscepti-
bility and reducing power. This matter has been discussed

368 LIBBIE H. HYMAX.

elsewhere (Hyman, '18) and further papers upon the subject will
appear shortly. Briefly it is believed that chemical differences
are as a rule responsible for these permanent differences in
potential and that such chemical differences arise in the final
analysis through differences in metabolic rate at different levels.

The existence of such a difference in electrical potential
along the axis of Tubularia was discovered by Mathews ('03).
Mathews found that the hydranth is negative to any region of
the stem and that distal levels of the stem are negative to
proximal levels. He correctly attributed these potential differ-
ences to differences in metabolic activity.

I have repeated these experiments on Tubularia and verified
all of Mathews' statements. A galvanometer constructed on
the principle of the D'Arsonval galvanometer and put out by
the Leeds and Northrup Company was used. Although not as
sensitive as some other types of galvanometers in use, the
instrument was found adequate for the purpose. Non-polariz-
able electrodes were employed, made in the usual way of the
glass tubes from medicine droppers, packed at the small end with
kaolin paste made with sea-water, and filled at the other end with
saturated zinc sulphate solution. Zinc rods, amalgamated with
mercury, dipped into the zinc sulphate solution, and were con-
nected by copper wires with the binding posts of the galvanom-
eter. Small rolls of hard filter paper were thrust into the ends
of the electrodes and the stems to be tested placed on these.
These filter paper rolls \vere kept soaked with sea-water. Al-
though such electrodes are made as nearly alike as possible,
there is almost always some difference of potential between them.
Such difference increased with use so that it was necessary to
renew parts of the electrodes or to make wholly new ones at
frequent intervals. The existence of a potential difference be-
tween the two electrodes of course makes it impossible to obtain
an absolute value of the amount of current originating from the
organism; but absolute values were not desired in the present
experiments. It was my purpose merely to discover which
regions of the organism were electronegative as compared with
other regions.

In performing the experiments isolated stems of Tubularia in

THE AXIAL GRADIENTS IN HYDROZOA. 369

perfect condition only were employed. They were placed across
the filter paper ends of the electrodes and the reading of the
galvanometer recorded. The stem was then reversed on the
electrodes and the reading again taken. The difference in the
two readings, particularly the difference in the direction of
swing of the indicator on the scale gives the desired information
about the electrical condition of the stems. Each reading was
repeated once, sometimes twice. When dead organisms are
tested in this way, the galvanometer gives the same reading
regardless of the position of the material on the electrodes.

The galvanometric readings obtained on Tiibularia are recorded
in Table II. In connection with this table it is necessary to
explain that the scale of the galvanometer used is printed in
red on the right side of the zero point and in black on the left
side. When the right electrode, which is connected with the
right hand binding post of the galvanometer, is negative, the
indicator swings to the right of the zero point and hence reads
on the red half of the scale; when the left electrode is negative,
the indicator reads on the left hand or black side of the scale.
In some cases, both readings may be on the same side of the zero
point but one farther to the right or left than the other. Left
and right refer, of course, to the hands of the observer as he sits
facing the instrument.

In the table, each number refers to one individual and all of
the data given under that number were obtained on one indi-
vidual. In Table II., a, are recorded the readings of the gal-
vanometer when the hydranth is compared with nearby portions
of the stem or with distant portions or with regions where
branches are present. Table II., b, gives the readings when
distal portions of the stem are compared with more proximal
regions or with far proximal regions or with proximal regions
bearing branches. In each table, the first column gives the
number of the individual, the second describes the material,
the third gives the reading in one position between regions not
very far separated, the fourth column the readings for the same
regions in the reversed position on the electrodes, the fifth and
sixth are the same as the third and fourth except that they
compare regions more widely separated or distal regions with
levels bearing branches.

370

LIBBIE H. HYMAN.

TABLE II., a.

ELECTRICAL GRADIENTS OF Tubularla; HYDRANTH COMPARED WITH NEARBY
PORTIONS OF THE STEM AND WITH FAR PROXIMAL PORTIONS OR REGIONS BEAR-
ING BRANCHES.

Each number refers to one individual. Scale of the galvanometer reads with
red figures to the right of the zero point, with black figures to the left of zero,
hence r means red, and b means black, and numbers accompanying r and b are
divisions on the scale. The right electrode is negative when the reading is on red
or farther to the right than before; and the left electrode is negative when the
reading is on black or farther to the left than before. Hydr., hydranth, br.,
region of branches, prox., proximal, dist., distal.

No.

Material.

Readings with Reference to Position of Material on
Electrodes as Designated in These Columns.

Hydr.

Right. Stem
Left.

Hydr. Left,
Stem Right.

Hydr. Hydr. Left,
Right, Far Far Prox.
Prox. or Br. or Br. Zone
Zone Left. Right.

I.
2.

3-

5-
6.
S.
10.
ii.

12.

13-
14.

15-

16.

17-
18.
19.

21.
23-

35-

Hvdr. and stem ....

5 r
3 r

2 b
2 b
o
2 r
2 r

0

4 r

5 r
o

0

2 r
2 r
o

2 b

3 r
o

0
0
0

i r

2 b
o
5 r
o
o
i r
13 r

3 r
5 r

0

6 r

Ib
lb
4b

2 b

15 &

10 b

5b
5b
8 b
8 b
3b
5b
6 b
9b
5b
6 b
6 b
5b
4b
4b
6 b
4b

4b

2 b
2 b

6 b
3b

2 b

18 b

4b
4b
14 b
I5j>

0
3 r
o
o
2 r
2 r

0
0
0

o

2 b

2 b

o
i r
i b
o

2 r
2 r

8 b
5b
4b
4b
6 b
6 b

2 b
2 b

5b
5b

o
o

5b
5b

3*
3 b

5b

2 b

Small hvdr. and stem

Hvdr., stem, and br. .

Ditto

Large hvdr., stem and br

Small hvdr. and stem

Large hvdr , stem and br

Large hvdr. and stem

Medium hvdr. and stem

Ditto. . .

Ditto

Ditto, and br

Ditto .

Small hvdr. and stem

Medium hydr stem and br

Small hvdr and stem

Very large hydr and stem .

ATedium hvdr and ^tem

Ditto

THE AXIAL GRADIENTS IN HYDROZOA.

371

TABLE II., a, Continued.

No.

Material.

Readings with Reference to Position of Material on
Electrodes as Designated in these Columns.

Hydr.
Rieht, Stem
~Left.

Hydr. Left,
Stem Right.

Hydr.
Right. Far
Prox. or Br.
Zone Left.

Hydr. Left.
Far Prox.
or Br. Zone
Right.

37-
38.
39-
40.

Ditto, and br

7 r
4 r
13 r
5 r
7 r

8 r

19 b
10 b

i r
o
o

18 b

2 r

2 b
o
o r

6 b

2 b
i b
6 b

Medium hvdr. and stem

Ditto, and br

Medium hvdr. and stem..

The tables give all of the readings which were made. There
have been no omissions or selection of data. Forty-two indi-
viduals in all wrere tested. The first forty of these came from
the same lot of Tubularia, collected on December 6, and tested
on December 7 and 8. The last two individuals came from
another lot of material collected on December 8 and tested on
the same day. Material was kept at a temperature of 10° C.
A number of different readings were commonly made on each
stem, various levels of the stem being tested in order to obtain
a picture of the potential differences along the whole organism.

The following conclusions may be drawn from the data pre-
sented in Tables II., a, and b:

1. The hydranth is always electronegative to nearby regions
of the stem. This is shown without exception in the twenty-
three cases given in Table II., a. The galvanometer invariably
reads to the right when the hydranth is on the right electrode,
and to the left when the position is reversed.

2. The difference between hydranth and distal regions of the
stem is greater in the case of larger hydranths and much less in
the case of small hydranths. Thus in nos. 2, 8, 17, and 19,
where small hydranths were used, the potential difference be-
tween hydranths and stem is 2 to 5 points of the scale; while
when medium or large hydranths are used, the differences are
much greater. These conditions are probably associated with
age.

3. Hydranths are usually more negative to distal portions of
the stem than to more proximal regions or regions where lateral

372 LIBBIE H. HYMAN.

TABLE II., b.

ELECTRICAL GRADIENTS OF Ttibularia.

Distal regions of stem compared with proximal or with far proximal or region
bearing branches. Otherwise as in table II., a.

No.

Material.

Readings with Reference to Position of Material
on Electrodes as Designated in These Columns.

Dist. Stem
Right, Prox.
Left.

Dist. Stem
Left, Prox.
Right.

Dist. Stem Dist. Stem
Right, Far Left, Far
Prox. or Prox. or
Br. Left. Br. Right.

I.
4-

7-
9-
13-

20.

22.
24-

25-
26.

27.
28.
29.

30.

31-
32-

33-

34-
36.

41.

42.

Hydr., stem and br

i b
i b
i b
o

0

i r

o
4b
5 b
3 b

2 b

2 b
4b
3 b
3 b
2 b

i b
i b
2 r
No potent
2 r

2 b
I b
0
0
0

o
i r
i r

No potent

12 r
12 r
7 r
8 r

5^

5 b
Sb
3b
3b
5b
3-Sb
2 r
o
5 b
10 b
8 b
5b
4b
5b
5b

3b
3b
4b
ial differem
i b
4b

2 b

3b

2 b
2 b

3 b

2 b
4b

ial differem

o

5?>
9b
Qb

2 b

2 b

0

I b
2 b

2 b

2 b

2 b

:e

i b
i b
:e

0

o

4b

4b

0

i b

0
0

i b
I b

i b

i b

2 r
2 r

Stem without hydr., hydr. regen-
erating .

Stem without hydr., like 4

Stem without hvdr., like 4.

Medium hvdr and stem

Ditto

Ditto

Ditto

Ditto and br .

Medium hydr. and stem

Ditto . ...

Ditto

Ditto

Ditto

Ditto

Ditto

Ditto, and br

Small hydr and stem . .

Medium hydr stem and br.,.

Hydr., very long stem

Like 41..

THE AXIAL GRADIENTS IN HYDROZOA. 373

branches are present. This is shown in numbers 3, 5, 6, 8, 10,
15, 16, 18, 37, and 40, Table II., a. Thus for example, in no. 3,
the difference between the hydranth and the distal stem is 15
divisions of the scale in the first trial, 12 in the second, while the
difference between the same hydranth and a more proximal region
of the stem where a lateral branch was present was 8 divisions in
both trials. In one case, no. 13, the region of branching was
negative to the hydranth; in another case, no. 39, there was
practically no potential difference between the hydranth and
the branching region. We therefore see that proximal regions
of the stem are more negative than distal, especially when they
bear branches. This is due to the fact that the hydranth domi-
nates only a certain length of stem, and beyond that length
physiological isolation has occurred with the formation of a
new individual, expressed by the development of lateral branches.
Such new individuals like the original one are electrically negative
apically.

4. Distal regions of the stem are nearly always electronegative
to nearby proximal regions. This was the case in 14 of the 17
cases tested in Table II., b. In two cases, nos. 28 and 34, there
was no potential difference between two such regions of the stem;
in one case, no. 13, the gradient was reversed, the distal region
being positive to the proximal region. Such cases as these three
account for the fact that occasionally, distal and proximal pieces
of the stem regenerate simultaneously, or that the proximal piece
may precede.

5. The potential difference between distal and proximal regions
of the stem is always very much less than that between hydranth
and distal regions of the stem.

6. The potential difference is usually slight or absent or may
be reversed between distal regions of the stem and far proximal
regions, or regions bearing branches. Of six cases tested, two
showed no potential difference (nos. 4 and 33, Table II., b);
in one case, the distal region was negative (no. i); and in the
other three cases, the far proximal region or branching region
was negative to the distal region (nos. 9, 25, 36). This verifies
what was said in paragraph 3. These far proximal regions are
really beginnings of new individuals and hence are more electro-
negative than the regions immediately distal to them.

374 LIBBIE H. HYMAN.

I therefore find, as Mathews did, that within the limits of a
single Tubularia individual, any distal region is electronegative
(galvanometrically) to any proximal region. Since electro-
negativity is usually associated with a higher rate of oxidative
metabolism in organisms, these experimental data constitute
strong evidence that there is a metabolic gradient along the
axis of Tubularia, that the apical end of this axis has the highest
rate of activity, and that this rate diminishes proximally.

E. DIFFERENCES IN RATE OF REGENERATION OF DISTAL AND
PROXIMAL PIECES OF EQUAL LENGTH.

A large number of experiments were performed with reference
to this point with the result that the apical pieces were found
to regenerate markedly faster than the basal pieces in practically
all cases. A few cases were observed in which the proximal piece
regenerated first.

i. Method of Procedure. — The method of cutting the pieces
was invariably as follows unless specifically stated otherwise.
Stems free from branches and filled with coenosarc throughout
their length were removed from the colony and placed on a glass
plate. The hydranth and the first millimeter or two of the stem
were then removed by a cut and discarded and the basal end
injured by removal from the colony also cut off and discarded.
The piece of stem was then cut into two equal halves, a distal or
apical half and a proximal or basal half. In most cases, unless
otherwise stated, such halves were 8-12 mm. long. Figure I
illustrates the method of cutting the pieces.

After cutting the pieces were handled in two different ways.
In the majority of the experiments all of the apical halves were
placed in one finger bowl and all of the basal halves in another
finger bowl. Such experiments are designated throughout this
paper as mass experiments. The number of oral hydranths
emerged in each finger bowl at a given time was then recorded.
In other experiments, which are designated as individual experi-
ments, each half was placed in a separate stender dish and the
time of emergence of the oral hydranth on each half stem recorded
as accurately as possible. In all cases the record was taken only
when the hydranth had emerged completely from the perisarc.

THE AXIAL GRADIENTS IN HYDROZOA. 3/5

Observations were made and the hydranths emerged recorded
every two to four hours during the daytime. No observations
were made during the night and hence there are in all of the
experiments gaps of from six to ten hours for each night period.
The first morning observation after such a gap is indicated in all
of the tables by an asterisk.

There has been no selection of experiments for presentation in
this paper. Practically all of the experiments performed are
presented.

2. Mass Experiments in June. — These experiments were per-
formed between June 16 and July 9, 1919. The material was in
excellent condition up to July i, when the pieces for the last
experiment were cut. There was a great abundance of material,
growing rapidly and containing hydranths of all sizes. Material
was always cut on the same day as collected since, as is well
known, Tubularia will not keep in good condition in the labora-
tory in the summer. The hydranths fall off within twenty-four
fours, new ones being subsequently regenerated; further the
ccenosarc either dies away in the basal regions or else retreats
to other parts of the colonies. The regenerating pieces were kept
in the laboratory during the earlier experiments; the tempera-
ture was naturally variable and as recorded in the daytime
ranged from 15° to 24° C., with probably lower temperatures at
night. The later experiments were placed in the refrigerator at a
constant temperature of 13° C.

The mass experiments performed in June are recorded in Table
III. As already stated, all of the apical halves were placed in
one finger bowl and all of the basal halves in another; the two
finger bowls were kept under the same conditions. At frequent
intervals the regenerating pieces were examined and the number
of oral hydranths emerged recorded. The record was taken
only when the hydranth had completely emerged from the top
of the ccenosarc and had spread its tentacles.

Details of these experiments not given in the table are as
follows. Experiment I was performed on slender stems; experi-
ments 2 on stout stems; the other experiments were with medium
sized stems although there was some variation in the diameter
of the stems. Experiments I, 2, 5, and 10 regenerated in the

376

LIBBIE H. HYMAX.

TABLE III.

fa
o

&

fa

o

71

H

J
<

ffi

J
«!
w

<i

fQ

a
z

«!
i-)
U

E
<

§

z

o

—I
H

Z
w

K
B«

fa a

O o

*jT C-]

o

C/3

Z

i— <

Q

s

o

Ix.

tf
w

z

d,
K

w

l/l

•=

KS

oooc.xo^.o^oaMvor..^ M ^^oooo c^« ^

a
v

«

C\f^cc O r1*^GO M t-i ro^i-H M ro^ro^^l"

o

•O

p

H

m

IT) LO IO UTi

SO *O t** t^* Is** f~* cc O\ O\ O\ O w i— i M c^ c^ ro *^t" ^O ^O QO ON

to

* * * *

a

^

M M M O) 0)

o
o

H

^

MvCOOroOO)<NLO

"2

a

M2 O t^- t^ 00 OO OO C^ O\
*

"3

to

s5

0 ° ° °° 2 2 M g S o! ot ^?

A

«

M ir; LO Tf IO

M C^ 01

"5 c

•P .0

E

K

^O ^D W C1) NO O f) ^3 *-O w O^ O

"OOt^OOOQ O\O\O\C t-n I-H r*^

* ' #

rt CTJ

o £

^

OOwrotot-i-ifo^ovoaOfoiOHro^-iot-oo

tS in

« "o

«

M M IOQO OxT-o^^oooooo O NO r*-

.S'S

H

t_

M

"OOOO t>r^cooO'COco o\O\C\O O M M M 01 -rj-io

S c

= p

*

CJ -i
d; "£

^

C 0 M M o] 0' M M 10 10 i- o. ^ r- 0, Cs 0 M M M ^ *t rt o r- ON

O !_'

(N (N (N 01 rorOf^f^O^^O^^roro

P (n

°co ce

a

d rcr^-ON^OGO cs ONO IDO r^-oo M M r^iooo o

«j (_>

"

UOl^lOlOlOlOlO LOj-U^lOtOi/^ LOUO

C.~

"o .5

CO k^

2

a

^^^.O.O^o 0000 ^^.= =00,0 aov0,o 0 «

n 'S

000-=l-xOvOON<MI>OOOM(N(n

H 01 CS CN CN

•sl

«

l-H M' M M O) N 0) 01

umbers
cated.

en

i-

a

'* *

--

g.S

^-

000000, 00.0=0 c^^^,,^..

2 P

a -a

S

M HH M CS CS C^ CS

c/j +->

n%
• E?

IO

in

a

^^ ^tt^ # ^ M

cu

P £
O "
^ CJ

^

1— 1 HH Ol N

gl

«

HH M M M CS

C P
0 |

7 >>

oo .a

uJ

a

UO IO
#° *

CO i—,

0 0

o >-

^

M 1— 1 M

p. . f"~l

5

«

in IT; j>.

3
C

o

'5

i

en

u

LO UO LO

f^* O* O C1) c*; LO t^* M t— i
*° ^ -jT

THE AXIAL GRADIENTS IN HYDROZOA. 377

laboratory at variable and generally moderately warm tempera-
tures, from 15° to 24° C. (day records). In experiment 6, the
pieces were placed in the refrigerator (temp. 13° C.) for the first
twelve hours and in experiment n for the first twenty hours
after section, and were then removed to laboratory temperature.
In experiments 15, 16, and 21 the pieces were kept in the re-
frigerator (temp. 13° C.) for the entire period of regeneration
as the weather had become unfavorably warm by this time.
There was some mortality, particularly among the basal pieces,
owing probably to the warm weather. In experiment 6, two
basal pieces were living but had not regenerated when the
experiment was concluded. In experiment 10, where fifty pairs
of pieces were cut, seven basal pieces died and four had failed to
regenerate when the experiment was discontinued. In experi-
ment 1 1 , three apical halves and twelve basal halves had died or
failed to regenerate when the experiment was discontinued. One
basal piece died in experiment 21.

The length of the pieces in all of the experiments recorded in
Table III. was 8-12 mm. It was not possible to find stems free
from branches long enough to give longer pieces in the summer
material, owing to the fact that the colonies are growing rapidly
and branching extensively at this season. The number of pieces
cut in each experiment depended on the number of healthy stems
of sufficient length available in the day's collection. Although
material was very abundant and large quantities of it were
brought in whenever desired, most of the colonies consisted of
stems so short as to be useless for the experiments.

3. Mass Experiments in December. — According to the state-
ments of Mr. Gray, head of the supply department at Woods
Hole, Tubularia is most abundant and in excellent condition in the
early summer reaching a climax in June. After that, as the
weather becomes warm, the colonies die away, the protoplasm
withdrawing into the perisarc and apparently passing into a
dormant state. In the fall, as the water becomes colder, the
colonies begin to grow again, reaching their height in November
and December, and then with still colder weather, once more
passing into the quiescent state, emerging in the spring. In
November, 1919, no Tubularia could be found at Woods Hole,

378 LIBBIE H. HYMAN.

in spite of diligent search by the collectors. It was, however,
obtainable early in December in fair abundance, and experiments
were performed upon it from December 6 to 16. The colonies
at this time were in excellent condition, branching freely and
growing rapidly. The general appearance of the material was
much the same as in June except that the hydranths attained a
larger size than in June and a few lots of material consisted of
very long stems, much longer than any observed in June. The
majority of the December material, however, was branching so
freely that most of the stems were relatively short and in some
cases it was necessary to cut pieces less than 8 mm. long. The
temperature of the running water in the laboratory in December
was 8° C. and it was therefore possible to keep the material for
two or three days in excellent condition. Two collections of
material were used in the December experiments; one collected
on December 6 was cut for experiments on December 6, 7, and 8;
the other, collected on December 8, was used on December 8
to 1 1 . All of the pieces were kept on the water tables in slowly
running water at a temperature of approximately 12° C., varying,
however, for slight periods from 10° to 14°.

The results of the mass experiments performed in December
are given in Table IV. In experiment 26, the pieces were about
5 mm. long; in experiment 27, 5-8 mm. long; and in experi-
ments 35, 44, and 47, 10—12 mm. long. The pieces were cut
as in Fig. I, except in the case of experiment 44, in which the
basal pieces were cut at the proximal end of long stems so that
some 10-15 mm. of stem was removed between the levels of the
apical and basal pieces in this experiment. There was no
mortality among the pieces. The temperature throughout was
12° C. ±2.

4. Conclusions from Mass Experiments. — The data given in
Tables III. and IV. permit us to draw the following conclusions
concerning the rate of regeneration of oral hydranths on apical
and basal pieces of the stem of Tubularia of equal length :

(a) Hydranths invariably emerge first on the apical pieces
and a considerable number of such pieces will have regenerated
before any of the basal pieces have produced a hydranth.

(b} At any given time there are in nearly all cases a greater

THE AXIAL GRADIENTS IN HYDROZOA.

379

number of apical pieces with oral hydranths than basal pieces.
The difference is always more marked in the early part of the
regeneration period; later the basal pieces may catch up with
the more tardy of the apical pieces with the result that the
number of regenerated basal pieces may in a few cases equal the

TABLE IV.

MASS EXPERIMENTS PERFORMED IN DECEMBER SHOWING RATE OF REGENERATION
OF APICAL AND BASAL HALVES OF STEMS OF Tubnlaria.

Hrs. means number of hours elapsed since cutting; a, apical half; b, basal half;
figures under a and b give number of hydranths emerged at time indicated; asterisk
indicates first morning observation.

26.

27.

35-

44-

47-

Hrs.

n.

b.

Hrs.

a.

b.

Hrs.

a.

b.

Hrs.

a.

b.

Hrs.

a.

b.

*6i

18

7

45

I

o

*6o

16

0

52

2

0

6l

I

o

63

19

9

Si

2

o

62

19

o

58

4

0

65

2

o

65

21

ii

53

8

2

64

22

4

61

8

o

67

4

o

67

21

14

55

15

3

66

23

9

*69

16

I

69

5

o

69

22

15

57

25

8

68

29

18

7i

18

8

73

12

3

7i

17

59

27

19

70

31

25

73

19

10

75

14

3

73

18

*68

40

40

72

35

29

75

22

16

*85

26

14

76

18

70

42

42

74

37

29

77

25

19

87

29

18

*84

20

72

48

44

77

4i

30

79

26

20

89

32

20

90

21

74

49

45

*S4

44

36

81

27

23

9i

33

26

*II2

22

76

5i

46

87

45

42

83

28

25

95

33

30

80

53

50

89

46

45

*93

30

27

97

34

31

83

54

5i

9i

47

47

105

28

99

35

33

*92

55

55

93

49

47

*ii7

29

101

36

33

94

57

56

95

49

49

119

30

*in

37

36

96

57

57

97

5i

49

119

37

98

58

57

99

5i

IOO

60

57

104

58

*ii7

60

number of regenerated apical pieces (exps. 27 and 35, Table IV.) ;
but in no case are the basal pieces in advance.

(c) In all cases the apical pieces complete their regeneration
first.

(d) When other factors are equal the rate of regeneration is a
function of temperature.

(e) When other factors are equal, the rate of regeneration is a
function of the diameter of the stem. More slender stems re-
generate more rapidly than stouter stems. Thus in experiment
I, Table III., the pieces were cut from slender stems bearing
small hydranths; those in experiment 2, same table, from stouter

380 LIBBIE H. HYMAN.

stems bearing larger hydranths, cut at the same time and from
the same lot of material. It is perfectly apparent that the more
slender pieces regenerate more rapidly, and this was also evi-
denced throughout all of my experiments. It is probable that
this relation of the rate of regeneration to the diameter of the
stem is connected with the age of the stem, but since one does
not certainly know that slender stems are younger than stouter
ones, the matter must be left open at present. Morgan ('06 6)
found that young stems regenerate more rapidly than old ones
and that when the hydranths are removed from the top and
lateral branches of a stem, the lateral branches regenerate first.
At any rate, these facts dispose of the suggestion which has
been made that apical pieces regenerate more rapidly than
basal pieces because they are of larger diameter and hence contain
more protoplasm. As a matter of fact it is the pieces of smaller
diameter which regenerate the more rapidly. Further in slender
stems there is no difference in diameter along the stem, and yet
the apical halves of such stems regenerate hydranths earlier
than the basal halves.

5. Individual Experiments. — These experiments were identical
with the mass experiments except that each piece was placed
in a separate dish and the number of hours required for it to
produce an oral hydranth recorded as accurately as possible.
The records of the four experiments of this kind which were
performed are given in Table V. Experiments 9 and 17 were
performed in June at room temperatures; experiments 29 and
45 in December at a temperature of 12° C. ± 2. Pieces were
8-12 mm. long except in experiment 29, where they were 5-8 mm,
long. There was some mortality in the June experiments but
none in December.

6. Conclusions from Individual Experiments. — The results of
these experiments lead to the same conclusions as previously
stated from mass experiments. Of 122 pairs of pieces in which
both pieces regenerated, the apical halves regenerated hydranths
first in in cases, or 91 per cent.; the basal halves first in 10
cases, or 8 per cent.; and the time of emergence of the hydranth
was practically the same in both pieces in one case. Cases
where the basal piece preceded in regeneration are indicated by

THE AXIAL GRADIENTS IN HYDROZOA.

381

TABLE V.

RECORDS OF THE TIME OF EMERGENCE OF INDIVIDUAL APICAL AND BASAL HALVES

OF THE SAME STEM.

Dagger calls attention to cases where the basal half emerged first; other abbre-
viations and symbols as before.

No.

Hours Since
Cutting.

No.

Hours Since
Cutting.

No.

Hours Since
Cutting.

No.

Hours Since
Cutting.

Exp. 9.

Exp. 17.

Exp. 29.

Exp. 45.

a.

b.

a.

b.

a.

b.

a.

b.

I

43

dead

I

40

*64

I

45

71

I

73

*9i

2

*34

*5S

2

77

86

2

103

*H4

2

*67

77

3

42

75

3

*39

*64

3

65

79

3

60

75

4

44

50

4

*39

53

4

*88

79t

4

73

77

5

41

dead

5

43

77

5

57

73

5

60

73

6

42

dead

6

*39

45

6

56

69

6

*9I

73t

7

38

48

7

*39

76

7

54

7i

7

79

*67t

8

42

61

8

45

67

8

75

*88

8

59

71

9

38

44

9

68

67t

9

*64

66

9

56

*67

10

40

50

10

*64

74

10

*64

72

10

59

71

ii

38

50

1 1

*64

dead

ii

56

74

ii

*H5

93t

12

42

48

12

70

dead

12

56

76

12

*67

95

13

46

50

13

71

*87

13

53

*64

13

72

74

14

42

48

14

45

*&4

14

*64

*8?

14

72

73

15

36

50

15

4 1

53

15

66

*87

15

*66

*H4

16

42

50

16

A I

dead

16

*64

68

16

*90

78t

17

36

46

17

dead

116

17

68

90

17

*66

74

18

53

dead

18

75

112

18

92

97

18

72

78

19

42

67

Aver.

53

73

19

66

*87

19

68

76

20

36

46

20

*64

70

20

*66

76

21

42

63

21

76

*87

21

72

76

22

36

46

22

70

96

22

68

78

23

42

46

23

70

96

23

68

80

24

38

40

24

68

*87

24

*66

76

26

42

44

25

*64

76

25

68

76

27

42

42

26

*64

72

26

68

*90

28

40

*58

27

68

*II2

27

68

72

29

42

63

28

90

95

28

78

74t

30

50

*58

29

73 104

29

68

78

31

48

63

30

66

76

30

58

70

32

65

67

31

72

*II2

31

*66

72

33

61

78

32

68

78

32

*66

76

34

63

62. st

33

114

99t

33

68

76

__35_

40

46

34

69

*86

34

68

78

Aver.

42

52

35

*86

9i

35

72

76

36

69

75

36

*66

92

37

75

*86

Aver.

69

77

38

75

*86

39

67

*86

40

*86

101

41

66

97

42

118

I03t

Aver.

7i

83

Total number of regenerated pairs . . .
Number of cases where a preceded . . .
Number of cases where b preceded . . .
Number of cases where a and b equal

. 122
in or 91%

10 or 8%

i

382 LIBBIE H. HYMAN.

a dagger in Table V. The average difference between the
number of hours required for the emergence of the hydranths on
apical and basal halves was 10 hours in exp. 9; 20 hours in exp.
17; 12 hours in exp. 29; and 8 hours in exp. 45. The individual
differences range from half an hour to more than forty hours.

It may be inquired why in a small percentage of cases the basal
piece precedes the apical piece in regeneration, and why there is
such a great variation in the difference between the time of
regeneration of the two pieces. It is highly probable that these
variations are related to the degree of physiological isolation
existent in the basal pieces before they were cut from the stems.
It has already been pointed out that a hydranth controls only a
certain length of the stem proximal to it and beyond that limit
a new individual arises which eventually expresses its presence
by the formation of a lateral bud. Now it is evident that such
new individuals must exist physiologically before they give
morphological expression of their existence by bud formation.
It has already been stated that in these experiments the longest
obtainable stems free from buds were used. Such stems are
the exception rather than the rule since the majority of the
material obtainable, particularly in the summer, will furnish
only a small proportion of long stems free from branches. It is
therefore obvious that the basal regions of such long stems must
be in various stages of the process of physiological isolation and
branch formation. The nearer such basal regions are to branch
formation the more rapidly will they regenerate when isolated
and those that are on the very verge of branch formation may
conceivably regenerate as rapidly as or even more rapidly than
more apical pieces. This matter is referred to again in connec-
tion with experiments on the rate of regeneration of basal pieces
cut below branches.

7. Remarks on the Temperature Coefficient. — It has generally
been accepted that when the rate of a biological process increases
two to three times with each ten degrees rise in temperature
that such a process is chemical in nature. It may be doubted
that this line of reasoning is strictly correct. The use of the
temperature coefficient to analyze the nature of a biological
process involves the unwarranted assumption that such processes

THE AXIAL GRADIENTS IN HYDROZOA. 383

may be purely chemical; but it is very doubtful that they ever
are solely chemical in nature, and, of course, equally doubtful
that they are ever the consequence of purely physical changes-
In all probability biological processes are neither complexes of
purely chemical reactions nor purely the resultants of physical
changes but they involve both types of changes occurring simul-
taneously and mutually interacting. On a priori grounds, how-
ever, it may be accepted that the chemical processes are of
paramount importance in living things, since, while substances
having physical properties similar to or identical with those of
protoplasm exist which are not alive, in no case do non-living
materials carry on the chemical reactions characteristic of proto-
plasm; further, the "signs of life" are chemical or of chemical
origin, and protoplasm in which the chemical reactions have
fallen to a low level is to all intents dead. Granting, therefore,
that chemical reactions play the most important roles in life
processes and that in many cases physical changes are insignifi-
cant, it may be valid to draw conclusions from the value of the
temperature coefficient. But it must always be borne in mind
that the chemical reactions which occur in living things are
subject to processes of regulation in the organism. The relation
of chemical changes to temperature is therefore in the organism
a variable quantity. Thus Behre ('18) found that the rate of
respiratory metabolism of Planaria is lowered when the animals
are maintained at a high temperature and raised when they are
maintained at low temperatures. The temperature coefficient
for the rate of respiration of Planaria is therefore not a fixed
value for a certain range of temperature but depends to some
extent upon the temperature at which the animals had been
living previous to the experimental test. Since such modifica-
tions or regulations are known for emulsoid colloids, their be-
havior at any given time depending upon the conditions to
which they had previously been exposed (phenomenon of hys-
teresis), it is possible that this ability of organisms to modify
the rate of processes presumably chiefly chemical with reference
to temperature is due to the colloidal substratum in which the
chemical reactions take place.

In Tubularia similar regulations to temperature are observable.

384 LIBBIE H. HYMAX.

The rate of regeneration of pieces of the stem of Tubularia is,
as has long been known, dependent in large part upon the tem-
perature and the temperature coefficient of this process is
described as corresponding to that of chemical reasions (Moore,
'10). The rate of regeneration is not, however, wholly dependent
upon the temperature at which regeneration occurs but is to
some degree affected by the temperature at which the particular
stems used had been living previous to their utilization. Thus
in the experiments recorded in Tables III. and IV., it can be
noted that summer material regenerates more slowly at 13° C.
than does winter material at 12° C. While other possible
explanations of this fact could be suggested it seems reasonable
in the light of other results along this line to suppose that this
is another case of acclimation to temperature; material living for
some time at low temperature has elevated the rate of its chemical
processes above that which would result if the material were
suddenly lowered to the same temperature from a higher tem-
perature— a procedure usually practised in experiments on the
temperature coefficient.

F. RATE OF REGENERATION OF DISTAL AND PROXIMAL PIECES

OF UNEQUAL LENGTH.

Banus refers to Child's experiments on pieces of unequal
length in which Child found that longer pieces will regenerate
slightly faster than shorter ones provided the factor of level is
eliminated by always making the apical pieces the shorter pieces.
Since apical pieces regenerate faster than basal pieces no con-
clusions could be drawn regarding the effect of length on the time
of regeneration unless the apical piece were the shorter. Banus
has "repeated" this experiment and claims that the longer piece
always regenerates first regardless of level. Here Banus has
again misrepresented Child's statements and he has not in
reality repeated Child's experiment. Child distinctly states that
changes in the length of the piece "produce only very slight or
no appreciable differences in time of emergence of oral hydranths
provided the length of the piece is above a certain minimum.
But with reduction in length below the minimum the appearance
of the hydranth is delayed and this retardation increases with

THE AXIAL GRADIENTS IN HYDROZOA. 385

further reduction in length." The apical pieces will therefore
regenerate later than the basal pieces only when they are reduced
below a certain minimum length. This minimum length is
very much less than any used in Banus's experiments. In order
to get the result mentioned by Child it is necessary to cut the
apical pieces as small as 2 mm. Yet Banns in "repeating"
Child's experiments has used no pieces less than 10 mm. in length.
In pieces as long as this, length makes very little difference;
according to Child's results and my own, the apical pieces will
regenerate oral hydranths earlier than the basal pieces, just as
when both pieces are of equal size. In his tables 3, 4, and 5,
Banus presents data on apical and basal pieces in which the
lengths of the pieces were: 10 and 20 mm., 10 and 30 mm., and
10 and 40 mm. Banus found that the longer pieces in all of
these cases, regardless of whether they are apical or basal,
regenerate oral hydranths slightly in advance of the shorter
pieces.

With these results and statements of Banus I am quite unable
to agree. There is some truth in the statement that a longer
piece will regenerate slightly faster than a shorter piece with
apical end at the same level. Yet in the case of apical pieces,
the difference between pieces 10 and 20 mm. long is very slight
indeed, in fact, practically nil; but it is plainly marked in shorter
pieces, say 5 and 10 mm. long. In the case of the basal pieces
the difference in time of regeneration between 10 and 20 mm.
pieces is somewhat greater but here it must be remembered that
the apical end of a basal piece 20 mm. long is in these experi-
ments at a level 10 mm. more distal than that of a basal piece
10 mm. long, and the factor of level again comes into play. In
all cases in pieces exceeding 5 mm. in length, the apical pieces
will in general regenerate more quickly than the basal pieces,
regardless of their relative lengths; and a basal piece twice as
long as an apical piece will still regenerate more slowly than the
apical piece, notwithstanding the effect of length.

I have repeated Banus's experiment on relatively long pieces
of unequal length and the results are given in Table VI. Three
pairs of experiments were performed. In one experiment*of each
pair the apical piece was half as long as the basal; in the other

386

LIBBIE H. HYMAN.

experiment, the apical piece was twice as long as the basal.
Pieces for the two experiments of each pair were cut from the
same lot of stems at the same time and kept under the same
conditions. All of the apical pieces were kept in one finger bowl
and the basal in another. The method of cutting the pieces for
such experiments is given in text-figure 2. All of these experi-
ments were performed in December at a temperature of 12° ± 2.
The length of the pieces is stated in the table. No experiments
were attempted with pieces in which the ratio of length was
I : 3 or i : 4, as it is difficult if not impossible to obtain a sufficient
number of unbranched stems of the requisite length.

The data given in Table VI. show quite clearly that an

TABLE VI.

RECORDS OF MASS EXPERIMENTS WITH APICAL AND BASAL PIECES OF UNEQUAL

LENGTH.

Columns under a and b record number of hydranths emerged at time indicated.

Exp. 32,
a = 4 — 6 Mm.,
b = 8 — 12 Mm

Exp. 33,

a = 8 — 12 Mm.,
/; = 4 — 6 Mm.

Exp. 37,
a = 8 — 10 Mm.,
6=15—20 Mm.

Exp. 38.
^ = 15 — 20 Mm.,
£=8— -10 Mm.

Exp. 42,
« = io — ii Mm.,
b=2o — 22 Mm.

Exp. 43,
a=2o — 22 ftlm..
/>=io — ii Mm,

Mrs.

a.

b.

Mrs.

a.

b.

Hrs.

a.

/..

Hrs.

a.

b.

Hrs.

a. b.

Hrs.

a.

b.

56

I

O

58

I

O

*59

18

0

*59

7 o

53

3 o

53

7 o

58

2

o

*6y

15

2

61

24

i

6l

II O

55

7 o

55

9 o

*67

II

9

69

16

4

63

35

3

63

i?

0

57

ii

O

57

10

0

69

13

ii

7i

18

5

65

4i

7

65

22

0

59

18

0

59

12

I

7i

15

16

73

21

7

67

47

13

67

24

3

62

20

0

62

17

2

73

I?

i?

75

22

U

69

49

18

69

28

10

*69

25

25

*69

26 18

75

2O

19

77

17

?i

49

25

71

28

14

72

26

26

72

28

24

77

21

20

79

19

73

50

27

73

30

18

74

26

27

74

28

26

79

22

22

81

20

76

30

76

32

21

76

26

28

76

29

27

84

23

22

84

21

*83

35

*83

35

34

78

26

29

78

29

29

*9i

23

23

*9i

22

86

38

86

37

38

82

27

30

82

30

29

94

24

23

88

40

88

38

45

84

29

84

29

104

24

90

4i

90

43

46

*96

30

*96

30

94

43

94

43

47

96

44

96

44

49

98

46

*io8

49 SO

*io8

47

114

50

no

48

120

49

*I32

50

G. ALTERATION OF THE REGIONAL DIFFERENCES IN RATE OF

REGENERATION.

apical piece will regenerate faster than ii basal piece of twice its
length, in pieces at least as long as 5 mm. Although the factor of
length is of some consequence, the factor of level is of vastly

THE AXIAL GRADIENTS IN HYDROZOA. 387

greater importance. In experiments 32 and 33, the longer pieces
in each case regenerate faster, but the effect of length does not
overcome the effect of level, the apical pieces in both experiments
regenerating first on the whole. The influence of length is most
marked in experiments 32 and 33, where short pieces were
employed. It is very little evident in experiments 37 and 38,
and 42 and 43, where pieces 10 mm. in length were employed.
In fact, in experiment 37, the apical pieces 10 mm. long regenerate
slightly faster than the apical pieces 20 mm. long; and in experi-
ments 42 and 43 there is practically no difference. We may
therefore say that length is of little consequence in pieces exceed-
ing 10 mm. in length, in agreement with Child's previous state-
ment and in contradiction to the claims of Banus. In all
experiments the longer basal pieces regenerate faster than the
shorter basal pieces, but it is probable that this effect is one of
level rather than of length, because the longer basal pieces have
their apical ends at a higher level than the shorter basal pieces.

It may therefore be concluded that in the case of apical and
basal pieces of unequal length, the apical pieces will still regen-
erate more rapidly on the whole regardless of whether they
constitute the shorter or the longer pieces, always provided that
their minimum length is 5 mm. The level at which the pieces
are cut is still the dominant factor in such experiments. In
pieces below 10 mm. in size, a longer piece will regenerate slightly
faster than a shorter one; but in pieces above 10 mm. length,
length is of practically no consequence. Long basal pieces in
such experiments regenerate faster than shorter basal ones mainly
because their apical ends are at a higher level than the apical ends
of the shorter pieces. These results are the contrary of those of
Banus whose experiments are invalidated owing to his erroneous
method of cutting the apical pieces as discussed at greater length
below.

The data already presented incontestably demonstrate that a
regional difference in rate of regeneration exists along the axis
of Tubularia, such that regeneration is the more rapid the nearer
the piece lies to the apical end. It may next be inquired whether
this regional difference is modifiable under either certain normal
conditions or under experimental conditions. To this inquiry

388 LIBBIE H. HYMAN.

an affirmative answer may be returned. It is possible to modify
or eliminate the regional differences in question. The various
methods by means of which this was attempted or accomplished
are discussed in this section.

1. The Effect of Cold. — A number of experiments were per-
formed in which both apical and basal pieces were exposed to a
lowered temperature for a number of hours after cutting or
during the entire period of regeneration. The pieces were
removed from room temperatures (approximately 20° C.) to the
temperature of the refrigerator (13° C.) for various periods of
time. Although such a proceeding invariably retards the rate
of regeneration, the differences between the regeneration of apical
and basal pieces were unaltered by such exposure to low tempera-
ture. These experiments are therefore included in Table III.
(exps. 6, n, 15, 1 6, and 21), as showing the typical difference
between the rate of regeneration of apical and basal halves.
It is highly probable, however, that with very low temperatures,
in the neighborhood of zero, the typical difference between pieces
of different level would be reduced or eliminated.

2. Effect of Using Material Kept in the Laboratory. — Two
experiments were performed in June upon material which had
been kept in the laboratory aquaria for a week preceding the
cutting of the pieces. As is well known under such circum-
stances, the hydranths of Tubularia fall from the stems and new
hydranths are subsequently regenerated. Such new hydranths
are smaller than and have a lower rate of activity than the original
hydranths; it is therefore to be expected that the regional
differences along the axis will be reduced in such cases. As
already stated only two experiments were performed as the
weather had become warm and little material was available.
The material for these experiments was collected on June 30
and cut on July 6. The pieces were kept in the refrigerator
(13° C.) throughout the regeneration period. The results are
given in Table VII. Experiment 21, Table III., furnishes a
control for these experiments. It will readily be seen that the
differences between the rate of regeneration of apical and basal
pieces are plainly reduced as a consequence of the depressing
effect of laboratory conditions upon the physiological axis of

THE AXIAL GRADIENTS IN HYDROZOA.

389

Tubularia. This experiment shows that the metabolic gradient
of Tubularia is not a fixed and permanent gradient in the stem
but is readily variable under the conditions of the animal's
environment. Experiments such as those of Banus in which no
account is taken nor any description given of the conditions of
the material or the environment do not therefore merit serious
consideration.

TABLE VII.

RECORD OF MASS EXPERIMENTS WITH APICAL AND BASAL PIECES OF EQUAL
LENGTH, THE PIECES BEING TAKEN FROM MATERIAL KEPT ONE WEEK IN-
LABORATORY CONDITIONS BEFORE CUTTING.

Temp. 13° C. Control, cxp. 21, Table III., in which the pieces were cut on
the same day as the material was collected.

Exp. 24.

Exp. 25.

Hrs.

a.

b.

Hrs.

a.

b.

72

I

0

72

4

O

*84

II

5

*84

9

3

87

12

9

87

n

6

94

22

16

89

13

9

96

22

18

94

22

18

*io7

33

28

96

23

23

no

33

30

*i07

27

26

114

35

33

no

28

28

117

36

33

114

29

28

119

36

33

117

29

30

123

37

35

119

30

*I46

36

158

37

3. The Effect of the Presence of Branches. — In the consideration
of the data on the electrical gradient in Tubularia it was pointed
out that the control of a Tubularia hydranth extends only for a
limited distance down the stem and that the stem beyond this
limit is more or less differentiated as another individual. This
differentiation, at first purely physiological, is later morpho-
logically apparent by the formation of a bud at the level of the
apical end of the new individual. The appearance of the bud
not only indicates the formation of a new individual but also
is an expression of a loss of control of the basal portions of the
stem by the original hydranth. It is therefore to be expected
that pieces taken above such lateral branches will have a lower
metabolic rate than pieces of the same level from unbranched
stems; and further that pieces taken below the branch, since

590 LIBBIE H. HYMAN.

they are near the apical ends of new individuals (the real apical
end being the hydranth of the branch) will have a higher meta-
bolic rate than ordinary basal pieces. Owing to the operation of
both of these factors it may be expected that the difference
between apical and basal pieces will be reduced when they are
cut above and below, respectively, the level of a branch. This
was found to be the case. Banus in his paper does not state
whether or not he used stems free from branches and did not
reply to inquiries on this point.

In preparing pieces for this kind of experiment, the following
procedure was usually adopted. Stems having one branch at
about the middle of the stem were selected, the terminal hy-
dranth, upper millimeter or two, and basal end cut off and dis-
carded as usual. An apical piece was then cut anterior to the
branch, and a basal piece of equal length posterior to the branch;
the small piece bearing the branch was discarded. As found by
Morgan and verified in my experiments, the stumps of lateral
branches left on pieces wTill regenerate hydranths more rapidly
than the distal end of such pieces, and these lateral hydranths
will then inhibit the formation of the terminal hydranth; hence
in experiments of this kind it is necessary to avoid using pieces
bearing stumps of branches. The method of cutting the pieces
in most of the experiments is illustrated in Fig. 3. In one
experiment, stems having two branches were selected, the apical
piece cut in front of the first branch, and the basal piece between
the two branches as illustrated in Fig. 4.

The results are presented in Table VIII. All experiments of
this kind were mass experiments. Experiments I2a and 126
were performed in June at room temperature, the remaining
experiments in December at 12° C. ± 2. The controls for these
experiments are indicated at the top of the table; such controls
were cut at the same time and from the same lot of material,
with the exception that they came from stems without branches,
and were kept under the same conditions. The pieces in experi-
ments 28, 30, and 34 were short pieces, 5-10 mm. long; those in
experiments 12 and 36, approximately 10 mm. long. The pieces
for all experiments except number 36 were prepared according
to Fig. 3; those for experiment 36 as in Fig. 4.

THE AXIAL GRADIENTS IN HYDROZOA.

391

The effect on the relative times of regeneration of apical and
basal pieces by cutting them above and below a lateral branch
is of course slight but nevertheless it is evident in most cases.
If the experiments given in Table VIII. are compared with those

TABLE VIII.

RECORDS OF MASS EXPERIMENTS WITH APICAL AND BASAL PIECES OF EQUAL
LENGTH, THE APICAL PIECES BEING TAKEN IN FRONT OF THE FIRST BRANCH
THE BASAL PIECES BELOW THE BRANCH.
Columns under a and b give number of hydranths emerged at time indicated.

Controls in Table III. and IV., exp. 10 for exp. 12; exps. 26 and 27 for exp. 28, 30

and 34; exp. 35 for exp. 36.

Exp. iza.

Exp. i2/>.

Exp. 28,

Exp. 30.

Exp. 34.

Exp. 36.

Hrs.

a.

b.

Hrs.

a.

b.

Hrs.

a.

b.

Hrs.

a.

b.

Hrs.

it.

b.

Hrs.

<*.

b.

37

I

O

33

I

0

53

I

O

54

I

0

56

I

0

*59

2

O

43

2

0

37

2

0

55

2

O

*62

4

0

58

3

0

63

4

O

45

3

O

*43

6

2

57

3

0

66 ! 5

2

*67

17

2

65

5

0

49

4

I

47

10

3

60

n

0

68

7

4

69

17

3

67

8

2

5i

4

3

49

13

6

*68

13

5

70

9

8

7i

17

8

69

n

4

*57

6

4

5i

16

9

70

14

6

72

9

9

73

22

9

71

14

IO

60

8

4

53

17

13

72

IS

7

79

n

10

75

25

14

73

15

II

66

9

5

55

18

15

74

16

9

82

ii

13

77

32

20

76

18

12

68

10

6

59

25

19

76

17

n

*90

12

13

79

35

26

*83

28

20

72

ii

7

61

28

20

78

18

12

93

13

14

81

38

27

86

35

23

74

n

8

*68

29

22

80

19

13

95

15

15

84

40

30

88

36

27

*8i

ii

9

7i

30

24

83

20 , 15

99

15

16

*9i

55

41

90

43

31

88

12

10

75

3i

24

*QI

22

18

*ii6

16

94

57

41

92

47

36

95

13

10

77

25

94

19

96

59

42

94

48

40

105

14

10

85

25

117

20

98

44

96

48

41

108

1 1

*92

26

119

22

IOO

47

98

48

43

no

12

97

27

1 06

1 5I

*io8

5i

46

IOO

28

*n6

58

no

48

107

29

122

59

114

49

I

I2O

Si

in Tables III. and IV., it will be found that in general more basal
pieces have regenerated in the experimental series when the
same number of apical pieces have regenerated in both control
and experimental series. This appears chiefly in the early part
of the regeneration period. A few such comparisons may be
pointed out; in making them it is necessary to select experiments
in which the total number of regenerating pieces is similar in
experiment and control, since the number of regenerated apical
pieces in the early stages of an experiment is greater relative to
the basal pieces, the greater the total number of pieces. In
experiment I2&, Table VIII., 2, 9, and 19 basal pieces have

392 LIBBIE H. HYMAN.

formed hydranths as compared with i, 6, and less than 15 basal
pieces in the control experiment, number 10, Table III., when
6, 1 6, and 25 apical pieces have regenerated in both cases.
Similarly, in experiment 30, for which experiment 26 is a control,
the regeneration of the apical and basal pieces is practically
simultaneous, a result which is never obtained when stems free
from branches are employed. Comparison of experiment 36,
Table VIII., with its control, experiment 35, Table IV., shows
the same effect; in the former case 12 basal pieces, in the latter
case no basal pieces have regenerated at the time when 18 apical
pieces have regenerated in both experiments. In other cases,
as in experiment 34, little difference from the control could be
observed. The decrease in the time difference between the
apical and basal pieces in these experiments is apparently largely
due to a delay in the regeneration of the apical pieces. This is
to be expected, since as already explained all basal pieces are
probably more or less isolated as new individuals, and hence
are slightly accelerated in both experimental and control series.
The "apical" pieces, on the other hand, in the present experi-
ments, since they are taken in front of the level of a branch, really
represent the basal end of the first zooid of the stem, and hence
are delayed in regeneration as compared with pieces similar in
position from stems where branches have not yet arisen and where
the hydranth still controls most of the length of the stem. In
regard to the basal pieces, it should further be pointed out that
the really high metabolic point of the new individuals formed at
the base of Tiibularia stems is in the hydranth of the branch,
and the basal piece itself below the level of the branch retains
only part of the increased metabolic rate after the bud has
formed.

Banus has presented one table in which he has compared the
rate of regeneration of three equal pieces, each 10 mm. in length,
from different levels of the same stem. I have not repeated these
experiments as they seem to be lacking in point. The reason
why the apical pieces in these experiments of Banus regenerate
more slowly than the middle pieces is doubtless, as in the case
of the other experiments reported in his paper, the consequence
of an erroneous method of cutting the apical pieces. That some

THE AXIAL GRADIENTS IN HYDROZOA. 393

basal pieces may precede the middle pieces is to be expected on
the basis of what has already been said. It should be obvious
that in stems as long as 30 mm., the length required for these
experiments, the basal regions must already be more or less
physiologically isolated as new individuals whether branches are
present or not. Therefore it may be expected that at least some
of these basal pieces will regenerate more rapidly than the
middle pieces. It is furthermore to be remarked that the
metabolic gradient is steepest near the hydranth and gradually
diminishes in slope down the stem; and it has never been claimed
by us that any marked axial difference exists along the basal
parts of the stems of hydroids. Indeed, we believe that in many
cases the gradient has disappeared in these basal regions, as
shown by the tendency for such levels of the stem to produce
numerous adventitious buds, irregularly arranged, while in the
more distal levels of the stem, bud formation proceeds in a very
definite and orderly manner.

4. The Effect of Depressing Agents. — It has been pointed out
by us on numerous previous occasions that a certain relation
exists between metabolic rate and depressing agents, such that
regions of higher metabolic rate are more affected by depressing
agents than regions of lower metabolic rate. If this general
statement is correct, and various lines of evidence establish its
accuracy, then it should be possible to reduce, eliminate, or
reverse the differences in rate of regeneration that normally
exist between apical and basal pieces. This is the case. Only
two depressing agents were employed, ethyl ether and potassium
cyanide. Apical and basal pieces of equal length were cut in the
usual way and both exposed to the same concentration of these
substances, made up in sea-water, for a certain length of time.
The pieces were then thoroughly washed in several changes of
sea-water and completed their regeneration in normal sea-water.

These experiments are presented in Tables IX. and X. In
Table IX. are given the results of all the mass experiments
performed with cyanide. They were performed in June, at
room temperature, except number 22, which was placed in the
refrigerator later. The concentration of cyanide used and the
number of hours during which the pieces were exposed to it are

394

LIBBIE H. HYMAN.

given in the table. In experiment 7, the concentration employed,
1/20000 mol., was too weak to produce any effect, but the effects
of i/ioooo and 1/5000 mol. solutions are very striking. The
rate of regeneration of both apical and basal pieces is retarded,
but that of the apical pieces, in accordance with the hypothesis,
is more retarded so that the basal pieces regenerate on the whole

the more rapidly.

TABLE IX.

RECORD OF MASS EXPERIMENTS ON THE RATE OF REGENERATION OF APICAL AND
BASAL HALVES WHEN BOTH ARE EXPOSED FOR A NUMBER OF HOURS AFTER
CUTTING TO THE SAME CONCENTRATION OF POTASSIUM CYANIDE.

Columns give numbers of hydranths emerged at hours indicated. Exp. 10
control for exp. 13; exp. 21 for exp. 22. See Table III.

Exp. 7.
KNC 1/20000 Mol. for 20 Hrs.
After Cutting.

Exp. 13.
KNC i/ioooo Mol. for 12 Hrs.
After Cutting.

Exp. 22.

KNC 1/5000 Mol. for 9 Hrs.
After Cutting.

Hrs.

a.

b.

Hrs.

a.

*•

Hrs. a.

b.

36

I

0

60

I

o

84

i

0

40

2

0

62

2

2

93

2

O

46

5

2

*69

3

7

put into refrigerator

48

ii

3

72 8 9

50

13

8

74 12

14

178

2 I

*58

18

16

76 13

16

205

7

3

61

22

18

78

15

20

2IO

8 7

64

23

20

80

17

20

216

8 9

67

21

83

20

23

*227

ii 13

73

23

85

24

24

229

ii 14

*93

34

36

231

12

16

IOI

35

40

237

13

20

106

35

41

*249 16

22

108

36

42

256

17

22

in

37

264

19

22

*I20

38

*273

23

23

276

24

23

280

24

26

283

24

29

287

26

29

*297

28

30

300

29

305

30

The results of the experiments with ether are given in Table X.
Experiments 14 and 20 are mass experiments; 18 and 46 give
records of the number of hours required for the regeneration of
each piece. Experiments 14, 20, and 18 were performed in June
at room temperatures, except that exp. 20 was kept in the
refrigerator for part of the time after regeneration had begun.
Experiment 46 was performed in December at a temperature of

THE AXIAL GRADIENTS IN HYDROZOA. 395

12° C. ± 2. It will be observed that in the case of the mass
experiments, the basal pieces on the whole regenerate more
rapidly than the apical pieces, although both are retarded as
compared with the controls. The individual experiments bring
out the same point. The number of deaths was considerable
but of 39 pairs in which both pieces regenerated, the basal pieces
preceded the apical pieces in 14 cases, or 38 per cent, as compared
with the result under normal conditions as given in Table V.,
where but 8 per cent, of the basal pieces precede the apical pieces.
A number of interesting points are brought out by these experi-
ments with ether and cyanide. In the first place the rate of
regeneration is greatly retarded. In the case of cyanide, where
different concentrations were employed, the retardation is pro-
portional to the concentration used. This retardation is evi-
denced by both kinds of pieces, the apical pieces being, however,
more retarded than the basal pieces, with the consequence that
the usual relation between the time of regeneration of apical and
basal pieces is reversed. Now there can be little doubt that the
rate of regeneration of pieces of Tubularia primarily depends
upon the rate of chemical processes in those pieces. The effect
of temperature upon the rate of regeneration is sufficient proof
of this. Therefore, since depressing agents retard the rate of
regeneration, it is impossible to doubt that they bring about this
effect by lowering the rate of chemical reactions in the pieces.
This is further evidenced by the fact that concentrations of these
reagents which are effective at room temperatures are entirely
without effect at temperatures of 12° and 13° C. Thus I per
cent, ether is very effective at 20° C. but has no effect at 12° C.
In order to alter the relations of apical and basal pieces at 12° C.,
it is necessary to use 2 per cent, ether. The same is true of
cyanide. When, therefore, the metabolic rate is already lowered
by low temperature, the action of depressing agents is diminished.
This further supports the statement made at the beginning of
this section that the action of depressing agents is related to the
rate of chemical activity of the protoplasm which is exposed to
them, and that such effects are greater the higher the metabolic
rate of the living material. The differential effect, therefore, of
ether and cyanide on the rate of regeneration of apical and basal

396

LIBBIE H. HYMAN.

TABLE X.

RATE OF REGENERATION OF APICAL AND BASAL HALVES WHEN BOTH ARE EXPOSED
FOR A NUMBER OF HOURS AFTER CUTTING TO THE SAME CONCENTRATION OF
ETHER.
Experiments 14 and 20, mass experiments; experiments 18 and 46, individual

experiments. Exp. 10, control for exp. 14; exp. 17 for 18 and 20; exp. 45 for 46;

in Tables III. and V.

Mass Experiments.

Individual Experiments.

Exp. 14, i <~-c Ether
for 12 Hrs. After
Cutting.

Exp. 20. i <7 Ether
for 12 Hrs. After
Cutting.

Exp. 18, i «7, Ether
for 15 Hrs. Atter
Cutting.

Exp. 46, 2 <~( Ether
for 17 Hrs. After
Cutting.

Hrs.

a.

b.

Hrs.

a.

b.

No.

rt.

b.

No. a.

b.

53

0

i

60

3

3

I

70

74

i dead dead

55

2

i

65

8

7

2

*87

US

2 dead *I37

59

5

4

67

8

12

3

*s-

90

3

dead dead

61

7

6

69

10

14

4

115

64f

4

dead

dead

*68

10

12

73

13

19

5

68

dead

5

9i

160

7i

10

14

75

U

19

6

74

94

6

IOO

pit

73

14

16

*83

16

25

7

92

dead

7

94-5

95

75

16

19

88

17

8

87

US

8 137

I25t

77

20

23

90

19

9

99

dead

9 73

88

79

22

24

92

20

10

dead

99

10 89

79t

81

25

28

ii

99

96 f

ii 113

127

83

26

29

12

74

7of

12 ; dead

125

*9I

30

32

13

dead

dead

13

127

77t

94

31

34

14

dead

87

14

dead

dead

99

33

35

15

87

dead

15 7i

9i

101

36

36

16

dead

76

16 100 7if

106

36

36

17

94

89t

17

95

*I37

*H5

37

36

18

96

dead

18

74

*89

129

37

37

19

90

*I07

19

*88

90

*i4O

37

38

20

63

63

20

68

dead

21

*88

112

22

90

IOO

23

*88

*88f

24

IOO

94t

25

*88

90

26

78

*n6

27

dead

US

.

•

28

75

89

29

89

77t

30

dead

77

31

69

dead

32

77

9i

33

dead

77

34

69

*iii

35

*m

87t

36

91 101

37

89

98

38

*in

Pit

39

7i

87

40

dead

dead

Total number of regenerated pairs ,

Number where a preceded

Number where b preceded

Number where a and b equal

• 39

.24 or 61%
. 14 or 38%

i

THE AXIAL GRADIENTS IN HYDROZOA. 397

pieces indicates very clearly that the apical pieces have a higher
rate of chemical activity. They are more affected by the de-
pressing agents and more retarded.

5. The Effect of Cutting the Distal End of the Apical Piece at the
Base of the Hydranth. — When questioned regarding his manner
of cutting the pieces for his experiments, Banus replied as follows
(I quote verbatim from his letter): "The most distal cut was
usually made as near as possible to the hydranth without includ-
ing any part of it. Other times more basal parts were used.
No difference in the results was found." To two subsequent
letter* requesting more specific statements concerning this matter
and asking for a diagram showing the exact relation of the most
distal cut to the base of the hydranth, Banus returned no replies.
The first sentence quoted leaves little doubt that Banus made his
most distal cut just below the base of the hydranth, therefore
including in the apical pieces, the little neck or stalk region of
the hydranths. The rest of Banus's statement is too vague to
merit any attention. What is meant by "more basal parts"?
How is one to know in the experiments reported by Banus in
which cases the distal cut was made at the base of the hydranth,
and in which cases more basally? Certain it is that in some of
Banus's pairs of pieces the apical piece emerges first, and in
others the basal piece. This indicates some great irregularity in
his method of procedure. Probably those cases where the apical
pieces emerged first are the ones in which "more basal parts
were used." In the absence of more definite information, specu-
lation is idle. We are here concerned with the fact that usually
the distal cut was made at the base of the hydranth.

I have performed three experiments in which the apical pieces
were cut in the manner usually employed by Banus and as
represented in text-figure 5. Such apical pieces include the
stalk of the hydranth. This stalk is incapable of regeneration.
It together with that portion of the coenosarc which occupies the
distal end of the perisarc dies away and disintegrates. This
process of death and disintegration of the apical end of apical
pieces cut in this manner naturally delays the regeneration of the
apical pieces, because regeneration does not begin until the end
of the piece has rounded off and become covered with a layer of

LIBBIE H. HYMAN.

cells. But this is not the only retarding factor in such pieces.
The coenosarc after the death of the apical end withdraws into
the perisarc leaving a short apical region of empty perisarc-
This empty perisarc crumples to a greater or less extent. There-
fore when the hydranth does regenerate it has to push out
through this empty region before it can unfold, and this of itself
would further delay the time of emergence of the oral hydranth;
but to make matters worse, the crumpling of the empty perisarc
renders it very difficult for the hydranth to push its way to the
surface. On account of all of these factors, the regeneration of
the apical pieces is very greatly delayed when they are cut in
the manner employed by Banus. In fact, in many cases, the
oral hydranth is so greatly retarded that the aboral hydranth
emerges first, and in a few cases, the oral hydranth never emerged
on such pieces, a complete reversal of polarity with disappearance
of the primordium of the oral hydranth having been observed.
Presumably Banus failed to notice whether oral or aboral hy-
dranths had emerged, but the two ends of such pieces are easily
distinguished by the bit of empty perisarc so that there is no
doubt of the correctness of my statements. Not only are the
oral hydranths of these pieces delayed but they are often ab-
normal in appearance; they are enlarged and distended, owing
probably to the pressure to which they are subjected in being
forced out through the crumpled perisarc, and their tentacles are
short and stumpy. They regulate to normal within a few hours
after they have emerged. In two or three cases, partially
doubled hydranths were produced.

The three experiments performed with pieces cut in the way
employed by Banus and as represented in Fig. 5 are presented in
Table XI. They were performed in December and regenerated
at a temperature of 12° C. ± 2. Experiment 31 consisted of
pieces 5-8 mm. long, the other experiments of pieces 10-12 mm.
long. In connection with experiments 39 and 49, the number
of both oral and aboral hydranths emerged on the apical pieces
at each observation is given. These records include of course
only those cases in which the aboral hydranths emerged first.
In some of these cases the oral hydranths subsequently emerged,
and this is indicated by the number in parenthesis which follows

THE AXIAL GRADIENTS IN HYDROZOA.

399

the number of pieces having aboral hydranths. In experiment
49, six such pieces had failed to give rise to oral hydranths when
the experiment was concluded and in three of these cases, no
primordia of the oral hydranths were present, the polarity having
been completely reversed.

The data given in Table XI. show in a very striking manner
that the regeneration of apical pieces cut so that their distal ends

TABLE XI.

RECORD OF MASS EXPERIMENTS ON THE RATE OF REGENERATION OF APICAL
AND BASAL PIECES OF EQUAL LENGTH WHEN THE DISTAL END OF THE APICAL
PIECES is TAKEN AT THE BASE OF THE ORIGINAL HYDRANTHS.

Exp. 31.

Exp. 39.

Exp. 49.

Mrs.

it.

b.

Hrs.

a

Oral.

a
Aboral.

6.

Hrs.

a
Oral.

a
Aboral.

6.

60

i

2

62

5

o

6l

0

I

*69

8

19

64

12

2

65

3

2

71

12

23

66

16

23

67

7

8

73

13

28

68

25

38

69

8

15

76

16

33

70

29

43

71

13

19

78

21

39

72

31

44

73

22

I

28

80

28

47

75

34

45

75

25

2d)

36

82

33

49

*82

42

47

*S5 31

4(l)

46

84

38

49

85

43 , 2

48

87

32

7d)

47

87

44

49

87

45 3

49

89

34

8(1)

49

*94

47

52

89

48

3

50

9i

35

10(2)

49

101

48

9i

48

3

5i

93

38

10(2)

49

103

49

95

49

3

Si

96

40

10(2)

49

105 50

*I07

50

3

52

98

40

10(2)

50

107

52

109

5i

3d)

*IIO

44

10(4)

in

52

3(3)

are just at the base of the original hydranths is markedly delayed
and that in the majority of cases, the basal pieces regenerate first.
It should be remarked that the delay is chiefly in the time of
emergence of the oral hydranths and not in its formation; for
the primordia of the hydranths in these apical pieces form in
advance as a rule of those of the corresponding basal pieces;
but these hydranths can not emerge as rapidly owing to the
fact that they must be pushed through the piece of empty
crumpled perisarc left by the death of stalk region.

The data in Table XI. furnish the explanation of Banus's
results. It is obvious that anyone who practices the method of
cutting the apical pieces described in connection with this table
and who mixes up such a method with procedures where "more

4-OO LIBBIE H. HYMAN.

basal parts are used" can expect nothing but irregular and
inexplicable results. By using such methods and failing to de-
scribe them it is possible to accumulate data which appear to
contradict everything that previous workers have obtained. As'
long as no one takes the trouble to inquire by what procedures
such data were obtained and as long as the author refuses to
furnish any information about his methods, such data might
stand on record indenfiitely in the scientific journals to puzzle
future investigators. I believe that I have conclusively shown
that Banus's data are completely invalidated by his experimental
method. This work and that of previous investigators — Driesch,
Morgan, Child, Stevens, and Alice — demonstrate incontestably
that in Tubularia when other factors are equal the rate of re-
generation of pieces of equal size depends upon the level which
they occupied in the intact stem; it is more rapid the nearer the
pieces lie to the original distal end of the stem. A metabolic
gradient exists in the stem of Tubularia which is the primary cause
of these regional differences in rate of regeneration.

H. SUMMARY.

1. This experimental work was undertaken as a reply to a
paper published by Banus ('18).

2. The existence of a metabolic gradient in the stem of Tubu-
laria is demonstrated in this paper in four different ways.

(a) Differential susceptibility of apical and basal regions of
the stem to ether and cyanide. Apical regions are more suscep-
tible.

(b) Differential capacity of apical and basal regions to reduce
potassium permanganate. The apical end of the organism has
the greatest reducing power.

(c) Difference in electrical potential along the stem. Apical
regions are electronegative (galvanometrically) to basal levels
within the limits of the individual. (At a certain distance from
the original hydranth of Tubularia a new individual is arising
and the apical end of this is likewise electronegative to regions
anterior to its level.) Since in general electronegativity is asso-
ciated in protoplasm with increased oxidative metabolism, this
difference in electrical potential along the stem of Tubularia is

THE AXIAL GRADIENTS IN HYDROZOA. 40!

evidence that distal levels have a higher metabolic rate than
proximal levels.

(d) Difference in the rate of regeneration of apical and basal
pieces. Work upon this point constitutes the bulk of the paper
and is summarized under the subsequent heads.

3. Apical halves of the stem of Tubularia regenerate oral
hydranths markedly faster than basal halves. In cutting such
pieces it is essential to discard the original hydranth and the
first millimeter or two of the stem; the remaining stem is then
cut into two equal halves. The difference between such halves
has been demonstrated by:

(a) Mass experiments in which all of the apical halves have
been placed in one dish, the basal in another. In such cases,
the number of apical pieces which have regenerated oral hy-
dranths is nearly always in excess, rarely equal to, and never
less than the number of basal pieces which have regenerated.

(b) Individual experiments, in which the number of hours
required for the emergence of the oral hydranth on each piece
was recorded. The apical pieces regenerated oral hydranths
first in 91 per cent, of the cases (122 pairs of pieces observed).

4. Apical pieces regenerate on the whole more rapidly than
basal pieces, even when the latter are twice as long as the apical
pieces. Such apical pieces must not however be less than 5 mm.
in length. In pieces over 10 mm. in length, length has very
little effect upon the time of regeneration; in pieces less than
10 mm. in length, the longer pieces regenerate faster than shorter
ones having their distal ends at the same anterior level but this
effect of length is not sufficient to overcome the influence of
level except in very short pieces (under 5 mm.).

5. The difference in rate of regeneration of apical and basal
pieces which exists under normal condition can be somewhat
reduced by using stems bearing branches and cutting the apical
piece above the branch and the basal piece below the branch.
Since the first branch marks the limit of the Tubularia individual,
the apical pieces above such branches are really the basal regions
of the principal Tubularia individual, and the basal pieces below
the branch are near the apical end of the second individual.
In consequence of these relations, the difference between the

4O2 LIBBIE H. HYMAN.

time of regeneration of such apical and basal pieces is less than
is the case when pieces are cut from corresponding levels of
stems without branches.

6. The difference in rate of regeneration of apical and basal
pieces of the stem of Tubularia can be reversed by putting both
sets of pieces for a certain time after cutting into appropriate
concentrations of depressing agents like cyanide and ether.
Under such circumstances the basal pieces regenerate in advance
of the apical ones on the whole. This is due to the fact that
depressing agents affect most strongly those regions having a
higher rate of chemical activity. Since apical pieces have a
higher metabolic rate than basal pieces, they are more affected
by the same concentration of depressing agent and hence their
regeneration is more retarded. In such cases the basal pieces
regenerate the more rapidly. That this explanation is correct
is further evidenced by the fact that the action of depressing
agents is greatly influenced by temperature. At lowered tem-
peratures a higher concentration of the agent must be employed
to obtain the same effect produced at higher temperatures by
lower concentrations.

7. These results are in accord with those obtained by a number
of previous investigators and are directly opposed to the results
presented by Banus. Banus claims that there is no difference
on the average between the time of regeneration of oral hydranths
on apical and basal pieces of the stem of Tubularia. Personal
communication with Banus has elicited the fact that his usual
method of cutting the apical pieces was erroneous. He cut them
in such a way that the distal end of the apical piece was taken
just below the base of the hydranth. In such cases, as shown
in this paper, the distal ends of the apical pieces die away and
the regeneration and time of emergence of the oral hydranth
on these pieces is greatly delayed. It is believed that Banus's
results are invalidated by such a method of procedure.

8. The results presented in this paper together with those of
others quoted in the paper show that the rate of regeneration of
pieces of Tubularia depends, when other factors are equal, upon
the level which those pieces occupied in the intact stem; it is
more rapid the nearer the pieces lie to the original distal end of

THE AXIAL GRADIENTS IN HYDROZOA. 403

the stem. A metabolic gradient exists in the stem of Tubularie
which is the primary cause of these regional differences in rata
of regeneration.

LITERATURE.
Banus, M. Garcia.

'18 Is the Theory of Axial Gradient in the Regeneration of Tubularia supported

by facts? Jr. Exp. Zool., Vol. 26, pp. 265-275.
Behre, Ellinor H.

'18 An Experimental Study of Acclimation to Temperature in Planaria

dorotocephala. BIOL. BULL., Vol. 35, pp. 277-317.
Child, C. M.

'07 An Analysis of Form Regulation in Tubularia. IV. Regional and Polar
Differences in the Time of Hydranth-Formation as a Special Case of
Regulation in a Complex System. Arch. f. Entw'lungsmech., Vol. 24,
pp. 1-28.

'iga Demonstration of the Axial Gradient by Means of Potassium Perman-
ganate. BIOL. BULL., Vol. 37, pp. 133-147.

'igb The Axial Gradients in Hydrozoa. II. Susceptibility in Relation of
Physiological Axes, Regions of Colony, and Stages of Development of
Certain Hydroids. BIOL. BULL., Vol. 37, pp. 101-125.
Driesch, H.

'99 Studien iiber das Regulationsvermogen der Organismen. II. Quantitative
Regulationen bei der Reparation der Tubularia. Arch, f . Entw'lungsmech. ,
Vol. 9, pp. 103-136.
Hyman, L. H.

'18 Suggestions Regarding the Causes of Bioelectric Phenomena. Sci., N.S.,

Vol. 48, pp. 518-524.
Loeb, J.

'91 Untersuchungen zur Physiologischen Morphologic der Thiere. I. Ueber

Hetermorphose.

'92 Idem. II. Organbildung und Wachstum.
Mathews, A. P.

'03 Electrical Polarity in the Hydroids. Am. Jr. of Physiol., Vol. 8, pp. 294-

300.
Moore, A. R.

'10 The Temperature Coefficient for the Process of Regeneration in Tubularia

crocea. Arch. f. Entw'lungsmech., Vol. 29, pp. 145-148.
Morgan, T. H.

'01 Regeneration in Tubularia. Arch. f. Entw'lungsmech., Vol. n, pp. 346-

381.
'05 Polarity Considered as a Phenomenon of Gradation of Materials. Jr. Exp.

Zool., Vol. 2, pp. 495-506.

'o6a The Physiology of Regeneration. Jr. Exp. Zool., Vol. 3, pp. 457-500.
'o6b Hydranth Formation and Polarity in Tubularia. Jr. Exp. Zool., Vol. 3,

pp. 501-515.
'08 Some Further Records Concerning the Physiology of Regeneration in

Tubularia. BIOL. BULL., Vol. 14, pp. 134-149.
Morgan, T. H., and Stevens, N. M.

'04 Experiments on Polarity in Tubularia. Jr. Exp. Zool., Vol. i, pp. 559-587.

THE GERM-CELLS OF CICADA (TIBICEN)
SEPTEMDECIM (HOMOPTERA).1

ELMER L. SHAFFER,
DEPARTMENT OF BIOLOGY, PRINCETON UNIVERSITY.

CONTENTS.

A. Introduction 405

B. Materials and Methods 406

C. Observations on Chromosomes AO7

1. Diploid chromosome groups 407

(a) Spermatogonia 407

(b) Follicle-cells 408

(c) Embryonic cells 410

(d) Somatic cells 410

2. Spermatocytes 411

(a) Growth stages 411

(b) Maturation tetrads 412

(c) Maturation Divisions 413

(d) Giant spermatocytes 414

3. Growth Stages and Synapsis in Oocyte 415

4. Chromatin Nucleoli in Oocyte 417

5. Discussion of Chromosomes 419

(a) Review 419

(b) Ring Tetrads 421

(c) Synapsis 422

D. Observations on Mitochondria 423

1. Mitochondria in Spermatogenesis 423

(a) Spermatogonia 423

(b) Spermatocytes 425

(c) Transformation of Spermatid 427

(d) Discussion 427

2. Nutrition of Egg 429

3. Mitochondria in Oogenesis 433

(a) Mitochondria in growth stages 433

(b) Yolk formation 435

(c) Discussion 438

4. Significance of Perinuclear Zone 441

(a) Relation to yolk-nuclei 441

(b) Origin of Mitochondria 443

E. General Considerations 446

1. Chromosomes 446

2. Mitochondria 448

F. Summary 450

1 A thesis presented to the Faculty of Princeton University, in candidacy for the
degree of Doctor of Philosophy.

404

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 405

A. INTRODUCTION.

The study presented here was done at Princeton University
during the last year. The material was collected during the
appearance of the ly-year locust (Cicada septemdecim) in the
vicinity of Princeton, N. J., during the spring of 1919 at the
suggestion of Professor E. G. Conklin. With such an interesting
life cycle as these insects have, it was thought that a cytological
study of their germ cells might reveal some important facts in
the history of the mitochondria and the chromosomes. Both
of these are constant and important structures of all cells, and
it is the opinion of the writer that no cytological study can be
complete which neglects either one or the other. It is only by a
correlated study of cytoplasmic and nuclear structures that we
can ever hope to solve the many perplexing questions in cell
economy. We use the almost mystic phrase: "interaction
between nucleus and cytoplasm" in many cases to cover our
ignorance concerning certain cell activities, but we are far from
knowing any of the specific actions and reactions between the
nucleus and cytoplasm. WTithin recent years the study of mito-
chondria of animal and plant cells has attracted many workers,
with the result that many cytologists have come to regard these
structures as of vital importance in cell activities. Attention,
which for many years has been centered on the nuclear activities,
has been drawn to a more intensive study of the cytoplasm and
its structures. All these studies have emphasized the importance
of such structures as mitochondria in relation to cell metabolism.
Moreover, although the chromosome hypothesis of heredity
seems to be firmly supported by a vast amount of evidence, yet
one group of cytologists (Meves, Benda, Duesberg, etc.) maintain
that the mitochondria also have a role as the bearers of hereditary
characters. While the evidence which these workers have
gathered is not of a convincing nature, nevertheless the facts are
worthy of careful consideration. There are some reasons for
believing that inheritance in some cases is through the cytoplasm,
and we must not loose sight of the fact that "cytoplasm as well
as nucleus is concerned in heredity and differentiation" (Conklin,
'16). However, whether or not such cytoplasmic structures as

406 ELMER L. SHAFFER.

mitochondria constitute the idioplasm is an entirely different
problem. Modern researches in genetics have shown that,
despite mutations, hereditary characters are relatively stable
and that the hereditary constitutions of organisms are definitely
organized; hence the idioplasm which is causal in the develop-
ment of the hereditary characters must similarly be stable and
highly organized.

Keeping in mind the "cell as a whole," I have studied the
chromosomes and mitochondria in the oogenesis and spermato-
genesis of Cicada, and my observations give added evidence to
the chromosomes as the idioplasmic substance, while there is no
evidence from an unbiased standpoint that the mitochondria
behave as idioplasmic substances. There is evidence, however,
that the mitochondria are intimately concerned in cell meta-
bolism.

Throughout this work I have had the constant advice and
encouragement of Prof. E. G. Conklin, and it is with great
pleasure that I here express my indebtedness.

B. MATERIALS AND METHODS.

The youngest specimens of Cicada obtained were those of the
second pupal stage about three weeks prior to their emergence
from the ground and their final moult into the imago. These
specimens were collected about the middle of April by digging
under trees in the vicinity of Princeton. The pupae were found
lying about a foot from the surface of the ground and were most
abundant several feet away from the tree trunks. In the testes
of such pupae, one finds most of the cells in the maturation stages
besides an abundance of spermatids, spermatozoa, and a few
spermatogonia. In the adult or imago, the testes are almost
completely filled with sperm except for a small number of
spermatogonia. After copulation, the testes are reduced to about
one tenth their former size and contain only a small residue of
spermatozoa, some degenerated cells and a few spermatogonia
which also show signs of degeneration.

There are two testes, each consisting of a great many radiating
ellipsoidal follicles which give the testes a berry-like appearance.
In the female there are two typical ovaries, each consisting of a

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 407

great number of ovarian tubules containing a great many oocytes.
The oldest oocyte of a second stage pupa is about one seventieth
the linear size of the oldest oocyte of the adult, which shows
the tremendous growth that takes place in a few weeks. It has
been estimated (Marlatt, 1898) that the female Cicada lays
between 400 and 600 eggs.

The male and female gonads were dissected out in Ringer's
solution and fixed in either Flemming's (strong), Bouin's,
Benda's, or Regaud's fixing fluids. Mitochondria were well
preserved by the Flemming, Benda and Regaud fluids, but were
either partially or wholly destroyed in Bouin's fluid depending
on the length of time the gonads remained in the fixing fluid.
Material fixed in Flemming's fluid (10 to 12 hours) was usually
the best for studying both chromosomes and mitochondria.
Sections were cut 8 to 10 micra in thickness and the stains used
were iron-haematoxylin (with and without counterstain), Benda's
crystal-violet and alizarin, and Altmann's fuchsin-methylene
green.

The developing eggs were also collected from time to time for
a study of the chromosomes of the embryonic cells. These were
fixed in either Bouin's or Carnoy's fluids and were imbedded by
the celloidin-paraffine method.

C. OBSERVATIONS.
i . Diploid Chromosome Groups.

(a) Spermatogonia. — The spermatogonia are found in the
proximal end of the ellipsoidal follicles of the testes. They form
a cap of cells at this end of the follicle containing the primary
and secondary spermatogonia. From the proximal end of the
follicle there proceeds a short narrow filament which contains
the spermatogonia of the multiplication stages. Mitotic figures
are quite abundant among the spermatogonia of the multiplica-
tion stages, but the metaphase plates are usually so crowded
that it is impossible to make accurate counts of the chromosomes.
On the other hand, the primary and secondary spermatogonia
rarely show cell divisions, but the metaphase plates when they
appear are very clear.

The male chromosome number is 19, which indicates that there

408 ELMER L. SHAFFER.

is present an unpaired sex element (Figs. I and 2). One pair of
the complex is strikingly larger than the rest, being in the form
of somewhat curved rods (Text-fig, i). This pair corresponds
to the " macrochromosome " pair described by Kornhauser ('14)
in Enchenopa.

Two other chromosome pairs (BB, CC, Text-fig, i) can also
be distinguished from the other chromosomes by their size,
being approximately half the size of the macro-chromosomes
(A A). The other 13 chromosomes show no size differences
which would enable us to arrange them in pairs or distinguish

TEXT-FIG, i. Spermatogonial chromosomes, showing the relative sizes of the
chromosomes; the macrochromosome pair, A A, the BB and CC pairs, and 13
other chromosomes which show no size differences.

them from each other. However, as will be later shown, the
AA, BB, and CC chromosome pairs are so characteristic in their
form and size that they can be recognized in all the diploid
groups. The size relations of the chromosomes of Cicada corre-
spond to those described by Kornhauser ('14) in Enchenopa.
Also in the Cercopidce (Homoptera), Boring ('13) has described
three pairs of chromosomes (A, B, C) which bear similar size
relations to the three pairs here described in Cicada.

The sex-chromosome cannot be identified in the spermatogonial
groups either by its size nor by any peculiarities in its staining
reactions. In the resting spermatogonia there is, however,
always present a single chromatin nucleolus (Fig. 13) which
probably is the persisting sex chromosome.

(b) Ovarian Follicle Cells. — Among the follicle-cells of the
ovary, mitotic figures are very abundant. I have found a great
many mitoses not only among the follicle-cells surrounding the
young growing oocytes, but even among the follicle-cells sur-
rounding the almost mature oocytes. I have searched for evi-
dences of amitotic division among these cells, which has often

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 409

been described, especially in the follicles surrounding mature
oocytes; but I have failed to find any strong evidence for the
occurrence of this method of nuclear division. The only indica-
tion that amitotic division may take place is found in the follicles
surrounding the old oocytes. Here the cells are usually bi-
nucleate and appear to be the end stages of amitotic divisions.
However, I have found many karyokinetic figures among such
cells and it seems reasonable to suppose that the binucleate
follicle-cells arise through the failure of the division of the cell
body following mitotic division of the nucleus.

All metaphase plates of dividing follicle-cells show 20 chromo-
somes (Figs. 3, 4, 5, 6, Text-fig. 2), and it is possible in all of
these to recognize the chromosome pairs AA, BB, and CC as in
the spermatogonial plates, by their size relations. Figure 12
is that of the late telophase of a dividing follicle cell in which

c 9

/ET*

«i«:.*

TEXT-FIG. 2. Two metaphase plates from ovarian follicle-cells showing the
A A, BB, CC pairs of chromosomes; note that the A A pair lies in the center of
the group.

one of the macrochromosomes (A) can be recognized in the
daughter cells. It will also be seen that there is present a
precocious longitudinal split of the chromosomes in preparation
for the next cell-division.

In the follicle-cells surrounding the older oocytes, the chromo-
somes do not have the thick compact appearance found in the
younger follicle-cells. The chromosomes- are usually thinner,
poor in chromatin content, and the respective chromosomes
appear somewhat longer, often showing an equational split
(Figs. 10, n). I am unable to account for this difference in
appearance of the chromosomes (compare Figs. 7, 10, n). In

4-IO ELMER L. SHAFFER.

the resting nuclei of the old follicle-cells, it is very noticeable that
they are very poor in chromatin content, all the basichromatin
being accumulated in one or two small masses (nucleoli). When
such cells prepare for mitosis, the chromosomes which are re-
constituted are correspondingly poor in chromatin. However,
the linin basis of the chromosomes is still present and conse-
quently the number and size relations of the chromosomes is
maintained. This, I believe, gives added evidence to the view
previously expressed (Shaffer, '20) that the linin is the morpho-
logically stable substance which maintains the chromosomal
organization and structure.

(r) Embryonic Cells. — Although a number of the developing
eggs of Cicada were collected, I was unable to obtain good
material for a study of the chromosomes due to the difficulties
in sectioning. The eggs were so full of yolk that it was possible
to cut them only by imbedding by the celloidin-paraffine method.

Figure 9 is that of a metaphase plate of a cell from the blasto-
derm showing 20 chromosomes, and hence of the female type.
The chromosome pairs AA, BB, and CC can be distinguished
as in the other diploid nuclei. While I have been unable to
make an exhaustive study of the chromosomes of the embryonic
cells, yet I have found no variations in the chromosome numbers
or in their size relations.

(d) Somatic Cells. — On dissecting the female locusts to remove
the ovaries, a number of round, brown-pigmented bodies re-
sembling eggs were found in the abdomen. On sectioning these
it was found that they were of a glandular nature and are perhaps
concerned in the secretion of adhesive materials for the eggs.
In a small cap of cells which lies at one end of these glandular
bodies, mitotic figures were found in abundance. There are
alwrays 20 chromosomes which show similar size relations to
the diploid chromosome groups previously described. After
mitotic division of the nucleus the cell-body fails to divide,
resulting in the formation of binucleate and multinucleate cells.
At the time of division of such multinucleate cells, typical tri-
asters are formed (Fig. 52). In the telophase of such a division,
the tripartite daughter nuclei reconstruct to form six separate
nuclei, the cell-body again failing to divide. At this time the

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 4! I

cytoplasm becomes active in the secretion of large globules and
although the nuclei increase in number, I have been unable to
find mitotic figures and it seems possible that the increase in their
numbers is brought about by amitosis.

2. Spermatocytes .

(a) Growth Stages. — Unfortunately the material wh'ch was
collected showed very few of the early growth stages of the
spermatocytes and consequently I was unable to make a detailed
study of the process of synapsis. Apparently the early growth
stages must take place some time before the month of April.
In the youngest pupae which I have collected, the only growth
stages of the spermatocytes which I have been able to find are
those of the pachytene-bouquet stage (Fig. 17) in which the
thick synaptic threads are polarized at one side of the nucleus.
Sections across the bouquet usually show 18 chromatic blocks
representing the end view of the threads. Since each loop has
been cut twice, this would indicate that there are 9 pachytene
loops. At the base of the polarized bouquet is usually found the
compact, deeply-staining nucleolus (Fig. 17, A") which is the
persisting sex-chromosome. This is undoubtedly the same com-
pact chromatic nucleolus found in the resting stages of the
spermatogonia (Fig. 13). As is usual, the bouquet is polarized
toward the pole of the cell containing the idiozome (Fig. 17, id.).
It has often been stated that the idiozome exerts some attraction
on the synaptic threads influencing their polarization. There is
no evidence to support this view and I am inclined to believe
that the same factors which determine at which point in the cell
the idiozome should lie also determines the polarization of the
synaptic threads.

Occasionally the pacyhtene threads show a longitudinal split,
and in such cases it is noticed that the chromatic granules
(chromomeres) of the two halves of the thread do not correspond
either in size or location. Consequently the longitudinal split
cannot be interpreted as an equational split, but is rather the
primary split or point of synaptic union. Usually there is
present a single loop of the bouquet which is much larger than
the other loops (Fig. 17, A A) and this undoubtedly represents

412 ELMER L. SHAFFER.

the synaptic condition of the macrochromosome pair of the
spermatogonial chromosomes.

(6) Tetrads and Maturation Divisions. — Stages of the early
prophases of the spermatocytes were quite abundant and it
was possible to follow the formation of the first maturation
tetrads. In the early prophases, the homologous threads show
a great variety of twisting about each other (Fig. 20). One
pair (A A, Fig. 20) is easily distinguishable from the others by
its large size and is derived from the spermatogonial macro-
chromosomes. Figure 18 represents the stages in the formation
of the definitive tetrad from this pair (Figs. 61, 62, 63, 64). In
the early prophases the homologous threads of the AA tetrad
are very long and twisted about each other. However, they
retain their connections at the ends, thus making the tetrad a
large ring if its twists were straightened out. The space en-
closed by this ring is the interchromosomal space which marked
the point of synaptic union of the threads. In the later prophase
stages, the large ring condenses, the threads become thicker retain-
ing their point of union at the ends and the interchromosomal
space becomes reduced in size until in the definitive maturation
tetrad it is reduced to a small oval slit between the two halves
of the ring (Figs. 18, 21, 62). In the first maturation metaphase,
the macrochromosome tetrad no longer appears in the form of a
ring, but rather in the form of a ring flattened at the poles, or
as two slightly bent rods whose concavities oppose each other.
In a similar way, the tetrad derived from the BB pair of the
spermatogonia goes through the formation of a ring tetrad
(Fig. 19), resulting in a tetrad similar to the macrochromosome
tetrad (A A), but approximately half its size. The other tetrads
show no ring formation; usually the homologous threads become
free at one of the synaptic ends, retaining their connection at the
other end, thus giving the appearance of two chromosomes
joined end-to-end (Fig. 20). The condensation of such tetrads
produces the typical dumb-bell form tetrad, with the narrow
portion of the dumb-bell marking the retained point of synaptic
union. Thus, if the point of synaptic union is retained at both
ends, rings like the A A and BB tetrads are produced; if the
synaptic union is retained at one end only, the dumb-bell type

THE GERM-CELLS OF CICADA (XIBICEN) SEPTEMDECIM. 413

tetrad is produced. Payne ('14) has described a different method
of ring-formation in Forficula. The two univalent chromosomes
are first joined end-to-end; while retaining this point of union,
the free ends come together by a bending process with the result-
ing formation of a ring each half of which represents a univalent
chromosome. A similar method of ring-formation has been
described by Sutton ('02) in Brachystola and by Davis ('08) in
several Orthopterans. In Forficula besides the "bending pro-
cess" of ring-formation, Payne describes ring-formation of the
type here described in Cicada.

(r) Maturation Divisions. — In the metaphase plates of the
first maturation division, the chromosomes are always grouped
in a characteristic manner (Figs. 23, 24, 53 to 57). There are
10 chromosomes in the metaphase plate, 8 of which are arranged
in a circle surrounding the macrochromosome tetrad (A A , Fig. 23)
which always lies in the center of the group. The sex chromosome
(Fig. 23, X) always lies outside this circle of chromosomes and
often does not lie in the same plane. Boring ('07) has figured the
chromosomes of a number of species of Homopterans in which
the sex-chromosome lies outside the group of autosome tetrads.
The constant position of the AA tetrad in the center of the
spermatocyte complex can be traced back to the diploid chromo-
some groups in which the macrochromosome pair shows a marked
tendency to lie in the middle of the metaphase plate with the
other chromosomes grouped about them (text-fig. 2). In a
previous paper (Shaffer, '20) it was pointed out that the char-
acteristic grouping of the chromosomes in the metaphase had its
explanation in the persistence of the interchromosomal linin
fibres (Fig. 24) which undoubtedly persist as a part of the chro-
mosomal architecture and maintain definite spatial relations
between the chromosomes. The evidence in Cicada seems to
support this view.

In the side-view of the metaphase, the AA and BB tetrads
are arranged on the spindle in the direction of the spindle axis.
Hence in polar views only a half of each tetrad can be seen (Fig.
23). In both the A A and BB tetrads the spindle fiber attach-
ments are median (Figs. 18, 19) or atelomitic (Carothers, '14),
and since the tetrads lie in the direction of the spindle axis, they

414 ELMER L. SHAFFER.

will separate in the anaphase at the point of synaptic union and
the division is reductional. The other tetrads have terminal
spindle fiber attachments, but I am unable to say whether they
divide reductionally or equationally. In the case of the CC
tetrad, there is evidence that the narrow portion of the dumb-bell
actually marks the point of synaptic union, and since separation
of the dyads in the anaphase occur at this point, the division is
also reductional.

The sex-chromosome (Fig. 25, X} usually lies on the outer
surface of the spindle. In the anaphase it usually lags behind
the other dividing chromosomes, sometimes appearing bipartite,
and passes undivided to one of the daughter cells (Figs. 27, 70).
As the dyads come into the late anaphase of the division, a
secondary (equational) split can often be seen (Fig. 27, A A}.

Following the first maturation division, there is no inter-
kinesis or construction of a telophase nucleus. The dyads again
become arranged in the metaphase, each showing the secondary
split. Figure 28 (also Fig. 58) is that of second spermatocytes
(daughter plates), one having 9 dyads the other having 9 dayds
plus the X-chromosome. It will be noted that the grouping of
the dyads is exactly similar to the grouping of the tetrads in the
first maturation division metaphase, namely 8 dyads arranged
in a circle around the macrochromosome dyad. In Notonecta,
Browne ('16) found that the chromosomes always assumed a
definite grouping in the metaphase, but the grouping was different
in the two maturation divisions. In the anaphase of the second
maturation division all the dyads divide and there are no lagging
chromosomes.

(d) Giant Spermatocytes. — It is quite common to find spermato-
cytes with double or more the number of chromosomes. Figure
59 is a photograph of such a spermatocyte in the metaphase
which has over twice the normal number of tetrads. These
giant spermatocytes develop normally and give rise to giant
spermatids and spermatozoa (Figs. 31, 33<r). The origin of these
giant cells may be traced back to the spermatogonia in which
there has been a failure of division of the cell-body resulting in
cells with double the diploid number of chromosomes. Such
cells are quite common among the spermatogonia of the multi-

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 415

plication stages. Wilcox ('95) has described giant spermatids
and spermatozoa in Cicada tibicen; he also finds that they are
derived from abnormal spermatogonia in which there has been
a failure of the division of the cell-body, and that their develop-
ment insofar as the spermatozoan is concerned follows a normal
course with the production of typical spermatozoa which merely
differ from the normal ones by their large size. It seems difficult
to ascertain the significance of these giant cells or whether there
is any possibility that they play a part in the fertilization of the

egg-

3. Growth Stages and Synapsis in Oocytes.

While it has not been possible to study the process of synapsis
in the spermatocytes due to the absence of the proper stages in
my material, in the oocytes I have been able to follow the various
growth stages in some detail.

The young oocytes in varying stages of growth are found at the
base of the nurse chamber or "Keimlager" (Fig. 34). They are
always distinguishable from the nurse-cells by their definite cell
outline, by the thread-like appearance of the chromatin as com-
pared to the granular nuclei of the nurse cells, and by the presence
of a definite mitochondria! zone. Only occasionally could oogon-
ial divisions be found in the ovaries of the youngest pupae, and
hence I have been unable to trace the chromosomes from the
last oogonial division into the early growth stages. The very
early growth stages (leptotene, etc.) are not found in the material
collected after the latter part of April but in the material col-
lected earlier, there is an abundance of oocytes in the pre-
synaptic stages.

Figure 35 is that of the nucleus of a young oocyte corresponding
to von Winiwarter's "protobroque" nucleus. The chromatin is
in the form of a network with the nodal points of the net staining
somewhat more deeply than the rest of the reticulum. On closer
analysis of this nucleus it is seen that the net-like appearance
of the chromatin is due simply to the optical effect of numerous
delicate chromatic threads crossing each other in all directions.
In the later stages the individual threads become more evident
and this stage (Figs. 36, 74) no doubt corresponds to von Wini-
warter's "deutobroque" nucleus. Gradually these threads as-

416 ELMER L. SHAFFER.

sume more and more a definite individuality, and the nucleus
then becomes filled with a number of very delicate leptotene
threads (Figs. 37, 75). These often show a distinct polarization,
usually being attached at one side of the nuclear membrane
with the free ends suspended in the nuclear sap. During this
stage there is also a marked tendency for the leptotene threads
to become associated in pairs and this pairing becomes more
marked in the later stages (Fig. 38). The threads are now
markedly polarized and it becomes quite clear that they are
actually pairing side-to-side (zygotene stage). This paired ap-
pearance of the threads is not an accidental one, for I have
observed it in a great many cells with great clearness. That it
is also not due to a longitudinal split of a single thread is evidenced
by the fact that the chromatin granules in homologous threads
do not lie at the same level. After this pairing has taken place
and the threads have become well polarized, they gradually
become shorter and thicker (Fig. 39), forming the typical pachy-
tene stage, the threads sometimes showing the primary split or
point of synaptic union. A typical pachytene bouquet stage
follows, in which the loops have both their free ends attached
at one pole of the spindle (Figs. 40, 76). Usually there is one
especially large loop (A A, Fig. 40) similar to the large loop
found in the bouquet stage in the spermatocytes, which un-
doubtedly represents the macrochromosome pair. Cross-sec-
tions of cells in the bouquet stage show the threads on end view
(Figs. 41, 77), and it is possible to count these. Usually there
are 20 such cut ends of the pachytene threads and since each
thread has been cut twice, we can deduce that there are ten
pachytene loops, and it is at once seen that this number corre-
sponds to the reduced number of chromosomes.

The release from the bouquet stage sets free the thick woolly-
looking pachytene threads (Fig. 42) and the primary split comes
clearly into evidence, and often the homologous elements become
separated along the synaptic line. As this separation con-
tinues, the homologous threads become twisted about each other
assuming the typical strepsistene condition (Fig. 43). The
separation may begin at either or both ends of the threads or
they may retain their union at the ends and separate along their

THE GERM-CELLS OF CICADA (xiBICEN) SEPTEMDECIM. 417

middle points. As the strepsistene stage advances, the threads
lengthen considerably, become more lightly staining, and as
they separate more widely their individuality becomes less and
less distinct. In the older oocytes, the chromatin of the germinal
vesicles appears more or less granular, much of it still retaining
its affinity for the basic stains. On close analysis of such nuclei
(Figs. 46, 47) much of the chromatin shows evidences of still
retaining, in part, the arrangement in threads, and I interpret
these as being the original conjugating threads of the early
growth stages which have been greatly expanded.

4. Chromatin Nucleoli.

I shall not attempt to review the literature concerning nucleoli
since there is considerable confusion of interpretation in this
field. It is evident that many of the different opinions expressed
in the literature have arisen from the fact that these structures
are not always homologous. Recently Nakahara ('18) has
advanced the view that in the oocytes of Perla, the nucleoli are
derived from the yolk-nucleus which consists of a dense granular
mass in the cytoplasm closely applied to the nucleus. As will
be shown later, Nakahara's "yolk-nucleus" is really an accumula-
tion of mitochondria, and it is difficult to see how these could
give rise to nucleoli.

In the oocytes of Cicada true chromatin nucleoli are found, and
whenever they are present take the basic stains. The interest
connected with chromatin nucleoli in the oocyte is bound up
with the possible homology of such structures with the chromo-
some nucleoli or persisting sex-chromosomes found in the growth
stages of the spermatocyte. Since one sex-chromosome of the
female is derived from the sperm, there is no a priori reason why
we should not expect to find such a body persisting in the stages
in the oocyte homologous to those stages in the spermatocyte
where it is found to persist as a nucleolus.

In the very young oocyte of Cicada (Figs. 35, 36, 73, 74), there
are usually present two chromatin nucleoli. In the leptotene
nuclei these are no longer found, nor can they be found in the
succeeding pachytene stages. However, in the strepsistene
nuclei, two deeply staining bodies again appear (Figs. 43, 78)

418

ELMER L. SHAFFER.

which are similar to those found in the protobroque and deuto-
broque (preleptotene) nuclei. In the later stages these bodies
become closely associated (Text-fig. 3, Figs. 81, 84) and persist
as definite chromatin nucleoli. A true plasmosome is also present
in these stages which becomes vacuolated and grows in size with
the age of the oocyte (pi., Figs. 44, 46). In the older germinal

$, r n

TEXT-FIG. 3. Oocyte of post-synaptic period, showing mitochondria arranged
in perinuclear zone, and two chromatin nucleoli (cr.n.) and a plasmosome (pi.).

vesicles, the two chromatin nucleoli may persist (Fig. 87), but
more commonly a single large chromatin nucleolus is found
together with six to eight smaller nucleoli (Fig. 46).

It is difficult to say whether the two chromatin nucleoli really
represent persisting sex-chromosomes and are therefore homol-
ogous to the chromosome nucleoli of the spermatocytes. There
is some evidence that this is the case. The disappearance of the
definite nucleoli during the synaptic stages (leptotene and pachy-
tene) leads me to believe that they are resolved into chromatic
threads which go through a synaptic process similar to the
autosome threads. The fact that there are ten pachytene threads
in the bouquet stage indicates that all the chromosomes are in
synapsis. The reappearance of the two chromatin nucleoli in
the strepsistene stages are perhaps brought about by a re-con-
densation of the synaptic threads representing the sex pair
resulting in the formation of the two compact chromatin nucleoli.

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 419

This view is supported by the fact that in those stages where the
nucleoli re-appear, they usually lie close to each other (Figs.
43, 78), and in the strepsistene stages they are more loosely
granular in appearance than in the later stages. The later
increase in number of the nucleoli in the germinal vesicle stages
may be an expression of the increased metabolic activity of the
nucleus during this period.

Wilson ('06) was unable to find chromosome-nucleoli in the
oocytes of Anas, Euschistus and other forms during the "contrac-
tion figure of the synaptic period" and he is inclined to doubt
the persistence of nucleoli in the oocytes homologous to those
found in the spermatocytes. As has been shown in Cicada,
it is quite probable that the sex-chromosomes go through synaptic
stages just like the autosomes and hence do not persist as compact
bodies in these stages as does the odd chromosome of the sperma-
tocytes. However, they may be present in the form of nucleoli
before and after the synaptic period. Foot and Strobell ('n)
have described a chromatin nucleolus in the oocytes of Protenor
which gives rise to two large idiochromosomes at the time of cell-
division. In Gelastocoris, Payne ('12) has described a chromatin
nucleolus which appears after the last oogonial division and
persists until shortly after synapsis. It later becomes reduced
in size and disappears, and Payne interprets it as having been
derived from the sex-chromosomes. The reason for the per-
sistence of the sex-chromosome in the spermatocyte and its non-
persistence in the oocytes during synapsis may lie in the fact
that in the former the odd sex element has no homologue with
which to pair, while in the latter the two sex-chromosomes are
homologous and synapsis becomes possible. However, with the
wide variations which these nucleolar structures exhibit in the
oocytes, it is not possible to make any generalization as to their
homology with the nucleoli of the spermatocytes.

5. Discussion of the Chromosomes in Homopterans.

(a) Review. — The earliest work on the chromosomes of the
Hemiptera homoptera is that of Wilcox ('95) on Cicada tibicen
in which he states that there are 12 chromosomes in the spermato-
gonia and 24 "elements" in the spermatocytes. The work of

42O ELMER L. SHAFFER.

Stevens ('050, '056) and of Morgan ('08, '09) on the Aphids
and Phylloxerans has dealt mainly with the problem of sex
determination. It was found that only spermatozoa bearing the
sex-chromosome are functional and consequently only females
are produced from fertilized eggs. Males are produced partheno-
genetically from eggs which during the maturation processes
reduce the chromosome number to one half the somatic number;
hence there is only one sex-chromosome remaining in the male.
Stevens ('06) and Boring ('07, '13) have studied the chromosomes
of over 20 species of Homopterans mainly in relation to sex
determination. Kornhauser ('14) has made a comparative study
of the chromosomes of two species of Enchenopa in which he
describes parasynapsis and pre-reduction.

In all the Homoptera studies thus far with one exception, the
sex-chromosomes consist of but a single (X) element which
persists as a chromosome nucleolus through the growth stages
of the spermatocyte and passes, undivided to one pole of the
first maturation spindle. The only exception to this is that
described by Kornhauser ('14) in Enchenopa binotata in which
there are two sex-chromosomes (XY). These behave very much
like the autosomes in the growth stages of the spermatocyte,
becoming resolved into synaptic threads and pairing in synapsis
much as the autosomes do. In the preleptotene stages the sex
pair is found persisting as nucleoli. In the strepsistene stages,
when the chromosomes become more diffuse and lightly staining,
the chromatic nucleoli again appear by a condensation of the
threads representing the sex pair. It will at once be noted that
this behavior of the sex pair in Enchenopa resembles the behavior
of the two chromatic nucleoli which I have described in the
oocytes of Cicada and I am, therefore, led to believe that the
chromatic nucleoli of the oocytes are homologous to the chromo-
some-nucleolus of the spermatocytes.

A common characteristic of the diploid chromosome complexes
of many Homopterans is the usual presence of a pair of large
rod-shaped chromosomes (A A pair of Cicada). Stevens ('06) has
shown these in Aphrophora, and Boring ('07) has figured these in
many of the Homoptera that she has studied. Kornhauser ('14)
had described them in two species of Enchenopa that he has

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 42!

studied and applies the name "macro-chromosome" to them.
Besides the macrochromosome pair, there are usually present one
or two chromosome pairs (the BB and CC pairs of Cicada) which
are somewhat smaller than the macrochromosome pair, but which
can easily be distinguished by their size from the other chromo-
somes. The presence of a pair of macrochromosomes and one
or two pairs of somewhat smaller chromosomes therefore seems
to be characteristic of the diploid chromosome groups of the
Homoptera.

(b] Ring Tetrads. — McClung ('14) has described two types of
ring tetrads in the Orthoptera which he terms (i) the Hippiscus
type, and (2) the Stenobothrus type. Both types of ring tetrads
are essentially the same as far as the formation and relation of
their elements (chromatids) are concerned. Half of each ring
represents a whole chromosome and each univalent chromosome
is secondarily split, so that each ring tetrad really consists of two
rings superimposed one upon the other. The difference between
the two types lies in their relation to the spindle-fiber attach-
ments and their position in the metaphase plate. In the Hip-
piscus ring type, the spindle-fiber attachments are at the synaptic
ends (terminal) and the rings lie flat in the metaphase plate.
Consequently polar views of the metaphase plate show the entire
ring, while side views show only half of the ring. In the anaphase
the two superimposed rings separate and consequently the divi-
sion is post-reductional. On the other hand, the rings of the
Stenobothrus type usually have median spindle-fiber attachments
and the ring lies in the metaphase in the direction of the spindle
axis. Only lateral views of the spindle would therefore show the
complete ring, while polar views show only half of the ring. In
the anaphase the two halves of the ring (which are univalent
chromosomes) separate as simple V's and the first maturation
division is pre-reductional.

In Cicada, the large ring of the early prophases is derived from
the macrochromosome pair, AA, of the spermatogonia. In the
spermatogonial divisions this pair has median spindle-fibre at-
tachments and the spindle-fiber attachments of the ring tetrad
derived from this pair are also median (Fig. 18). In the meta-
phase the ring is arranged on the spindle in the direction of the

422 ELMER L. SHAFFER.

spindle axis with the sister chromatids directed toward the same
pole. Polar views of the metaphase (Fig. 55) show only half
the tetrad, but oblique views show the relation clearly (Fig. 57).
Thus the rings of Cicada are of the Stenobothrus type and they
divide reductionally in the first maturation division.

(c) Synapsis. — Although I have been unable to study the
process of synapsis in the male germ cells, my observations on
the oocytes indicates that the chromosomes conjugate para-
synaptically, and there is no reason for supposing that it might
be different in the spermatocytes. From Fig. 38 it is quite
evident that the leptotene threads pair side-to-side, but the
essential point is whether each leptotene thread actually repre-
sents a single univalent oogonial chromosome. Wenrich ('16)
has shown in Phrynotettix that the leptotene threads are derived
from the telophase chromosomes of the last spermatogonial
division by a process of unravelling of the chromatic blocks
contained within separate chromosomal vesicles. In Cicada I
have been unable to trace the oogonial telophase chromosomes
into the early growth stages of the oocyte, but there is some
indirect evidence that each leptotene thread is derived from a
single oogonial chromosome. In the late leptotene stages, the
threads are well polarized and all extend in one direction through
the nucleus. I have studied a great many such nuclei in cross-
section and it is possible to count the number of leptotene threads
cut on end view. Such counts usually approximate the diploid
chromosome number (20) and points to the fact that each lepto-
tene thread is derived from a single chromosome.

As has been noted before (page 417), a casual study of the
germinal vesicles of the oocytes does not indicate any persisting
individuality of the chromosomes; however, careful and minute
analysis of such nuclei reveals the fact that the synaptic threads
are still present but in a much-expanded and diffuse condition.
Similarly in the male germ cells following the strepsistene stages,
the synaptic threads appear to loose their individuality, becoming
diffuse and widely expanded ("confused stage" of Wilson). This
diffuse stage in the spermatocytes, although of relatively short
duration, is no doubt homologous to the "germinal vesicle"
stage of the oocyte, since both occur at homologous periods in

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 423

the male and female germ-cells, namely, after synapsis and before
the appearance of the definitive maturation tetrads.

D. OBSERVATIONS ON MITOCHONDRIA. 1
i. Mitochondria in Spermatogenesis.

(a) Spermatogonia. — Mitochondria are present in the sperma-
togonia of all stages in the form of granules usually localized
at one end of the cell, usually the end directed toward the cyst
cavity (Fig. 13). This is quite the usual location of mitochondria
in spermatogonia, but as I have shown in Passalus (Shaffer, '17),
the mitochondria may be diffusely spread in the cytoplasm.
There is no indication that the mitochondria of the spermato-
gonia ever assume the form of filaments. During mitosis the
mitochondria are spread along the outer spindle-fibers and by the
time of the telophase and the cell-constriction, the mitochondria
become grouped as granular masses lying above the daughter
nuclei.

The amount of cell-degeneration taking place in the testes of
Cicada was so striking as to call for more than mere casual
observation. Sometimes an entire half of a follicle would be
filled with cells in various stages of degeneration. This degenera-
tion is usually found only among the spermatogonia prior to the
commencement of the growth stages of the spermatocyte, which
seems to be a critical time for the cells. Once they begin the
synaptic processes, they apparently go through to maturation
normally. It is quite noticeable that cell-degeneration is more
abundant in the adult testes than in the testes of pupae.

1 I shall continue to use the term "mitochondria" to denote those cytoplasmic
structures, regardless of their form, which are preserved by special reagents
(destroyed by acetic acid, etc.) and which take specific stains. I see no advantage
in the employment of the term "chondriosome" as recently urged by Duesberg
('19). The term "mitochondria" is not only more commonly used in this country
but also conveys the idea of their nature quite as well as "chondriosome." It can
be equally well applied to those structures when filar in form as when granular and
it seems to be etymologically as proper as the term "chondriosome." The various
names which have been proposed to designate these structures (chondrioconts,
plastosomes, plastidules, karyochondria, etc.) are based either upon their supposed
origin or else upon their morphology. Since there is still much confusion as to
their origin and since their morphology may vary at different times in the cell-cycle,
it seems best not to commit ourselves to the use of new terms but to employ the
original name, "mitochondria," of Benda.

424 ELMER L. SHAFFER.

In studying the process of degeneration it is at once evident
that the mitochondria are concerned in the process. The first
noticeable change in the degenerating cells is that the mito-
chondria become larger and more deeply staining and the nucleus
often becomes polymorphic (Fig. 14). While the mitochondria
have increased in size, they have decreased in numbers and it
is quite evident that the large mitochondria have grown in size
by an agglutination of the smaller normal ones. I have de-
scribed a similar process in the degenerating spermatogonia of
Passalus (Shaffer, '17), but was unable to follow in detail the
succeeding stages of degeneration. In Cicada, the further pro-
cesses in the degeneration involve a continued agglutination or
coalescence of the mitochondria which results in the formation
of large bodies of globular form staining intensely (Fig. 15).
This process is continued (Fig. 16) until finally the degenerated
cell becomes a deeply staining mass with the nucleus barely vis-
ible. Often the large globules show vacuoles which I interpret
as being due to a partial dissolution of their substance by the
reagents used.

Scott ('16) found that in experimental phosphorus poisoning
of white mice the mitochondria are the first elements of the
acinus cells of the pancreas to show any pathological changes.
They begin to loose their characteristic filamentous form and
finally agglutinate in large compact masses in the cell. "The
mitochondria in these agglutinated masses fuse to form droplets
possessing the characteristic properties of lipoids" (Scott, 'i 6,
p. 251), hence developing a fatty degeneration of the pancreas.
The significance of these observations of Scott is at once apparent
in connection with the degeneration of cells above described in
Cicada. The behavior of the mitochondria in the degenerating
spermatogonia of Cicada is essentially similar to their behavior
in the fatty degeneration of the pancreatic cells and I am conse-
quently of the opinion that the degeneration of the spermatogonia
is a fatty degeneration.

Recently, Athias ('19) has studied the mitochondria in the
interstitial cells of the ovary of the bat, Vespertilion idee. The
mitochondria are present here in the form of granules and fila-
ments, which gradually become transformed into fat globules.

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 425

These represent "sans doute le produit de secretion des cellules
interstitielles." Athias is of the opinion that the mitochondria
are transformed into lipoid globules by a change in their chemical
constitution. " On observe tres souvent des images qui montrent
que le produit graisseux doit resulter d'un changement chimique
de la substance mitochondriale; on peut suivre, en effet, les
differentes phases de la transformation des plastosomes en gout-
telettes de graisse" (page 195).

What the significance of such cellular degeneration, as found
in the testes of Cicada, can be is difficult to say, since it occurrs
regularly and is moreover found in almost all insect testes. It
may possibly be that these degenerated cells in some way supply
nutriment to the spermatozoa, as suggested by Wieman ('10)
in the case of degenerated cells in the testes of Leptinotarsa, and
are hence homologous to the nurse cells of the ovaries. I have
often noticed deeply staining bodies, somewhat resembling the
degenerated cells, lying among the spermatozoa; but other than
this there is no evidence that the degenerated cells are a source
of nutriment for the spermatozoa.

(b) Spermatocytes. — Since the spermatocytes of the early
growth period are not present in my material, it has been im-
possible to follow the mitochondria from the last spermatogonial
division into the spermatocytes. However, in the pachytene
stages (Fig. 17) mitochondria are present in the form of filaments
which sometimes appear granular. As will be noticed they are
most abundant at the pole of the cell where the idiozome is
located, which is the usual localization of the mitochondria in
the spermatocytes, but which is by no means general. In
Passalus (Shaffer, '17) I have found that the mitochondria of
the early spermatocytes are in the form of diffusely spread
granules and are often most abundant in a zone immediately
surrounding the nucleus.1 Undoubtedly, the filar mitochondria

1 Duesberg ('18) has questioned these observations stating (foot-note, p. 138):
"it is characteristic, even if not quite general, that the male auxocytes have their
chondriosomes accumulated at one pole of the nucleus around the idiozome."
Duesberg's ('10) own figures (Figs. 51, 52, 53) of the spermatocytes of the guinea-
pig seem to contradict this, for he describes the mitochondria as being spread
diffusely in the cytoplasm. "Us (mitochondria) cessent d'etre groupes exclusive-
ment autour de 1'idiozome, pour se repandre dans toute le cytoplasme" (p. 65).

426 ELMER L. SHAFFER.

of the spermatocytes are genetically related to the granular
ones of the spermatogonia. There is no evidence that the
mitochondria go through a process of pairing during the growth
stages; on the contrary they increase in numbers. This increase
in the numbers of the mitochondria is a most important point
in an understanding of their nature and significance. Such an
increase in numbers of mitochondria during the growth period
has been explained by some workers (Goldschmidt, Buchner,
etc.) as being due to an elimination of material from the nucleus
into the cytoplasm (chromidial theory). On the other hand,
another group of workers insist that mitochondria are always
derived from pre-existing mitochondria by a process of autonom-
ous division. Wilke ('13) has described such a process in
Hydrometra. As will be shown later, it is quite possible that
the mitochondria may arise in another way.

As the first maturation spindle is forming (Fig. 21), the mito-
condria begin to migrate from one pole of the spindle towards
the opposite pole and finally surround the entire spindle by
the time of the metaphase (Fig. 22). As the chromosomes begin
to divide in the anaphase, there is no indication of any division
of the mitochondria (Fig. 25). In the late anaphase, when the
cell-constriction begins to appear (Fig. 27), the mitochondria
begin to divide, so that when the cell-constriction is complete,
the daughter cells (second spermatocytes) each contain approxi-
mately equal amounts of mitochondria (Fig. 28). The view of
autonomous division is not supported here and it is quite evident
that their division is due to their separation by the cell-con-
striction.

In the second maturation division, the behavior of the mito-
chondria is the same as in the first maturation division ; they
surround the spindle peripherally and become divided by the
cell constriction (Fig. 29), so that the daughter cells (spermatids)
receive equal amounts of mitochondria. As is usual in the
insects, the mitochondria of the spermatid become resolved into
a compact body, the Nebenkern (Fig. 30, N), which usually
shows a lighter peripheral area. It is interesting to note that

There are many other cases in which the mitochondria are not localized at one
pole of the nucleus in the spermatocyte, and while such localization is common,
it is not general as Duesberg maintains.

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. \2

in the giant spermatocytes (Fig. 59) the mitochondrial content
is-much greater than in the normal cells and the giant spermatids
derived from such giant spermatocytes have Nebenkerns which
are correspondingly larger than the normal ones (compare Figs.
30 and 31).

(c) Transformation of the Spermatid. — Besides the deeply stain-
ing Nebenkern, there are also present in the cytoplasm of the
spermatids, an acrosome-sphere, a centrosome and often a
chromatoid body (Fig. 30). The latter is found in some of the
spermatogonia and some of the spermatocytes, passing undivided
to one of the daughter cells at the time of mitosis. During the
transformation of the spermatid, the chromatoid body is cast off
in the elongating tail, as has been described by Wilson ('13) in
Pentatoma.

The centrosome of the spermatid is derived from the centro-
some of the second maturation division. In the late anaphase
(Fig. 29), the centrosomes lie close to the chromatin masses at
both poles and in the telophase, when the spermatid nucleus is
reconstructed, it can still be seen closely adhering to the surface
of the nuclear membrane (Fig. 30). The axial filament begins
to grow from the centrosome and pierces the Nebenkern (Fig.
32). The division of the Nebenkern into two halves lying on
each side of the axial filament is not quite so clear as in the case
of other insects. In the further elongation of the spermatid

•

(Fig. 33), the Nebenkern becomes drawn out, as the axial filament
grows, into two narrow filaments forming a sheath around the
axial filament.

What I have termed the "acrosome-sphere" is a derivative of
the spindle; but wrhether it represents the "sphere" material or
whether it is a portion of the mitosome could not be determined.
It is a rather large compact body and stains intensely with
haematoxylin. (Figs. 30, 32). As the spermatid goes through
the stages of transformation, the acrosome-sphere becomes some-
what compressed and forms a typical acrosome at the head of
the spermatozoon.

(d} Discussion. — Leaving aside for the present the question of
the origin of the mitochondria, let us turn to consider some of the
facts concerning the behavior of the mitochondria in spermato-

428 ELMER L. SHAFFER.

genesis. Not only do they have a characteristic behavior which
seems to be quite general in the insects, but they are present
in all generations of the male germ cells. There are many dis-
crepancies in the literature on this point, some workers maintain-
ing that the mitochondria disappear at various times. Buchner
('09) has described a disappearance of the mitochondria during
mitotic division in the spermatogonia of Gryllolalpa, but on the
contrary Duesberg ('10), working on the same form, has shown
that the mitochondria do not disappear at that time. Wilke
('13) states that in Hydrometra the mitochondria are absent from
some of the spermatogonia and in some cases from the spermato-
cytes. I am inclined to believe that the disappearances of
mitochondrial structures is in all cases due to improperly fixed
material. In fact, there is some evidence from Wilke's figures
that this is the case. In the spermatocytes of Hydrometra,
Wilke describes a deeply staining perinuclear zone of the cyto-
plasm in which are located "yolk-spherules" (Dotterkugeln).
These spherules at first homogeneous, begin to show the appear-
ance of threads within them and finally definite mitochondria
are set free in the cytoplasm, being formed out of the substance
of the "Dotterkugeln." His results, however, do not indicate
that a reversal of this process might not be taking place. I have
often seen in material which has been improperly fixed bodies
resembling Wilke's "Dotterkugeln," which are produced by an
artificial agglutination of the mitochondria. As I shall later
attempt to show, certain methods of fixation may produce a
variety of changes in mitochondrial structures ranging from a
slight distortion of their form to their complete disappearance.
I merely wish to urge the fact here that mitochondria are constant
cytoplasmic structures of the cell and when proper technical
methods are employed, they can be demonstrated at all periods
in the cell-cycle.

The role of the mitochondria in spermatogenesis is difficult
to interpret on the basis of their morphological behavior. The
formation of the compact Nebenkern from the filar mitochondria
may be an indication of some chemical change in the mitochondria
or it may be merely an expression of the compactness which the
other cell elements, notably the nucleus, show in the transforma-

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 429

tion of the spermatid into the spermatozoon. O. Vander Stricht
calls the Nebenkern of the spermatid a "vitelline body." Wild-
man ('14) ascribes a nutritive function to the mitochondria
(karyochondria) of the spermatid of Ascaris. The view has
often been expressed that the mitochondria of the spermatozoon
are concerned in locomotion, since they usually form a part of
the tail. On the basis of the behavior of the mitochondria in
spermatogenesis, we can only say that, just like the acrosome or
the axial filament, they form a definite structural element of the
spermatozoon, and it may be that their only significance is
bound up with their function as an organelle of the spermatozoon.
This view is contested by Meves, Duesberg and others on the
basis of their observations on fertilization in certain forms where
the mitochondria of the sperm apparently enter the egg and can
be traced into the cleavage cells. If this behavior of the mito-
chondria can be found to be of general occurrence in fertilization,
their significance from the point of view of heredity and develop-
ment becomes of utmost importance. It may be said that the
evidence for this is still in a very unsatisfactory state. I shall
reserve for later discussion the questions bearing on this problem,
until added observations on the mitochondria during oogenesis
may be presented.

2. Nutrition of the Egg.

Without entering into a detailed description of the structure
of the ovary of Cicada, I wish merely to mention some facts in
regard to the nutrition of the egg during the growth period. The
study of insect ovaries dates back over a century, many of the
works being concerned with the question of the origin of the
various cell-elements, viz., germ-cells, nurse-cells, epithelial cells
and follicle-cells. Much of this work is of little value for the
solution of these problems, since the study has been usually
confined to that of the adult ovaries, and it is evident that only
an embryological as well as a histological study of the ovary can
enable us to come to any definite decisions. As pointed out by
Hegner ('14), the origin of the nurse-cells and germ-cells may be
different in different forms. In the case of Miastor, Hegner states
that the nurse-cells are mesodermal in origin, while in the
Chrysomelid beetles "the nurse-cells in the ovaries seem to be

43° ELMER L. SHAFFER.

of germ-cell origin" (Hegner, '14, page 119). In Dytiscns,
Giardina ('01) has shown that germ-cells and nurse-cells arise by
differential divisions of a stem-cell. In Cicada, I am unable to
ascertain the origin of the cell-elements of the ovary from a
study of the pupal and adult ovaries. The differentiation of the
cells must take place at a comparatively early time in the long
life cycle of the insect.

There are two ovaries, each consisting of a great many ovarian
tubules, the ovaries of the adult being much larger than those
of the pupae, due to the presence of a large number of mature
eggs. Figure 34 is that of a longitudinal section through an
ovarian tubule of an adult recently emerged from the pupal
case and it will at once be seen that it is a typical Hemipteran
ovarian tubule. At the proximal end of the tubule is the narrow
end-filament which is undoubtedly of a ligamentous nature
helping to support the ovaries in the abdomen. The tubule
may be divided into three zones depending on the character of
the cells present. At the proximal end is the nurse chamber,
containing all the nurse-cells. These stain very deeply and it
is impossible to make out their cell walls. The chromatin of
their nuclei is in the form of diffusely spread granules, and usually
there is present a chromatic nucleolus and a true plasmosome
(Fig. 45, n.c.). At the base of the nurse chamber (Keimlager)
are found numerous young oocytes in various stages of synapsis.
Proceeding distally from the nurse chamber are the older oocytes
of the post-synaptic stages, those farthest away from the nurse
chamber being the oldest and largest (Fig. 34, ooct. 2). In this
region are also found the follicle-cells (f.c.} which begin to form
definite follicles around the oocytes. As the young oocytes
begin to migrate distally from the base of the nurse chamber,
they still retain protoplasmic connections with the cytoplasm of
the nurse-cells, resulting in the formation of pseudopod-like
projections from the oocytes, the egg-strings (e.s.), by means of
which nutriment is passed from the nurse chamber to the oocytes.
Even old oocytes in which yolk is beginning to form still retain
connection with the nurse chamber by means of the egg-string
(Fig. 47). In the ovaries of young pupae the egg-strings appear
simply as cytoplasmic protrusions of the oocytes, but in the

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 43!

adult ovaries, the egg-strings loose their cytoplasmic appearance
and seem to be of a fibrous structure (Figs. 34, 45, 47, e.s.).
This fibrous appearance of the egg-strings has often been figured
in the ovaries of other insects, but I am inclined to doubt its
normality; it seems rather that the fibrous appearance is due
to fixation and really represents a compression of the cytoplasmic
processes of the oocytes.

In the nurse chamber, the nurse-cells are arranged in what
appears to be a syncytium, and only occasionally can cell-walls
be distinguished. In the young pupse, the egg-strings of the
oocytes pass through the region of the follicle-cells into the nurse
chamber and seem to fuse and become continuous with the cyto-
plasm of the nurse-cells. In the adult ovaries when all the cells
are at the height of their functional activity, the egg-strings
become much enlarged assuming the fibrous appearance above
described. In the nurse chamber, the egg-strings end in a
central fibrous mass (Figs. 34, 71, 72, i.n.c.), the substance of
which is continuous with the egg-strings and from which they
lead to the oocytes. The nurse-cells immediately in the region
of the central plasmatic mass stain deeply and show evidences of
degeneration. Within the plasmatic mass can be seen many
nurse-cell nuclei in various stages of disintegration (Figs. 72, 34,
i.n.c.). The nuclei become smaller, staining intensely, and finally
become broken down and the products of their disintegration
can be seen passing down the egg-string into the oocyte (Figs.
45, 71). The nurse-cells are thus ingested by the protoplasmic
process of the oocytes which are probably furnished with some
substances (enzymes) enabling them to digest the nurse-cells.
In the adult ovaries, at the height of the breeding season, the
ingestion of the nurse-cells has taken place to such an extent
that almost half of the nurse-cells have disappeared and the
central plasmatic mass has grown in size containing a great
many ingested nurse-cells. The growth of the central plasmatic
mass is undoubtedly correlated with the disappearance of the
nurse-cells in this region of the ovary.

It is a noteworthy fact that the ingestion of the nurse-cells
takes place at a time when there is a rapid growth of the oocytes.
The oocytes at the beginning of the synaptic processes have large

432 ELMER L. SHAFFER.

nuclei surrounded by a small amount of cytoplasm containing
mitochondria. As the synaptic period progresses the oocyte
grows somewhat in size, chiefly by an increase in the size of the
nucleus and a slight increase in the cytoplasmic volume. After
the synaptic processes have been completed, the cytoplasmic
volume increase rapidly, so that the oocyte becomes over four
times its linear size before yolk begins to form. After the cyto-
plasmic volume has reached its maximum, the yolk-building
process commences and at the end of this period the almost
mature oocyte is about three times its linear size at the beginning
of the process. This great increase in size is not due to any
increase in cytoplasmic volume, but due to the great enlargement
of the yolk spherules. It is doubtful whether the cytoplasmic
volume increases at all after the beginning of the yolk formation.
The ingestion of the nurse-cells by the oocytes in Cicada is
perhaps homologous to the ingestion of the cells in the ovaries of
other forms, the classical example of which is found in Hydra.
As is well known, in Hydra one cell in the ovary grows large by
ingesting the other cells in the ovary and becomes the functional
oocyte. In Dinophiliis, Conklin ('06) has described the fusion
of from 25 to 30 oogonia to form a single large (female-producing)
egg. In dona and other Ascidians, the test-cells are ingested
by the oocytes and remain in the egg, perhaps aiding in the
elaboration of nutrient materials, but are cast out of the egg
prior to fertilization. In the ovaries of the certain insects (e.g.,
Dytiscus) of the type in which the oocytes are supplied with a
separate group of nurse-cells, the process of absorption of the
nurse-cells as the oocyte grows has been often described. Kor-
schelt ('86) has described in the ovaries of Notonecta and Redumus
a disintegration of nurse-cell nuclei in the nurse chamber through
the action of the oocytes, with the production of a central space,
" Plasmatische Raum," in the nurse chamber, which is free from
nuclei. Foot and Strobell ('n) have also described a similar
plasmatic area free from nuclei in the ovaries of Protenor, but
are not inclined strongly toward Korschelt's view that in this
region the nurse cells disintegrate and furnish nutriment for the
oocytes. In Cicada a study of Figs. 34, 45, 71, 72 will show that
there is no doubt that Korschelt's view is correct, for, not only

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 433

can nurse cells be found in various stages of disintegration, but
the products of such disintegration can actually be seen to pass
down the egg-strings to the oocytes. From Fig. 45, the impres-
sion may be gathered that the material derived from the dis-
integrated nurse cells passing down the egg-string is accumulated
in a zone about the nucleus. Such, however, is not the case and
as will be shown later, the deeply staining perinuclear zone
consists of mitochondria and in material fixed in Bouin's fluid,
the mitochondria disappear while the granules in the egg-string

are preserved.

3. Mitochondria in Oogenesis.

(a) Mitochondria in Growth Stages. — The young oocytes of
Cicada lying at the base of the nurse-chamber are at all times
distinguishable from the nurse-cells not only by their character-
istic nuclei, but also by the presence of definite aggregations of
granular mitochondria which are characteristically localized in
the cells.

During the entire synaptic period of the oocyte, the mito-
chondria are found lying in the cytoplasm as a cap of deeply
staining granules closely applied to the nuclear membrane at
one pole (Plate V.). As will be seen from Figs. 38, 39, 40, the
pole of the nucleus at which the mitochondria are found always
corresponds to the pole of the nucleus towards which the synaptic
threads are polarized. It will at once be seen by comparing
Figs. 17 and 40, that the mitochondria of the oocyte and sperma-
tocyte are localized in the cytoplasm at homologous positions,
namely at the side of the nucleus where the idiozome, the sphere
or centrosome is probably located. In the oocytes, however,
no centrosome or sphere can be distinguished and similarly
Montgomery ('n) states that in Euschistus "it is not clear
\vhether this body (idiozome or sphere) has any homologue in
the oocytes," etc.

The mitochondria of the oocytes of Cicada are always in the
form of granules, never assuming the filar form as found in the
spermatocytes. During the later synaptic stages (pachytene
stage, Fig. 40) there is an increase in the amount of mitochondria,
but their localization at one pole remains constant. In the
post-synaptic stages we find that, although the mass of mito-

434 ELMER L. SHAFFER.

chondria is still found at one pole of the nucleus, they gradually
completely encircle the nucleus. After this period, when the
synaptic threads have become diffuse and the "germinal vesicle"
is established, the mitochondria increase greatly in numbers
and are found localized in a zone immediately surrounding the
nucleus (text-fig. 3, and Figs. 83, 84).

This perinuclear zone of mitochondria is sharply delimited
from the rest of the cytoplasm. Faure-Fremiet ('08) has de-
scribed the mitochondria of the oocytes of Julus as being similarly
arranged in a perinuclear zone whose cytoplasmic limit is defi-
nitely marked by the presence of a membrane formation. I can
find no evidence for the presence of such a limiting membrane in
Cicada. The mitochondria continue to increase greatly in
numbers in the perinuclear region up until the stage in which
the cytoplasmic volume of the oocyte is greatest (Fig. 85).
Throughout the period in which mitochondria are forming in the
perinuclear zone, the nuclear membrane remains intact at all
times as is evidenced in material improperly fixed in which the
nuclear membrane has shrunken away from the cytoplasm as
shown in Fig. 85. Neither is there any evidence that any nuclear
materials are discharged or extruded into the cytoplasm at any
time. There are many descriptions in the literature of bodies
in the cytoplasm of oocytes which have been derived from the
discharge of nuclear materials. Goldschmidt and his pupils have
maintained that the mitochondria are derived from the passage
of nuclear materials into the cytoplasm, but there is no evidence
that this is the case in Cicada. Vejdovsky ('12) has described
in the oocytes of Aphrophora large deeply staining globules and
vacuoles (compare my Fig. 42). He does not relate these struc-
tures to the mitochondria in any way, but indicates that these
nucleolar-like bodies in the cytoplasm are derived by a "nucleoli-
zation" of the chromosomes and a casting-out of the resulting
nucleoli into the cytoplasm. He is of the opinion that the
vacuoles represent the escaped nuclear sap, there being no nuclear
membrane present at this time. In the first place, it seems very
doubtful that the nuclear membrane does disappear at any time
in the resting nucleus. Secondly, if there is a nuclear membrane
present it is difficult to see how such solid bodies as nucleoli

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 435

could pass through the membrane unless they be in a fluid, diffus-
ible state, and if so they would become diffused when they
entered the cytoplasm. Cases of "extruded nucleoli " have often
been mentioned in the literature, but it must be said that many
of the interpretations are exceedingly doubtful. In Cicada,
when the ovaries have been fixed for a considerable length of
time (24 hours) in Bouin's fluid, the mitochondria entirely dis-
appear, leaving occasionally traces of their dissolution in the form
of cytoplasmic vacuoles in the region of the nucleus (Fig. 82).
By fixing the ovaries in Bouin's fluid for varying lengths of time
(5 to 12 hours) a variety of peculiar structures may be found in
the cytoplasm which are all referrable to the various stages and
degrees of dissolution of the perinuclear zone of mitochondria
(Figs. 39, 41, 42, 80, 81, 82). When the ovaries are fixed for ten
hours in Bouin's fluid, a zone of the cytoplasm is found around
the nuclei of the oocytes which takes the plasma stain intensely
(Fig. 42). Within it may be found vacuoles and deeply staining
bodies resembling those structures which Vejdovskv ('12)
figures in Aphrophora. In studying the effect of acetic acid of
fixing fluids upon the mitochondria, I am led to believe that
many of the peculiar cytoplasmic bodies which have been de-
scribed and figured in germ-cells under various names such as
extruded nucleoli or plasmosomes, vacuoles, idiozomes, spheres,
yolk-spherules (Dotterkugeln), etc., are the result of imperfect
fixation of mitochondrial substances. O. Vander Stricht ('04)
has shown a similar effect of reagents in the distortion and dis-
solution of mitochondria in the "couche vitellogene" of the
oocyte of the bat.

(b) Yolk-formation. — Throughout the growth period of the
oocyte, the mitochondria increase greatly in numbers and con-
tinue to be located in the well delimited perinuclear zone. At
the time when the cytoplasmic volume of the oocyte is at its
maximum (Fig. 85) the zone of the mitochondria occupies ap-
proximately a third of the cytoplasmic volume. After this stage
in the growth of the oocyte, the perinuclear mitochondrial mass
begins to loose its well-defined zonal limits and the granules
become dispersed in the cytoplasm towards the periphery of the
oocyte (Fig. 86). The migration of the mitochondria from the

436 ELMER L. SHAFFER.

perinuclear zone to the periphery of the cell continues, until in
the older oocytes all the mitochondria have been localized in the
cortical region of the oocyte (Fig. 47). A similar centrifugal
migration of the mitochondria from the perinuclear zone toward
the cell periphery has been described by Payne ('16) in the
oocytes of Gryllotalpa, by Faure-Fremiet ('08) in the oocytes of
Julus, by Govaerts ('13) in the oocytes of the beetles Trichiosoma
and Cicindella, and by many other workers wrho have studied
the mitochondria in the oocytes.

At this time the cytoplasm of the oocyte is of a hyaline, homo-
geneous appearance, except for the mitochondrial granules located
in the peripheral zone. In somewhat older oocytes (in which
yolk-spherules have not as yet formed) numerous vacuoles appear
in the cortical zone of the cytoplasm where the mitochondria
are located. These vacuoles are at first very small in size and
within them can be seen the mitochondrial granules. That the
granules in the vacuoles are actually mitochondria is proved by
the fact that they respond to all the specific stains and are only
found in material which has been fixed according to the mito-
chondrial technique. The further history of the mitochondria
and the vacuoles of the cortical layer is concerned with the
process of formation of the yolk-spherules. The mitochondria
within the vacuoles begin to disintegrate, often showing small
vacuoles within themselves. At this time the vacuoles containing
the disintegrated mitochondria appear very much like numerous
small nuclei. These structures are no doubt similar to the
"pseudo-nuclei" of Blochmann. Korschelt ('89) has described
similar "pseudo-nuclei" in the oocytes of several insects, and
he considers them as follicle-cell nuclei which have migrated into
the oocyte much as the test-cells of the Ascidians migrate into
the oocyte to aid in the elaboration of nutrient materials.
Hegner ('15) has studied similar "secondary nuclei" in the
oocytes of Camponotus (Hymenoptera) and has also given a
satisfactory review of the literature dealing with these structures.
However, he is of the opinion that the "secondary nuclei" arise
as buddings from the oocyte nucleus and that this process is,
hence, comparable to chromatin diminution processes in other
forms (e.g., Miaslor, Ascaris). Hegner finds that the "secondary

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 437

nuclei" surround the oocyte nucleus increasing in numbers pre-
sumably at the expense of the oocyte nucleus. Later they
become scattered toward the periphery of the oocyte, increasing
in numbers and finally disappearing. According to Blochmann
the "pseudo-nuclei" may become incorporated in the yolk-
spherules. From this general behavior of the "pseudo-nuclei"
and the "secondary nuclei" of Hegner I am convinced that in
all cases they represent stages in the transformation of mito-
chondria into yolk. Loyez ('08) has shown that these structures
are not nuclei in any sense of the word and similar to Govaerts
('13), she has shown their relation to mitochondria and yolk-
formation.

Figure 48 shows the various stages in the development of the
yolk-spherules from the substance of the mitochondrial granules
contained in the vacuoles. The substance of the vacuoles takes
the plasma stain lightly at first, and as the vacuole grows in
size the mitochondrial granules disappear and the substance of
the vacuole begins to take the plasma stain more deeply. These
plasma staining bodies grow considerably in size and often a
small area of their substance begins to show a marked affinity
for the basic stains. This basic staining area grows in extent and
finally the entire body, which is none other than a yolk-spherule,
takes the basic stain intensely. The relative size of the yolk-
spherules is enormous as compared to the size of the mitochon-
drial granules from which they have been derived, and this
indicates, together with the changes in the staining reactions,
that a series of chemical reactions take place during this trans-
formation. The mature egg is filled with these large deeply
staining yolk-spherules, between which lie small granules which
give the characteristic mitochondrial staining reactions similar
to the mitochondria described by Duesberg ('08) in the egg of
the bee. Apparently not all of the mitochondria are transformed
into yolk; or perhaps new ones are being formed which have as
yet not been transformed into yolk.

The transformation of mitochondria into yolk has been de-
scribed by Loyez, Russo, Faure-Fremiet, Govaerts, Hegner,
Lams, Vanderstricht, and others. L. and R. Zoja have described
a transformation of the "plastidules fuchsinophiles" into yolk

ELMER L. SHAFFER.

in the egg of Helix. R. Vander Stricht ('n) has described in the
oocyte of the cat a "couche vitellogene" or perinuclear ring of
mitochondria which grows in size during the early growth stages
of the oocyte and which later becomes dispersed toward the
periphery of the cytoplasm and there gives rise to yolk. O.
Vander Stricht ('94) has also described a "couche vitellogene"
(of mitochondrial nature) in the oocytes of the bat, which in the
young oocytes is arranged in the perinuclear zone, but in the older
oocytes it becomes diffused in the cytoplasm. This diffused
substance forms the " pseudochromosomes " of the oocytes of
the adult ovaries and from these are derived the yolk bodies.
Wildman ('13) has described two kinds of cytoplasmic inclusions
in the spermatocytes of Ascaris, the "karyochondria" and the
"plastochondria," both of which are derived from nuclear ma-
terial. The "karyochondria" become transformed into yolk
granules which later fuse to form the "refractive" body of the
spermatid. The "plastochondria" have a negative behavior
and take no part in spermiogenesis. Recently Gajewska ('19)
has described a perinuclear ring of mitochondria in the oocytes
of Triton and indicates the relation of this to yolk formation.

I have cited only a few of the cases in which the mitochondria
appear to have a genetic relationship to yolk formation, and as
will be shown later, this relation is more general than is usually
supposed. The chemical processes involved in the transforma-
tion of mitochondria into yolk cannot involve complicated chem-
ical changes since the mitochondria themselves are of a phos-
pholipoid nature and closely allied chemically to yolk.

(c) Discussion. — In almost every work dealing with oogenesis
and the study of oocytes, mention is made of certain cytoplasmic
inclusions which are either of a granular, globular or filar nature,
or else assume larger proportions and appear as single compact
bodies. Various names have been applied to these structures,
such as "Dotterkern" or yolk-nucleus, yolk-matrix, "couche
vitellogene," "corps de Balbiani," pseudochromosomes, extruded
nucleoli, "zona plasmatica perinucleare," etc. In nearly all
these cases it will be noted that these special portions of the
cytoplasm are involved in the process of yolk elaboration, and I
shall attempt to point out some of the homologies existing be-

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 439

tween the above-named structures and the perinuclear zone of
mitochondria as found in the oocytes of Cicada. First, however,
it may be said that the occurrence of a special zone of mito-
chondria immediately surrounding the nucleus is by no means
uncommon. Faure-Fremiet ('08) has described a perinuclear
ring of mitochondria in the oocytes of Julns, as does Payne ('77)
in the oocytes of Gryllotalpa, Vejdovsky ('12) in the spermato-
cytes of Diestramena, Schaefer ('07) in the spermatocytes of
Dytiscus, Shaffer ('17) in the spermatocytes of Passalus, etc.
In fact in almost every case where the mitochondria have been
studied in the early growth period of the germ-cells, they have
been found in close spatial relations to the nucleus.

In the early literature dealing with insect oogenesis, there is
almost always figured granules or deeply staining areas of the
cytoplasm whose homology with the perinuclear ring of mito-
chondria as described here in Cicada becomes at once evident.
Korschelt ('89) describes in the oocytes of Dytiscus deeply
staining portions of the cytoplasm in the region of the nucleus
which he interprets as representing nutrient materials derived
from the nurse-cells, and into which the nucleus of the oocyte
sends amoeboid processes. Marshall ('070) describes small
nuclear-like bodies around the nuclei of the oocytes of Polistes.
These increase in number and later migrate to the periphery of
the oocyte, but Marshall offers no explanation as to their sig-
nificance. The same author (Marshall, '076) figures a deeply
staining granular zone around the nucleus of the oocyte of Platy-
phlax (Hymenoptera). Often this granular mass is cone-shaped
extending toward the egg-string (compare my Fig. 45). In the
older oocytes, he describes the appearance of deeply staining
bodies which lie scattered in the cytoplasm (when fixed in Flem-
ming's fluid) ; but their further history is not traced. McGill
('06) describes a deeply staining perinuclear zone of granules in
the oocytes of Plathemis and A nax, which she considers as the
"yolk-nucleus." It later breaks away from the nuclear wall and
becomes scattered in the cytoplasm. Hegner ('15) describes a
perinuclear zone of granules in the oocytes of Camponoius which
he states "resembles chromatin in some respects and may repre-
sent chromatin which has passed through the nuclear membrane

440 ELMER L. SHAFFER.

into the cytoplasm" (page 508). More recently, Nakahara
('18) has described in the oocytes of Perla a deeply staining
mass of granules in the cytoplasm closely surrounding the nuclear
membrane, which he calls the "yolk-nucleus." This later breaks
away from the nuclear membrane, becomes insignificant and
finally disappears. Nakahara is of the opinion that the nucleoli
of the oocyte nucleus are derived from the passage of material
from the "yolk-nucleus" into the oocyte nucleus, and hence
nucleoli are extra-nuclear in origin. These interpretations of
Nakahara are hardly excusable since at this late date he has
had the advantage of a large mass of data bearing on mitochon-
dria, and it is evident that he has not made a proper study of
the literature. Many similar cases of almost riotous interpreta-
tions of mitochondrial structures might be mentioned, many of
them inexcusable in light of the recent studies on the mitochon-
dria. As has been before stated, many of these erroneous inter-
pretations are based upon material prepared without regard for
the special technical processes involved for the demonstration of
mitochondrial structures. Perhaps the best example of this is
the work of Giardina ('04). This writer has studied the oocytes
of a number of insects (Mantis, Periplaneta, Stenobothrus, Gryllus]
and the gastropod, Helix arvensis. In all of these forms, Giardina
has described a special zone of the cytoplasm immediately sur-
rounding the nucleus to which he assigns the name "zona
plasmatica perinucleare." At times this zone may appear
granular, striated, vacuolated, homogeneous, etc., and is always
sharply delimited from the rest of the cytoplasm by a membrane.
After considerable discussion as to its physical state, Giardina
comes to the conclusion that it is a formation in situ of the cyto-
plasm under the action of substances from the nucleus, and that
it acts as an intermediary between the nucleus and the cytoplasm
in the nutrition of the egg. From a study of Giardina's figures
and a comparison with the oocytes of Cicada fixed in Bouin's
fluid for varying lengths of time (Figs. 39, 41, 42), it becomes at
once evident that the "zona plasmatica perinucleare" of Giardina
is nothing more than poorly-fixed mitochondria arranged in a
perinuclear zone. Duesberg ('12) is of this opinion as is also
Faure-Fremiet ('10).

THE GERM-CELLS OF CICADA (lIBICEN) SEPTEMDECIM. 44!

It thus seems that perhaps it is quite general in the insects
that the mitochondria of the oocyte are first arranged in a
definite zone around the nucleus during the period in which
they increase in numbers, and in the older oocytes the perinuclear
zone of mitochondria becomes dispersed towards the periphery
of the oocyte and becomes concerned in the process of yolk-
formation.

4. Significance of the Perinuclear Zone of Mitochondria.

(a] Relation to Yolk-nuclei. — As I have before indicated, the
occurrence of a perinuclear zone of mitochondria is quite com-
monly met with especially in the female germ-cells during the
growth period. I shall not attempt a lengthy review of the
literature in this regard since we already have the excellent
review of Duesberg ('12), but I wish to point out some of the
significant facts which have a bearing on the problems presented
here. I wish merely to discuss the origin and homologies of
the yolk-nucleus or "corps de Balbiani," mention of which has
been made in almost every work in oogenesis which has appeared
in the past two decades. As has been shown in Duesberg's
('12) review, much of the recent work in this connection has
demonstrated that these bodies (yolk-nuclei) are in many cases
partly constituted of mitochondrial substance. O. Vander
Stricht ('04) has shown that the "couche vitellogene" surround-
ing the nucleus of the oocyte of the bat is of a mitochondrial
nature, and von Winiwarter and Sainmont ('09) have described
a zone of granular mitochondria around the nucleus of the
oocyte in the kitten, which they homologize with the yolk-
nucleus and "couche vitellogene" of Vander Stricht. Recently,
Gajewska ('19) has shown that in Triton the yolk-nucleus is made
up of three substances: "Zuerst eine Anhaufung von Ergasto-
plasmatischer Substanz (Ergastoplasma-Dotterkern), dann ein
Mitochondrienkonglomerat (Dotterkern als Kornerkonglomerat)
und endlich ein Haufen von Fettkugeln und Eiweissplattchen.
Die Muttersubstanz fur den Dotterkern ist der perinukleare
Ring (couche vitellogene)" (p. 116).

The fact which I wish to emphasize here is that in all cases the
yolk-nucleus has its beginning in the young oocyte in the form

442 ELMER L. SHAFFER.

of granular masses closely applied to the nuclear membrane,
and these masses grow in amount during the period in which
they are close to the nucleus. This is clearly shown in the early
works on the yolk-nucleus such as Munson ('99) in Lim-ulus,
Crampton ('99) in Molgula, Calkins ('95) in Lumbricus, Van
Bambeke ('95) in Scorpcena scrofa, Vander Stricht ('98) in the
human oocytes and in Tegnaria, and by others of the early
workers on these structures. In almost all these cases, the close
association of these granules with the nucleus and the similarities
in the staining reactions with that of the chromatin have given
the earlier workers the opinion that the granular masses represent
extruded nuclear material. In many of these cases, the granular
masses surround the centrosome forming a compact body with
sharp outlines.

The formation of the typical yolk-nucleus (corps de Balbiani)
as found in the Arachnids has been described by Vander Stricht
('98, '04) and Faure-Fremiet ('10). Here, as in other cases, the
granules (mitochondria) forming the yolk-nucleus are first found
in a perinuclear zone. Within this granular mass is found a
deeply-staining vesicle, and later the mitochondria gather about
it in concentric layers forming the typical compact Arachnid
yolk-nucleus, which usually shows a ray-like structure. The
relation between the yolk-nucleus and the attraction-sphere has
often been noted. I have previously called attention to the fact
that the mitochondria are usually (although not always) localized
at the pole of the cell at which the centrosome, sphere or idiozome
lies. In the formation of the compact Arachnid yolk-nucleus,
the mitochondria gather about the centrosome-sphere as a center,
the relation between the two being only spatial. Duesberg
('12) has noted this relation and applies the term "idiozom"
or "centrotheka" to the central vesicle of the yolk-nucleus and
points to the fact that this conception corresponds to Henneguy's
definition of the "corps de Balbiani": " Un corpuscle centrale
entoure d'un zone d'aspect homogene on finement granuleux."

It thus seems that in oocytes the perinuclear ring of mito-
chondria may behave in two ways; it may become dispersed in
the cytoplasm and give rise to yolk in the periphery of the cell
(as seems to be the case in the insects), or the mitochondria of the

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 443

perinuclear zone may become massed about the centrosome,
forming the compact "corps de Balbiani" which later disinte-
grates in the formation of yolk (as in the Arachnids, etc.).

(b) Origin of the Mitochondria. — There are two facts of im-
portance to be noted in the foregoing discussion: (i) the quite
usual presence of mitochondria in a zone of the cytoplasm im-
mediately surrounding the nucleus; (2) the presence of this
perinuclear zone at a time when the mitochondria are increasing
greatly in number. One of the most important questions bearing
on the nature and role of the mitochondria is bound up with the
mode of their origin and increase in number, and there has been
considerable controversy regarding these questions. According
to the view of Meves, Bouin, Duesberg and others, the mito-
chondria have no "de novo" origin, but are always derived from
pre-existing mitochondria by a process of division. The chief
evidences to support this view are: (i) the constant presence of
mitochondria in all cells at all times; (2) the mitochondria within
the cell are actually distributed to the daughter cells at the time
of mitosis; (3) in some few cases (Meves in Ascaris, Duesberg in
Ciona and Apis) the mitochondria of the fertilized egg have
been traced into the embryonic cells. In the first place, it may
be said that the omnipresence of mitochondria in all living cells
may be interpreted upon an entirely different basis as I shall
attempt to show later. There is, of course, a genetic continuity
of mitochondria to a certain degree; the mitochondria of a cell
are certainly carried into the daughter cells at the time of
mitosis, but the view that individual mitochondria give rise to
"homologous" mitochondria of succeeding cell-generations is
entirely without evidence. Duesberg ('18) found that the yellow
ooplasm of the Ascidian egg was rich in mitochondria, and, using
the results of Conklin ('05) who traced yellow ooplasm of the egg
into the embryo, consequently maintained that the mitochondria
were genetically continuous from the fertilized egg to the em-
bryonic cells. While there is no question that the yellow ooplasm
is continuous, yet this is far from establishing that the individual
mitochondria are continuous.

Opposed to the "genetic continuity" hypothesis of the mito-
chondria, we have the "chromidial" hypothesis developed by

444 ELMER L. SHAFFER.

Goldschmidt ('09) and his students. As has been before men-
tioned, there have been many descriptions of cytoplasmic bodies
whose origin has been attributed to material extruded from the
nucleus. According to Goldschmidt and others (Buchner, Jor-
gensen, Schaxal, Wasilieff, Popoff, etc.), mitochondria are derived
from the chromidia and ultimately from the chromatin of the
nucleus, and are hence similar to Hertwig's "chromidia." The
fact that the mitochondria lie at the pole of the nucleus where the
idiozome lies and toward which the synaptic threads are polarized,
has been taken by Buchner and Wasilieff to be the place where
mitochondria arise by the emigration of chromatic materials
from the nucleus into the cytoplasm. Buchner ('09) derives
the mitochondria of the spermatocytes of Gryllus from the
material of the sex-chromosome, which in the bouquet stage
becomes vacuolated and shows evidences of disintegration.
While chromatin from the nucleus may at times come to lie in
the cytoplasm (e.g., chromatin diminution processes in Ascaris
and Miaslor, etc.), yet there is no strong evidence that it may
give rise to the mitochondria. It is difficult to see how escape of
chromatin from the nucleus could account for the tremendous
increase in the mitochondria as found during the growth period
of the oocytes of Cicada. Furthermore, the difference of behavior
of chromatin and mitochondria towards fixing fluids and specific
stains indicates that they are of a totally different chemical
nature (see Cowdry '16, p. 426).

According to the view of Vejovsky ('07, '12) the mitochondria
are cytoplasmic structures having their origin in the "regressive
modification" (Duesberg, '12) of the sphere material. Mont-
gomery ('n) expresses the opinion that "it is probable that they
(mitochondria) are produced by either idiozome or nucleus or by
a joint action of both" (p. 787). Although it is quite usual that
the mitochondria are found lying close to the idiozome or sphere,
there is no conclusive evidence that they have their origin in
or from the sphere material.

We have now discussed the three views prevalent regarding
the origin of the mitochondria. From a study of the mito-
chondria of Cicada and from a review of much of the literature
bearing on the subject, still another view presents itself which

THE GERM-CELLS OF CICADA (lIBICEN) SEPTEMDECIM. 445

has hitherto been only occasionally expressed in the literature.
This is what might be called the "interaction theory," and
according to it the mitochondria are neither self-perpetuating
structures, nor are they derived from the chromatin of the
nucleus, nor by a disintegration of the sphere material. Accord-
ing to this view the mitochondria are cytoplasmic differentiations
which arise through specific chemical actions of the nucleus
upon materials in the cytoplasm. I have before emphasized
the almost universal presence of a particular zone of mitochondria
immediately surrounding the nucleus at certain times in the
cell-cycle, and I believe this perinuclear zone to be one of the
best morphological demonstrations of an "interaction" between
nucleus and cytoplasm. The presence of this perinuclear zone
of mitochondria during the period in the oocyte when the mito-
chondria are increasing in number as the cytoplasmic volume is
also growing by assimilation of products from the nurse-cells,
supports the view that the mitochondria are. being built up by
nuclear action on substances in the cytoplasm. After the maxi-
mum amount of mitochondria has been elaborated, the peri-
nuclear zone is dissipated and the mitochondria become diffusely
spread in the cytoplasm followed by their transformation into
yolk.

Montgomery ('n) and Browne ('13) have also expressed views
that the mitochondria may arise as a result of the interaction
between nucleus and cytoplasm.

The view of mitochondrial continuity has no more than weak
circumstantial evidence to support. Granting that the mito-
chondria of the spermatozoa are brought into the egg and that
the cells of the embryo and adult all possess mitochondria, it is
far from establishing the fact that the mitochondria of these
cells are derivable from those of previous cell-generations which
go back to the fertilized egg. Mitochondria do arise, as such,
de novo in the cells by the chemical actions of the nucleus on the
cytoplasm immediately surrounding it, although there may be
a limited amount of mitochondria carried over from the previous
cell-divisions.

Just wrhat portion of the cytoplasm is acted upon by the nucleus
is another matter. In the oocytes of Cicada, nutriment is

44^ ELMER L. SHAFFER.

brought in from the nurse chamber through the egg-string and
assimilated by the cytoplasm; the nucleus exerts a chemical
influence (enzyme?) on these products whereby the mitochondria
are differentiated in the cytoplasm immediately surrounding the
nucleus. According to this view it is possible to explain the
presence of mitochondria in all cells (animal and plant) at all
times, for all cells are constantly receiving nutriment which is
taken into the cytoplasm (assimilated) and then acted upon by
the nucleus resulting, among other things, in the elaboration of
mitochondria. It thus seems that the mitochondria arise as
differentiated parts of the cytoplasm through specific chemical
(enzyme) reactions of the nucleus upon the products of assimilation
of the cell. What the significance of such cell structures in the
cell economy may be is quite another problem.

E. GENERAL CONSIDERATIONS.
i. Chromosomes.

My study of the chromosomes of Cicada, I believe, presents
added evidence to the already large body of facts bearing on the
individuality of the chromosomes and my observations indicate
a persisting chromosomal organization which is constant through-
out the cell-cycle. I have studied the metaphase plates of
hundreds of cells, germinal and somatic, and have found no
variations in either chromosome number, the relative sizes of
the chromosomes or their characteristic grouping. The only
exceptions to this is found in the giant spermatocytes and the
multinuclear cells in the adhesive gland of the female, where the
increase in chromosome number is due to the suppression of the
division of the cell-body at mitosis resulting in the formation of
polyvalent chromosome groups. In such polyvalent cells, we
can still recognize double, triple, quadruple, etc., sets of each of
the chromosome pairs.

There have appeared at various times in the cytological
literature discrepancies of chromosome numbers in certain species,
purporting to show that chromosomes vary in number and cannot
be regarded as persistent structures of the cell. McClung ('17)
has dealt ably and at length with such criticisms and has par-
ticularly concerned himself with the work of Delle Valle. Ac-

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 447

cording to Delle Valle, chromosomes are not constant structures
of the cell and their number varies as ordinary fluctuating
variations, since the number in a particular cell depends upon
the mean size of the chromosomes, the amount of chromatin
in the nucleus being constant (McClung, '17, p. 548). I have
already (page 409) called attention to the remarkable difference
between the metaphase chromosomes of the young follicle-cells
and those of old follicle-cells (compare Figs. 7, 10, 11). In the
old follicle-cells, the chromosomes are somewhat longer, thinner
and poor in chromatin constitution. The nuclei from which such
chromosomes are derived are much poorer in basichromatin
than the nuclei of the young follicle-cells, usually having only a
single small mass of basichromatin. Nevertheless, as will be
seen from Figs. 10 and n, the chromosome number remains
constant and their relative sizes are similar to the chromosomes
of other diploid groups; their only difference seems to be that
they possess less chromatin. It is at once evident that this
condition cannot be interpreted on the basis of Delle Valle's
hypothesis. It also brings to light the fact that the material
concerned in maintaining the chromosome number and size is
not the chromatin of the nucleus, which is apparently more
or less variable in amount, but the underlying structural basis
of the chromosomes, namely the linin. In a previous paper
(Shaffer, '20), I have emphasized the importance of the linin as
being responsible for the architecture and organization of the
chromosomes and for the maintenance of the stability of the
nuclear elements. The chromomeres of the chromosomes are
linearly arranged in a definite order (Wenrich, '16) which is
maintained constant through the agency of the linin, the struc-
tural basis. Besides this, the linin is also concerned in the
movements and localization of the chromatic elements of the
nucleus. Some of the most fundamental problems of the cell
are bound up with such phenomena, namely, wrhat determines
how the chromomeres shall be arranged in the chromosome, or
by what agency are homologous chromosomes brought together
in synapsis. "As long as 'conjugation' of the chromosomes is
dealt with as though they were entities with independent power
of movement, instead of the processes back of it, the super-

ELMER L. SHAFFER.

structure of theory must remain as unwieldy as at present"
(Kingsbury, '12, p. 48). As I have before pointed out, it is
possible that we may find an explanation for such movements of
the chromatic elements in the linin ground-work of the nucleus.
Homologous chromosomes have linin connectives running be-
tween them and it is possible that their union in synapsis is
brought about by a contractility of these interchromosomal linin
fibers, very much as the spongioplasm acts in localizing substances
in the egg (Conklin, '17). The linin is, therefore, the persistent
material of the nucleus, while the chromatin may be variable
in amount in certain phases of the cell-cycle.

(&) Mitochondria. — The function or the role of the mitochon-
dria in the cells of animals and plants still remains one of the
unanswered cytological problems of to-day. A few cytologists
have insisted that the mitochondria are idioplasmic materials
which have a role in the transmission of hereditary characters
similar to the chromosomes. According to this view, the chro-
mosomes bear the determiners for the generic or racial characters
of the organism, while the mitochondria bear the determiners
for the specific or individual hereditary characters. From what
little we do know of "cytoplasmic inheritance" it seems that the
reverse is true (Conklin, '17) and that the larger orientations
of development are fixed by the cytoplasm, particularly that of
the egg. The idioplasmic view of the mitochondria was de-
veloped from the fact that they were found in all cells at all
times and that they behaved characteristically during mitosis,
becoming equally distributed to the daughter cells. These facts
seemed to indicate a persistence and continuity of the mito-
chondrial substance through the cell-cycle which would fit in
with the view of their idioplasmic nature. As I have before
indicated (p. 445) there is no basis for maintaining their genetic
continuity, but rather that "new" mitochondria may be formed
in the cell without any relation to previously existing mito-
chondria.

According to another view the mitochondria may become
transformed into certain histological elements of the cell, such
as muscle and nerve fibrillte, collagenic fibrils, and certain of
the glandular secretions (pancreas, thyroid, etc.). Without

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 449

entering into a -detailed discussion of the "histogenetic" view of
the mitochondria, it may be said that the evidence is far from
being convincing. From what we know of the chemical nature
of the mitochondria, it becomes difficult to understand how
they may become transformed into structures so different
chemically. According to Cowdry ('16, p. 435), "it is apparent
that the doctrine of an actual chemical transformation of mito-
chondria into substances of diverse constitution is weak."

As to the chemical nature of the mitochondria, practically all
workers agree that they are combinations of lipins with varying
amounts of albumin (phospholipins). The transformation of
the mitochondria into the yolk-spherules of the egg at once
indicates their lecithin nature. N. H. Cowdry ('17) has given
a summary of the more important data bearing upon the chemical
nature of the mitochondria. Lowschin ('13) has been able to
make mitochondria artificially in lecithin and albumin solutions.
These mitochondria behave in every way (form, solubility,
fixation and staining) like true mitochondria of organic cells.
Russo ('12) has described an increase in the number of mito-
chondria of the oocytes of the fowl following injections of
solutions of lecithin.

While we are beginning to know something about the chemical
nature of the mitochondria, we are far from knowing their role
in the physiology of the cell. That they bear an important
relation to metabolism is conceded by many workers, and the
class of chemical compounds to which the mitochondria are allied
chemically (the lipins) have recently been emphasized in bio-
chemical works as being intimately concerned in metabolic
processes. In fact, Mathews ('15) believes that the phospho-
lipins are the most important substances in organic matter.

In the oocytes of Cicada it is quite clear that the mitochondria
are related to the nutritive metabolic processes and that they are
actually transformed portions of the products of assimilation.
According to Cowdry ('17), in plants the "mitochondria are
concerned in the formation of chlorophyll, and thus the very
existence of the plant depends upon them." Maclean ('18) fed
one group of hens on a normal diet and another group on a diet
free from fats and lipins and concluded that the "lipins play an

45° ELMER L. SHAFFER.

important, if as yet unknown part in the history of fat meta-
bolism" (p. 172).

Some workers (Kingsbury, '12, Cowdry, '17, etc.) have ex-
expressed the view that the mitochondria are concerned in the
processes of cell-respiration, but there has not been sufficient
experimental work to support this view. Maclean ('18) points
out some very interesting relations between lipins and oxydative
processes and mentions the work of Stanewitch who found that
wheat embryos treated with solvents which extracted most of
the lipins showed a respiration energy which was lower than
normal.

It would, I think, be premature to make any definite statement
as to the function or significance of the mitochondria in organisms
until we know more of the biological significance of the lipins.
Maclean says (p. 107):

"From what has already been said regarding the unsatis-
factory state of our knowledge of the lipins, it follows that their
exact function in the animal and vegetable economy is neces-
sarily obscure. Their great importance is proved by their
general occurrence in every cell, but little or no direct experi-
mental proof indicating their specific function has yet been
obtained. When we consider the obscurity in which the chem-
istry of the lipins has been shrouded and the fact that even now,
in many cases, is not satisfactorily established, it is easy to
understand that many of the properties and functions ascribed
to these bodies are based on little more than the imagination."

For the present, all we can say regarding the function of the
mitochondria is that they form the major part of the lipoid
constitution of organic cells, and when we know more of the
biological significance of lipins, we shall know more concerning
the role of the mitochondria in cell economy. Obviously, further
progress in this direction must come along experimental lines.

F. SUMMARY.

1. The chromosome number in all male diploid groups of
Cicada (Tibicen) septemdecim is 19, and in the female diploid
groups 20.

2. The diploid chromosome groups are characterized by the

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 45!

presence of one large pair of chromosomes (the macrochromosome
pair, AA} and two pairs of somewhat smaller chromosomes
(BB, CC}. The other 13 chromosomes show no size differences.

3. There is no variation in chromosome number or in their
form and arrangement in any of the diploid groups studied.

4. In the spermatocytes there are two ring tetrads of the
Stenobothrus type, which are derived from the A A and BB
chromosome pairs of the spermatogonia. These tetrads divide
reductionally in the first maturation division.

5. An odd chromosome is present which persists as a nucleolus
in the growth stages of the spermatocyte. It passes undivided
to one pole in the first maturation division and divides in the
second division.

6. The tetrads are always grouped characteristically in the
metaphase of the first maturation division and the same grouping
is found in the metaphase of the second spermatocytes.

7. The synaptic stages in the oocyte were studied, giving
evidence that the chromosomes pair side-to-side (parasynapsis).

8. Two chromatin nucleoli are present in the preleptotene
stages of the oocyte. These disappear in the synaptic stages and
reappear in the post-synaptic stages. These nucleoli are inter-
preted to represent the two sex-chromosomes of the female
wrhich go through a synaptic phase like the autosomes.

9. Mitochondria are found in the spermatogonia in the form
of granules localized at the end of the cell bordering on the cyst
cavity.

10. The first evidences of a degeneration of a spermatogonium
is shown by an agglutination of the mitochondria. As degenera-
tion continues the mitochondria continue to agglutinate forming
large lipoid globules in the cell, evidently yielding a fatty de-
generation.

11. In the spermatocytes the mitochondria are filar; they
surround the spindle peripherally at the time of the maturation
divisions and become divided by the cell-constriction. The
mitochondria of the spermatid form the round compact Neben-
kern which later becomes drawn out as a sheath surrounding the
axial filament of the spermatozoon.

12. The ovaries of Cicada are typically Hemipteran in struc-

452 ELMER L. SHAFFER.

ture. The egg-strings of the oocytes pass up into the nurse
chamber and serve to carry the nutrient materials to the oocytes.
In the ovaries of the adult, the nurse-cells are ingested at the
upper end of the egg-string and their ingested products pass down
into the oocyte as nutrient materials.

13. Mitochondria are found in the young oocytes as a deeply
staining mass of granules lying in the cytoplasm at one pole of
the nucleus. In the later stages the mitochondria gradually
extend around the nucleus forming a perinuclear zone of mito-
chondria.

14. The mitochondria increase greatly in numbers during the
postsynaptic stages still retaining their perinuclear arrangement.

15. When the cytoplasmic volume of the oocyte has reached
its maximum, the perinuclear arrangement of the mitochondria
becomes lost and the mitochondria become dispersed toward
the periphery of the oocyte.

1 6. At the periphery of the oocyte, the mitochondria become
transformed into yolk-spherules. First vacuoles are found sur-
rounding the mitochondria, these structures resembling the
"pseudo-nuclei" of Blochmann. The substance of the vacuoles
at first takes the plasma stain lightly and as it grows in size, it
becomes more and more deeply staining. The globules increase
greatly in size and show a marked affinity for the basic stains.

17. The relation betwreen the perinuclear zone of mitochondria
and such structures as yolk-nuclei, etc., is pointed out.

1 8. The zone of the cytoplasm immediately surrounding the
nucleus is taken to be the locus in the cell of the formation of
mitochondria through the chemical action of the nucleus upon
the products of assimilation taken in by the cytoplasm.

LITERATURE CITED.
Athias, M.

'19 Recherches sur les cellules interstitielles de 1'ovaire des Cheiropteres.

Arch, de Biol., T. XXX.
Boring, A. M.

'07 A Study of the Spermatogenesis of Twenty Species of the Membracidae,

Jassidae. Cercopidae and Fulgoridse. Jour. Exp. Zool., Vol. IV., No. 4.
'13 The Chromosomes of the Cercopidae. BIOL. BULL., Vol. XXIV., No. 3.
Browne, E. N.

'13 A Study of the Male Germ Cells in Notonecta. Jour. Exp. Zool., Vol. 14,
No. i.

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 453

'16 A Comparative Study of the Chromosomes of Six Species of Nolonecta.

Jour. Morph., Vol. 27, No. i.
Buchner, P.

'09 Das accessorische chromosome in spermatogenese und ovogenese der

Orthopteren, etc. Arch. f. Zellf. Bd. 3.
Calkins, G. N.

'95 Observations on the Yolk-nucleus in the Eggs of Lumbricus. Trans. N. Y.

Acad. of Sci., Vol. XII.
Conklin, E. G.

'02 Karyokinesis and Cytokinesis, etc. Jour. Phila. A. N. S., Vol. 12.

'06 Sex Differentiation in Dinophilus. Science, Vol. XXIV., p. 294.

'05 The Organization and Cell-lineage of the Ascidian Egg. Jour. Phila.

Acad., N.S., Vol. 13.

'17 The Share of Egg and Sperm in Heredity. Proc. Nat. Acad. Sci., Vol. 3.
'17 The effects of Centrifugal Force on the Eggs of Crepidula. Jour. Exp.

Zool., Vol. 23.
Cowdry, E. V.
'16 The General Functional Significance of Mitochondria. Am. Jour. Anat.,

Vol. 19, No. 3.
Cowdry, N. H.

'17 A Comparison of Mitochondria in Plant and Animal Cells. BIOL. BULL.,

Vol. XXXIII. , No. 3.
Crampton, H. E.

'99 Studies upon the Early History of the Ascidian Egg. Jour. Morph.,

Vol. 15, Supp.
Davis, H. S.
'08 Spermatogenesis in the Acridae and Locustidae. Bui. Mus. Comp. Zool.

Harvard Coll., Vol. 53.
Duesberg, J.

'08 Sur 1'existence des mitochondries dans 1'oeuf et 1'embryon d'Apis mellifica.

Anat. Anzeig., Bd. XXXII.
'10 Nouvelles recherches sur 1'appareil mitochondrial des cellules seminales.

Arch. f. Zellf., Bd. 6, Heft. i.
'12 Plastosomes, "Apparato reticolare interno," und Chromidial-apparat.

Ergeb. der Anat. u. Entwick., Bd. XX.
'18 Chondriosomes in the Testicle-cells of Fundulus. Am. Jour. Anat., Vol. 23,

No. i.
'19 On the Present Status of the Chondriosome Problem. BIOL. BULL., Vol.

XXXVI., No. 2.
Faure-Fremiet, E.

'08 Evolution de 1'appareil mitochondrial dans 1'oeuf de Julus terrestris. C. R.

Soc. de Biol. T. 64.
'10 Etude sur les mitochondries des Protozoaires et des cellules sexuelles.

Arch. d'Anat. Micr. T. n.
Foot, K. and Strobell, E. C.

'n Amitosis in the Ovary of Protenor belfragei and a Study of the Chromatin

Nucleolus. Arch. f. Zellf., Bd. 7, No. 2.
Gajewska, H.

'19 Ueber den sogenannten Dotterkern der Amphibien. Arch. f. Zellf. Bd. 15,
ist Heft.

454 ELMER L. SHAFFER.

Giardina, A.

'04 Sull 'esistenza di una speciale zona plasmatica perinucleare nell'oocite.
Publicazioni de Laboratorio di Zoologica e. di Anat. Comp. dell Universita
di Palermo.
Goldschmidt, R.

'09 Das skelett der Muskelzelle von Ascaris nebst Bemerkungen ueber Chromid-

ialapparat der Metazoenzelle. Arch. f. Zellf., Bd. IV.
Govaerts, P.

'13 Recherches sur la structure de 1'ovaire des Insectes. Arch, de Biol., T.

XXVIII.
Hegner, R. V.

'14 The Germ Cell Cycle in Animals. Macmillan, N. Y.
Kingsbury, B. F.

'12 Cytoplasmic Fixation. Anat. Rec., Vol. 6.
Kornhauser, S. I.

'14 A Comparative Study of the Chromosomes in the Spermatogenesis of

Enchenopa, etc. Arch. f. Zellf., Bd. XII., No. 2.
Korschelt, E.

'86 Ueber die Enstehung und Bedeutung der verschiedenen Zellenelemente des

Insektovariums. Zeit. f. Wiss. Zool., Bd. XLIII.
'89 Beitrage zur Morphologic und Physiologic des Zellkernes. Zool. Jahrb.,

Bd. 4.
Lowschin, A. M.

'13 " Myelinformen " und Chondriosomen. Ber. Deutsch. Bot. Gesell., Bd. 31.
Loyez, M.

'08 Les "noyaux de Blochmann " et la formation du vitellin chez les Hymen-

opteres. C. R. Assoc. Anat. 10 Reunion, Marseilles.
Maclean, H.

'18 Lecithin and Allied Substance. The Lipins. Monographs on Biochem-
istry, London.
McClung, C. E.

'14 A Comparative Study of the Chromosomes in Orthopteran Spermatogenesis.

Jour. Morph., Vol. 25, No. 4.
'17 The Multiple Chromosomes of Hesperotettix and Mermiria. Jour. Morph.,

Vol. 29.
McGill, C.

'06 The Behavior of the Nucleoli during the Oogenesis of the Dragon-fly, etc.

Zool. Jahrb., Bd. 23.
Marlatt, C. L.

'98 The Periodical Cicada. U. S. Dept. Agri. Div. Ent., Bull. 14, N.S.
Marshall, W. S.

'oya Contribution towards the Anatomy and Embryology of Polistes pallipes,

II. Zeit. f. Wiss. Zool., Bd. 86.
'o7b The Cellular Elements of the Ovary of Platyphlax designatus. Zeit. f.

Wiss. Zool., Bd. 86.
Mathews, A. P.

'15 Physiological Chemistry. Wm. Wood and Co., N. Y., 1040 pp.
Montgomery, T. H.

'n The Spermatogenesis of an Hemipteran, Euschistus. Jour. Morph., Vol.
22, No. 3.

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 455

Morgan, T. H.

'08 The Production of Two Kinds of Spermatozoa in the Phylloxerans, etc.

Proc. Soc. Exp. Biol. and Med., Vol. 5, No. 3.
'09 A Biological and Cytological Determination of Sex Determination in

Phylloxerans and Aphids. Jour. Exp. Zool., Vol. VII, No. 2.
Munson, J. P.

'99 The Ovarian Egg of Limuliis. Jour. Morph., Vol. 15.
Nakahara, W.

'18 Some Observations on the Growing Oocytes of the Stone-fly, Perla im-

margenata Say, etc. Anat. Rec., Vol. XV.
Payne, F. W.

'12 A Further Study of the Chromosomes of the Reduviidae. Jour. Morph.,

Vol. XXIII.

"14 Chromosomal Variation and the Formation of the First Spermatocyte
Chromosomes in the European Earwig, Forficula (sp.?). Jour. Morph.,
Vol. XXV., No. 4-
'16 A Study of the Germ Cells of Gryllotalpa borealis, etc. Jour. Morph., Vol.

28, No. i.
Russo, A.

'12 Aumento dei granuli protoplasmic!, etc. Arch. f. Zellf., Bd. 8.
Schafer, F.

'07 Spermatogenese von Dytiscus. Zool. Jahrb., Bd. 23.
Scott, W. J. M.

'16 Experimental Mitochondrial Changes in the Pancreas in Phosphorus

Poisoning. Amer. Jour. Anat., Vol. 20, No. 2.
Shaffer, E. L.
'17 Mitochondria and other Structures in the Spermatogenesis of Passalus

cornutus. BIOL. BULL., Vol. XXXII., No. 5.
'20 A Comparative Study of the Chromosomes of Lachnoslerna. BIOL. BULL.,

Vol. 38, No. 2.
Stevens, N. M.

'053 A Study of the Germ Cells of Aphis rosx, etc. Jour. Exp. Zool., Vol. 2,

No. 3.
'osb Studies in Spermatogenesis with Especial Reference to the "Accessory "

Chromosome. Carneg. Inst. Wash. Pub. No. 36, Part 2.

'06 Studies in Spermatogenesis. Part 2. Carneg. Inst. Wash. Pub. No. 36.
Sutton, W. S.

'02 On the Morphology of the Chromosome Groups of Brachystola magna.

BIOL. BULL., Vol. 4.
Vander Stricht, O.

'98 Contribution a 1'etude du noyau vitellin de Balbiani dans 1'oocyte de la

femme. Verhand. der Anat. Gesell., Jena.
'04 La couche vitellogene et les mitochondries de 1'oeuf des Mammiferes.

Verhand. der Anat. Gesell. Jena.
Vander Stricht, R.

'n Vitellogenese dans 1'ovule de chatte. Arch, de Biol., T. XXVI.
Van Bambeke, Ch.

'95 Contribution a 1'histoire de la constitution de 1'oeuf. II. Arch, de Biol.,
T. XIII.

45^ ELMER L. SHAFFER.

von Winwarter, H. et Sainmont, G.

'09 Nouvelles recherches sur 1'ovogenese et 1'organogenese de 1'ovaire des

Mammiferes (chat). Arch, de Biol., T. XXIV.
Wenrich, D. H.

'16 The Spermatogenesis of Phryonlettix magnus with Especial Reference to
Synapsis and the Individuality of the Chromosomes. Bull. Mus. Comp.
Zool. Harvard Coll., Vol. LX., No. 3.
Wieman, H. L.

'10 A Study in the Germ Cells of Leptinotarsa signaticollis., Jour. Morph., Vol.

21, No. 2.
Wilcox, E. V.

'95 Spermatogenesis of Caloptenus femur-rubrum and Cicada tibicen. Bull.

Mus. Comp. Zool. Harvard Coll., Vol. XXVII.
Wildman, E. E.

'13 The Spermatogenesis of Ascaris megalocephala with Special Reference to

Two Cytoplasmic Inclusions, etc. Jour. Morph., Vol. 24, No. 3.
Wilke, G.

'13 Chromatinreifung und Mitochondrienkorper in der spermatogeneses von

Hydromelra pahidim. Arch. f. Zellf., Bd. X.
Wilson, E. B.

'06 Studies on the Chromosomes, III. Jour. Exp. Zool., Vol. 3, No. i.
'13 A Chromatoid Body in Penlatoma. BIOL. BULL., Vol. XXIV., No. 6.

THE GERM-CELLS OF CICADA (TIBICEN) SEPTEMDECIM. 457

ABBREVIATIONS.

A A, macrochromosome pair.
BB, CC, smaller chromosome pairs,
f., Centrosome.
cy., cytoplasm of oocyte.
cr., chromatoid body.
cr.n., chromatin nucleolus of oocyte.
e.s., egg-string of oocyte.
f.e., follicular epithelium.
id., idiozome.
I.n.c., ingested nurse-cells.
2V., Nebenkern of spermatid.
n.c., nurse-cell.
ncl., nucleus of oocyte.
ooct., oocyte.
pi., plasmosome.
5., spindle derivative.
X., sex-chromosome.

All drawings were made at table level with the aid of a camera lucida. Plates
I, II, III, V (inc.) were made using a 1/12 oil immersion objective and a No. 12
ocular. They have been reduced }± in reproducing them here.

458 ELMER L. SHAFFER.

DESCRIPTION OF FIGURES.
PLATE I. (Fics. i TO 12).

FIGS, i, 2. Spermatogonial chromosome groups, showing macrochromosome
pair (A A), the smaller chromosome pairs (BB, CC) and thirteen other chromo-
somes showing no size differences.

FIGS. 3, 4, 5, 6, 7, 8. Metaphase plates of follicle-cells of ovary, showing 20
chromosomes (female type). The chromosome pairs (AA, BB, CC) are easily
distinguishable. Fig. 8 is late prophase.

FIG. 9. Metaphase plate from embryonic cell, showing 20 chromosomes hence
of female type.

FIGS. 10, ii. Metaphase plates of ovarian follicle-cells (20 chromosomes) from
follicles surrounding old oocytes. Note that the chromosomes are poor in chro-
matin content; in Fig. 11 note the precocious longitudinal split of each chromosome.

FIG. 12. Late telophase of a follicle-cell division in which the macrochromosome,
A, is recognizable in the daughter cells.

BIOLOGICAL BULLETIN, VOL. XXXV II

cc-

Ar-

-A

-•A

2

C1"'

3U

fi

•s

A'—

rZ

-c

•A

3

CC:

.4--.

r — B

"M

10

C.-JR-

^W^; .,

11

,--A

-A

12

ELMER L. SHAFFER.

460 ELMER L. SHAFFER.

PLATE II. (Fics. 13 TO 22).

FIG. 13. Secondary spermatogonium with characteristic chromatin nucleolusi
chromatoid body, and granular mitochondria localized at one end of the cell.

FIG. 14. Spermatogonium in early stage of degeneration. Mitochondria are
larger due to agglutination, and nucleus is polymorphic.

FIGS. 15, 16. Later stages in degeneration of spermatogonia. Mitochondria
continue to agglutinate and form large lipoid globules.

FIG. 17. Pachytene bouquet stage of spermatocyte, showing persisting sex-
chromosome (A'), the macrochromosome loop (AA) and the filar mitochondria
localized about the idiozome (id.).

FIG. 18. Stages in the condensation of the ring tetrad of the macrochromosome
pair (A A).

FIG. 19. Similar stages in the formation of the small ring tetrad of the BB
chromosome pair.

FIG. 20. Early prophase of the first maturation division.

FIGS. 21, 22. Formation of first maturation spindle; mitochondria enveloping
the spindle.

BIOLOGICAL BULLETIN. VOL. XXXVIII.

PLATE II.

•

;

'.••;*

n

jDoooe

A A

18

I 0

--•cr.

ELMER L. SHAFFER.

462 ELMER L. SHAFFER.

PLATE III. (Fics. 23 TO 33).

FIGS. 23, 24. Metaphase plates of first sparmatocytes showing characteristic
grouping of chromosomes; macrochromosoine tetrad in center surrounded by a
circle of autosome tetrads with the sex-chromosome (X) lying outside the group.

FIGS. 25, 26. Successive sections of same cell. Anaphase of first maturation
division showing separation of dyads of the .4.4, BB, CC tetrads. Mitochondria
surrounding the spindle.

FIG. 27. Late anaphase of first maturation division. Sex-chromosome lags
behind and appears bipartite; macrochromosome dyads (.4, .4) showing secondary
split. Mitochondria are divided by cell-constriction.

FIG. 28. Daughter plates of second spermatocyte, one with 9 dyads the other
with 9 dyads plus the sex-chromosome (X). All the dyads show the secondary
split.

FIG. 29. Late anaphase of the second maturation division. Centrosomes
adhere closely to the daughter nuclei; chromatoid body has passed to one of the
daughter cells. Mitochondria are divided by the cell constriction.

FIG. 30. Normal spermatid. Centrosome (c) adheres closely to nuclear
membrane.

FIG. 31. Giant spermatid, with all the structures of a normal spermatid,
except that they are much larger.

FIG. 32. Transformation of the spermatid. Elongation of the Nebenkern
pierced by axial filament. Cr., chromatoid body passing out into tail.

FIG. 33. Stages in spermiogenesis. Elongation of Nebenkern to form sheath
around axial filament.

BIOLOGICAL BULLETIN, VOL. XXXvlll

PLATE III.

A A

CC BB

23

* ^3 »
• •

24

|ff \ 4*^ * A

2 7

s-~.

31

er---

TlV

c r:

29

\

v

i

•--_

ELMER L. SHAFFER.

464 ELMER L. SHAFFER.

PLATE IV. (Fie. 34).

FIG. 34. Longitudinal section through proximal end of ovarian tubule, X 500.
Nurse-cells (n.c.) in nurse chamber with deeply staining cytoplasm. At base of
nurse chamber are found young oocytes in various stages of synapsis. I. n.c.,
region where ingestion of nurse-cells takes place. Products of this ingestion are
seen passing down egg-string (e.s.) to the older oocyte (ooct. 2), which also shows a
perinuclear zone of mitochondria.

BIOLOGICAL BULLETIN, VOL. XXXVII

PLATE IV

^ eft.

n.c.

ELMER L. SHAFFER.

466 ELMER L. SHAFFER.

PLATE V. (Fics. 35 TO 43).
Synaptic Stages in the Oocyte.

FIG. 35. Nucleus of very young oocyte at beginning of growth period. Chro-
matin in form of a delicate network; two chromatin nucleoli are present (Proto-
broque nucleus).

FIG. 36. Deutobroque nucleus. Chromatin beginning to loose net-work ap-
pearance and individual threads become recognizable; two chromatin nucleoli
present. Mitochondria in crescentic area closely applied to nuclear membrane.

FIG. 37. Leptotene stage. Polarization of threads; chromatic nucleoli absent
Mitochondria spreading around nucleus.

FIG. 38. Zygotene stage. Leptotene threads are pairing parasynaptically.
Mitochondria completely surround the nucleus.

FIG. 39. Pachytene stage. Bouin fixation, showing dissolution of mitochondria .

FIG. 40. Pachytene bouquet stage. Macrochromosome loop recognizable
as largest loop.

FIG. 41. Section across bouquet stage showing pachytene threads on end view,
20 in number. Bouin fixation, showing effect on mitochondria.

FIG. 42. Release from bouquet stage; primary split evident in threads. Bouin
fixation, 10 hours, showing artefacts produced by dissolution of mitochondria.

FIG. 43. Strepsistene stage. Synaptic threads separating and twisting about
each other; two chromatin nucleoli are again present. Mitochondria increasing
in numbers in perinuclear zone.

BIOLOGICAL BULLETIN, VOL. XXXVIII

*• '*

,-
* f "'.' ./

' . ;

cr.n.

.

/ ••

••'

•

mm

••> •»
^Tvl

..

/ v * -'>> .*c^-v

/

Q 8
t5 o

36

39

' i^m^j

-

ELMER L SHAFFER

ELMER L. SHAFFER.

PLATE VI. (Fics. 44 TO 48).

FIG. 44. Portion of nucleus (nd.) and cytoplasm (cy.) of oocyte showing migra-
tion of mitochondria from perinuclear region toward the periphery of the cell.
X 1,200.

FIG. 45. Ingestion of nurse-cells by upper end of egg-string (e.s.). Products
of ingestion are seen passing down egg-string into cytoplasm of oocyte. Note
perinuclear arrangement of mitochondria.

FIG. 46. Germinal vesicle of old oocyte showing plasmosome (pi.) and several
chromatic nucleoli. X 800.

FIG. 47. Portion of oocyte and its follicle. Mitochondria arranged in periphery
of cytoplasm. Persistence of egg-string. X 600.

FIG. 48. Stages in the transformation of mitochondria into yolk-sperules.

BIOLOGICAL BULLETIN VO . X>XNI.

i! '£ *i

^|p£?

'•"'"-•JCS'-J"' . ^^k

' ./":?; V

*.w ^'. ** r • £ ^

J

i. /i. c.

.<"*•»-

" -t^

•• I '.

tv^;

i'"-
^t5- *

XA\

198?

t>L li^ff

$*$&

%&v»*

"TI.C,
"-ooct.

• -e. s.

O '•

© V

» . -

V 0

J. c.

/.-

• «

.o

•-

'• j A -V •" * v

• flSsKpry

•• v >.«»'ST;. --

^

• » • •

' • » .,. ^

ELMER L. SHAFFER.

47O ELMER L. SHAFFER.

PLATES VII, VIII, IX.

Photomicrographs taken at a magnification of about 1,000 diameters, except
Nos. 71 and 72 which have been magnified 500. The reproductions here have not
been reduced.

PLATE VII. (Fics. 49 TO 70).

FIGS. 49, 50, 51. Metaphase plates of follicle-cells of ovary, showing 20 chromo-
somes among which the macrochromosome pair can be distinguished. Fig. 51
is at a greater magnification.

FIG. 52. Tripolar spindle in one of the cells of the adhesive gland in the female.

FIGS. 53, 54, 55, 56. Metaphase plates of first spermatocytes from four different
animals, each showing 10 bivalent chromosomes which are similarly grouped in
each case.

FIG. 57. An oblique section through the metaphase plate of the ist maturation
division, showing the two halves of the macrochromosome in the center of the
complex.

FIG. 58. Daughter plates cf second spermatocytes, one with 9 dyads, the other
with 9 dyads plus the sex-chromosome. Note that the grouping of the chromo-
somes is the same as in the first spermatocyte.

FIG. 59. Giant spermatocyte with a great many bivalent chromosomes and
a large amount of mitochondria.

FIG. 60. Anaphase of first maturation division showing mitochondria sur-
rounding spindle.

FIGS. 61, 62, 63. Various forms of ring tetrad (macrochromosome tetrad)
in the early prophases.

FIG. 64. Macrochromosome tetrad in late prophase.

FIG. 65. Late prophase of first spermatocyte, showing character of tetrads.

FIG. 66. Anaphase of first maturation division from smear preparation, showing
separation of macrochromosome dyads.

FIG. 67. Ring tetrad in prophase of first spermatocyte; from smear prepara-
tion.

FIG. 68. First maturation division showing separation of macrochromosome
dyads.

FIG. 69. Anaphase of first maturation division showing separation of the
dumb-bell shaped tetrads.

FIG. 70. Late anaphase of first maturation division, showing lagging sex-
chromosome.

BIOLOGICAL BU.LETIN, VOL. XXXVIII.

4 9

50

52

53

» *.

;*•

55

**»

57

•>

64

65

66

•

£* /V

6/

I

•*•

••*

V

70

L. SHAFFER

472 ELMER L. SHAFFER.

PLATE VIII. (Fics. 71 TO Si).

FIGS. 71, 72. Portion? of the nurse chambers of the ovaries, showing fibrous
appearance of central plasmatic mass in which the egg-strings of the oocytes end.
Nurse-cells in various stages of ingestion may be seen and the products of their
disintegration may be seen passing down the egg-strings into the oocytes. The
oocytes are distinguishable by their deeply staining perinuclear zone of mito-
chondria.

Figs. 73 to 75 are various stages of synapsis in oocyte.

FIG. 73. Protobroque nucleus at beginning of synaptic period. Two chromatic
nucleoli are present.

FIG. 74. Deutobroque nucleus of oocyte. Treads become more evident;
two chromatic nucleoli are present.

FIG. 75. Leptotene stage. Polarization of the threads. Chromatic nucleoli
are absent.

FIG. 76. Pachytene bouquet stage.

FIG. 77. Section across bouquet stage, showing 20 threads on end view.

FIG. 78. Strepsistene stage. Reappearance of the two chromatin nucleoli.

FIG. 80. Oocyte from ovaries fixed in Bourn's fluid (10 hours) showing effect
of acetic acid on mitochondria. Vacuoles, globules and nucleolar-like structures
are found in cytoplasm due to the partial dissolution and agglutination of the
mitochondria.

FIG. 81. Oocyte after fixation in Benin's fluid for six hours, showing perinuclear
zone of mitochondria only partially destroyed. Two chromatin nucleoli are
present in the nucleus.

BIOLOGICAL BULLETIN, VOL. XXXVIII.

I

. v;~v

i

7-4

*

76

ELMER I . SHAFFER

474 ELMER L. SHAFFER.

PLATE IX. (Fics. 82 TO 87).

FIG. 82. Oocyte with its egg-string, from ovaries fixed in Bouin's fluid 15
hours showing the complete disappearance of mitochondria.

FIGS. 83, 84. Oocytes from Flemming fixed material showing characteristic
arrangement of mitochondria in a perinuclear zone sharply delimited from the
rest of the cytoplasm.

FIG. 85. Oocyte at the period when the cytoplasmic volume has reached its
maximum. The mitochondria still surround the nucleus, but the zone is not so
sharply delimited. Shrinkage of the nucleus away from the cytoplasm shows the
nuclear membrane to be intact.

FIG. 86. Older oocyte in which perinuclear arrangement of mitochondria is
lost, the granules becoming scattered toward the periphery of the cytoplasm.

FIG. 87. Typical germinal vesicle of almost mature oocyte, showing two
chromatic nucleoli.

'-'

J,.

« , ,

i

+

86

• !

ElMER L. SHAFFER

MBL WHOI LIBRARY

IdH 17KA 1

\