Full text of "Journal of morphology"

-I M-

iSC

^J***>

vJKPHULUO

lUv» i

Busdirv Institution, )'., .

iSTAR 1

Charles Otis Whitman
1901

Enlargement from a group photograph

JOURNAL

MORPHOLOGY

Founded by C. O . Whitman

EDITED BY

J. S. KINGSLEY

Tufts College, Mass.
with the collaboration of
Gary N. Calkins T. H. Montgomery

Columbia University University of Pennsylvania

W. M. Wheeler William Patten

Bussey Institution, Harvard University Dartmouth College

Edwin G. Conklin

Princeton University

VOLUME 22
1911

THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY
PHILADELPHIA

/ / 9^

WAVERLY PRESS

BALTIMORE, U S. A.

PREFATORY NOTE

In 1909 a number of friends of Professor Charles Otis Whitman
planned a volume of the Journal of Morphology as an acknowl-
edgment of the debt of American science to him as the founder
and editor of the Journal.

A committee, consisting of Frank R. Lillie, Edwin G. Conkhn
and Thomas H. Morgan, was appointed to receive contributions
of articles from his former students and his associates of the
Marine Biological Laboratory. It was decided that the numbers
for the year 1911 should constitute a 'Whitman Volume.'

Professor Whitman died in 1910 before the first number was
issued and the volume becomes a memorial to one who had a wide
influence in the elevation of biological science.

As Whitman's ideals were broader than mere morphology, so
the volume in the scope of its contents overlaps that field on all
sides.

The articles which were accepted by the committee were so
numerous and extensive that not all of them could be published
during the present year, although the volume is far larger than
usual; consequently some of them will appear, with proper
acknowledgment, in the next volume of the Journal of Morphology.

THE

CHARLES OTIS WHITMAN

MEMORIAL VOLUME

EDITORIAL COMMITTEE

Frank R. Lillie
Edwin G. Conklin
Thomas H. Morgan

TO THE MEMORY OF

CHARLES OTIS WHITMAN

1842-1910
THIS VOLUME IS DEDICATED

HIS PUPILS AND ASSOCIATES

CONTENTS

1910

No. 1. MARCH

Bennet M. Allen. The origin of the sex-cells of Amia and Lepidosteus.

Twenty-seven figures 1

Leo Loeb. The cyclic changes in the ovary of the guinea pig 37

Edmund B. Wilson. Studies on chromosomes. VII. A review of the

chromosomes of Nezara; with some more general considerations. Nine

figures. One plate 71

C. B. Davenport. The transplantation of ovaries in chickens HI

W. J. Moenkhaus. The effects of inbreeding and selection on the fertility,

vigor and sex ratio of Drosophila ampelophila 123

G. H. Parker. The mechanism of locomotion in gastropods. One figure 155

No. 2. JUNE

C. M. Child. The regulatory processes in organisms 171

Lorande Loss Woodruff. Paramecium aurelia and Paramecium caudatum.

One figure ' 223

E. A. Andrews. Male organs for sperm-transfer in the crayfish, Cambarus

affinis; their structure and use. Four plates and thirty-one text figures. . 239

Wallace Craig. Oviposition induced by the male in pigeons 299

William Morton Wheeler. The ant-colony as an organism 307

Oilman A. Drew. Sexual activities of the squid, Loligo pealii (Les). Four

plates 327

Frank R. LiLLiE. Studies of fertilization in Nereis. I. The cortical "changes

in the egg. II. Partial fertilization. One double plate 361

W. E. RiTTER and Myrtle E. Johnson. The growth and differentiation of the

chain of Cyclosalpa affinis (Chamisso). Four plates 395

Oscar Riddle. On the formation, significance and chemistry of the white

and yellow yolk of ova. Three plates . 455

Xll CONTENTS

No. 3. SEPTEMBER

Charles W. Hargitt. Some problems of eoelenterate ontogeny. Three

plates and three text figures 493

Victor E. Shelford. Physiological animal geography. Nineteen figures. . . 551
R. M. Strong. On the olfactory organs and the sense of smell in birds. Two

plates and four text figures 619

Henry H. Donaldson. On the regular seasonal changes in the relative

weight of the central nervous system of the leopard frog. Five charts. . . 663
Ralph S. Lillie. The physiology of cell-division. IV. The action of salt

solutions followed by hypertonic sea-water on unfertilized sea-urchin

eggs and the role of membranes in mitosis. Three figures 695

ThOs. H. Montgomery, Jr. The spermatogenesis of an hemipteron, Eus-

chistus. Five plates— 147 figures 731

Winterton C. Curtis. The life history of the Scolex polymorphus of the

Woods Hole region. Thirteen figures 821

No. 4. DECEMBER

H. H. Newman and J. Thomas Patterson. The limits of hereditary control
in armadillo quadruplets : A study of blastogenic variation. Five figures
and eight plates 855

Charles Zeleny. Experiments on the control of asymmetry in the develop-
ment of the serpulid, Hydroides dianthus. Seven figures 927

William A. Locy. Anatomical illustration before Vesalius. Twenty-three

figures 945

S. J. Holmes. Minimal size reduction in planarians through successive re-
generations 989

E. H. Harper. The geotropism of Paramoecium. Five figures 993

Elliot Rowland Downing. The formation of the spermatophore in Areni-
cola and a theory of the alternation of generations in animals. Four
plates and seven text figures 1001

Biography, Charles Otis Whitman. Five portraits xv

■e in Chi

■■'}.;m.'\

iitury, ai/i»)iig
■ :ufather o^ '•
i iipr (born ' i^ chiidi

married Jiiii T68I

I. wo SODS, -l8(},iflj| a .a loeeal.Til ?o EKSiiboffl .sqxJ-nll «■ int>ii ^oemejielnS

born in Wovcest^
His early lifev
moved to Waterford i(
stock. He attended tl
ford "and fitted for '
teaching winters to o\y
penses.'^ "He early fi(
»v}iile here (Woodstocl'
v'cry fine collection oi ■

lu the preparation
■■ Mrs. Whitman, Dr.
'of. Cornelia M. Clap >

ixm

Charles Otis Whitman
1857

Enlareotneat from a tin-type. Kindness of Professor E. S Morse

CHARLES OTIS WHITMAN

FRANK R. LILLIE

Charles Otis Whitman was born December 14, 1842, in Wood-
stock, Maine, He died at his home in Chicago, December 6, 1910.
On his father's side^ he was descended from Jacob Whitman, who
was a resident of Bridgewater, Massachusetts, whence he emi-
grated to Buckfield, Maine. Three sons of Jacob Whitman
settled in Woodstock, Maine, about the beginning of the nine-
teenth century, among them Joseph (born September 30, 1783),
the grandfather of the subject of our sketch. Whitman's
father (born February 19, 1821) was the eighth of ten children;
his mother (born December 12, 1823) was Marcia Leonard,
daughter of Solomon Leonard, also of Woodstock. Whitman
married Emily Nunn of Peru, Ohio, in August, 1884, and had
two sons, Francis, born in Milwaukee, Wisconsin, and Carroll,
born in Worcester, Massachusetts.

His early life was spent in Woodstock, though his father re-
moved to Waterford for a while, subsequently returning to Wood-
stock. He attended the town schools in Woodstock and Water-
ford ''and fitted for college at Norway and other academies,
teaching winters to obtain the means for paying his school ex-
penses." "He early developed a taste for natural history, and
while here (Woodstock) and a boy, he procured and mounted a
very fine collection of the birds of Maine. So artistically pre-

^ In the preparation of this brief biography I have been indebted for information
to Mrs. Whitman, Dr. Wallace Craig, Prof. H. H. Donaldson, Prof. E. S. Morse,
Prof. Cornelia M. Clapp, George T. Little, the librarian of Bowdoin College, Mrs.
Cornelia Fletcher Day (Westford, Mass.), Mrs. Sarah H. Trumbull (Beverly,
Mass.), Mrs. Helen Keith Frost (Westford, Mass.), Secretary of the Boston School
Committee, Headmaster English High School, Boston, Mass., Prof. E. L. Mark,
Edward Phelps AUis, Jr., Dr. Reinhardt Dohrn, and others. I must also express
my indebtedness to E. G. Conklin and T. H. Morgan, for criticism of the manu-
script. The section dealing with Whitman's scientific work was prepared by
T. H. Morgan, E. G. Conk in, J. Percy Moore and others — See foot-note p. xlvii.

XVI CHARLES OTIS WHITMAN

pared where they, and so naturally mounted, that they attracted
much attention among ornithological students." ^

It would be interesting to know more of his early life but Pro-
fessor Whitman rarely spoke of it, though he referred at times to
work on his father's farm. In reply to the question whether
he had been interested in natural history as a boy, he replied to
Professor Wallace Craig that he judged he must have been, be-
cause of his persistence in getting his grandfather to tell hunting
stories. He never tired of the stories, and often walked a mile
to have an evening of them; his grandfather was very kind in
always telling these when asked. He also said that he kept
pigeons as a boy, and was fascinated by them and sat and watched
them by the hour, intensely interested in their feeding, their
young, and in everything that they did.

We thus get a distinct though undetailed view of a boyhood
spent on a New England farm, an education acquired by dint of
labor and self sacrifice, and of an original interest in natural
history, shown in his observation of pigeons and his collection
of the birds of Maine.

He entered Bowdoin College as a sophomore in September, 1865,
and graduated with the degree of Bachelor of Arts in July, 1868.
The college curriculum of this time was the usual course of re-
quired studies with much emphasis on the classical languages,
some study of modern languages and of mathematics, the elements
of philosophy and a variety of sciences taught no doubt mainly
from text-books. The influence of this classical education re-
mained with him all his life, and was no doubt responsible for
the views that he entertained in favor of the requirement of Latin
for college education. There was certainly little to stimulate
his interest in the field in which he subsequently won distinction.
His membership in the Greek Letter Society Delta Kappa Epsi-
lon, in the Athenaeum Society (literary), and in the Philologian
Society (debating) may help to indicate his social and intellectual
interests at this period. At his graduation he ranked about
ninth in a class of twenty-three. The title of his commence-

2 See Lapham, William B. "History of Woodstock, Me., with family sketches
and an appendix." Portland, Stephen Barry, Printer, 1882.

BIOGRAPHICAL SKETCH XVll

ment oration ''Free Enquiry" indicates already an unfettered
mind.

On graduating from Bowdoin, Whitman was appointed prin-
cipal of Westford Academy in Westford, Massachusetts. He
began to teach there on December 16, 1868, and remained until
the spring of 1872. He must have taught a great variety of sub-
jects, to judge by the catalogue of 1872, as there was a four years'
course involving mathematics, English, Latin, Greek, French,
geography, book-keeping, history, natural philosophy, chemistry,
mental philosophy, astronomy, physiology, and botany, and there
were but two assistant teachers and ninety pupils in 1871-72.
However, he continued his interest in birds and taxidermy, and
the library of the Academy still has a good collection of Westford
birds prepared by a lady whom Whitman instructed in the art
while there; it also contains a fine specimen of one of the largest
of Maine loons set up by Whitman himself.

During the school year 1871-72 Whitman substituted in the
Enghsh High School in Boston, Massachusetts, and was regularly
appointed sub-master in September, 1872. At that time the
departmental system had not been introduced into the school and
he taught general high school subjects. He remained with the
school until the summer of 1875.

While in Boston Whitman came under the influence of Louis
Agassiz, and was one of the fifty students who, in July and
August, 1873, attended the Anderson School of Natural History
founded by Agassiz on the island of Penikese. Here he met
Professor E. S. Morse, who was an instructor under Agassiz, a
circumstance which had a great effect in Whitman's later life,
leading to his call to the University of Tokyo as related further
on. Professor Morse was much attracted to him by the beauti-
ful and accurate way in which he drew the lower forms of life,
particularly the Ascidian Perophora, on which Morse himself
was working at the same time. Morse and Whitman remained
the best of friends throughout life, and at Whitman's invitation
many years later Morse delivered several lectures at the Marine
Biological Laboratory.

XVlll CHARLES OTIS WHITMAN

Louis Agassiz died in December, 1873, and the Penikese school
was opened again in 1874 for the last time by his son, Alexander
Agassiz. Whitman was again one of the privileged fifty who
worked there, though ninety other applicants had to be refused
admission for lack of accommodations. The Penikese school
started a tide of biological work at the sea-shore in American which
ebbed indeed for a while, but began to flow again in the decade
of the eighties and has been running stronger ever isnce. No doubt
the germ of the Marine Biological Laboratory, Whitman's most
significant scientific enterprise, was implanted there in Whitman's
heart and in the hearts of others. Doctor Craig states that
Whitman felt that he got his first start in scientific zoology from
Agassiz, but that he did not really get under way until he worked
with Leuckart on Clepsine in Germany. Asked by Doctor Craig
(in August, 1910) what he thought of Agassiz's method, Whitman
replied that he did not think much of it at first but that as time
went on he thought more and more of it. ''We are apt to do
the work for the student too much. What we should do is to
set him a problem and let him work it out."

In 1875 Whitman decided to go to Germany to study natural
history. Apparently he had not yet decided to abandon his
career as teacher in the high school, for he left open the possibility
of returning to his position after a year's absence. He sailed in
July, 1875, and settled in Leipzig. From there he wrote to his
successor in Westford Academy, Mr. William E. Frost, May 28,
1876: ''Mr. Seaver (the head-master of the Enghsh High School
at that time) says he will secure my re-election and another
year's absence if possible. I have not much doubt of his ability
to do this. At any rate I shall remain another year." But
when 1877 arrived he was not yet ready to leave and he decided
to remain a part, at least, of a third year. In 1878 he received
the degree of doctor of philosophy from the University of Leipzig,
and sailed for America in July of the same year, although he still
wished "to remain a little longer in Deutschland, but the Fates

5 Letter to Mr. Frost, Jul.v 6, 1878.

BIOGRAPHICAL SKETCH XIX

In 1878 he published his first scientific paper ''The Embryology
of Clepsine" in the Quarterly Journal of Microscopical Science,
vol. 18, pp. 215-315. This was his doctor's thesis; in many
respects it was a very notable, indeed epoch-making work. It
was the first time that the primordia of any ectodermal organs
had been followed to individual cells, and that the cleavage process
itself had been adequately interpreted as a process of 'histoge-
netic sundering.' He laid emphasis on the existence of embryonic
axes in the unsegmented egg, and anticipated to a considerable
extent views that did not receive adequate recognition until
the period of study of ' cell-lineage' began about fifteen years later.

On his return to America he was appointed Junior Master, first
grade, of the English High School in Boston, teaching English,
and resigned in 1879. It is evident that now for the first time,
at the age of nearly thirty-seven, he had irrevocably decided to
devote himself entirely to zoology, for by his resignation he burned
his bridges behind him. He received an appointment as fellow
in biology in Johns Hopkins Uni-^^/?rsity for 1879-80, but he did
not enter on the fellowship, having in the meantime accepted the
chair of zoology in the University of Tokyo. He sailed for Yoko-
hama, August 21, 1879. On the voyage he made observations
on the flight of flying fish which he described in the American
Naturalist, vol. 14, 1880, maintaining that their course through
the air is actual flight.

WHITMAN IX TOKYO

With his appointment to the University of Tokyo in 1879,
begins Whitman's real influence as a teacher and organizer in
zoology. He was then nearly thirty-seven years of age and had
passed through a most varied preparation for his life-work. He
had pushed on without haste but without rest, always carrying
with him his original and vital interest in living things since he
first studied pigeons as a boy, in spite of the necessity of earning
a livelihood by teaching school. He thus came to his chosen
life work in full maturity with a mind broadened by varied expe-
riences, yet with actual boyish enthusiasm and interest, that
never left him throughout fife.

XX CHARLES OTIS WHITMAN

The work in zoology in the Imperial University of Tokyo was
first organized by Professor E. S. Morse, who was invited from
abroad in 1877. ^ He remained there two years and was succeeded
in 1879 by Professor Whitman. Professor Iwakawa states that
Professor Huxley was first invited by Professor Morse to accept
the chair as his successor. Professor Huxley wrote that for years
he had been desirous of studying biology in oriental countries
and that the present call from Tokyo was the best chance he could
ever have; however, he regretted that the dechning condition of
his health would not allow him to accept. Professor Morse
states in a letter that while instructor at Penikese he had known
Whitman and was much impressed by the beauty and accuracy
of his work ; his experience as teacher in Boston was also a recom-
mendation; so Professor Morse secured Whitman's call to the
chair of zoology in Tokyo and it was accepted. Professor Whit-
man remained in Japan for two years until 1881. He had only
four students, but as all became professors of zoology in the Im-
perial University he may be xu -,tly regarded (as Dr. Takahashi
states) as the father of zoolo'^Ty in Japan. Professor Iwakawa
says that Professor Whitman's teaching really laid the foundation
of modern zoology in Japan.

It is impossible to reproduce the tone of affection and reverence
in which these reminiscences are written by two of his original
pupils, Iwakawa and Ishikawa, and a later student during the
Chicago period, Takahashi. Professor Iwakawa says, "Once he
was my teacher while he was in Japan and since then until today
I have been paying respects and admiration both for his character
and for his work in biology." "I am constrained by what I
regard as a duty to him to let others get a glimpse of what I
knew him to be while he was with us in Tokyo" — and the whole
tenor of his reminiscences is one of affectionate admiration and
devotion. ''Professor Whitman's attitude of mind toward his

* In the Magazine of Zoolog}-, published by the Zoological Society of Japan,
Tokyo, vol. 23, no. 269, March 15, 1911, there appear three articles on Professor
Whitman, the first by Professor Tomotaro Iwakawa, the second by Professor
Chiyomatsu Ishikawa, and the third by Dr. Katashi Takahashi. For the trans-
lation of these articles I am indebted to Dr. Shigeo Yamanouchi. They form the
basis of the following account.

Charles Otis Whitman

1882

From Lapham's History of Woodstock, Maine,

BIOGRAPHICAL SKETCH XXI

pupils was such as a mother toward her son." Professor Ishikawa
writes in a similar spirit throughout. He says, "On receiving
the tidings of Professor Whitman's death I am very much sur-
prised and bitterly mourned." "I mourned bitterly in the recol-
lection that those delightful days we had together shall never
again be realized, but have now become a memory." "The work
he has done during his life still remains and will be remembered
forever." Takashahi says, "As he was the teacher of our pro-
fessors, he will be justly regarded as our father of zoology in Japan.
I feel as if I had lost my grandfather because of his being the
teacher of our professors and because of his cherished kindness
shown to me as a father might have shown to his son during my
stay in his laboratory in the University of Chicago."

The following incident, as related by Professor Iwakawa and
translated by a Japanese friend, is worth quoting:

For the purpose of making bird specimens for the museum, the Uni-
versity secured two government licenses in hunting seasons and the
licenses were handed to the Zoological Department for the use of the
students.

To make the specimens was one of purposes of hunting and the
other end seemed to eat flesh of birds. One Saturday, a number of
pigeons was brought to our laboratory and the next day being Sunday,
some of us came to the laboratory to have the share of feast. Dr.
lijima dissected the birds. A fire shovel was cleansed and put lard on
and then flesh; then put into the stove to fry. Salt, sauce, knife and
fork were ready and the party waited to have the flesh cooked. Dr.
Sasaki had belatedly come. Being he was an hearty eater, the party
who were already there refused to add him into the company and so all
went to Dr. Whitman's office and locked the door from inside so that he
could not get in. Unexpectedly Dr. Whitman came, in spite of that
the day was Sunday. He put his slippers on as usual and tried to get
in his office by the door he used to enter. To his surprise the door was
locked. He came over to our laboratory and there Dr. Sasaki sat alone.
Professor Whitman tried to open the door that leads to his office from
our laboratory. Again to his surprise the door was also locked. Dr.
Sasaki being left very much uneasy, called out, 'Professor Whitman
has come!' The party inside the door, including Dr. lijima, believing
that Professor Whitman would never come on Sunday and that the
warning might be Dr. Sasaki's stratagem to induce the party to open
the door, took the alarm easy and were chattering over quite noisily.
Then there was heard a voice from outside, 'Who are in the room?'
Evidently that was Whitman's. Frightened all at once the party fled

XXU CHARLES OTIS WHITMAN

to a court from the door which leads to the corridor. The party now in
the court, sent a spy to look after what Professor Whitman was doing
and assured that he, taking a few things, has gone home. As the coast
was clear, the party returned to the laboratory. Dr. Sasaki who had
been left alone in the laboratory, had already helped himself alone with
the fried birds that were left in the stove, and was sitting quite satis-
fied. He spoke smilingly to the party, when they entered in. 'I was
very much embarrassed indeed ! The handle of the shovel was peeping
outside the mouth of the stove. Fried birds were making cooking noise
inside and a tempting odor was ejecting from within. But our teacher
had not asked a bit. Simply he said, ' who are that ran out of the room?'
So I answered, 'I do not know.' 'At any event I thank you for your
kind feast.' Dr lijima seemed very much disappointed; as yet the case
being as such that Dr. Sasaki only cannot be blamed at, and he spoke
to Dr. Sasaki, 'You lucky fellow.' The following day we were in a con-
stant concern and reluctantly expected some sort of punishment on our
conduct of the previous day. On the contrary Professor Whitman
didn't even ask a word about what thus happened the previous day.
Professor's Whitman's attitude of mind toward his pupils v/as such as
mother toward her son.

There were but two rooms devoted to the department of zool-
ogy in the Imperial University in Whitman's day; literature and
apparatus were very scanty and Whitman first introduced modern
laboratory equipment and methods in microscopical technique.
His four students were Ishikawa in the first year class, lijima and
Iwakawa in the second year, and Sasaki in the third year. There
was an assistant, a janitor and two artists, all of whom were kept
busy collecting and drawing leeches. It was characteristic of
Whitman that he should set each of his four students to work on
a special problem for research, even Ishikawa in his first year of
zoology.

Hard work was the order of the day in Whitman's laboratory;
he set the example himself, and the students, who lived in a
dormitory near by, often worked until midnight. Twice a day
Whitman consulted with each student about his work. In the
absence of a University biological library Whitman placed his
own journals and books at the disposal of the students and aided
them in translating German and French. He kept each man
close to the study of his individual problem and deplored the wast-
ing of time spent on other subjects. From time to time he
delivered lectures on special topics and as a general course he
expounded Spencer's Principles of Biology.

BIOGRAPHICAL SKETCH XXHl

At the end of two years each of the four students had a paper
ready for pubhcation and Professor Whitman presented them to
the Journal of the College of Science of the Imperial University
for publication; the officials in charge of the journal replied that,
as the journal was organized for the purpose of publishing the
researches of professors, any theses of students worth publishing
should be published under the name of the professor. This
aroused Professor Whitman's indignation, and he withdrew the
papers, remarking that he would never again present papers to
the University for publication. He then sent three of them to
the Quarterly Journal of Microscopical Science where they were
published. Sasaki's paper on Salamander was published after
Whitman's departure in the Journal of the Science College.

This incident seems to have been the beginning of an estrange-
ment between Whitman and the administration of the University,
which was aggravated by the inability or unwillingness of the
administration to accede to many of his requests for more ade-
quate equipment for the department. The period of his appoint-
ment having come to an end in 1881, the University requested
him to remain, but the proposal was refused and in August, 1881,
he left Japan without bidding formal farewell to the University.
He published a short brochure entitled "Zoology in the University
of Tokyo" shortly before he left, but that he had no intention of
wantonly hurting Japanese susceptibilities is evident, as Professor
Ishikawa states, from a thorough study of it. He had made a
close study of the Japanese and he discovered and pointed out
their most obvious weak points in an honest and essentially
friendly fashion. ''Professor Whitman loved Japan and sym-
pathized with the Japanese. That his love and sympathy
poured forth to the Japanese in a degree far surpassing any ever
shown to us was marvelously evidenced at the time of the Russo-
Japanese War" (Ishikawa). Indeed, those of us who were with
Whitman at this time knew that he could not have suffered
more keenly in the misfortunes or rejoiced more in the triumphs
of his own country. This was fully realized by the Japanese
people and the slight unpleasantness of his departure was soon
forgiven.

XXIV CHARLES OTIS WHITMAN

IN EUROPE

Professor Whitman left Japan in August, 1881, and from Novem-
ber 11, 1881, to May 2, 1882, he worked at the Zoological Station
of Naples as guest of Professor Dohrn. His sojourn in the Zoolog-
ical Station laid the foundation of an everlasting friendship with
Dohrn, and, when he left, Professor Dohrn gave him a testimonial
recommending him strongly to some professorship. While at
Naples Whitman studied the embryology, life-history, and classi-
fication of Dicyemids and wrote a paper on the subject, published
in January, 1893, which is still the standard work of reference.

Whitman had thus come under the influence of three of the
great leaders of his time in zoology, Agassiz, Leuckart and Dohrn.
His original bent in the direction of the natural history of birds
was diverted by these experiences towards the study of marine
life and lower organisms, but later on he returned to his original
interests in birds, particularly pigeons, with a mind deepened by
intimate acquaintance with the fundamental problems of biology.

After leaving Naples he went to Leipzig where he remained until
the middle of September, engaged among other things in pre-
paring his Naples work for publication. On September 1, 1882,
he wrote to Mr. Frost: ''I leave for America on the 15th of
September and shall go to Leonard's (Newton Highlands) and
shall hope to see you somehow or somewhere. I have not yet
decided where to spend next winter. There is some possibility
of my going to Johns Hopkins — though nothing definite yet.
Have just finished manuscript of work done in Naples, and a
portion is already in print."

AT HARVARD

In the autumn of 1882 he was appointed Assistant in Zoology
at the Museum of Comparative Zoology of Harvard University,
and held this position until 1886. In the spring of 1883 he went
to Key West, Florida, to secure for Mr. Alexander Agassiz material
with which to complete Mr. Agassiz's monograph on ''The Por-
pitidae and Velellidae." Though he spent six weeks there he did
not meet with success. In the summer of 1883 he worked at

BIOGRAPHICAL SKETCH XXV

Mr. Agassiz's Newport Laboratory on the development of pelagic
fish eggs. Some of the results, worked up later in Cambridge,
were published with Mr. Agassiz in two papers (1884 and 1889).
In the first paper the origin of the periblast was correctly described
for the first time, a most important contribution in view of the
confusion of opinions on this subject. During this summer he
met his future wife, Miss Emily Nunn, who was also working
at Mr. Agassiz's laboratory.

The years 1883-1886 were productive years; during part of
this time Whitman edited the department of ''Microscopy" of
the American Naturalist. He worked out and published his
papers on ''A Rare Form of the Blastoderm of the Chick" ('83),
''External Morphology of the Leech" ('84), "On the Develop-
ment of Some Pelagic Fish Eggs" ('84), "Segmental Sense-Organs
of the Leech" ('84), "The Leeches of Japan" ('86), "The Germ-
Layers of Clepsine" ('86), and some minor papers. He also
prepared and published his book on "Methods of Research in
Microscopical Anatomy and Embryology" ('85) (see Bibliog-
raphy).

THE LAKE LABORATORY AND THE FOUNDING OF THE JOURNAL OF
MORPHOLOGY

From 1886 to 1889 Whitman acted as director of the Lake
Laboratory at Milwaukee, Wisconsin,^ founded by Edward
Phelps Allis, Jr. Mr. Allis had decided to start a laboratory for
biological and related research and Whitman was recommended
to him as a proper person to take charge. There followed a con-
ference in which the plans and purposes of the laboratory were
discussed, and Whitman then presented the need of an American
journal for publication of zoological research, pointing out that
American workers were obliged either to present their papers to
some scientific society or to send them for pubhcation to some one
of several European journals. Whitman then asked Mr. Allis
if he would consider the publication of such a journal, in con-
nection with the laboratory. He was asked to submit figures and

^ I am indebted to Mr. Allis for some of the information on which the following
statements are based.

XXVI CHARLES OTIS WHITMAN

plans, and it was finally arranged that he should come to Mil-
waukee, take charge of the laboratory, to be known as the Lake
Laboratory, «and also edit with the cooperation of Mr. Allis, a
journal to be called the Journal of Morphology. The journal
was to be a model of publications of the kind.

Whitman may not have been the first to realize the need of
establishing a journal of zoological and anatomical science in
America, but he was the first to possess sufficient courage, energy
and influence to set about realizing the need. He was fortunate
indeed to find a man of scientific attainments and enthusiasm
with an ample and liberal purse to support him in this under-
taking. In the introduction to the Journal Whitman wrote,
"The mixed character and scattered sources of our publications
are twin evils that have become intolerable both at home and
abroad. The establishment of the Journal of Morphology may
not be the death blow to these evils; but there is hope that it will,
at least, relieve the more embarrassing difficulties of the present
situation."

In its make-up both scientific and typographical, the Journal
of Morphology was a model of what a research publication should
be, and it did much to coordinate zoological research in America,
to give it a worthy setting, and to make it better known abroad.
Eighteen volumes were published between 1887 and 1903, always
at considerable financial loss, and its publication was then sus-
pended for a while in spite of Whitman's efforts to secure the
needed support. The American Journal of Anatomy and The
Journal of Experimental Zoology, begun in the period of su"spen-
sion of the Journal of Morphology, did not, however, suffice
for the growing needs of zoological and anatomical science, and
the Journal of Morphology was taken up again by The Wistar
Institute of Anatomy and Biology in Philadelphia, in 1908, and
its publication has continued ever since. As Professor Mall
says, "The Journal of Morphology served as a model for many of
our scientific journals, both biological and medical, which have
come into existence during the past twenty-one years. The im-
portance of sound scientific journals to anatomical and zoological
science is now clear to all, and both anatomists and zoologists

BIOGRAPHICAL SKETCH XXVll

owe to Professor Whitman a debt of gratitude for having been the
pioneer in this field" (Anatomical Record, vol. 2, 1908, p. 381).

In 1898, realizing the need of some means for more rapid publi-
cation than was afforded by the Journal of Morphology, Whitman
started the Zoological Bulletin with the cooperation of W. M.
Wheeler. The idea was to afford means for the rapid publication
of shorter articles and preliminary notices dealing with investi-
gations in zoology which required only simple illustrations. The
Bulletin was therefore published monthly. It was intended to
be a companion serial to the Journal of Morphology. After the
publication of two volumes the name was changed to the Bio-
logical Bulletin and it was transferred to the Marine Biological
Laboratory as its official publication.

At the Lake Laboratory Whitman was associated with Edward
Phelps Allis, the founder, Howard Ayers, William Patten, A. C.
Eycleshymer, and some others. The work of the laboratory was
research work in morphology, especially embryology. Whitman
himself began investigations on Amia and Necturus, but though
he carried some of this work quite far, but little of it was ever
published. His scientific activity during this time may be in-
ferred from the hst of publications covering the period 1886 to
1889.

AT CLARK UNIVERSITY: 1889-1892

In 1889 Whitman accepted a call to the chair of zoology in the
newly founded Clark University of Worcester, Massachusetts.
Professor G. Stanley Hall of Johns Hopkins University had sought
to establish with the aid of Jonas Clark of Worcester, a strictly grad-
uate and research institution, which should accomplish all that the
Johns Hopkins University had set out to do in elevating the stand-
ard of scholarship in America, but without the hindrance of under-
graduate instruction. Whitman met there with thoroughly con-
genial conditions and associates. President Hall had assembled
a small but remarkable group of scientific men, all animated by
the same high ideals of scholarship. They were unencumbered
with undergraduate instruction, provided with fairly adequate
means for research, and they seemed destined to realize the fine
aim that President Hall had set before them.

XXVlll CHARLES OTIS WHITMAN

Whitman's teaching career, interrupted since he left Tokyo
eight years before, was now resumed, and continued to the time
of his death. A small body of research students was attracted
to him, who carried on their work in Worcester during the aca-
demic year and at the Marine Biological Laboratory in Woods
Hole during the summer. Whitman's laboratory was a paradise
to the properly qualified research worker. There was practically
no set instruction and the student's liberty was complete in all
respects, but a spirit of hard work and complete absorption in
the fundamental problems of biology prevailed. The problems
of biology were the true topics of the day and, when the zoologi-
cal club met, such subjects as Darwinism and Lamarckism were
discussed with a fire and enthusiasm comparable to the most
intense political or rehgious controversies. The main business
of each student was his research problem, a secondary business
was the preparation of some subject set for presentation at the
zoological club, and the animated discussion of fundamental
problems of biology prevented too much narrowness. Students
read much and thought much because they had both time and
inclination, and were not subject to trivial academic demands.

Whitman had a great respect for the intellectual independence
of his students. He set them worthy problems but left the work-
ing out to the student ; he was at the same time their severest and
most friendly critic. He maintained their courage through
difficulties, rejoiced with them in their discoveries, and always
acknowledged their complete ownership in their results. He
required convincing proof of each statement, and one could feel
sure that whatever passed him would stand. He was completely
loyal to them in all relations, and it is characteristic that the main
event which finally induced him to resign and move to Chicago
was an act of the administration which he regarded as an injustice
to one of his students. He was not alone in his displeasure with
the administration, though the causes were various and the de-
partments of physics and chemistry, zoology, anatomy, neurology,
and palaeontology of the new University of Chicago were organ-
ized by seceders from Clark University in 1892.

BIOGRAPHICAL SKETCH XXIX

PROFESSOR WHITMAN AND THE MARINE BIOLOGICAL LABORATORY

The organization of the Marine Biological Laboratory was a
response to the same demand that established and maintained
a marine laboratory on the island of Penikese in 1873 and 1874.
In his address at the opening of the Marine Biological Laboratory
Professor Whitman said:

The Annisquam Laboratory, the immediate predecessor of this, was
organized to serve the same ends as the Penikese School, and the forces
there engaged have simply been supplemented and transferred to the
new Marine Biological Laboratory of Woods Hole, with such changes
only as circumstances have rendered necessary. It was through the
generous support and active cooperation of the Woman's Education
Association of, Boston that Professor Hyatt was able to maintain the
Laboratory at Annisquam, and the same Association initiated and car-
ried through the movement that has given us this Laboratory.

In 1886 efforts were made by the Association to place the Annis-
quam Laboratory on an independent and broader foundation.
A circular letter sent to many of the leading biologists of the
country received encouraging repHes and accordingly a prelim-
inary meeting was held on March 5, 1887, in the library of the
Boston Society of Natural History. A committee was there
organized to perfect plans for the organization of a permanent
sea-side laboratory, to elect trustees and to devise ways and means
for collecting the necessary funds. The committee met with
sufficient success for a modest beginning and accordingly in March,
1888, the Marine Biological Laboratory was formally incorpor-
ated with ten members. Seven trustees were chosen at a meeting
of the Corporation held the same month. In June, 1888, the
Trustees issued a circular in which they announced the poUcy
of the Laboratory to support instruction as well as research,
and invited the cooperation of the universities and colleges of
the country. Professor Whitman's appointment as director of
the Laboratory was also announced in this circular.

This brief account of some facts in the early history of the
Marine Biological Laboratory may suffice to show the origin of
Professor Whitman's connection with the institution. He found
a local organization that planned to become national in scope,

XXX CHARLES OTIS WHITMAN

to enlist the cooperation of colleges and universities throughout
the country and to provide for research and instruction in biology.
The location of the Laboratory was also fixed and the first build-
ing erected at Woods Hole. Although the incorporators were all
residents of Boston, yet they had provided for a national organ-
ization by offering each institution invited to cooperate the privi-
lege of naming five members each of the Corporation during the
term of cooperation. Apparently, Professor Whitman had noth-
ing to do with the original statement of these principles, but
after his appointment as director, at least, he became their chief
exponent and developed them to a much greater extent than the
original incorporators had intended, so that the Corporation soon
came to have a large and nation-wide membership, and the Board
of Trustees was enlarged to include 12 members in 1890, 17 in
1892, and 21 in 1895. The membership of the Corporation grew
by leaps and bounds, and rapidly became representative of the
entire country, as the practice was followed for some years of
inviting all who worked at the Laboratory to become members.
The attendance at the Laboratory was 17 in 1888, 44 in 1889,
47 in 1890, 71 in 1891, 110 in 1892, 199 in 1895; and the number
of institutions represented was 13 in 1888, 29 in 1889, 32 in 1890,
31 in 1891, 52 in 1892 and 85 in 1895.

The early years of the Laboratory were years of grfeat pros-
perity; to accommodate the growing tide of workers an L was
added to the original building in 1890; in 1892 a building equal to
the original Laboratory in size was added to form the third side
of a quadrangle, and two separate buildings, one for botany and
another for a lecture hall and research rooms were added by 1896.

Whitman's part during this period of rapid material develop-
ment was to furnish the spirit and develop the ideals of the
institution. It is obvious that the idea of cooperation had a pri-
mary practical significance in the minds of the original trustees,
to secure support for the new institution. Though he did not
lose sight of its practical significance, the idea of cooperation was
transformed by Whitman into an ideal of a scientific democracy,
which furnished a motive for loyalt}^ and devotion such as rarely,
if ever, existed in a scientific enterprise, so that the development

BIOGRAPHICAL SKETCH XXXI

of the Laboratory became a kind of cult to a large and influential
body of naturalists. Whitman not only awakened this spirit,
which was compounded of devotion to himself as well as to the
ideal which he represented, but he kept it alive, and more than
once, by refusing to compromise any fraction of the fundamental
idea for immediate practical advantage, he saved the principle
from extinction. That the Laboratory today is still a scientific
democracy is due entirely to Whitman's uncompromising devo-
tion.

In his first report Professor Whitman states, ''The new Lab-
oratory at Woods Holl is nothing more, and, I trust, nothing
less, than a first step towards the establishment of an ideal bio-
logical station, organized on a basis broad enough to represent
all important features of the several types of laboratories hitherto
known in Europe and America." Thus he formed great plans for
the germinal institution. He early maintained that in such an
ideal biological station it was essential that all biological interests
should be represented, and accordingly successively added depart-
ments of botany, physiology and embryology to the original
zoology, each with its side of research as well as instruction. But
the variety of work that has been welcomed at Woods Hole can-
not be included even within these broad divisions. Professor
Whitman had most catholic interests in biology and it is remark-
able in what fundamental ways he comprehended the problems
of each division. The association of workers in different fields
of biology has been one of the most helpful and stimulating fea-
tures of the Station.

The Marine Biological Laboratory was designed for instruction
as well as research. The original circular opens with these words:
"The Trustees of the Marine Biological Laboratory earnestly
desire to enlist your cooperation in the support of a sea-side lab-
oratory for instruction and investigation in biology." Instruc-
tion was in fact placed first, not only in the opening sentence but
throughout the circular. However, the Laboratory started out
at once under Whitman as primarily a research institution, and
in his address at the opening of the Laboratory, July 17, 1888,
he said:

XXXll CHARLES OTIS WHITMAN

111 every attempt hitherto made to combme the two chief interests
here represented, instruction has been the object of first concern. Now
the only way to keep the distributive function efficient and active is to
unite it in proper relations with the productive function. The Labor-
atory (i.e., the side of investigation) is the creative agent — -the source of
all supplies; the school is merely the receiver and distributor. Any
attempt to combine the two which ignores or reverses these relations
must end in disappointment and failure.

In the fifth annual report Professor Whitman states:

The two functions of instruction and investigation have worked ad-
mirably together, each growing stronger in the success of the other.
We have endeavored to keep the two properly balanced, but I think
we have nearly reached the limit of our capacity for instruction with
our present space and means. We already see that to tax our teaching
force much more would not tend to improve the side of investigation.

In the eighth annual report for the year 1895 Professor Whitman
again returns to this theme:

Our instruction and investigation have been inspired by a common
purpose, and thus kept in such relations that each has added to the
strength of the other, and added more and more with every stride for-
ward. If instruction has increased, it is chiefly due to the stimulating
influence of investigation; if investigation has gained, it is because in-
struction has multiplied workers. Mutual service is the bond of union,
but the union is not merely one of coordination, in which the two ele-
ments are simply balanced one against the other; it is one of a more vital
order, in which each is servant and only one is master. All our classes face
in one direction — towards original work — and all our activities, sympa-
thies and interests are dominated by the spirit of research. Does that
render our instruction less efficient? Just the contrary. It fills with
life and purpose, makes students more earnest, dignifies the work of the
teachers, and wins their best effort. Moreover, it re-enforces the service
of the regular staff by contributions from every member of the investi-
gating departments.

Farther on:

What does instruction mean for us? It means, not wholly, but pre-
eminently, preparation for original work, and much of it is especially
designed for the benefit of investigators, not beginners only, but for
specialists who are independent workers.

It will be plain,^ I trust, that we are not cultivating two antagonistic
functions, between which we have to carefully guard the balance, lest
one may prosper at the expense of the other. There can be no excess in
either direction, for every gain, whether on one side or the other, is
a gain not only for the part but also for the whole.

BIOGRAPHICAL SKETCH XXXlll

These extracts explain Whitman's position with reference to
the functions of instruction in a primarily research institution.
His ideas seem to have been sound, if we may judge from the
experience of twenty-three years, during which the two have
existed side by side with mutual advantage.

During the third session of the Laboratory Whitman organized
the evening course of Biological Lectures which has proved ever
since one of the stimulating features of the Laboratory life. In
his report for this session Whitman outlines the idea as follows :

These were not intended to take the place of systematic lectures, such
as are given in the regular courses of instruction; they stand rather for
the higher and the more general needs of the science. Their leading
purpose, if I may be permitted to define it more with reference to the
possibilities of its future development than to its present attainment,
was to meet the rapidly growing need of cooperative union among special-
ists. Specialization has now reached a point where such union appears
to be an essential means of progress. Specialization is not science, but
merely the method of science. For the sake of greater concentration
of effort, we divide the labor; but this division of labor leads to inter-
dependence among the laborers, and makes social coordination more and
more essential. This is the law of progress throughout the social as
well as the organic world. An organism travels towards its most per-
fect state in proportion as its component cell-individuals reach the limit
of specialization, and form a whole of mutually dependent parts. Sci-
entific organization obeys the same law. As methods of investigation
improve, specialization advances, and at the same time the mutual de-
pendence of specialists increases. Isolation in work becomes more and
more unendurable. Comparison of results, interchange of views and
ideas, and a thousand other advantages of social contact, become of
paramount importance to the highest development.

In such considerations may be found the leading motive for this
course of lectures. While directed in the main to the higher needs of
investigators, they deal, as a rule, with subjects of present and quite
general interest to beginners. In general, it may be said that the
authors undertake to set forth what has been accomplished in their
special fields of research, to give the conclusions of the best work and
thought, to point out general bearings, and to state the problems that
await solution.

The educational value which such lectures may be presumed to have,
and the consideration that through them the aims, the needs, and the
possibilities of biological work might, in some measure, be made better
known to the public, especially to those whose liberal benefactions have
enabled the Laboratory to carry forward its work, suggested the propri-
ety of publication.

XXXIV CHARLES OTIS WHITMAN

At various times these lectures, which have sometimes taken
on a spirit of some formahty, have been supplemented by informal
discussions following lectures dehvered by investigators before
classes, especially the class in embryology during the early years,
and later in physiology; at other times research seminars have
been formed for the distinct purpose of discussing and criticising
work presented by the investigators; and at all times in the his-
tory of the Laboratory free and informal discussion between
investigators of their work in progress has been a characteristic
feature in the laboratory life. In all this the steady and sane
influence of Whitman was at work. All coveted discussions with
Whitman; he had a most sympathetic interest in all work going
on in the Laboratory, and deep insight into the fundamental
problems. One frequently discovered after unburdening one's
self in response to his sympathetic attitude that he had thought
out the problem in question more thoroughly. But his courteous
and honest attitude always saved such a situation from being
painful. He exercised in these ways a steadying influence on the
investigations of others, for he was never hurried into following
a mere fashion in research.

The social life of the Laboratory in Whitman's time was simple
and sincere. He had a horror of all formality and met everybody
on a plain and equal footing. His hospitality usually took the
form of small dimiers particularly well cooked and served, with
not more than half a dozen guests usually. He was a most charm-
ing host, gracious and self-effacing. The conversation usually
turned on some scientific subject and he had the knack of making
the others talk, and it was considered quite a triumph for the
others to draw him out. He sustained relations with his stu-
dents both at Woods Hole and elsewhere, that can only be de-
scribed as fatherly. He often helped thjem financially, and stood
by them with the greatest loyalty in securing positions. To the
respect that all his students felt for his scholarship and ability
was added the love and devotion that they owed to the best of
friends.

No account of Whitman's relations to the Marine Biological
Laboratory would be complete which failed to describe his con-

BIOGRAPHICAL SKETCH XXXV

duct in various crises of the history of the institution. The essen-
tial character of the man comes out better probably in its mingled
elements than in any other known relations. But this account
must necessarily be incomplete and partial to the extent that
Whitman is the subject, and not the Laboratory. Up to about
1895 the relations of the Director and trustees seem to have been
on the whole cordial, in spite of minor difficulties. But the
rapid growth of the Laboratory imposed financial burdens of no
shght amount. In 1890 an 'L' was added to the original build-
ing; in 1890 a new wing was built; in 1893-4 a new dining hall and
kitchen were erected, and the present botanical laboratory.
The expenses of these additions was met by numerous contribu-
tions from friends and by a loan of $3,500 secured by a mortgage
upon the property of the Laboratory, and an unsecured loan of
$3,000 from one of the trustees.

The Boston trustees themselves felt great satisfaction in the
rapid growth of the Laboratory. In 1894 they could say: ''The
only serious perplexities of the last year have been the result of
its rapid growth and prosperity;" the Laboratory had in fact
become self supporting so far as current expenses were concerned.
It was important, however, to meet the outstanding loans for
new buildings and the following appeal was issued:

Reluctant as the trustees were to incur expenses which would make it
necessary, in this time of financial stress, to ask help from the friends of
the Laboratory, yet, in the opinion of many, to have checked the growth
of the institution at this stage, by turning away desirable students and
investigators, would have inflicted a permanent injury. We ask, then,
from those whose conviction of the value of such a Laboratory has
helped to bring it to its present condition of prosperity, still further aid
in its future development (from the Trustees' Report to the Corporation
for the year 1894).

But the enlargements, great as they had been, were still in-
adequate to the growing demand. In proposing the further
enlargement which Professor Whitman felt to be necessary to
provide for the growth of the Laboratory, he was hampered by
the reluctance of some, at least, of the trustees, to incur further
indebtedness. A new building was needed of the size of the orig-
inal laboratory to provide a lecture hall and more rooms for inves-

XXXVl CHARLES OTIS WHITMAN

tigators at an estimated cost of $3,000. Professor Whitman
organized the investigators of the Laboratory into a Biological
Association to work for the needed building. This Association
pledged SI, 500 towards the cost of the new building, and the trus-
tees finally agreed to secure an equal sum. The building was erected
in 1896, and has been fully occupied ever since, thus justifying
Whitman's estimate of the needs of the Laboratory. But this
plan left the debt for previous buildings still outstanding. '^ Sub-
sequent events showed that Doctor Whitman raised the whole of
the $3,000, besides the money needed for equipment, and the trus-
tees did not as a body raise anything; although a few individuals
who were supporters of Doctor Whitman and his policy raised a
few hundred dollars" (from ''A Reply to the Statement of the
Former Trustees of the Marine Biological Laboratory," 1897,
p. 8).

While it is perhaps undesirable to revive old controversies, yet
it seems needful in justice to Doctor Whitman, to state the issues
of the years 1896-1897, with the dispassionateness which fourteen
elapsed years should furnish. It was never true that a majority
of the board of trustees lost confidence in, or were out of sympathy
with Doctor Whitman ; but a minority of the board, who neverthe-
less constituted the governing element by virtue of their original
membership and residence in Boston where all the meetings were
held, were much displeased with him for not listening respect-
fully enough to their motives of caution, and for his dominance
in Laboratory affairs. The existence of a. small deficit in the
operating expenses of the year 1896 led them to declare that the
Laboratory should not be opened in 1897, unless a sum of $2,000
were raised not later than April 15. This sum was much in excess
of the deficit and the vote was not taken until February 5, 1897.
An offer on the part of one of the trustees, Mr. L. L. Nunn, to
bear any added deficit resulting from operations of 1897, was re-
fused. The trustees raised the sum of $1,140 by April 12, and
the treasurer reported on May 5 that there was a balance in the
treasury of $735.55; there was also about $670 accumulated
interest in funds available for any purpose the trustees might
approve. The deficit in the meantime had melted away. The

October 10, 190S. Photograph by R. AI. Strong

1908. Photograph by KenjiToda

Charles Otis WTiitman

BIOGRAPHICAL SKETCH XXXVll

announcement of the 1897 session was therefore very late and the
attendance suffered seriously in consequence of the inimor that
had spread that the Laboratory would not be opened that year.

A meeting of the board of trustees was held at Woods Hole
on August 6, 1897, and at this meeting a majority of the members
present, who were favorable to Whitman, voted to call a special
meeting of the members of the Corporation to be held in Boston
on August 16 for the purpose of considering changes in the by-
laws. The purpose of the proposed changes was (1) to provide
that the annual meeting of the Corporation should be held in
Woods Hole instead of in Boston, and in August instead of Nov-
ember, and to increase the quorum so as to secure a more repre-
sentative attendance and avoid local control, and (2) to change
the body of the trustees from a body practically self-perpetuating
to an elective body, elected by the Corporation in four groups, one
such group to be elected each year for a period of four years, and
thus avoid the old practice of the simultaneous annual election
of all members.

At this meeting about eighty-seven members of the Corporation
recorded their names with the clerk, and it was estimated that
there were about twenty others present who did not do so. It
was the largest and most representative meeting of the Corpora-
tion ever held up to that time. The program as outlined was
unanimously adopted.

This amounted to no less than a revolution in the government
of the Laboratory, and the action was promptly followed by the
resignation of seven out of the nine members of the board of
trustees resident in Boston and its vicinity. Six of these and one
other trustee then drew up a statement which was primarily an
attack on the Director, Professor Whitman, which they published
in Science, October 8, 1897. To this statement a complete reply
was made in the more dignified, but less permanent, form of a
separate pamphlet by a committee of three of the trustees who
stood by the Director (''A Reply to the Statement of the Former
Trustees of the Marine Biological Laboratory," Boston, Alfred
Mudge and Son, Printers, No. 24 Franklin Street, 1897.) This
reply and the facts that two-thirds of the board of trustees stood

XXXVlll CHARLES OTIS WHITMAN

by Professor Whitman, that the places of the 'former trustees'
were taken by well-known naturahsts, and that the progress of
the Laboratory was not seriously interrupted even by so serious
a controversy, constitute a sufficient vindication for Whitman.

This struggle was unfortunately necessary to estabUsh the
national, representative and democratic character of the institu-
tion, a character that grows with the years and which commands
the loyalty and devotion of the present members, both of the
Corporation and of the board of trustees.

Once again it was necessary for Whitman to take a firm stand
to maintain the fundamental ideals of organization of the Labor-
atory. This was when the newly organized Carnegie Institution
of Washington offered in 1902 to take the Laboratory as a depart-
ment. This would have permanently solved the difficult prob-
lem of maintenance, but Whitman was convinced that it would
destroy the representative democratic character of the institu-
tion, although every possible concession to the existing form of
organization was generously offered by the Carnegie Institution.
In this opinion he stood nearly alone, but none the less firmly,
and it was his insistence that finally brought about a delay of the
decision with an annual grant of $10,000 a year for a period of
three years (1903-1905) from the Carnegie Institution in the
form of a subscription to twenty work rooms. At the end of
this period a very notable petition signed not only by all members
of the laboratory, but also by a large number of representative
naturahsts, for the continuation of the temporary arrangement
was not granted by the trustees of the Carnegie Institution, and
the original proposal lapsed. The independence of the Labora-
tory had been maintained, but it was apparently as far from a
stable basis of financial support as ever.

Following this. Whitman gradually withdrew from active par-
ticipation in the management of the Laboratory, although he
retained the title of Dhector until 1908. However, he no longer
attended meetings, and was even absent from the Laboratory
for two successive seasons, 1904 and 1905. The house which
he had occupied at Woods Hole burned down in the winter of 1905-
1906; and, fearing that this would make his return impossible

BIOGRAPHICAL SKETCH XXXIX

his friends raised a sum of S3, 000 by subscription and the
property was bought and the house restored and presented to
Whitman. This very signal mark of love and appreciation on the
part of his friends, indicating as it did so clearly their desire to
remove every obstacle that prevented his presence among them,
touched Whitman most deeply. He was present again at the
Laboratory in the sessions of 1906 and 1907, but never again,
except for a brief visit of two or three days in 1909.

His gradually increasing engrossment in the study of heredity
and evolution in pigeons may be assigned as the principal cause
of his withdrawal from residence at the Laboratory. For many
years he transferred his large collection of birds from Chicago to
Woods Hole and back again each summer. He always suffered
some losses of valuable birds, even when the railroad companies
allowed him to take his birds as excess baggage and to attend to
them en route. However, when this permission was refused
and they had to come by express and might be delayed over an
extra night, the losses became more serious. Indeed, the trans-
fer became an intolerable burden, and he relinquished his charge
of affairs at Woods Hole rather than curtail his own research,
an eminently characteristic choice.

In 1908 he tendered his resignation; his letter and the reply
thereto follow:

To the Trustees of the Marine Biological Laboratory, Woods Hole,

Mass.

Gentlemen :

This year has brought the twenty-first birthday of the Marine Bio-
logical Laboratory. For these many years you have continued to honor
me with the directorship of the Laboratory. In late years I have so
far drifted out of office and out of use that a formal resignation at this
time can scarcely be more than an announcement of the fact accom-
plished. The time has arrived, however, when a reorganization seems
to be imperatively demanded, and as a prelude thereto, I must ask you
to accept this note as a somewhat belated announcement of my resig-
nation of the office of director.

Let me take this opportunity to thank you one and all very heartily
for the cordial support you have extended to me.

Respectfully,

CO. Whitman.

Xl CHARLES OTIS WHITMAN

August 13, 1908.

The Corporation and Trustees of the Marine Biological Laboratory,
in accepting the resignation of the Director, Professor CO. Whitman,
have ordered to be put upon their records and to be forwarded to Doctor
Whitman the following minute :

The Corporation and Trustees desire to express to the retiring Direc-
tor their regret that he finds it necessary to withdraw from the active
directorship of the laboratory, and their appreciation of the inestimable
value of his services. Since the establishment of the Laboratory at
Woods Hole twenty-one years ago, he has been continually its Direc-
tor and he has to a very large extent guided its growth and development.
He has stood for the principles of cooperation and independence which
have made the laboratory unique in character and truly national in its
reputation and influence. His high ideals and his generous apprecia-
tion of the work of others have been an inspiration to the many biolo-
gists, who, during these years, have attended the laboratory.

The corporation and trustees desire that the retiring Director niay
continue to serve the laboratory as honorary director and trustee and
that his presence at the laboratory may continue to be an inspiration
in the future as in the past.

Professor Whitman's reply was as follows:

To the Corporation and Trustees of the Marine Biological Laboratory,
Woods Hole, Mass.

Ladies and Gentlemen: Your action of August 13, in which you
express a desire to have me serve the laboratory as 'honorary director
and trustee' is in itself alone an all-sufficient reward for whatever services
I have rendered as Director. Your goodwill is the all-important
recompense, and no title that you could confer could add to the weight
of your approbation. In fact, titles belittle the spirit. Let me have the
latter without the former — -without title or office of any kind. Please
respect this wish and believe me, as ever, a sincere and devoted friend
of the Laboratory.

Respectfully and cordially,

C. O. Whitman.

The report of the trustees to the Corporation bearing on Pro-
fessor Whitman's resignation and on his services to the Laboratory
expresses so well what many others feel that it is appropriate to
quote it in large part:

Professor Whitman's resignation as Director of the Marine Biological
Laboratory, after twenty-one years of service in that' position, impres-
sively recalls the inestimable value of his services in the establishment
and development of this institution. If we have today one of the lead-

BIOGRAPHICAL SKETCH xll

ing marine laboratories of the world, we owe it in large part to him. The
interest of almost every member of this board of trustees and of the
corporation was enlisted through his efforts, and the splendid influence
which the Marine Biological Laboratory has had upon the development
of biology in this country is traceable ultimately to him.

His connection with the laboratory began at a time when it had neither
permanent home, recognized standing, nor scientific ideals. Some of the
leading biologists of this country felt that it could not compete as a
research station with the U. S. Fish Commission Station, backed as
the latter was by the resources of the government, and that its chief
field of usefulness must be as a summer school. Whitman thought
otherwise, and by his real greatness as a scientist, his untiring energy
and enthusiasm, his splendid ideals and his unfailing faith and courage
he made it from the start the principal center in America for biological
research.

From start to finish his ideals for the laboratory were these: (1) A
national center for research in every department of biology; (2) a lab-
oratory founded upon the cooperation of individuals and institutions;
(3) an organization independent in its government and free to follow its
natural course of growth and development. For these ideals he has
labored consistently and persistently year after year, sometimes with a
disregard of present advantage, to be gained by the sacrifice of one or the
other of these ideals, which cost him friendships which he highly prized.
At one particular crisis he wrote : ' If I have made any enemies through
unkindness or injustice, I am sincerely sorry for it ; but if I have made any
because I have stated my conviction on the question before us I can
afford to part with all friends who are made enemies for such a cause.'
His faith in the ultimate achievement of these ideals was so great that
he chose rather to sacrifice present good than, as he believed, the future
welfare of the laboratory; and his plans for the laboratory were so great,
while current resources were so small, that he was frequently charged
with being impractical. But it is only fair and just to recognize how
much was accomplished by adherence to these ideals and to what an
extent the spirit and success of the laboratory are due to them.

Woods Hole is indeed a national center for research in several branches,
if not in every department, of biology. Whitman had the wisdom to
see that biology could progress only as a whole. 'The great charm of a
biological station,' he wrote, 'must be the fullness with which it repre-
sents the biological system. Its power and efficiency diminish in geo-
metrical ratio with every source of light excluded.' To zoology, which
was the only subject represented at first, he added botany and physiology
and he strove to make Woods Hole a center in each of these departments.
He was one of the first to insist upon adequate provision for experimen-
tal work. He was, we believe, the first in this country to plan and plead
for a biological farm for the study of problems of heredity and evolu-
tion. He desired to make Woods Hole a center for the comparative
study of anatomy, pathology and psychology. Some of these lines of
work have since been taken up and largely developed elsewhere, but if

Xlii CHARLES OTIS WHITMAN

Whitman could have had the necessary .support in his plans they would
have been centered at Woods Hole. This need of a national center of
research in every department of biology is still before the laboratory
as a living issue, and although this grand concept has so far failed of
complete realization, who can say how much the laboratory owes to
this catholicity of spirit of its director, how much biology as a whole
owes to this splendid ideal?

If the laboratory was to be truly national, Professor Whitman be-
lieved that it must be founded upon the cooperation of individuals and
institutions; no one man nor institution, however great, could accom-
plish this purpose. He recognized that common ideals must form the
basis of such cooperation, and he sought to bring into close connection
with the laboratory every person and every institution that shared these
ideals with himself. With these ideals, and by means of his o^vn per-
sonal charm and scientific abilities. Whitman secured the cooperation
of many of the younger biologists of the country. There was thus de-
veloped at Woods Hole a center for research work in biology which has
had few equals in the history of the world. By his own work, as well
as by his appreciation of the really fundamental problems of biology, he
has set a very high standard for the scientific work of the laboratory,
and by his kindness, sincerity, and generosity he has called forth similar
qualities in others, so that it has been characteristic of Woods Hole,
as of few other laboratories at home or abroad, that a spirit of genuine
cooperation and mutual helpfulness prevails. Who that experienced
it can ever forget the inspiration and enthusiasm of those early years of
the laboratory? Who of us can forget the cordial appreciation and
generous encouragement which we received from Professor Whitman?
Some of us feel that we there incurred a debt of gratitude to him which
we can never fully repa3^ Since those early years other laboratories have
arisen and other duties have drawn men away from Woods Hole, but
the Marine Biological Laboratory never loses its charm for those who
have worked there, and this charm will continue as long as the spirit
of cooperation, which Whitman instilled into it, prevails.

Finally, Professor Whitman stood for the complete autonomy of the
laboratory. Although aid might have been had more than once from
universities and institutions by surrendering the independence of the
laboratory, he steadfastly and consistently refused to do this, even
though in doing so he had to face the opposition of almost all the mem-
bers of the board of trustees and the corporation. There is still a dif-
ference of opinion as to the expediency of this stand, but there is prob-
ably no question as to the desirability of the autonomy. If the labor-
atory can obtain endowments such as to provide for its present and
future needs and to insure its independence we shall all greatly rejoice,
but whether it shall succeed in this aim or not, we are probably all
agreed that this much at least of Professor Whitman's ideal must be
maintained, viz : that the laboratory must be left free to grow and develop
as its own needs and the interests of science demand.

BIOGRAPHICAL SKETCH xliii

These are the ideals which Professor Whitman succeeded in making
part and parcel of the Marine Biological Laboratory and which we count
among our most valuable possessions. To those who measure the suc-
cess of an institution by the size of its buildings or endowments, his
efforts at Woods Hole may seem in large part to have failed, but those
who realize that ideals are the motive forces of the world, that life con-
sists not in abundance of possessions but in abundance of service, that
science is not paraphernalia but knowledge — ^these will not fail to recog-
nize the great value of the work Professor Whitman has done for the
Marine Biological Laboratory and for the whole science of biology.

To these words of appreciation there is but little to add. It
may be that the Marine Biological Laboratory is Whitman's
most enduring monument, as it was his chief work of organiza-
tion. But the principles will endure eternally, whatever the life
of the particular expression they have been given in the Labor-
atory, and the fact that Whitman was the chief champion of these
ideals and that he gave them visible and effective expression is
one of his chief claims to affectionate and reverent remembrance.

THE AMERICAN MORPHOLOGICAL SOCIETY

Professor Whitman was the leader in the three most important
organizations for the advance of zoology in America during the
time of his active life: in 1887 he founded the Journal of Mor-
phology, in 1888 he became director of the Marine Biological
Laboratory, and in 1890 he took the leading part in the founda-
tion of the American Morphological Society. A circular was
sent out October 16, 1890, calling on those interested to unite
in the formation of an Association of Morphologists ''in connec-
tion and affiliation with the American Society of Naturahsts,"
which shall hold stated meetings during the Christmas vacation,
at which special and general morphological problems may be
brought forward and discussed. Attention was directed in the
circular to the scientific isolation of zoologists in America, and
the advantages of their cooperation in such a society. The com-
mittee signing this call consisted of C. 0. Whitman (chairman),
Henry F. Osborn, E. B. Wilson, E. G. Gardiner, and J. Playfair
McMurrich. The first meeting was held in Boston, December
29, 1890. Dr. E. B. Wilson was elected chairman for the meet-

Xliv CHARLES OTIS WHITMAN

ing. Whitman was then elected president for the next meeting,
and was re-elected during, the following three meetings. In
1902 the name of the society was changed to ''The American
Society of Zoologists" and it is still our dominant zoological
society.

AT THE UNIVERSITY OF CHICAGO

We have departed from the chronological order of events in
thus sketching Whitman's various activities. In 1892 Whitman
moved from Clark University to the University of Chicago,
taking with him the major part of his department and all his
students. Professors Mall, Donaldson and Baur also came at
the same time from Clark University to the University of Chicago
and took part with others in the formation of a department of
biology of which Whitman was head. After the first year,
however, the department was broken up into separate depart-
ments of zoology, anatomy, neurology, physiology and palaeon-
tology. Concerning this event Professor Mall says (The Resig-
nation of Professor Whitman as Director of the Marine Biological
Laboratory at Woods Hole, Mass., Anat. Record, vol. 2, no. 8,
November, 1908).

When the University of Chicago was founded in 1893, Professor Whit-
man was made head of the biological department, which in its organiza-
tion was unusually strong on the anatomical side. It was planned at
the beginning to divide the department as soon as circumstances would
warrant, and with the very rapid growth of the University this took
place within a year. Then the anatomical department was established
coordinate with those of zoology and botany. This proved to be the
most important step in the organization of anatomical departments in
America, and for it we are largely indebted to Professor Whitman.

Whitman was in fact mainly responsible for the unusually
comprehensive organization of the biological departments in the
University of Chicago, and for their establishment in a single
group of buildings, thus rendering possible a degree of mutual
support and cooperation among the biological sciences, the full
possibihties of which have not been realized even in Chicago up
to this day. Whitman thus carried out in Chicago as far as
possible the same form of organization that he planned for Woods

BIOGRAPHICAL SKETCH xlv

Hole, involving the representation of every branch of biological
knowledge, so as to bring the conibined forces to bear on the
fundamental problems of biology. It was possible in Chicago to
proceed along such ideal hues, for the institution was unhampered
by history or tradition or by fixed location of departments estab-
lished in remote neighborhoods.

The department of zoology under Whitman was primarily a
research department and the members of his staff were selected
primarily for their standing as investigators. Whitman, him-
self, taught graduate students exclusively, for the most part
candidates for the doctor's degree. He lectured but once a week
and not always regularly; but each lecture was a finished essay,
and in a way, a piece of original work. He never attempted to
present what students could find in books. He consulted about
once a week with each student on his research problem, and was
a very rigorous and strict critic, but he tended more and more as
time went on to let each student work out his own salvation. It
often became necessary for the student to seek him at his house
for consultation about his work, but such a consultation was
always well worth while, as Whitman would leave his own work
and play the part of host most delightfully, as well as that of
teacher.

The details of departmental administration were very irksome
to Whitman, but on the fundamental principles of administra-
tion of the department he was firm as a rock and quite uncom-
promising. Even trifling details would often seem to him con-
trary to correct ideals, and then the matters of administration
loomed large and were rigorously decided.

Whitman's students who took the degree of Doctor of Philos-
ophy under his instruction, with their present academic standing,
were the following:

1. At Clark University:

Hermon Carey Bumpus, Business Manager, University of Wisconsin.
William Morton Wheeler, Professor of Entomology, Bussey Institution,
Harvard University.

Edwin Oakes Jordan, Professor of Bacteriology, University of Chicago.

xlvi CHARLES OTIS WHITMAN

2. At the University of Chicago :

Herbert Parlin Johnson, Associate Professor of Bacteriology, Medical De-
partment of St. Louis University.

Frank Rattray Lillie, Chairman of the Department of Zoology, and Profes-
sor of Embryology, University of Chicago ; Director, Marine Biological Laboratory.

Albert Chatjncey Eycleshymer, Professor of Anatomy, Director of Anatomi-
cal Department, St. Louis University.

William Albert Locy, Professor of Zoology, Northwestern University.

Howard Stedman Brode, Professor of Biology, Whitman College, Walla Walla,
Washington.

Cornelia Maria Clapp, Professor of Zoology, Mount Holyoke College.

Agnes Mary Claypole (Mrs. Robert O. Moody).

Albert Davis Mead, Professor of Comparative Anatomy, Brown University,
Providence, R. I.

Charles Lawrence Bristol, Professor of Zoology, New York University.

Samuel J. Holmes, Assistant Professor of Zoology, University of Wisconsin.

John P. Munson, Normal School, EUensburg, Wash.

Emily Ray Gregory, Professor of Biology, College for Women, Constantinople,
Turkey.

Aaron Louis Treadwell, Professor of Biology, Vassar College.

Michael Frederick Guyer, Professor of Zoology, University of Wisconsin.

Elliot Rowland Downing, School of Education, University of Chicago.

Wilhblmina Entemann Key, Lombard College, Galesburg, 111.

Ralph Stayner Lillie, Instructor in Comparative Physiology, University of
Pennsylvania.

Virgil Everett McCaskill, President State Normal School, Stevens Point,
Wisconsin.

John McClellan Prather, Teacher of Zoology, Central High School, St. Louis,
Mo.

Eugene Howard Harper, Instructor in Zoology, Northwestern University.

Bennett Mills Allen, Assistant Professor of Anatomy, University of Wis-
consin.

William J. Moenkhaus, Professor of Physiology, University of Indiana.

Charles Dwight Marsh, United States Department of Agriculture.

John William Scott, Instructor in Zoology, Kansas State Agricultural College,
Manhattan, Kansas.

Charles Zelbny, Associate Professor of Zoology, University of Illinois.

Lynds Jones, Professor of Ecology, Oberlin College.

Horatio Hackett Newman, Associate Professor of Zoology, University of
Chicago.

James Francis Abbott, Professor of Zoology, Washington University, St. Louis.
Mo.

Victor Ernest Shelford, Instructor in Zoology, University of Chicago.

Oscar Riddle, the University of Chicago.

Charles Henry Turner, Sumner High School, St. Louis, Mo.

Frank Eugene Lutz, Curator, American Museum of Natural History, New
York City.

BIOGRAPHICAL SKETCH xlvU

George Washington Tannreuther, Assistant in Zoology, University of Mis-
souri.

Wallace Craig, Professor of Philosophy, University of Maine.

Charles Christopher Adams, Instructor in Ecology, University of Illinois.

James Thomas Patterson, Adjunct Professor of Zoology, University of Texas.

Mary Blount, Teacher, University High School, and Assistant in the Depart-
ment of Zoology, University of Chicago.

Katashi Takahashi, Professor of Zoology, Gakushinin College, Tokyo, Japan.

Marian Lydia Shorey, Instructor in Zoology, Milwaukee-Downer College,
Milwaukee, Wisconsin.

H. L. WiEMAN, Assistant Professor of Zoology, University of Cincinnati.

George William Bartelmez, Associate in Anatomy, University of Chicago.

WHITMAN'S SCIENTIFIC WORK«

Professor Whitman's scientific work covered an unusually
broad field. He had a distinct predeUction for monographic
treatment, and even while engaged on special problems concern-
ing a particular animal or group of animals little escaped his obser-
vation or his note-book. Thus his work on leeches was in the
first place embryological, but he soon turned to anatomy and
taxonomy, and to problems of behavior, all of which are treated
in his various publications. Later, in his work on the develop-
ment of Amia and Necturus, but little of which has been pub-
lished, he also paid particular attention to problems of anatomy
and behavior. And again, his work on pigeons concerned not
only problems of heredity and evolution, but he made an exhaus-
tive study of the taxonomy of the group, and the problems of
their behavior were constantly in his mind. He also assigned stu-
dents problems on their development, and a number of papers were
published by Guyer, Harper, Blount, Patterson, and Bartelmez
in this field. Another illustration was his plan for a monographic
study of Arenicola by cooperation of a group of his students,
parts of which have been published.

His own publications fall mostly in the subjects of embryology,
comparative anatomy and taxonomy, animal behavior, and evo-

® This section has been prepared by E. G. Conklin, Albert P. Mathews, T. H.
Morgan, J. Percy Moore, and Oscar Riddle. The writer of the main body of the
biography has merely prepared the introductory paragraphs.

Xlviii CHARLES OTIS WHITMAN

lution and heredity, an extraordinary breadth of field for a modern
zoologist. But whatever subject he touched he illuminated.
He was slow to publish, not because of lack of industry or results,
but because he was determined to examine the subject to the
bottom, and to be sure of his view-point. He rarely had occa-
sion to correct any published statement, and even less rarely,
perhaps, to change in any radical way a point of view to which
he had once committed himself. Work of such classical distinc-
tion could not be very abundant. He has left a large amount of
unpublished material, especially bearing on pigeons, though there
is much that dates from an earher period. The statements that
follow concerning unpublished material are on Dr. Riddle's
authority.

Embryology

In his first scientific pubhcation, ''The Embryology of Clep-
sine," Professor Whitman at once took rank as a great zoologist.
Although this paper was his doctor's dissertation, it was unusually
mature, and showed in striking manner the qualities which char-
acterized all of his later work, viz., patience and accuracy in
observation, great power of logical analysis, and a firm hold on
large problems. Indeed so fundamental and comprehensive was
this work that almost all of Professor Whitman's later embryo-
logical work is foreshadowed in it, while it furnished the stimulus
for a large amount of work on the organization of the egg and its
cell lineage, which was done later by his associates and students
at Woods Hole.

Even at this early date (1878) his conclusions as to the organi-
zation of the egg and the formation of the embryo were fundamen-
tally the same as his later views. He observed the bilateral
symmetry of the egg of Clepsine before cleavage; he studied
carefully the cleavage of the egg and observed the formation and
subsequent history of the ectoblasts, mesoblasts, neuroblasts,
and entoblasts; he described the growth and concrescence of the
germ bands of Clepsine, and compared them with the growth and
concrescence of the germ ring of fishes. His conception of the
fundamental problem of all development is expressed in these
words: "In the fecundated egg slumbers potentially the future

BIOGRAPHICAL SKETCH xlix

embryo. While we cannot say that the embryo is predehneated
we can say that it is predetermined" — words almost precisely like
those used by him twenty years later when deahng with this subject.

In 1887 he pubhshed another paper on this general subject,
entitled "A Contribution to the History of the Germ Layers of
Clepsine" in which he extended his former observations on the
cleavage, orientation of cleavage planes, origin of teloblasts and
germ bands, origin of the mesenteron, and the origin of the ecto-
diferm and its products. And in the same year in a paper on
''Ookinesis," he came back to the phenomena of maturation and
fecundation, which he had treated in his first paper, and gave
a very suggestive and comprehensive review and analysis of these
phenomena.

Whitman's general point of view regarding the problems of
development are made particularly plain in his biological lectures
and addresses. In his address before the Zoological Congress of
the Worlds Columbian Exposition, on "The Inadequacy of the
Cell Theory of Development" he discussed these problems in
a striking and suggestive manner. At this time the work of
Wilson and others had just shown the possibihty and importance
of tracing individual blastomeres throughout the development
from the time of their appearance until they give rise to particular
portions of the embryo, the cleavage thus constituting "a visible
mosaic;" shortly before, Roux had shown that the cleavage of
the frog's egg was a ''mosaic work;" at the same time the work of
Driesch and other experimentalists was leading to a directly oppo-
site opinion. In this conflict of opinion Whitman took a strong
and independent position, basing his conclusions not merely on
comparative embryology but also upon the comparison of proto-
zoa and metazoa. He protested against the view that organiza-
tion is the product of cell formation, and insisted that "organi-
zation precedes cell formation and regulates it." He contrasted
the Cell-doctrine with what might be called the Organism-doc-
trine. He insisted that, "an organism is an organism from the
egg onward, quite independent of the number of cells present,"
that cleavage is not a process by which organization arises, but
that organization precedes cleavage. "The test of organization

1 CHARLES OTIS WHITMAN

in an egg does not lie in its mode of cleavage, but in subtile form-
ative processes. The plastic forces mould the germ-mass
regardless of the way it is cut up into cells."

At the same time he showed clear insight into the results of the
experimental embryologists. In this same address he says,
"The formation of a whole from a part .... no more
disproves the existence of a definite organization in the case of the
egg than in the case of hydra." i^nd in the preface to the volume
of Biological Lectures for 1894 he strongly contests the view that
''developmental mechanics" has explained or can explain
vital phenomena, without reference to the historical develop-
ment of the organism. As to mechanism and vitalism he says,
"There is no warrant for the assertion that hfe is something differ-
ent from, and independent of, matter and energy. That is the
mistake of vitalism. On the other hand there is no warrant in
decomposition for identifjdng dead mechanism with living
mechanism." "The ultimate mystery is beyond the reach of
both mechanism and vitaUsm; let pretensions be dropped and
approximation to truth will be closer on both sides."

The influence of Professor Whitman's work on the science of
embryology and on the scientific and philosophical problems con-
nected with development was profound. His work was careful,
critical, consistent. He reached conclusions only after most
painstaking observations and mature dehberation, and when he
had once made up his mind he was not easily moved. Others
might' be "tossed about by every wind of doctrine," but he stood
unmoved and unshaken, having confidence in his own observa-
tions and reflections and refusing to doubt his conclusions until
he himself had seen and felt equally strong evidence against
them. As a result of this he was unusually stable and consistent,
and while his later work shows that his ideas were constantly
enlarging with new evidence, yet there was little in his earlier
work which needed correction.

Professor Whitman's work on the early development of the tele-
ost egg, published with Alexander Agassiz (1884 and 1889), is a
fine example of careful discriminating embryological investiga-
tion. The cleavage especially was studied with great care, and

BIOGRAPHICAL SKETCH ll

the history of the marginal cells led to the solution of "one of the
cardinal questions in the early development of the teleostean
fishes, namely the precise origin of what His and others have called
the 'parablast.' " The authors showed that the nuclei of the peri-
blast, as they termed this layer, are derived entirely from the mar-
ginal cells of the blastodisc, and the fallacy of ''free cell-forma-
tion" and of separate origin of the ''parablast and archiblast"
thus received its quietus. Their hope that a similar result would
be reached for other meroblastic vertebrate ova has since been
reahzed.

Professor Whitman's interest in the Hirudinea, aroused and
fostered during his occupany of a table in Leuckart's laboratory
at Leipzig, continued unabated until his labors were so tragically
terminated. That this interest was no narrow one, but traversed
much of the depth as well as the length and breadth of biological
science, is much less apparent from a list of the titles of his papers,
than from a perusal of their contents. Under such unassuming
titles as ''The Leeches of Japan," or "Description of Clepsine
plana," we find treated, not merely specific and faunal details
but the much larger questions of the formulation of a new system
for standardizing specific descriptions, the morphological basis
of generic groupings and the origin, evolution, ecology, adapta-
tions, morphology, classification and paths of dissemination of the
land leeches. The latter small group of animals particularly
appealed to his naturalist's spirit, not as the disgusting pests that
they had seemed to most previous describers and tropical travelers,
but as the keen-sensed, facile winners in a competitive struggle
for Ufe that must have been of the utmost severity for animals
that have departed so widely from the habits and environment
of their ancestors.

Although Professor Whitman's contributions to the adult
morphology of the Hirudinea were not numerous, numbering
only three major papers and seven or eight preliminary sketches,
polemics and reviews, they have exerted a great and lasting influ-
ence. Taken together they form an exhibit of the catholicity of
his interests and exempUfy the soundness of his scientific ideals.

Hi CHARLES OTIS WHITMAN

Whitman was ever keenly alive to the importance of a many sided
view of animal life and was equally sympathetic to the work of
the field naturalist and student of ecology and habits, of the sys-
tematist, morphologist and physiologist in all their multiform
specialties, requiring only that their work should be thoroughly
and honestly done.

His earlier papers especially indicate a real and vital interest
in systematic zoology and the species question. To whatever
wide ranging speculations his studies eventually led him, they
received their primary impetus through his efforts to find a satis-
factory basis for the discrimination and definition of the species of
leeches. Indeed, his very first paper after the publication of
his Inaugural Dissertation on the "Embryology of Clepsine," was
on ''A new species of Branchiobdella" (B. pentadonta) which he at
that time, in common with other zoologists, regarded as a leech.

To such good purpose did he labor that the history of the sys-
tematic study of the leeches may fairly be divided into a pre-
Whitmanian and a Whitmanian period. Notwithstanding that
he often rode rough-shod over many of the most sacred traditions
of systematic zoology, and notwithstanding that his actual deter-
minations of species and genera have sometimes proven unfor-
tunate, nevertheless Professor Whitman discovered the criteria
by which evolution and specific radiation in the leeches may
best be measured and expressed, and he set a standard for specific
description that has since been the guide and model to all the best
workers in this field. Out of the chaotic condition that Whitman
found order has been wrought, largely through the use of the tools
that he devised. It cannot be denied that he imported new light
and life to the subject, nor that all later students of leeches have
felt the vivifying influence of his ideals.

The key to Whitman's successful analysis of the external mor-
phology of leeches is his discovery (first announced in two pre-
hminary papers pubhshed in 1884 and subsequently repeatedly
reverted to and expanded) of the segmental sense organs to which
he later appHed Haeckel's term 'sensillae.' These little, whitish,
translucent dots, visible on many living leeches, had been noticed
occasionally by earlier observers and Ebrard had even suggested

BIOGRAPHICAL SKETCH 1111

that they might have a respiratory function, but their real nature
and significance as sense organs had passed quite unsuspected.

Whitman showed that these sensillae are groups of tactile cells
which differ from the scattered epidermal sense cells chiefly
in their aggregation into groups arranged syimnetrlcally according
to a definite plan on one ring (now designated the neural or sen-
sory annulus) of every somite and that they are serially homolo-
gous with the eyes, which differ from them chiefly in the acquisi-
tion of visual cells (Glaskorper of Leydig) and pigment cup.

During the two years (1879-1881) which he spent in Japan
as Professor of Zoology at the University of Toklo, Professor
Whitman collected material for the description of the leeches of
that country. Under his direction most beautiful and accurate
colored drawings were made of the species of the several families,
but unfortunately the first part only, treating of the Hirudinidae
or ''ten-eyed leeches," was ever pubhshed. In this paper Whit-
man's method of analytical description, based primarily upon the
metamerlc arrangement of the sensillae and the somite composi-
tion thus made evident, was first fully applied to the family Hiru-
dinidae. In the course of somewhat elaborate descriptions of
the Japanese land leech (Haemadypsa japonica), the Japanese
medicinal leech (Hlrudo nlpponlca) and three species of Leptos-
toma (earlier named Microstoma and now known as Whitmanla)
the somites are analysed successively and compared as respects
their constituent elements. Comparisons are made with several
other European, American and Asiatic genera and the previously
unknown or only vaguely apprehended fact brought to light that
the metameres of the different genera of ten-eyed leeches are very
differently developed as regards the number and size of constitu-
ent rings, particularly toward the extremities of the body. These
observed differences in the number of rings constituting the so-
mites are correlated with differences in the mode of life of the
animals. In addition to their morphological value, these descrip-
tions and the accompanying illustrations are models of beauty and
accuracy, than which no more satisfying have been published
before or since.

liv CHARLES OTIS WHITMAN

Much attention is given to the description and illustration of
the complex color patterns and, although Whitman perceived
that in each species all patterns are variants of a single fundamen-
tal one, the structural basis underlying them in the arrangement
of the muscles and other organs was left to be discovered by his
student Arnold Graf, whose tragic end at Woods Hole, Professor
Whitman so keenly felt.

The discovery of the neurology of eyes and sensillae was fully
elaborated and the structure of these organs minutely described.
They were compared to the lateral sense organs of the Capitelli-
dae and the lateral line organs of fishes. The precise determina-
tion of the limits and composition of the somites that his method
required, led Professor Whitman to the formulation of definite
views regarding the nature and history of the metameres, namely:

1. That the neural or sensory annulus is the first of the
somite.

2. That each gangUon of the central nervous system supplies
the first three rings of one somite and the last two rings of the
immediately preceding somites or equivalent portions of somites
having less than five rings, and consequently that there is a
lack of correlation between definite neuromerism and definT
itive metamerism.

3. That the quinque-annulate or typical complete somite of
the middle region is primitive, and that there is a progressive
reduction or abbreviation of the somites toward the ends of the
body that is correlated with specialization in other respects.

It must not be understood that Whitman neglected the internal
anatomy. On the contrary he was assiduous in the dissection
and description of the several organ systems, but except in the case
of the nervous system, he added comparatively Uttle to our knowl-
edge.

In a later largely controversial article 'Some New Facts about
the Hirudinea,' written in reply to criticism. Whitman forcibly
reasserts his views and brings many new facts to their support.
In this paper his strong leaning toward the annelid theory of the
origin of vertebrates, to which reference is frequently made in
later papers, is indicated. He also extends his former opinion

BIOGRAPHICAL SKETCH Iv

of the homology of the segmental organs with the lateral line
organs to include the ear and even the eye of vertebrates, which
he believed to have had their origin in organs similar to the sen-
sillae of leeches.

Three years later ('Description of Clepsine plana,') we find
Whitman with all his enthusiasm applying the same criteria of
metamerism and the same methods of analysis to the external
morphology of the Glossiphonidae. This he designated the type
somite and derived from it both abbreviated somites having the
number of rings reduced to less than three and supplemented
somites with the number multiplied to more than three.

In support of these views he also appealed to the facts of em-
bryology and instanced many cases of ring multiplication, incip-
ient or advanced, in various genera of leeches. Each somite
of the leeches' body, to a considerable degree, undergoes an inde-
pendent individual development, the nature and extent of which
is correlated with the physiological demands to which it is sub-
jected. But never did Whitman carry out his view to its logical
consequences and see, as others have, that the uniannulate and
biannulate somites existing at the ends of the body of nearly all
leeches are steps toward the elaboration of the triannulate as
the latter is toward the quinque-annulate somite.

The relation of sensillae and eyes became clearer and the proof
of their homology was buttressed by many facts. Cases of dual
sense organs, composed partly of superficial cells bearing tactile
hairs and partly of clear visual cells situated more deeply along
the course of the optic nerve, and all gradations from typical
sensillae with which one or two visual cells and perhaps a little
pigment may be associated, to complete eyes with only a trace of
tactile cells were described, the latter leading by easy gradations
to the strictly visual organs of Hirudo. In some cases every step
in the transition was found in the successive somites of a single
leech, as in Clepsine hollensis.

Strong embryological evidence was brought to bear, especially
in a paper in the Zoologischer Jahrbiicher for 1893, as showing the
common origin of both kinds of sense cells from common prolifera-
tions of the ectoderm, also that the segmental sense organs are

Ivi CHARLES OTIS WHITMAN

more primitive than the scattered sense cells (goblet cells, labial
sense cells and Bayer's organs) and that they cannot have been
derived from aggregations of. the latter, as had been maintained
by Maier and Apathy.

It is most fitting that Whitman's last important paper relating
to the leeches should have appeared in the "Festschrift zum sieben-
zigsten Geburtstage" Leuckarts who h^d guided his early inter-
est in the group. This memoir on ''The Metamerism of Clep-
sine," is the culmination of ^^^litman's work on metamerism.
More than any previous work it is concerned with the nervous
system and as an example of complete morphological analysis
has few equals among papers dealing with invertebrates. The
elements of the central neuromeres and peripheral nerves are cor-
related one by one with such external features as annuli, sensillae
and eyes throughout the body and especially in the simpler somites
at the two extremities. The presence in these terminal segments
of every morphological element is determined and accounted for
and the conclusion reached that complete homodynamy exists
throughout. The earlier determination of twenty-six somites in
the body of all leeches anterior to the caudal sucker, and of seven
in the sucker was confirmed.

Metamerism is traced to the extreme tip of the anterior end,
where, however, there is a cephalic region in which the dorsal
halves alone of the somites are represented and in which as a
consequence, there is a delayed embryonic development. There
is no non-metameric residuum at the anterior end and, although
acknowledging that embryology furnishes some evidences of the
presence of a rudimentary apical organ and remnants of a pair of
head kidneys, Whitman denies that there is any element here
other than, or added to, the first metamere. There is no pro-
soma in the sense of Hatschek of an unpaired, non-metameric
and premetameric region opposed to the segmented region begin-
ning with the larval mouth. Whitman also contends that these
facts confirm the opinion long held by himself in common with
Leuckart and others that metamerism originated in multiplica-
tion by fission.

BIOGRAPHICAL SKETCH Ivll

Animal Behavior

Two papers were published by Dr. Whitman on Animal Be-
havior, one as a Woods Hole lecture, the other in the Monist.
There was also a short letter, commenting on an article of Pro-
fessor Lankester's on the origins of intelhgence, pubhshed in the
Chicago Tribune. Of these, the lecture entitled "Animal
Behavior" is the most important. It shows him at his best and
is one of the ablest of his papers, which is equivalent to saying
that it is one of the most admirable papers on this subject. Its
style is clear, interesting, and direct; and it may be taken as a
model by every investigator. Nowhere else is reasoning more
solid and sound, or comment more illuminative. No other of
his papers illustrates better the qualities of his genius: the selec-
tion of a fundamental problem; painstaking study; publication
only after years of observation and reflection ; skill in laying bare
the simple basis of an apparently complex group of phenomena ; a
grasp of the subject in all its bearings; and the use of the compara-
tive or phyletic method of attack.

He considers in this paper the fundamental questions of the
origin of instinct and intelligence as illustrated by a study of the
behavior of three kinds of animals upon which he worked at
different periods of his life: the leech, Necturus, and the pig-
eon.

He begins with the simplest acts of Clepsine; its deceptive
quiet when disturbed, a quiet of intense rigidity; and its rolhng
into a ball after feeding when it detaches itself from its host. He
shows that the latter instinct can never have been acquired as
a habit which has been stereotyped as an instinct, because Clep-
sine feeds but twice or three times in its life. • ''If the view here
taken be correct," he says, "the instinct of rolhng into a ball is
not a matter of deliberation at all, but merely the action of an
organization more or less nicely adjusted to special conditions
and stimuh. Intelhgence does not precede and direct, but the
indifferent organic foundation with its general activities is pri-
mary; the special behavior or instinct is built up by slowly modify-
ing the organic basis."

Iviii CHARLES OTIS WHITMAN

Similarly, the instinct of capturing food exhibited perfectly
by the youngest Necturus is innate. The pause before seizing
the bait is part of a very old primitive mechanism found is fishes
and finally developed into the ^pointing' of a dog. Its object
is to fix the aim. The timidity of the young Necturus is also
innate and not the result of painful experiences.

"We have taken a very important step in our study when we
have ascertained that behavior, which at first sight appeared to owe
its purposive character to intelhgence, cannot possibly be so
explained, but must depend largely, at least, upon the mechanism
of organization. The origin and meaning of the behavior ante-
date all individual acquisitions and form part of the problem of
the origin and history of the organization itself." "We see
at once that behavior does not stand for a simple and primary
adaptation of a pre-existing mechanism to a special need. As the
necessity for food did not arise for the first time in Necturus,
the organization adapted to securing it must be traced back to
foundations evolved long in advance of the species. The retro-
spect stretches back to the origin of the vertebrate phylum .

The point of special emphasis here is that instincts
are evolved, not improvised, and that their genealogy may be as
complex and far-reaching as the history of their organic bases."
This passage should be read by all physiologists who, of all biol-
ogists, are most given to neglecting phylogeny in their explana-
tions.

While instinct thus comes before intelligence and not after
it, as many have believed, some intelhgence was implied by the
inhibition of instinctive acts in Necturus by fear. To chnch his
argument that instincts are evolved like structures and are not
inherited habits, he turns to two instincts cited by Romanes as
clear evidence of their origin in habits; the tumbling and pouting
of pigeons. An examination shows that the rudiments of these
instincts are to be found in all species of pigeons. Again the value
of the phyletic method of study is illustrated by the 'brooding'
instinct of birds. This he shows to be the evolution of an instinct
shown even in fishes, which hover over the nest and drive away

BIOGRAPHICAL SKETCH lix

intruders. Even Clepsine sticks more firmly than usual when it is
over its eggs.

He sums up this part of the paper with a few general state-
ments: ''Instinct and structure are to be studied from the com-
mon standpoint of phyletic descent and that not the less because
we may seldom, if ever, be able to trace the whole development of
an instinct. Instincts are evolved, not involved ....
and the key to their genetic history is to be sought in their more
general rather than in their later and incidental uses." ''As the
genesis of organs takes its departure from the elementary struc-
ture of protoplasm, so does the genesis of instincts proceed from
the fundamental functions of protoplasm."

Taking up now the origin of intelligence he says, "Since instinct
supplied at least the earlier rudiments of brain and nerve, since
instinct and mind work with the same mechanisms and in the
same channels, and since instinctive action is gradually super-
seded by intelligent action, we are compelled to regard instinct
as the actual germ of mind." "We are apt to contrast the extremes
of instinct and intelligence — to emphasize the bhndness and inflex-
ibility of the one and the consciousness and freedom of the other.
It is like contrasting the extremes of Ught and dark and forgetting
all the transitional degrees of twihght." "Instinct is blind;
so is the highest human wisdom blind. The distinction is one
of degree. There is no absolute bhndness on the one side, and no
absolute wisdom on the other. Instinct is a dim sphere of light,
but its dimness and outer boundary are certainly variable ; intelli-
gence is only the same dimness improved in various degrees. ' '

To show how instinct becomes less fixed and choice appears he
cites the behavior of three species of pigeons, the passenger pigeon,
the ring dove and the common pigeon. If the egg is removed
from under the wild pigeon and placed on the side of the nest, the
bird returning to the nest remains a moment or two only and then
leaves the nest not to return. The ring dove, on the contrary,
after some time of perplexity, will return one egg into the nest
leaving the other out; the common pigeon, after a longer period
of perplexity and uneasiness, will put both eggs back into the nest.
There is here a gradual loss of precision of the instinct to leave the

Ix CHARLES OTIS WHITMAN

disturbed nest when the rigor of natural selection is relaxed. The
common dove is not so much an automaton. Choice begins to
appear. "With choice no new factor enters, but only plasticity,
so that the pigeon becomes capable of higher action and is encour-
aged and even constrained by circumstances to learn to use its
privilege of choice." ''This little freedom is the dawning grace of
a new dispensation in which education by experience comes in
as an amelioration of the law of elimination. This slight amena-
bility to natural educational influences cannot of course work
any great miracles of transformation in a pigeon's brain, but it
shows the way to the open door of a freer commerce with the
external world, through which a brain with richer instinctive
endowments might rise to higher achievement."

"Superiority in instinct endowments and concurring advantages
of environment would tend to liberate the possessors from the
severities of natural selection; and thus nature, like domestica-
tion, would furnish conditions inviting to greater freedom of
action, and with the same result, namely, that the instincts
would become more plastic and tractable. Plasticity of instinct
is not intelligence, but it is the open door through which the greater
educator, experience, comes in and works every wonder of
intelligence."

Evolution

No account of Whitman's work would be complete without
reference to his essay on "Evolution and Epigenesis," not only
because this essay reveals him in one of his most thoughtful
moods, but because the essay defines very sharply Whitman's
attitude toward one of the profound questions of the time — a
question that was then engaging the best thought and work of
all serious biologists. His keen critical sense is here shown to
advantage, his independence of thought led him to break some of
the idols of the day, and his thorough understanding of what had
been written and was being written at the time, all conspire to
make the essay a permanent contribution to our knowledge. He
succeeds as few others have done in holding the fine balance be-
tween the two extremes of thought represented by the terms pre-

BIOGRAPHICAL SKETCH Ixi

formation and epigenesis. He defines his position in the follow-
ing words: "I should perhaps say at the outset that I have no
theory of development either to announce or to defend. It is of
more importance just now to have well-defined standpoints and
clear ideas of guiding principles. The possibihty, not to say
probability, that the egg is from the beginning of its existence as
an individual cell definitely oriented has as yet received but little
attention." "The drift of opinion, as it seems to me, is neither
back to the standpoint of Harvey and WolfT nor to that of Bonnet
and Haller, but towards a new standpoint which seeks to avoid
the errors and blend the truth of the old hypotheses." Whit-
man's standpoint is summed up in the two following quotations:
''The indubitable fact on which we now build is — ^the ready-formed
living germ, with an organization cut directly from a pre-existing
parental organization of the same kind. The essential thing here
is — actual identity of germ organization and stirp organization."
And again: ''Let this organization stand for not more than our
neo-epigenesists freely concede, namely, that original constitution
of the germ, wh'ich predetermines its type of development — let
it stand for nothing more than that and obviously the standpoint
rises to an altitude scarcely dreamed of in the philosophy of Har-
vey and Wolff."

Whitman devoted much time to the study of Bonnet's theory
of evolution. If one asks why he should have thought it worth
while to give so much attention to a discredited theory of the
eighteenth century, the answer is first that he wished clearly to
point out the error of those "who imagine that they see in recent
theories of development a renaissance of Bonnet's evolution"
and, second, that "if our theories of development are carrying us
back to the standpoint reached by the evolutionists of the last
century it is a matter of more than historical interest." His
conclusion is "That the old and the new evolution are based on
antithetical conceptions which exclude each other at every point."
"The old evolution (preformation) was the greatest error that
ever obstructed the progress of our knowledge of development,
[f our examination has helped to clear the mist that obscured
important distinctions we have not labored wholly in vain."

Ixii CHARLES OTIS WHITMAN

Dicyemids

In 1882 Whitman published his paper on the Dicyemids. He
confirmed much of the earher work of van Beneden, but made out
a different relation between the nematogene and the rhombo-
gene individuals. He discovered certain facts that led him to
conclude that the germagene of van Beneden arose from fertihzed
eggs while all other individuals arose parthenogenetically. Whit-
man's conclusion in regard to the relationships of the Dicyemids
is the same as that to which Hartman has come in his recent mono-
graph.

Pigeons

Those who are familiar with "Whitman's work regard his
studies and experiments on pigeons as his greatest achievement.
He died at the very moment when he believed that he had reached
a point where he was prepared to publish the results of this
extensive and exhaustive investigation. Only on a few occasions
(see list of publication 1904-1907) has he stated in briefest out-
line a few of the principal conclusions he had reached. In a paper
read at the Universal Exposition in St. Louis in 1904; in his address
before the International Congress of Zoologists in Boston, 1907;
and in the Bulletin of the Wisconsin Natural History Society,
1907, Whitman has expressed himself clearly and forcibly on
certain fundamental questions of evolution. His address of
1907 has not been printed but the substance of that address is
found in his other writings.

The dominant feature of Professor Whitman's long and still
unpublished work on inheritance and evolution lies in its inten-
sive and extensive attack upon the nature of a specific character.
In the 90's he wrote: ''It is to a comparative and experimental
analysis of specific characters that we must look for knowledge
of the phenomena of heredity and variation." And again, in
1904, in summarizing the results of many years of study of one
such character he wrote as follows :

''In tracing the origin and genesis of a single character we meet
the leading questions in the evolution of species. First and fore-
most the question as to the nature of the initial stages. Did the

BIOGRAPHICAL SKETCH Ixiii

character arise as a variation de novo, or as a progressive modifi-
cation of a pre-existing character? If de novo, did it spring
suddenly forth, with some decisive advantage in the struggle
for existence? Or did it appear as one of many minute changes,
and by some happy chance get a start that gave it the lead in
future development? In other words, did it begin as a discon-
tinuous variation, sport, or mutation? Or did it arise cumula-
tively, as a continuous development? If it originated by modi-
fication of an earher character, was it at first a sudden, sport-like
departure? Or was it a slow and continuous transformation,
of a progressive or retrogressive nature?

"Then we come inevitably to the deeper question, which natural
selection only partially penetrates — the question how variation,
multifarious and undirected, without the aid of design or a de-
signer, can advance to such definite and wonderful achieve-
ments as specific characters."

Whitman's devotion to the task of learning a specific character
knew no bounds; it heeded neither time, personal sacrifice, nor
the difficulties which the ensemble of hfe processes creates when a
particular process with which the biologist would become familiar
is examined. But, he was ever ready and eager to attend to
each and every perturbation of the system, from whatever
extrinsic source, if its analysis might lead directly or indirectly to
a better measure of realities in his own main sphere of study.
It thus happens that along the pathway which he has blazed
into the central problems of evolution are to be found many
discoveries in the fields of instinct, animal behavior, fertility, cor-
relative variation, the nature of sex, etc.

Having selected color-pattern in pigeons as supplying a satis-
factorily small group of specific characters easily accessible to
study, he first set about determining which patterns are the more
primitive and which the higher and more recent ones, the facts
being determined through a most painstaking search for the con-
vergent testimony of the most various kinds of evidence. Here
his uncompromising ideal of an intensive and extensive study
of a character, his own exceptional mastery of the broad field of
zoology, the eighteen years of unbroken and devoted study that

Lxiv CHARLES OTIS WHITMAN

he gave to this work inevitably led him to results of great impor-
tance.

A general survey was made of the color-patterns of nearly six
hundred wild species, and of nearly two hundred domestic races
of pigeons. Large numbers of genera and species from all parts
of the world were brought to the breeding pens of his yard.
With indefatigable patience the plumage patterns of the living
birds were studied; the sequence of pattern in the plumages
from young to old was accurately observed ; and thoughtful experi-
ments were devised to bridge the gap between the moults, and thus
displace apparent discontinuities with visually realized continui-
ties. The primitive pattern of many diverse orders of birds was
also ascertained, and the general primitive basis of color-marking
in all birds — the 'fundamental bars' were discovered.

The direction of the evolution as it was indicated by all these
studies was, moreover, again and again retested by evidence of an
entirely different sort. Such characters as voice, behavior, and
fertility were separately subjected to similar appropriate vigor-
ous comparative and breeding tests to learn whether the resulting
data would parallel each other and that furnished by the exten-
sive study of the color-pattern. Only when, by all these means
and others, he had accumulated a vast amount of reliable, consis-
tent and convergent testimony as to where the various genera and
species stand in the phylogenetic series, did Professor Whitman
permit himself to feel that he was reading aright the history of the
specific characters of the pattern. And it is a very real monument
to his scientific greatness, that, not until he knew all this of the
character with which he was working, and much besides, would
he write as much as one line concerning it.

In his yards were hybridized nearly forty wild species of pig-
eons, most of these crosses being made here for the first time.
The results of continued breeding of the simple and complex
hybrids from these forty pure wild species, and of several domestic
races, furnish a mass of most remarkable data. The conclusions
from these data being at the same time checked and supported by
the results of other lines of study on the same material.

BIOGRAPHICAL SKETCH IxV

In consequence Professor Whitman's work presents a great
body of searchingly self-critical and reliable conclusions, and these
conclusions unquestionably lead far into constructive evolution-
ary theory. For his material he believed he had demonstrated
beyond doubt the reahty and regnancy of definitely directed
variation, i.e., of orthogenesis, as the method of evolution. He
has accumulated and presented the most weighty evidences for
continuity as against discontinuity in the phenomena of varia-
tion, inheritance and evolution. He has thrown new hght on the
nature and meaning of 'mutants;' such 'mutants' at any
rate as occur among pigeons. He accomplished in 1903, and con-
tinuously since then, the remarkable result which in Mendelian
terms may be spoken of as the control or determination of the
dominance of sex and color. ^

His work was most bountifully and beautifully illustrated,
this feature having occupied many years of the undivided atten-
tion of excellent artists. Even in the unfinished parts, however,
the outlines of the work are so bold and its details of data are
so clear as a result of the polishing process to which he, who
was the very spirit of clarity and accuracy, subjected them, that
time and care will enable others to arrange most of the results
in a form that will still carry conviction to the reader.

Whitman's conception of Orthogenesis, and his attitude toward
the mutation theory is stated in the following paragraph:

' ' Among the rival theories of natural selection two are especially
noteworthy. One of these is now generally known as ortho-
genesis. Theodore Eimer was one of the early champions of
this theory, basing his arguments primarily upon his researches
on the variation of the wall-lizard (1874-81). Eimer boldly
announced his later works on 'The Origin of Species' (1888),
and the 'Orthogenesis of the Butterflies' (1897), as furnishing
complete proof of definitely directed variation, as the result of the
inheritance of acquired characters, and as showing the utter ' impo-
tence of natural selection J Elmer's intemperate ferocity toward
the views of Darwin and Weismann, coupled with an almost

' Unpublished data.

Ixvi CHARLES OTIS WHITMAN

fanatical advocacy of the notion that organic evolution depends
upon the inheritance of acquired characters, was enough to pre-
judice the whole case of orthogenesis. Moreover, the contro-
versial setting given to the idea of definitely directed variation,
without the aid of utility and natural selection, made it difficult
to escape the conclusion that orthogenesis was only a new form
of the old teleology, from the paralyzing domination of which Dar-
win and Lyell and their followers had rescued science. Thus
handicapped the theory of orthogenesis has found little favor out-
side the circle of Elmer's pupils."^

''The second of the two theories alluded to is the mutation
theory of Hugo, de Vries. The distinguished author of this
theory .... maintains, on the basis of long continued
experimental research, that species originate, not by slow gradual
variation, as held by Darwin and Wallace, but by sudden salta-
tions, or sport-like mutations. According to this theory, two
fundamentally distinct phenomena have hitherto been confounded
under the term variation. In other words, variation, as used by
Darwin and others, covers two classes of phenomena, totally dis-
tinct in nature, action, and effect. Variation proper is defined
as the ordinary, fluctuating, or individual variation, and this is
held to be absolutely impotent to form new species.

''Granting that the position with respect to the mutants obtained
from the evening primrose {Oenothera Lamarckiana) is unassail-
able, does it follow that all new species have arisen by mutation,
and that continuous variation has never had, and never can have,
anything to do with the origin of species?

"Plausible as is the argument and impressive as is the array of
evidence presented, I can but feel that there are reasons which
compel us to suspend judgment for a while on this pivotal point
of the mutation theory."

Whitman objected strongly to the implication that a variation
tendency must be considered to be teleological because it is not
orderless.

* Whitman — The Problem of the Origin of Species, Congress of Arts and Science,
Universal Exp. 1904.

BIOGKAPHICAL SKETCH Ixvii

"I venture to assert that variation is sometimes orderly, and at
other times rather disorderly, and that the one is just as free from
teleology as the other. In our aversion to the old teleology, so
effectually banished from science by Darwin, we should not for-
get that the world is full of order, the organic no less than the
inorganic. Indeed what is the whole development of an organism
if not strictly and marvelously orderly? Is not every stage, from
the primordial germ onward, and the whole sequence of stages,
rigidly orthogenetic? If variations are deviations in the direc-
tions of the developmental processes, what wonder is there if in
some directions there is less resistance to variation than in others?
What wonder if the organism is so balanced as to permit of both
unifarious and multifarious variations? If a developmental
process may run on throughout life (e.g., the lifelong multiplica-
tion of the surface-pores of the lateral-line system in Amia),
what wonder if we find a whole species gravitating slowly in one or
a few directions? And if we find large groups of species all
affected by a like variation, moving in the same general direction,
are we compelled to regard such ' a definite variation-tendency'
as teleological, and hence out of the pale of science? If a designer
sets limits to variation in order to reach a definite end, the direc-
tion of events is teleological; but if organization and the laws of
development exclude some lines of variation and favor others,
there is certainly nothing supernatural in this, and nothing which
in incompatible with natural selection. Natural selection may
enter at any stage of orthogenetic variation, preserve and modify
in various directions the results over which it may have had no
previous control."

The particular evidence in favor of orthogenesis on which
Whitman rested his case is found in the origin of the bars on the
wings of the wild pigeons and on the wings of many domesticated
birds.

''The rock pigeons (Columba livia) present two very distinct
color-patterns; one of which consists of black checkers uniformly
distributed to the feathers of the wing and the back, the other of
two black wing-bars on a slate-gray ground. These two patterns
may be seen in almost any flock of domestic pigeons.

Ixviii CHARLES OTIS WHITMAN

The inquiry as to the origin of these patterns involves the main
problem of the origin of species, for the general principles that
account for one character must hold for others, and so for the
species as a whole Darwin raised the same question, but did not
pursue it beyond the point of trying to determine which pattern
was to be considered original and how the derivation of the other
was to be understood. Darwin's explanation was so simple and
captivating that naturalists generally accepted it as final. It is
but fair to state that Darwin's conclusions did not rest on a com-
parative study of the color-patterns displayed in the many wild
species of pigeons. Accepting the view generally held by natura-
hsts, that the rock pigeons must be regarded as the ancestors of
domestic races, the question was limited to the point just stated."

Between the checkered and the barred types many intermediate
stages may be found in different individuals. But which way is
the series to be read, from checkers to bars or from bars to check-
ers? Whitman finds an answer to this question in the evidence
from experiments, from development, and from a comparative
study of the Columbidse.

"As an experiment, we may take one or more pairs of pure-bred,
typically barred pigeons, and keep them isolated from checkered
birds for several years, in order to see if the young ever advance
toward the checkered type.

"Another experiment should be tried for the purpose of seeing
what can be done by working in just the opposite direction. In
this case we take checkered birds, selecting in each generation
birds with the fewer and smaller checkers, and rejecting the others,
in order to see if the process of reduction can be carried to the
condition of three, two, and one bar, and finally, to complete
obliteration of both checkers and bars, leaving the wing a tabula
rasa of uniform gray color.

"If these experiments are continued sufficiently far, it will be
found from the second experiment that a gradual reduction of
pigment to the extreme conditions named can be comparatively
easily effected, and that the direction of reduction will always be
the same, from before backward; while, from the first experiment,

BIOGRAPHICAL SKETCH Ixix

it will be seen that it is hopeless to try to advance in the opposite
direction, from the bars forward to the checkered condition. No
variations will appear in that direction, but such as do appear
will take the opposite direction, tending to diminish the width of
the bars and to weaken their color. It is in this way that we must
account for the existence of some fancy breeds in which the bars
have been wholly obliterated. The direction of evolution can
never be reversed.

"I have tried both experiments for eight years, and as both tell
the same story as to the direction of variation, I am satisfied that
further experiments will not essentially modify the results."

After tracing wing bars of diverse kinds to checkers, the origin
of the checkers was traced from a still earlier and universal avian
character.

''It consists of a single dark spot occupying the centre of the
exposed part of each feather. In the course of evolution, this
spot has been divided into two lateral spots by the disappearance
of pigment along the shaft, beginning at the apex of the feather
and advancing gradually inward. The old Turtle-Dove charac-
ter thus passes by a continuous process of division into the Rock
Pigeon pattern, consisting of two checkers on each feather, more
or less completely separated. The evidences showing such a
gradual transmutation are still to be seen, and in such profusion
as to wholly exclude doubt. Hundreds of species have been
formed in this simple way, leaving no room for the claim of sud-
den, nontransitional mutations.

. ''The transitional stages between the Turtle-Dove pattern and
the checkered pattern of the Rock pigeons, are exhibited not only
as we pass from one species to another, but often as we advance
from the juvenal to the adult plumage; and frequently they may
be seen in different parts of one and the same individual plumage.

''A still older character than the Turtle-Dove spot is seen in the
cross-bars, or fundamental bars, that appear to mark all feathers of
all species of birds. These bars were first noticed in pigeons in the
summer of 1903, and were soon found to be common to all species
of pigeons and birds in general. From these fundamental feather-
bars or their secondary derivatives, a multitude of specific char-

IXX ^ CHARLES OTIS WHITMAN

acters have been evolved by gradual modification. The continu-
ity in the evolution of some of these characters can be experi-
mentally demonstrated. The Uttle Diamond Dove (Geopelia
cuneata) of Australia, owes its small white spots (two in each
feather) to these bars. The transitional stages connecting the
spots with the bars are not wholly given in passing from the juvenal
to the adult plumage. But if we pluck a few of the juvenal
feathers at suitable intervals, their places will be filled by new
feathers of different ages, and in this way we may get the stages
intermediate between the bars of the young and the spots of the adult.
Thus we see that the adult pattern, which normally appears to
come in as a striking mutation, by a single jump, is only an end-
stage in a continuous process of differentiation. So it is every-
where. Suppression of stages in ontogeny looks like saltations;
but whenever we can get at the history of the character, we find the
continuity comes to light."

PERSONAL CHARACTERISTICS

We have described principally the outward events of a life
that was not lacking in incident. Many traits of character shine
through these events, but the history of his inner life is far from
being comprised in such a sketch, and there is no one competent
to write it. With Whitman, more than with most men, one felt
that the inner life was the dominant factor, that it was genuine,
deep and worth knowing; the outward events were more or less
accidental; he would have created similar effects under a totally
different set of external circumstances.

In person he was of medium height, of stout build and good
color in his maturity, though thin and pale in his last years; his
hair was snow white from young manhood, his eyes were direct
and piercing, his forehead high, broad and noble; he wore a beard
and a heavy mustache that somewhat concealed the rather thick-
lipped mouth. His bearing was always erect and dignified; his
dress was simple and sufficiently conventional, but he entirely
eschewed the ceremonial dress and was only once seen in academic
cap and gown, and I believe not at all in evening dress.

BIOGRAPHICAL SKETCH Ixxi

Whitman's life was simple and studious; it was passed almost
entirely between his house and his laboratory. A large part of
his work, since 1891 certainly, was done at home, and from about
1895 when he began the study of pigeons, by far the major part.
He gradually collected a large number of species of pigeons from
all parts of the world, and in the latter part of his life the collection
comprised some 550 individuals representing about thirty species.
His house was surrounded by pigeon cotes, and he always had
some birds under observation indoors, so that the cooing of doves
was for years a dominant sound in his house. He took care of
the birds for the most part himself, though he usually had the
assistance of one or two maids. He thus actually lived with his
birds constantly, and very rarely was absent from them even for
a single day. He made observations and kept notes on all aspects
of the life and behavior of each species, as well as of such hybrids
as he was able to produce. He always had one Japanese artist
at work continuously drawing pigeons, and for several years
two — Hyashi and Toda. Thus he accumulated an immense
amount of material for his magnum opus, which, however, he was
not permitted to finish. It is hoped that a considerable portion
of his work may be available for posthumous publication, owing
to the self-sacrificing labor that has been put on it first by Dr. R.
M. Strong, and then by Dr. Oscar Riddle, one of his students,
who is now devoting his entire time to editing the manuscripts.

For many years Whitman carried his pigeons with him to Woods
Hole in June and back again in September, as already related.
But the burden became intolerable, especially as he always bore
the entire expense of his pigeon work personally. He was finally
obliged to relinquish the annual trip to Woods Hole and all the
cherished associations of the Marine Biological Laboratory.
His work probably flourished better under these conditions, but
it is to be feared that his health suffered from too close applica-
tion to work and from lack of variety in his life.

With all of his close application to his study, he was neverthe-
less most devoted to his friends; he was always pleased to see
them, and would spend hours in conversation with them as though
he had no other concern in the world. Although he was no smoker

Ixxii CHARLES OTIS WHITMAN

he would always offer cigars and cigarettes, and would frequently
light a cigarette himself, — which burned however mostly in his
fingers, — to heighten the spirit of hospitality. He rather fre-
quently invited his students or members of his department and
other friends to dinner, and then his usually simple meal was
changed to a more elaborate repast.

Dr. A. P. Mathews, one of his close friends, writes thus of
him,^

It is not, however, of his work as a scientist upon which I wish to
dwell, but rather to recall his personality that the memory of it may
remain always with us. His white hair; his kindling, eager, but thought-
ful eyes; his tender, gentle smile; his reticence of speech; his considera-
tion for others; his generosity and courage; his hospitality and gracious-
ness as a host; these endeared him to us all. We shall never forget his
simple, unassuming, modest manner; his encouraging sympathy; his
ripe and sane judgment. If when he was alone he lived simply, the
absorbed student of science, when with his guests in his home he was
the embodied spirit of hospitality.

His great influence as a teacher is due in part to his fine example and
noble ideals, and in part to his habit of picking out young men, who
showed any love for science, inviting them to his home, drawing them
out, encouraging them and giving them his friendship. Many of them
he helped financially, and all of those fortunate to work near him owe
him a debt of gratitude for his sympathy and inspiration. Probably
no teacher in zoology since Louis Agassiz has exerted so great an influ-
ence on young men.

He was not a faultless man, but his faults were the outcome of
his ardent, ideal, uncompromising disposition. He once said to
the writer, about the year 1906, that he felt he had been too un-
compromising in his beliefs. But it is questionable whether his
life would have been so valuable, had his disposition been more
pliable. The mood in which he made this remark was a rare one,
and it is to be doubted that even had it been more common he
could have overcome his native tendency. This quality of course
made him enemies who sometimes did not hesitate to express
unfavorable opinions in a more or less open manner. But those,
who knew Whitman best, know well that he never sought any
small personal advantage, and that any appearance of neglect of
small matters was due entirely to his absorption in higher con-

9 Science. N. S., vol. 33, no. 837, pp. 56-58, Januarj' 13, 1911.

BIOGRAPHICAL SKETCH Ixxiii

siderations. He had the courage of his convictions and rarely,
if ever, avoided an issue, turned from an opponent, or shunned
a fight.

Although Professor Whitman pubHshed relatively few papers,
he nevertheless occupied a commanding position in science.
Some of the reasons have already been indicated. His ''eye was
single and his whole body was therefore full of light;" his devotion
to scholarship was never open to the slightest shadow of suspicion.
He was continuously engaged in his personal research which dealt
with the most fundamental problems of biology, and he had
accumulated vast stores of data, which we hoped he would live to
publish himself. But apparently he could never satisfy himself
with reference to the fundamental problems on which his mind was
fixed; the grand consummation of his work had not come, and he
could not reconcile himself to the publication of more or less frag-
mentary pieces of work. His published papers, mostly short,
are models of condensed thought, written in a fine, polished,
characteristic style. No less care was devoted to the form than
to the substance, and some of his papers will certainly endure as
classics of the biology of his time

It was, however, not only his publications, but also his work
with his journal, his laboratory and his students, his constant
helpful association with other workers, and the example of his
austere and studious life that brought him recognition. He never
permitted himself to be distracted by the confusion of modern
life, social or academic, nor diverted from his steadfast purpose
by clamor for quick results.

SICKNESS, DEATH AND BURIAL

For several years before his death Professor Whitman suffered
considerably from indigestion, and lost much flesh. He was,
however, in better health than usual in the fall of 1910. A sudden
cold wave came on about December 1, 1910, and Whitman spent
the entire afternoon in his yard putting his birds into their winter
quarters. In his zeal for his pigeons he forgot about himself.
The next morning he was found in a state of coma, and pneumonia

Ixxiv CHARLES OTIS WHITMAN

rapidly developed and brought his life to a sudden and unexpected
termination on December 6. He, himself, had looked forward
to at least ten more years of active study, for apart from his
dyspepsia, which he had learned to control very well, his general
health was excellent. He thus died unprepared, with work that
he had carried on up to the last moment in an unfinished condition.

Memorial services in his honor were held in Convocation Hall
of the University of Chicago on December 8, and the same evening
his body was taken to Woods Hole in charge of a committee of
four appointed by the University, and his two sons. The inter-
ment took place in the lot of the Marine Biological Laboratory
in the Episcopal Cemetery at Woods Hole in the presence of a
small company of scientific friends and colleagues, who came to
Woods Hole for this purpose, and some of his friends of the
village. His grave lies almost within sight of the institution which
he had loved so well, overlooking the harbor.

At the annual meeting of the Corporation of the Marine Biolog-
ical Laboratory held in Woods Hole, August 8, 1911, the entire body
adjourned and marched to the grave of Professor Whitman, where
a memorial address was read and a wreath placed on the grave.
Thus those who were unable to attend the interment paid their
last respects to the memory of the dead leader.

LIST OF PROFESSOR WHITMAN'S PUBLICATIONS

1878 The embryology of Clepsine. Quar. Journ. Micr. Sci., vol. 18, pp. 215-315.

1878 Ueber die Embryologie von Clepsine. Zool. Anz., Bd. 1, p. 5.

1878 Changes preliminary to cleavage in the egg of Clepsine. Proc. Amer. Assoc.
Adv. Sci., vol. 26, pp. 263-270.

1880 Do flying fishes fly? Amer. Nat., vol. 14, pp. 641-653.

1881 Zoology in the University of Tokyo.

1882 Japanese aquatic animals living on land. Amer. Nat., vol. 16, pp. 403-405.
1882 Methods of microscopical research in the Zoological Station in Naples.

Amer. Nat., vol. 16, pp. 697-706; 772-785.
1882 Ibid: Journal de Micrographie, 6, pp. 558-565; pp. 18, 89 and 188, vol. 7
(French translation of preceding paper).

1882 A new species of Branchiobdella. Zool. Anz., pp. 636-637.

1883 A contribution to the embryology, life-history, and classification of the
Dicyemids. Mitth. Zool. Sta. Neapel, vol. 4, pp. 1-89.

1883 Treatment of pelagic fish eggs. Amer. Nat., vol. 22, pp. 1204-5.

BIOGRAPHICAL SKETCH IXXV

1883 A rare form of the blastoderm of the chick and its bearing on the question
of the formation of the vertebrate embryo. Quar. Journ. Micr. Sci., vol.
23, pp. 376-397, and Proc. Bos. Soc. Nat. Hist., vol. 22, pp. 178-79.

1884 External morphology of the leech. ProcAmer. Acad. Arts and Sci., vol.
20, pp. 76-87.

1884 On the development of some pelagic fish eggs. Proc. Amer. Acad. Arts

and Sci., vol. 20, pp. 23-75 (with A. Agassiz).
1884 Segmental sense organs of the leech. Amer, Nat., vol. 18, pp. 1104r-1109.

1884 The connective substance of Hirudinea. (Review) Amer. Nat., vol. 18, p.
1070.

1885 Methods of research in microscopical anatomy and embryology. VIII +
255 pp. Boston, S. E. Cassino and Co.

1885 Means of differentiating embryonic tissues. Amer. Nat., vol. 19, pp. 1134-
1137.

1886 Osmic acid and Merkel's fluid as a means of developing nascent histological
distinctions. Amer. Nat., vol. 20, p. 200.

1886 The leeches of Japan. Quar. Journ. Micr. Sci., vol. 26, pp. 317-416.

1886 Germ layers of Clepsine. Zool. Anz., Bd. 9, pp. 171-176.

1887 A contribution to the history of the germ layers in Clepsine. Jour. Morph.,
vol. 1, pp. 105-182.

1887 Ookinesis. Jour. Morph., vol. 1, pp. 228-252.

1887 Biological instruction in universities. Amer. Nat., vol. 21, pp. 507-519.

1888 The seat of formative and regenerative energy. Jour. Morph., vol. 2, pp.
27-49.

1888 The eggs of Amphibia. Amer. Nat., vol. 22, p. 857.

1888 Some new facts about the Hirudinea. Jour. Morph., vol. 2, pp. 585-599.

1888 Address at the opening the Marine Biological Laboratory July 17. First
Annual Report of the Mar. Biol. Lab. Boston, pp. 24-31.

1889 The development of osseous fishes. 2. The pre-embryonic stages of devel-
opment, (with A. Agassiz). Mem. Mus. Comp. Zool. Harvard College,
vol.. 14, pp. 1-56.

1889 Report of the Director of the Marine Biological Laboratory for the first
session, 1888. First Annual Report of the Mar. Biol. Lab. Boston, pp. 14-20.

1890 Report of the Director of the Marine Biological Laboratory for the second
session 1889. Second Annual Report of the Mar. Biol. Lab. for the year
1889. Boston, pp. 27-34.

1890 Report of the Director of the Marine Biological Laboratory for the third
session, 1890. Third Annual Report of the Mar. Biol. Lab. Boston, pp.
17-23.

1891 Report of the Director of the Marine Biological Laboratory for the fourth
session, 1891. Fourth Annual Report of the Mar. Biol. Lab. Boston, pp.
14-29.

1891 Description of Clepsine plana. Jour. Morph., vol., 4, pp. 407-418.

1891 Spermatophores as a means of hypodermic impregnation. Jour. Morph.,

vol. 4, pp. 361-406.
1891 Specialization and organization, companion principles of all progress;

The most important need of American biology. Biol. Lectures, M. B. L.,

pp. 1-26, Boston.

Ixxvi CHARLES OTIS WHITMAN

1891 The naturalist's occupation. 1. General survey. 2. A special problem.
Ibid., pp. 27-52.

1892 Metamerism of Clepsine. Festschrift Rudolph Leuckart, pp. 385-395.
1892 Artificial production of variation in types. Science, vol. 19, p. 227.

1892 Report of the Director of the Marine Biological Laboratory for the fifth
session, 1892. Fifth Annual Report of the Mar. Biol. Lab. Boston, pp.
18^7.

1893 A marine biological observatory. Pop. Sci. Monthly, vol. 42, pp. 1-15.
1893 A marine observatory the prime need of American biologists. Atlantic

Monthly, pp. 808-815.

1893 The inadequacy of the cell theory of development. Jour. Morph., vol. 8,
pp. 639-658, and in Biol. Lect. 1893.

1893 A sketch of the structure and development of the eye of Clepsine. Zool.
Jahrb., Abth. Anat. u. Ont., vol. 6, pp. 616-625.

1893 The work and the aims of the Marine Biological Laboratory. Biol. Lec-
tures, Woods Hole, pp. 236-241. Boston.

1893 General physiology and its relation to morphology, Amer. Nat. vol. 27,
pp. 802-807.

1894 Breeding habits of the three triclads of Limulus. Amer. Nat. vol. 28, pp.
544-545.

Prefatory note Biol. Loct., Woods Hole, 1894, pp. ii-vii.

1894 Report of the Director to the Trustees of the Marine Biological Laboratory
on the Work of the Sixth Session, 1893. Sixth Annual Report of the Mar.
Biol. Lab. for the year 1893. Boston, pp. 21-31.

1895 Evolution and epigenesis. Biol. Lectures, Woods Hole, 1894, pp. 205-224,
Ginn and Co.

1895 Bonnet's theory of evolution. A system of negations. Ibid: pp. 225-240.

1895 The palingenesia and the germ doctrine of Bonnet. Ibid: pp. 241-272.

1896 The egg of Amia and its cleavage, (with Eycleshymer). Jour. Morph.,
vol. 12, pp. 309-356.

1896 Report of the Director of the Marine Biological Laboratory for the Seventh
and Eighth Sessions, 1894-1895. Eighth Annual Report of the Mar. Biol.

Lab. for the year 1895. Boston, pp. 17-83.

1897 The centrosome problem and an experimental test. Science N. S. vol. 5,
1897, pp. 235-236.

1878 Lamarck and a perfecting tendency. Science N. S. vol. 7, p. 99.

1898 Some of the functions and features of a biological station. Science, N. S.
vol. 7, pp. 37^4, 1898. (Presidential address toSoc. Amer. Nat. 1897, but
not delivered.)

1898 Animal behavior. Biol. Lectures, Woods Hole, pp. 285-338.

1899 Myths in animal psychology. Monist, vol. 9, pp. 524-537.
1899 Apathy's grief and consolation. Zool. Anz., pp. 196-197.

1902 The impending crisis in the history of the Marine Biological Laboratory,
Science, vol. 16, pp. 529-533.

1902 A biological farm for the experimental investigation of heredity, variations
and evolution, and for the study of life histories, habits, instincts and intel-
ligence. Biol. Bull., vol. 3, pp. 214-224.

BIOGRAPHICAL SKETCH IxXvii

1904 Natural history work at the Marine Biological Laboratory, Woods Hole.
Science, vol. 13, pp. 538-540.

1904 Hybrids from wild species of pigeons crossed inter se and with domestic
races. Research Seminar, M. B. L., Biol. Bull., vol. 6, pp. 315-316.

1904 The origin and relationship of the rock pigeons as revealed in their color-
patterns. Biol. Bull. vol. 6. pp. 307-308.

1906 The problem of the origin of species. Congress of Arts and Sciences, St.
Louis Exposition, 1904, Boston, vol. 5, pp. 41-58.

1906 The origin of species. The introduction and abstract of a lecture delivered
before the Nat. Hist. Soc. December 20, 1906. Bull. Wis. Nat. Hist. Soc.
vol. 5, pp. 6-14.

1907 Cheques and bars in pigeons and the direction of evolution. Agricultural
Magazine, vol. 5, no. 6, pp. 174-182.

THE ORIGIN OF THE SEX-CELLS OF AMIA AND
LEPIDOSTEUS

BENNET M. ALLEN

From the Department of Anatomy, University of Wisconsin

TWENTY-SEVEN FIGURES

INTRODUCTION

There has been an increasing amount of attention given in the
last few years to a study of the origin and migration of the sex-
cells of the vertebrates. The number of forms in which this
subject has been studied is being constantly extended. While
much conclusive work has been done, upon the history of these
cells in the elasmobranchs, and an equal amount in tracing them
in the teleosts, up to this time, they have never been studied in
the ganoids. This work was begun over two years ago, and was
reported at the 1908 meeting of the Association of American
Anatomists in New York. (Allen, '09.)

The material for the present work is obtainable in great plenty
within less than half a mile of the grounds of the University of
Wisconsin. Since the breeding habits of these two fishes have
been thoroughly treated by other writers, it is not necessary to
redescribe them. Telleyesniczky's bichromate-acetic fluid and
Zenker's fluid have been used as fixing agents and have proved in
every way satisfactory. One secret of obtaining good sections
is to secure most thorough infiltration by placing the material
in a solution of paraffine in turpentine. While turpentine has a
bad reputation, no deleterious effects were noted in the course
of the work. Paraffine sections were made without difficulty
7m and 10/^ thick, and were stained, for the most part, in haem-
alum and orange G. Heidenhain's iron haematoxylin stain was

JOnKNAL OF MORPHOLOGY, VOL. 22, NO. 1
MARCH, 1911

2 BENNET M. ALLEN

sometimes used for the later stages, but showed no superiority
over the haem-alum. In the earher stages of development it
could not be used at all, owing to the deep stain that it gives the
yolk material.

With the abundance of material and the amount of time given
to the work, it was possible to make a careful study of a large
series of stages, much larger than it has been found necessary
to use in the preparation of this paper.

It is not necessary to enter into a detailed account of the
earlier work upon the origin of the sex-cells, because that has
already been done in earlier writings. Since the author's articles
on the sex-cells of Chrysemys and of Rana, some important papers
have made their appearance, which, with one exception, bear
out in a most gratifying manner the conclusions expressed by the
writer in the two papers mentioned above and in somewhat less
confident manner in the earlier writings of Wheeler, Woods and
Beard.

These papers will be discussed in the light of the facts set
forth in this paper in the last part of this article, since they are
to be considered in a more or less controversial manner.

OBSERVATIONS UPON LEPIDOSTEUS OSSEUS

Lepidosteus, 4 '^nm. total length. Cells which appear to be sex-
cells lie in the ventral portion of the single layered gut entoderm.
They can be but dimly distinguished from the other cells of the
entoderm among which they lie. They have a more spherical
shape than the other entoderm cells, never being flattened as the
neighboring cells frequently are. Another difference lies in the
fact that the sex-cells contain more and decidedly larger yolk
spherules than do the adjoining entoderm cells. Unfortunately
these differences are masked by the large quantities of yolk
found in the entoderm at this stage. This is true to so great an
extent than one can not be certain at this stage as to the identity
of the cells in question. At this period the hind gut has a much
greater diameter than it has at later stages. At a point one-
quarter the distance from its cranial to its caudal end it has a

SEX-CELLS OF AMIA AND LEPIDOSTEUS 3

dorso-ventral dimension of .24 mm. and a transverse diameter
of .20 mm.

Lepidosteus 6.8 mm. total length. In a similar part of the hind
gut of a specimen 6.8 mm. in length the dorso-ventral dimension
of the gut endoderm is .084 mm. while the transverse dimen-
sion is .056 mm. The total length of the hind gut in this later
stage is 1.70 mm. as compared with .76 in the 4 mm. embryo.
It is seen that there has been a very decided diminution in the
diameter of the gut, and, furthermore, that this is out of proportion
to the increase in length. It is correlated with a thickening of the
gut wall, due to the drawing together of the component cells.

In the 4 mm. stage the gut entoderm was composed of a single
layer of cells, while in the 6.8 mm. stage under consideration its
lateral and ventral portions are made up of two, and in some places
three irregularly arranged layers of cells, while the dorsal wall
is made up of a single layer as in the 4 mm. stage. Two series
chosen from several of this stage may be taken as showing typical
differences. Both show an advance over the preceding stages
in the greater ease with which the sex-cells may be distinguished.
This is due to the contrast between the ordinary entoderm cells
in which a considerable amount of the yolk material has been
absorbed and the more rounded yolk-filled sex-cells. In neither
embryo has the process of sex-cell migration commenced. This
is clearly evident in one, while in the other there are a very few
scattered yolk-filled cells of rather problematical character in
the loose mesenchyme above and at the sides of the gut entoderm.
One striking individual difference between • these two specimens
is found in the fact that while in one the sex-cells have retained
their primitive position in the ventral and lateral portions of the
gut entoderm, in the other they have migrated up into its dorsal
portions. Although it is somewhat difficult to establish with
absolute certainty this migration from the ventral to the dorsal
portions of the gut entoderm owing to the difficulty of distin-
guishing the sex-cells in the preceding stages, the individual
differences in this regard observed in this stage, together with
the fairly reliable observations upon earlier stages seem to point
to an actual migration of this character.

BENNET M. ALLEN

Lepidosteus 7.3 mm. total length. In the single series of 7.3
mm. embryos the sex-cells in general still occupy the ventral and
lateral portions of the gut entoderm, having come to lie in the
dorsal wall at only a few points, especially toward the cranial
end of the hind gut. A very few sex-cells are found to have
migrated into the loose mesenchyme between the gut entoderm
and lateral plates of mesoderm, occupying positions in it lateral
and dorsal to the latter. These migrant cells are merely the
precursors of a general migration which does not become con-
spicuous until the embryo has reached a length of 8.5 mm.

TABLE 1
Number of sex-cells in Lepidosteus

NUMBER

Entoderm

Mesoderm

mm.

6.8
8.6

No count

9.2

136

9.3A

9.3B

((

311

9.3 C

425

10.7

133
125

674

S07

Int.

Root

12.0

104

163

180

179

751

14.1

128

222

197

153

737

17.0

No count

262

235

18.0

171

173

24,0

147

154

Int. — Intestine.

Root — Root of mesentery.

R. — Right sex-gland.
L. — Left sex-gland.

Lepidosteus, 8.6 vim. total length. Passing over several interme-
diate stages studied, the conditions found in a specimen 8.6 mm.
long may be described. At this stage the lateral plates of meso-
derm are just beginning to split and to form the coelomic cavities
(fig. 7). The interval between the plates is filled with loose
mesenchyme, of which that portion lying between the gut ento-
derm and the aorta will later be condensed by the apposition of
the lateral plates of mesoderm in formation of the mesentery.

SEX-CELLS OF AMIA AND LEPIDOSTEUS 5

It will be seen by comparing the figures drawn to scale that the
mesentery is at this stage not only relatively but actually much
thicker than it is during later stages. The accompanying table
serves to show the number of sex-cells and their distribution in
certain stages.

In this specimen 73 sex-cells have already migrated out of
the gut entoderm into the surrounding mesoderm tissues. While
.most of them have migrated upward into the loose mesenchyme
and splanchnopleuric mesoderm of the anlage of the mesentery,
a few have passed laterally into the portions of the splanchno-
pleuric mesoderm that enter into the formation of the intestinal
wall. While the migration of the sex-cells is seen to be well
under way at this stage, the great majority of them still remain
in the dorsal portion of the gut entoderm. Very few, indeed,
are to be found in the ventral half at this stage. The sex-cells
of the gut entoderm are easily distinguishable from the other
entoderm cells; the latter have lost very nearly all of their yolk
material and have become cylindrical in shape. These features
stand out in sharp contrast to the large yolk content and spherical
form of the sex-cells.

The migration of sex-cells from the gut entoderm into the
mesenchyme dorsal or lateral to it may be clearly seen at a few
points, as illustrated in figs. 7 and 8. They retain for the most
part their spherical form, but cases like that shown in fig. 7
can be readily found. The shape of this sex-cell clearly indicates
the mode of progression. They, undoubtedly, pass through the
loose network of mesenchyme by an amoeboid movement, how-
ever slow or intermittent it may be.

In this stage sex-cells are found in the hind gut from its cranial
end to within .2 mm. of the cloaca, a distance of 2.6 mm.

Lepidosteus 9.2 mm. long. The number of sex-cells that have
migrated out of the gut entoderm is 136 in this specimen. The
number of these is still increasing but solely by migration from
the entoderm, since there is no evidence of division of the sex-
cells during these stages of sex-cell migration.

In this stage the coelomic cavities have appeared in the dorsal
portion of each lateral mesodermal plate and the mesentery is

6 BENNET M. ALLEN

consequently much more clearly defined. In three specimens
9.3 mm. long the following counts of sex-cells outside of the gut
entoderm were made:

A = 41 B = 311 C = 425.

A and C are extreme cases, indicating that the process of
migration is an irregular one in point of time. The mesentery
in specimen A is .46 mm. wide.

In specimen C those sex-cells destined to migrate out of the
entoderm have for the most part already done so, while in A,
an embryo of the same stage, the process is just beginning.
The coelome is least developed in A and furthest advanced in C.
This would indicate that the extent to which the migration of
sex-cells has been carried on is correlated with the degree of
development of the mesentery, resulting from the enlargement of
the coelomic cavities. WTiile these three specimens belong, no
doubt, to slightly different stages of development, they were very
carefullj^ matched as to length, and are most certainly of the same
age.

Lepidosteus 10.7 mm. total length. At this period of develop-
ment, the mesentery is well formed, being much thinner (.18 mm.)
than in the 9.3 mm. stage. This results in its possessing a denser
texture (fig. 9). The great majority of the sex-cells are scattered
through the mesentery, showing no definite arrangement; but
lying for the most part in the mesenchyme enclosed between the
somewhat denser splanchnic laj'ers of mesoderm. A few are
found in the mesodermal layers of the intestine, while a fairly
considerable number have remained in the gut entoderm. At
this time such sex-cells as are found in any but the dorsal wall
of the intestine, at its junction with the mesentery, are most
probably destined to remain in their present positions. A few
of the sex-cells have migrated to places immediately dorsal and
lateral to the root of the mesentery. The latter may be con-
sidered to have reached the sex-gland anlagen, although their
position relative to the root of the mesentery will be shifted, as
we shall see, in the later stages, probably by a general shifting
of the tissues in which they lie.

SEX-CELLS OF AMIA AND LEPIDOSTEUS 7

The number of sex-cells which have migrated from the ento-
derm is found to be 674. It can be fairly taken as the number
destined to undergo migration from the entoderm in this par-
ticular individual. Those still remaining in the entoderm num-
ber 133.

A few scattered sex-cells are found as far forward as the cranial
end of the hind gut. The latter is 3.41 mm. in length. Opposite
to the cranial portion of the hind gut the sex-cells are rather
sparse, increasing in number as one follows the series caudally.
The}'- become most numerous a short distance caudal to a point
two-thirds the distance from the cranial to the caudal end of the
hind gut. The last one in the mesoderm is found at a distance
of .46 mm. from the cloaca, and the last one in the entoderm
lies at a point .27 mm. from the cloaca.

Lepidosleus 12 mm. total length. At this period migration of
the sex-cells has progressed to the point where most of them have
reached their final positions. They are still to be seen in the
entoderm. This number (125) is quite close to that (133) of
the similarly situated sex-cells of the preceding stage. The
density of the mesodermal tissues surrounding the gut entoderm
makes it seem quite unlikely that any more sex-cells could mi-
grate into them from this source.

The distribution of the sex-cells is as follows:

392 outside of sex-gland anlagen.

Gut entoderm 125

Mesoderm of intestinal wall and
Mesentery 104

Root of mesentery between sex-
gland anlagen 163

Right sex-gland 180l

Left sex-gland 179/ ^59 ^ sex-gland anlagen.

Total 751

The table may be allowed to speak for itself. The sex-gland
anlagen grade into one another by an intermediate region at
the root of the mesentery. More or less arbitrary limits had to
be assumed to distinguish between these three regions. In
later stages, illustrated by the 17 mm. stage, fig. 11, we shall see

8 BENNET M. ALLEN

that the sex-cells undergo lateral migration, either apparent or
real, so as eventually to lie at some distance on each side of the
median point.

The narrowest portion of the mesentery is at about one-quarter
the distance from its origin to its insertion. Its minimum width,
as measured here amounts to .028 mm., thus showing a great
reduction as compared with the 10.7 mm. stage. This reduction
in width is shared by the entire mesenterj', certain regions remain-
ing broad only on account of the enclosed blood vessels. No
doubt the migration of the sex-cells out of the mesentery is in
large part responsible for this, but a considerable share of it must
be ascribed to the fact that there has been a tendency for the
tissues to become more compact.

The total length of the hind gut at this stage has reached 4.03
mm. Sex-cells are found in the entoderm at its cranial end, and
from there extend to within 0.62 mm. of the cloaca. The dis-
tribution of sex-cells within the sex-gland anlagen is somewhat
more restricted, since they extend from a point 0.31 mm. caudal
to the beginning of the hind gut, to a point 1.00 mm. cranial to
the cloaca. The}' are rather sparse at these two extremes.

As in the preceding stages, there is no clear evidence of division
of the sex-ceUs, although one can not be absolutely certain upon
this point. While at this time many are free from yolk material,
others show but little diminution of it. It is true that the sex-
cells are often found arranged in clusters, but there is no evidence
to show that these are due to repeated division of a parent cell
rather than to a tendency for them to congregate through mutual
attraction. ^Yhsit the nature of this attraction might be, we do
not know; but it might well be akin to that influence which causes
the sex-ceUs to migrate toward the sex-gland anlagen from their
source. Similar clusters of sex-cells were found in earty stages
in Chrysemys.

Lepidosteus 14-1 inm. total length. Little radical change is to
be seen in this stage. The sex-cells were counted and gave the
following results:

SEX-CELLS OF AMIA AND LEPIDOSTEUS

Gut entoderm 128

Mesoderm of intestinal wall and

mesentery 222

Root of mesentery between S. G.

anlagen 37

Right sex-gland 197 \

Left sex-gland 153 /

Total 737

387 outside of sex-gland anlagen.
:350 in sex-gland anlagen.

There is a strikingly close correspondence between the results
of the count hi this specimen and those in the preceding one.
Attention may be called to the fact that in this specimen a
materially greater number of sex-ceUs is found in the right sex-
gland than in the left. At the same time there is a very close
correspondence in the total number of sex-cells that have reached
the sex-gland anlagen as compared with the total number in the
12.0 mm. stage (359).

Lepidosteus 17 mm. total length. In this specimen those sex-
cells destined to occupy the sex-glands are seen to have migrated
some distance to each side of the root of the mesentery, fig. 11.
Their position relative to the root of the mesentery and to the
Wolffian duct varies at different points along the sex-gland
anlage. In the most cranial portion of the latter they lie just
medial to the Wolffian duct. As one follows the sex-glands
caudally, the sex-cells are found to lie closer and closer to the
mesentery, being situated midway between the latter and the
Wolffian duct in the middle region of the sex-gland anlage. The
most caudally situated sex-cells lie close to the root of the mesen-
tery.

In this and the succeeding stages the intestine had become so
voluminous as to make the counting of the sex-cells in its walls
very difficult and inaccurate. It is in fact not easy to distin-
guish them from the cells of the gut entoderm because of their
rather small size and their entire lack of yolk material at this
stage.

The total number of sex-cells in the sex-gland anlagen of this
specimen is rather high, there being 235 in the left sex-gland and
262 in the right. The total number is 497.

10 BENNET M. ALLEN

The slightly greater number of sex-cells in the sex-glands of this
specimen as compared with that in the previous ones is of little
significance. It most certainly does not indicate that there has
been any extensive division of them. In a previous work upon
Chrysemys, (Allen '07), it was shown that there was an extreme
amount of individual variation in the number of sex-cells. This
variation in Lepidosteus is relatively slight compared with that
observed in Chrysemys. In a specimen slightly older than this
stage (18 mm.) there were 171 sex-cells in the right sex-gland,
and 173 in the left one, the total number, 344, being not far from
the average.

In these two stages, 17 and 18 mm., the sex-cells usually occur
singly, although in places they are aggregated into clumps so
thick as sometimes to show as many as five or six in a section
of one of the sex-glands. Whether the sex-cells occur singly or
in clumps, they are surrounded by peritoneal cells which con-
tribute materially to the formation of the ridge-like anlage of
the sex-gland.

Lepidosteus 21+ mm. total length. In a specimen of this length,
fig. 12, there is no essential advance in the development of the
sex-gland. There were 147 sex-cells in the right sex-gland, and
154 in the left one. The total number, 301, is distinctly below
the average.

Comparison with other forms leaves no room for doubt as to
the identity of these sex-cells. Since the aim of this paper is
merely to trace out their origin, we will not follow them through
later stages in their history, but will describe the conditions found
in a specimen 110 mm., in length, fig. 13. A complete series of
sections through the sex-gland region of this specimen was not
made, so it is impossible to give a full account as to the number of
sex-cells and general condition of the sex-gland at this time.
In running through the series one is struck with the sparseness
of the sex-cells. Never are more than two or three to be found
in a single section, and often none at all. This would lead one
to infer that there has been little or no multiplication of the sex-
cells even at this late stage of development.

SEX-CELLS OF AMIA AND LEPIDOSTEUS

A glance at table 2 shows that there is a general tendency to a
reduction in the average size of the cell body in the later stages.
This may be due to the absorption of the contained yolk material.
There is no marked change in the size of the nucleus.

TABLE 2

Dimensions of sex-cells of Lepidosteus

CELL BODY

Stage

Nucleus

LARGEST

SMALLEST

AVERAGE

mm.
8.6

6.04

15,10

12.08

13.74

9.3

6.04

18.12

12.08

14.95

10.7

6.04

15.10

11.32

13.59 .

14.0

5.81

12.08

9.06

10.27

17.0

6.04

13.59

9.06

11.63

24.0

5.81

9.06

7.55

8.65

110.0

6.53

14.50

9.22

12.40

AMIA CALVA

Amia 4- fnm., total length. In the text figure A is shown a
transverse section of an Amia larva of this stage. It will serve
as a starting point from which we shall proceed to consider still
earlier stages in tracing out the earliest phases in the origin and
migration of the sex-cells. The section shown is taken "just
anterior to the hind gut, the gut entoderm being clearly marked
by its greater thickness and dorsal curvature. The cavity of the
intestine at this point opens into the large sub-germinal cavity.
The extra embryonic portions of the entoderm, i.e., those which
do not form part of the anlagen of the alimentary tract and its
appendages can logically be divided into four different regions :
(1) The roof of the sub-germinal cavity which is distinguishable
from the gut entoderm, as indicated; (2) The layer forming the
floor of the sub-germinal cavity; (3) The peripheral layer of
entoderm lateral to the sub-germinal cavity (peripheral entoderm) ;
(4) The central yolk mass, or vitellus (vitelline entoderm). In
the first three of these regions the cells are arranged in a single
layer. They are characterized by the fact that the yolk spherules
of the component cells are distinctly smaller than are those of

BENNET M. ALLEN

the vitellus, their diameter being from one-quarter to one-half
of that of the typical spherules to the vitellus. In the latter
cells are scattered a few of these smaller yolk spherules; but
the distinction between the first three divisions of the entoderm
and the vitellus is a very sharp one.

Perjph. C

Text figure A

In connection with this distinction it is interesting to note
that the yolk spherules along the cleavage planes that cut through
the vitellus are found to belong to this small type. It is easy to
see that if the vitellus were cut up into cells as small as those
comprising portions 1, 2, and 3 of the entoderm, the thickness
of the layers of small spherules which form merely a border to
the large cells would be so great as to comprise the entire body of
the more finely divided ones. This difference in the size of the

SEX-CELLS OF AMIA AND LEPIDOSTEUS 13

yolk spherules is then probably associated with the difference
in the size of the cells. The peripheral entodermal layer which
we have designated as division three is interrupted latero-
ventrally by blood vessels lying in the mesoderm.

The lateral plates of mesoderm have long since broken away
from the mesoblastic somites. Their inner margins lie at some
distance to each side of the median line. While there is the
slightest tendency in places for the splanchnic and somatic layers
of mesoderm to split apart along the medial margins of the lat-
eral plates, the remainder of the lateral plates show no indication
of a splitting, even in the arrangement of the nuclei. It is, how-
ever, quite probable that such a plane of cleavage is already laid
down. This is shown by the sex-cells (text fig. A) being imbed-
ded in the lateral plates. When the somatopleure and splanchno-
pleure separate later, these will be found to lie in the coelomic
cavity, being for a time merely adherent to the coelomic sur-
face of the medial portion of the somatic mesoderm. One can
fairly assume that during the period of migration, represented by
fig. 5, the sex-cells push their way between the two layers of
mesoderm following the potential cleft that separates them.

Text fig. A is very suggestive, as it shows sex-cells situated at
intervals from a point just beyond that at which the roof and
floor entoderm join the peripheral entoderm. The path of their
migration is thus clearly marked out. In this figure it should be
noted that the most laterally situated sex-cell lies in the ento-
derm, while all of the others are clearly imbedded in the lateral
plates of mesoderm as already indicated. In but one or two of
the many specimens examined were there any sex-cells found in
the roof or gut entoderm. They arise in the peripheral entoderm
from which they migrate into the lateral plates of mesoderm and
through them to their medial borders, whence, as I shall later
show, they pass into the sex-gland anlagen after the formation of
the coelomic cavity.

The total number of sex-cells found in the mesoderm of the
specimen of this stage was 87. Of these 40 were found on the
right side and 47 on the left. Text fig. A will indicate their dis-
tribution.

BENNET M. ALLEN

Table 3 serves to show for purposes of comparison the num-
bers of sex-cells found in different specimens of Amia.

TABLE 3
Number of sex-cells in Amia calva

A.GE

SPECIMEN

NUMBER

3P SEX-CELLS IN MESODERM

L. Total

Hours.

mm.

132

None

132

None

None 1

132

None

132

None

137

4 1 11

137

147

17 31

147

11 33

147

18 66

147

28 67

147

26 76

155

3.0

34 49

3.4

41 i 103

3.5

107

3.7

4.0

40 ■

5.0

6.6

6.0

7.0

7.6

9.1

11.4

15.0

16.0

23.7

102

This stage is a convenient starting point from which to proceed
in the study of earlier stages.

SEX-CELLS OF AMIA AND LEPIDOSTEUS 15

A7?iia 3.7 mm. total length. The conditions are, in the main, quite
similar to those found in the 4 mm. stage. In one of the two speci-
mens (B) in which the sex-cells were counted there were 92 sex-
cells in the mesoderm and 10 in the entoderm. Although this
total number of 102 is greater than the number found in the 4
mm. stage (87), yet, as shown in table 2, no significance is to
be attached to this on account of the great individual variation
in the number of sex-cells observed, not only in Amia, but also
shown by the author to be so obvious in the turtle, Chrysemys.
In A of this stage, 72 sex-cells were found, 42 on the right and
30 on the left side.

Amia 3.5 mm. total length. Two larvae of this stage were stud-
ied. It was rather difficult to measure the specimens accurately,
owing to the fact that the caudal portion of the body free from
the yolk has a strong ventral bend. It can be straightened out
only in later stages. The two specimens of this length were taken
from the same nest and both are distinctly younger than the pre-
ceding, yet they showed decided differences from one another in
the positions occupied by the sex-cells, probably owing to the
fact that this, in all likelihood, is the period of their most active
migration. In specimen A the sex-cells are quite numerous in
the portion of the lateral plate of mesoderm, which lies imme-
diately above the border of the subgerminal cavity. They occur
in fair numbers in the mesoderm between this region and a point
one-half the distance from this point to the median edge of the
lateral plate of mesoderm. Only three were found nearer the
median line than this. Of these, one had scarcely passed the
midway point, one was still some distance from the median edge
of the lateral plate, while one had actually reached that point.

In specimen B of this stage a large proportion of the sex-cells
have reached the median edge of the lateral plate of mesoderm
of each side. This is especially noticeable on the right side. The
conditions in this specimen approach those described for the 4 mm.
stage but do not show quite such an advanced condition, owing
to the fact that a larger proportion of sex-cells are scattered along
the outer portions of what we may call the sex-cell path. There

BENNET M. ALLEN

were noted two or three instances in which the sex-cells were
migrating from the peripheral entoderm into the mesoderm.

Amia, 3 mm., total length; 155 hours. In a specimen of 3.0 mm.
total length, the free caudal portion has but recently separated
from the vitelline mass, and has attained a total length of .56 mm.
By comparison with a number of embryos of 132, 137, and 147
hours old, the age of this embyro was estimated to be very close
to 155 hours. This estimate was made by counting the number of
sections passing through the posterior part of the embryo free from
tJie yolk mass. Sufficient numbers of embryos were used to give
a fairly accurate determination, there being seven specimens of
the 147-hour, three of the 137-hour, and two of the 132-hour
stages studied.

TABLE 4

The numbers of sex-cells in each were as follows:

RIGHT SIDE

LEFT SIDE

TOTAL

In A

39
59

In B

107

There were 49 sex-cells counted in the 3 mm., 155-hour embryo.
This, it will be seen, is decidedly below the average and yet the
number is greater than that found in the 5 mm. stage and in
the much later 16 mm. specimens.

Only two of the sex-cells have migrated a very short distance
along the lateral plate of mesoderm, beyond a point overlying
the lateral boundary of the subgerminal cavity; the remainder
of them all lie lateral to it. It will thus be seen that they show
a much earlier phase of migration than that observed in the 3.5
mm. embryo, not only as regards the number that have migrated
into the mesoderm, but likewise in the distance through which
they have travelled in their journey in that layer toward the sex-
gland anlagen.

Amia, 11^7 hr. stage. That there is a great amount of individ-
ual variation in the rapidity with which this migration from the
peripheral entoderm to the lateral plates of mesoderm is accom-
plished may be readily seen by referring to the numbers counted

SEX-CELLS OF AMIA AND LEPIDOSTEUS 17

in the mesoderm of seven specimens of the 147 hour stage. These
specimens were all taken from the same nest and kept in the
same dish, so there can be but very slight difference in their ages,
due, if it exists, to the small difference in the time at which the
eggs were laid. It will be seen that the total number of sex-
cells in the entoderm in these specimens varies from 22 to 76.
The latter number is not only greater than that observed in the
3 mm., 155 hour stage, but almost equals that counted in many
specimens of older stages after migration has been completed,
as, for instance, the 11.4 mm. and 15 mm. stages (see table).
In this stage clearly defined sex-cells can be seen in the peripheral
entoderm just below the lateral plates of mesoderm, figs. 16 and
17. These cells are distinguishable from the other entoderm
cells among which they lie, by the greater size of their contained
yolk granules as contrasted with the small size of the yolk granules
in the other cells that make up this layer. The difference is
further marked by the more rounded form of the sex-cells. Com-
parison of these sex-cells in the peripheral entoderm shows them
to be identical with other more clearly defined sex-cells in the
mesoderm. Of this identity there can be no question, and it is
equally clear, from a study of later stages, that these cells, having
once migrated into the lateral plates of mesoderm, pass unaltered
along the latter to come finally to rest in the sex-gland anlagen.
There can be no doubt about the origin of the sex-cells from the
entoderm. A number of cases were observed in which the sex-
cells were actually in process of passing from the peripheral ento-
derm into the lateral plates of mesoderm.

At this stage, sex-cells have a wide distribution in the periph-
eral entoderm, being scattered through a region extending
from a point opposite to the region where the blood cells originated
to the junction of the peripheral, sub-germinal and roof entoderm.
In three specimens of the 137 hour stage, conditions are quite
similar to the foregoing. In these embryos the number of sex-
cells ranged from 11 to 29. It will be seen that the maximum
number of sex-cells counted in this stage is greater than the
minimum number of the 147 hour stage, although in all three of
these 137 hour embryos, the caudal end of the embryo, that part

JOURNA.L OF MORPHOLOGY, VOL. 22, NO. 1

18 BENNET M. ALLEN

that has been Hfted off the yolk, is decidedly shorter than in any
of the 147 hour specimens.

Amia, 132 hr. stage. In four specimens of the 132 hour stage,
the caudal end of the embryo was just ready to undergo separa-
tion from the yolk. Only in one of them had this really com-
menced, the separated portion having reached a length of but
20)U. Not one of these four specimens showed a single sex-cell in
the mesoderm. There can be no question upon this point because
they could be very readily detected if present. In the 137 and
147 hour stages those that migrated into the mesoderm stand out
most clearly and sharply from the surrounding mesodermal cells.
The points of difference between the two kinds of cells are very
striking and unmistakable. The sex-cells on the one hand are
large, spherical, have sharply defined boundaries, and are filled
with large oval yolk grains; while the mesodermal cells are small,
flattened, syncytial, and contain a very few minute yolk granules.

It is very much more difficult to trace the earlier history of the
sex-cells in the peripheral entoderm, owing to the slight differences
that may be taken as criteria in distinguishing them from the
neighboring entoderm cells. Numbers of cells with all the
characteristics of sex-cells are found just beneath the anlagen
of the blood masses. This stage is just before the development
of blood vessels within the embryo, and the blood-forming cells
occur in the form of two sharply limited bands, one on each side
of the embryo and at some distance lateral to it. Here and
there, sex-cells are found in the peripheral entoderm, medial to
these areas; but clearly defined cases of this sort are rather rare
as compared with the large number seen in this region a little
later in the 147 hour stage. It is quite likely that many of these
sex-cells are overlooked at this stage owing to the fact that the
neighboring entodermal cells contain rather large yolk grains at
this time, while those seen in these cells in the 147 hour stage
are much smaller than at this stage.

It is quite possible that the sex-cells may migrate medially
in the entoderm from an entodermal source beneath the blood
anlagen to various points between this region and the edge of
the sub-germinal cavity. It is possible that a large proportion

SEX-CELLS OF AMIA AND LEPIDOSTEUS 19

of them may have developed in the peripheral entoderm through-
out this entire extent. On the other hand, it is also possible that
sex-cells may migrate up into this region from the central ento-
derm beneath.

We have traced the history of the sex-cells from the 4 mm.
stage where they are readily identified by any one who has had
any experience in observing these cells, back to the earliest stage
at which they are distinguishable in the entoderm. We shall
now follow them up to the period when they are enclosed in the
definitely formed sex-glands and finally to the stage at which
they are found to have begun to increase in number.

Amia 5 mm., total length. Passing from the 4 mm. stage to
the next represented in our series, 5 mm., we find that the sex-
cells have made but little progress in their migration toward the
median edge of the lateral mesodermal plates. The total number
of sex-cells counted in this stage was surprisingly small, being
43 as compared with 87 in the 4 mm. stage. This difference in
number is probably due to individual variation. The hind gut
has materially lengthened, being 1.3 mm. in length, compared
with .88 mm. in the 4 mm. stage. There has been a corre-
sponding increase in the length of the region over which the sex-
cells are distributed. In the 4 mm. stage they extend from a
point 0.06 mm., in front of the beginning of the hind gut, cau-
dally to a distance of 0.35 mm. In the 5 mm. stage that we are
considering, this region begins at the same point relative to the
hind-gut and extends caudally for 0.50 mm., one isolated sex-
cell being found at a distance of 0.57 mm. behind the cranial limit
of their distribution.

In the more caudal portion of this region the splanchnic and
somatic layers of mesoderm have begun to separate to form the
coelome. This separation does not at first lead to the formation
of a continuous cavity but rather to a series of isolated, some-
what rounded cavities. Further caudad, the coelome becomes
more and more completely developed, appearing as a large cav-
ity on each side.

Amia 6 mm., total length. At this time the first sex-cells appear
in the splanchnopleure just at the entrance of the hind gut.

20 BENNET M. ALLEN

The first sex-cells in the somatopleure are found in the sex-gland
anlagen a short distance (0.04 mm.) behind this point. The sex-
cells are distributed somewhat irregularly from the cranial end
of the hind-gut to a point 0.90 mm., caudad to this point and
there are a few scattering sex-cells still further caudad than this.

The coelome is apparent as a continuous cleft on either side
of the hind-gut along the entire extent of the region occupied
by the sex-cells. The majority of the sex-cells are to be found
in the dorso-medial extremity of the coelome, i.e., near the
root of the mesentery. A few lie lateral and ventral to the
intestine. The coelomic cleft has not as yet become wider than
the diameter of the average sex-cell and we consequently see
them usually bridging across it, fig. 18. In no case have they
penetrated into the somatic mesoderm as we find them doing
later. One sex-cell was found in the gut-entoderm, whither it
may have migrated from the mesoderm. It is, on the other hand,
quite possible for it to have migrated in the entoderm in the man-
ner of sex-cell migration in the turtle. This is a point of minor
significance and an occurrence which is at best very infrequent.

Amia, 7 mm., total length. Up to this time, the mesentery
has been only potentially present, the two lateral plates of meso-
derm being in contact above the gut-entoderm. Now, however,
we find that it has begun to elongate and become thin. This is
naturally correlated with the increase in the extent of the coelome,
fig. 19. Two well defined sex-cells are found in the gut-entoderm,
0.06 mm., cranial to the opening of the hind-gut. These are to
be interpreted in the same way as the cell in the entoderm men-
tioned above. The first sex-cell occurring in the mesoderm is
found 0.08 mm. caudad to the beginning of the hind-gut. The
sex-cells are distributed through a region extending from a point
immediately back of the opening of the hind-gut to a point 1.05
mm. behind it, with a few scattering ones behind these. The
total number of sex-cells is 74.

Amia, 9.1 stage. Sex-cells first appear .18 mm. cranial to
the opening of the hind-gut. They extend from this point to a
point 1.59 mm. caudad to this, giving a total extent of 1.67 mm.
The total number of sex-cells counted at this stage amounted to

SEX-CELLS OF AMIA AND LEPIDOSTEUS 21

76. Of these all were in the sex-gland anlagen except three; one
of which occurred in the gut-entoderm and two in the parietal
peritoneum. I am inclined to consider it unlikely for these mis-
placed sex-cells to reach the sex-glands. One is struck, however,
with the great difference in the relative number of misplaced sex-
cells in Amia as compared with Lepidosteus. This may be
apparent rather than real, owing to the possibility that in Amia
large numbers of them may have failed to migrate from the ento-
derm into the mesoderm during early stages. Owing to the diffi-
culty of certainly distinguishing sex-cells in the entoderm from or-
dinary entoderm cells, it was quite impossible to make any count
of those left behind in migration. All but a very few, however, that
reach the mesoderm succeed, as we have seen, in reaching the sex-
gland anlagen. A considerable number of cells seen in the ento-
derm in later stages contain small yolk spherules and show other
points of resemblance to sex-cells. In this stage the mesentery
has become quite lengthened and the coelome very large. The
sex-cells have penetrated into the root of the mesentery, fig.
20.

The sex-cells, with rare exceptions, still contain large quanti-
ties of yolk material. In these exceptional cases a finely granular
appearance gives at least the suggestion of small unstained yolk
spherules. The yolk appears in the shape of particles varying
in size from small granules up to large lemon-shaped pieces quite
as large as those with which the cells of the yolk entoderm are
so completely filled.

Amia 11. j^ mm., total length. The sex-cells are fairly numerous
over a region 1.85 mm. in length, beginning at a .point 0.06 mm.
back of the yolk stalk and ending at a point 0.85 nam. cranial
to the cloaca. Two isolated sex-cells are found caudad to the
point named, one of them occurring very close to the cloaca.
Their total number in this embryo is eighty- two. The sex-cells
have much the same characteristics as in the previous stage.

This stage is marked by a decided increase in the length of the
mesentery and by a decrease in the size of the yolk-sac, which is
now but 0.7 mm. in diameter and is greatly hollowed out to
form a portion of the intestinal wall.

22 BENNET M. ALLEN

While the sex-cells of the 9.1 mm. stage are imbedded in the
mesoderm at the root of the mesentery and always close to the
median line, they are found in the 11.4 mm. stage to occupy a
position a short distance on each side of this point. Not only
have they moved laterally, but they have also protruded into the
body cavity, accompanied by a few mesoderm cells which are
intercalated between them, fig. 21, and surround them with a
thin peritoneal investment as well.

Amia 15 mm., total length. In this stage the sex-cells extend
over a distance of 2.70 mm. in the caudad 0.50 mm. of which
they are very sparse. The sex-glands protrude further into the
body cavity than in the preceding stage, and the ligament of
attachment becomes narrower. The genital ridge is very much
lower in the gaps between sex-cells than it is in the sex-cell
regions. In spite of the fact that it may be very low for quite
a distance, it is continuous throughout. The genital ridges
diverge quite widely at their cranial ends, approaching the median
line at a point .4 mm. caudad to their point of commencement.

The sex-cells have almost uniformly used up their contained
yolk material, although a few scattered ones are still closely
packed full of them. The sex-cells in specimen A, numbered 28
on the right side and 49 on the left, the total number being 77.
The number of sex-cells in specimen B was 38 on the right side
and 45 on the left, the total being 83.

Amia 16 mm. long. In two 16 mm. larvae, conditions very
similar to those of the 15 mm. stage were found. None of the
sex-cells contained yolk material in a sufficiently large amount
to be clearly recognizable. The striking thing about these two
specimens is the very small number of sex-cells present, 33 in one
case and 39 in another. There is no indication of degeneration
or of a failure to migrate to the proper positions.- The case
seems to be similar to one cited in Chrysemys, both being due to
individual variation.

These two specimens were taken from the same brood and no
doubt had the same parentage. Another 16 mm. specimen taken
from a different brood showed 09 sex-cells, a number not
very far below the maximum. From this fact, and from the

SEX-CELLS OF AMIA AND LEPIDOSTEUS

total absence of any indication of degeneration of sex-cells in these
or earlier stages, I feel convinced that this small number does not
indicate any tendency to degeneration of sex-cells.

Amia, 23.7 mm. total length. In the next stage studied, 23.7
mm., the sex-cells numbered 102. Here again there is no evidence
of a change in the number of sex-cells originally present. The
number, although somewhat high, is exceeded by some of the
specimens of very much earlier stages. There is no evidence of
sex-cell division nor of any degeneration.

Amia, 1^0 mm. total length. At this stage the sex-gland is
elongated oval in transverse section. It has become bent over
in such a way that the proximal edge is medial and the free edge

TABLE 5

Dimensions of sex-cells of Amia

STAGE

NUCLEUS

CELL BODY

Hours

137

7.10

18.03

147

6.71

18.70

mm.

3.7

6.45

21.88

5.0

6.51

17.80

9.1

8.00

14.96

11.4

7.48

11.59

15.0

7.48

12.64

16.0

7.74

14.06

23.0

7.22

14.20

lateral in position. The mesodermal cells have increased greatly
in number. The peripheral cells have become arranged into a
somewhat poorly defined layer, while the sex-cells lie in the in-
terior of the sex-gland. No attempt was made to determine the
time at which the sex-cells begin to divide, or to study the further,
development of the sex-glands.

Measurements of the nuclei and cell bodies of the sex-cells
gave the following averages, two diameters being measured in
each of five sex-cells chosen at random in each stage.

Although the number of cells measured in each stage is hardly
sufficient to justify one in considering these average dimensions

24 BENNET M. ALLEN

to have any high degree of accuracy, I feel that we are quite
justified in concluding from these figures that: (1) there is a fair
decrease in the size of the cell-body as development proceeds,
and (2) that there is a slight increase in the size of the nucleus.
The decrease in the size of the cell-body is probably due to
the absorption of the yolk material with which the sex-cells are
so richly filled during the earlier stages. No good explanation to
account for the slight apparent increase in size of the nucleus
presents itself.

DISCUSSION OF RESULTS

We can not consider this work as completed without making
a comparison between the sex-cells and the other cells of the
embryo. This subject will first be taken up in Amia where we
have traced the sex-cells back to earlier stages than in Lepidosteus.
It has already been pointed out that the sex-cells, as first seen in
the peripheral entoderm, are to be distinguished only by the size
and arrangement of the yolk spherules. The nuclei bear a close
resemblance to those of surrounding cells of the same size, while
the larger nuclei of larger cells show many points of similarity to
them. In all except the earliest stages studied, these nuclei are
quite rounded. The chromatin appears in the form of slender
strands that take a peripheral position in the nucleus. There
is invariably a plasmosome present and rarely two of them. In
the 147 hour stage the nuclei of the sex-cells bear a resemblance
not only to those of the neighboring cells but also to those of the
gut entoderm. In fact, many nuclei of the mesoderm show simi-
lar characteristics.

After development has gone a little further, as in the 3.4 mm.
and 4 mm. stages, the mesodermal nuclei and those of the gut
entoderm are found to have become smaller and are more deeply
stained than those of the sex-cells and peripheral entoderm. In
all of these later stages, which include 5 mm., 6 mm., 9. 1 mm., 11.4
mm. and 16 mm. larvae, these differences are found to increase.
Although the sex-cells undergo a migration from the peripheral
entoderm into the lateral plates of mesoderm and through the
latter to the sex-gland anlagen, they still bear a close resemblance

SEX-CELLS OF AMIA AND LEPIDOSTEUS 25

to certain cells of the peripheral entoderm. This not only in-
volves a similarity of the nuclei but of the dimensions of the cell
bodies. This is true even after the sex-cells and the correspond-
ing cells of the peripheral entoderm have lost their yolk through
absorption.

In the stage of 11.4 mm., the yolk mass has been greatly
reduced (figs. 25 and 26). Only here and there about its periph-
ery are cells to be found with well defined outlines. The great
mass is syncytial, with large nuclei of varying size scattered here
and there. While these nuclei of the vitelline mass are much
larger than the sex-cell nuclei, they bear a close resemblance to
the latter. The nuclei of the well defined peripheral cells are prac-
tically identical in size and appearance with those of the sex-cells.

While the similarity between sex-cells and between these two
classes of cells is not so marked in Lepidosteus as in Amia, yet
it appears to be equally true. In the 17 mm. stage (figs. 14 and 15)
the yolk mass is still of fair size. There is a layer of peripheral
entoderm that is largely made up of cells with clear boundaries,
whose nuclei are similar to those of the sex-cells in respect to the
presence and character of the plasmosome and in the form and
distribution of the chromatin material. In many cases these
nuclei are larger than those of the sex-cells; but many are found
which are quite as small. These grade into the very large
nuclei of the syncytial vitelline entoderm.

At this stage the tissues of the body have taken on their dis-
tinctive characters and their component cells have undergone in
many cases a high degree of specialization. This emphasizes
strongly the similarity between the sex-cells and the cells of the
peripheral entoderm.

As we pass back to earlier stages, such as those of 9.3 mm.,
5.9 mm., etc., we still find this similarity between these types of
cells, although the nuclei of all the body cells tend to show greater
and greater similarity to one another in the earlier stages. For
instance, it becomes quite difficult to distinguish the nuclei of the
gut entoderm cells from those of the sex-cells. Even the nuclei
of the Wolffian ducts show quite a close resemblance to the sex-
cell nuclei during the early stages of development.

26 BENNET M. ALLEN

There are two ways of viewing the similarity that the sex-
cells of Amia and Lepidosteus bear to these cells of the peripheral
entoderm. The well defined cells of the peripheral entoderm
might be interpreted as sex-cells that have failed to migrate into
the lateral plates of mesoderm. It would then remain to give an
explanation of the resemblance that the nuclei of these cells bear
to the nuclei of the vitelline entoderm and to account for the
intermediate types of nuclei by which they grade into one another.

The other view of this problem is to consider sex-cells, periph-
eral entoderm cells, and vitelline entoderm cells as slightly differ-
entiated blastomeres, dating from an early stage of development,
and to consider the similarity that they bear to the cells of
the peripheral entoderm as due to the fact that they too have
remained in a relatively slightly differentiated condition. This
view seems the more probable of the two. It is by no means a
new one, having been advanced by Nussbaum in 1880.

It would be rash in the extreme to claim that the sex-cells
might not differ in some essential chromosomal characters from
the cells of the peripheral entoderm which they so closely resemble,
and yet careful study has failed as yet to show any real differences.
While such differences may exist, these cells all have much in
common with one another.

In a recent paper by A. P. Dustin ('07), this author gives a new
view of the origin and movements of the sex-cells of Triton alpes-
tris, Rana fusca and Bufo vulgaris. Since his view is so greatly
at variance with my own, it will be necessary to review this work
in some detail. He begins with an account of the sex-cells of
Triton, and stress is laid upon this form, the author showing a
strong tendency to bring his studies upon Rana and Bufo into
line with his work upon Triton.

He first recognizes the anlage of the sex-cells in the medial
portions of the lateral plates of mesoderm in the 3 mm. larva of
Triton. They occur only in the caudal half of the body and
involve only those parts of the lateral plates of mesoderm lying
medial to the Wolffian ducts. In the early stages these cells are
filled with large yolk spherules and do not greatly differ from the
mesodermal cells that surround them. At a later period the sex-

SEX-CELLS OF AMIA AND LEPIDOSTEUS 27

cell anlagen are pushed together in the median line, between the
aorta and the roof of the archenteron. They fuse into a median
longitudinal rod of cells lying just above the dorsal root of the
mesentery. By this time the sex-cells have lost their yolk
material and have, to a large extent, assumed their definitive
character. During these stages the number of the sex-cells has
increased from one hundred to one hundred and fifty, occasional
mitoses being observed. Soon after this stage of the median
anlage (9 mm.) has been reached, the sex-cells migrate laterally
to their final positions on each side of the root of the mesentery.
At the stage of 14 mm., a large number of them degenerate, leav-
ing only 60. A second generation of sex-cells soon begins to form
from a source entirely different from the first, namely, from a
transformation of ordinary peritoneal cells. Dustin is, in this
regard, quite in accord with Bouin who expressed similar views
regarding Rana. Dustin considers somewhat more briefly the
corresponding stages in Rana and Bufo. Here he finds what he
considers to be a substantially similar source of origin of the sex-
cells, namely the medial borders of the lateral plates of meso-
derm. An incredible feature of his account is the statement that
the lateral sex-gland anlagen contain no sex-cell at all comparable
in size to those of the yolk-filled entoderm, at the period imme-
diately prior to their union in the median line. Dustin would
have us believe, nevertheless, that these selfsame sex-cells show
a close resemblance to the entoderm cells immediately after this
union of the lateral anlagen, and this in spite of the fact that both
of these stages of development are so close together that the
embryos upon which he made these observations were all of the
same length. His own statement is as follows :

" Au moment ou les ebauches paires separees par une sorte de clivage
des lames laterales du mesoblaste se sont rapprochees de la ligne mediane,
les cellules sexuelles futures passent par une serie de transformations
cytologiques a la suite desquelles elles auront presque les caracteres des
cellules de I'hypoblaste vitellin. Les dimensions des corps cellulaires
augmentent dans de fortes proportions; les grains vitellins deviennent
beaucoup plus nombreux et plus volumineux; ils se colorent mieux par
I'orange G. Par le fait de I'augmentation du nombre des plaquettes

28 BENNET M. ALLEN

vitellines, le noyau, souvent refoule a la peripherie de la cellule, pre-
sente a sa surface une serie d'encoches lui donnant un aspect herisse
(p. 476).

He finds the number of sex-cells in Rana to increase gradually,
from 75 in the 8 mm. stage to 90 in the 15 mm. stage, at which
time sex-cells begin to be formed by the transformation of ordi-
nary peritoneal cells. Simultaneously there is a degeneration
of sex-cells which is overbalanced by this process of transforma-
tion.

In criticism of the above views I wish, first of all, to admit the
possibility that Dustin may be perfectly correct in his account
of the origin of the first line of sex-cells from the lateral plates
of mesoderm in Triton. His account of this feature is circum-
stantial and rather convincing. His account of a transformation
of peritoneal cells into sex-cells during later stages is by no
means so easy of acceptation. His figures to demonstrate this
are not convincing.

His counts of sex-cells are not given in any circumstantial
detail and there is no indication as to whether the number of sex-
cells recorded for any given stage is the result of a count of the
sex-cells in one specimen or in several. One can not be blamed
for being skeptical of the value of such counts if made upon but
one specimen of each stage, when so few stages are chosen to
demonstrate general processes of degeneration and new formation.
Such a process can only be established by a count of the sex-
cells of numerous specimens.

I wish to express my complete disbelief in the first appearance
of the sex-cells in the lateral plates of mesoderm of Rana and
Bufo in the manner described by Dustin. In my paper upon
''An Important Period in the History of the Sex-Cells of Rana
pipiens" ('07) I showed that the sex-cells migrate upward from the
median dorsal portion of the gut entoderm at the time when the
two lateral plates are pushing together to the median line in
the process of forming the mesentery. Attention was called to
the resemblance that this process bears to an actual pinching
off of the mass of sex-cells by the inner margins of the plates of

SEX-CELLS OF AMIA AND LEPIDOSTEUS 29

mesoderm. As pointed out in my article, the lateral plates of
mesoderm, examined immediately before their approximation in
the median line, show no cells which, as regards size or yolk con-
tent, in the least compare with the sex-cells.

It is especially gratifying to me to find support for my views
in two recent papers. In one of these Kuschakewitsch ('08),
referring to my paper of a few months before, stated: ''Der Ver-
fasser hat die Abschniiring von Dotterzellen langs der dorsalene
Sagittallinie des Dottersackes im hinteren Telle des Rumpfes
beobachtet und die Theilname dieser Dotterzellen am Aufbau
einer kompakten Mesenterial-anlage festgestellt, die Bouin
(1900)) als ''ebauche genitale primordiale" aufgefasst hatte.
Wie aus meiner Schilderung der entsprechenden Vorgange in
der Normalreihe von Rana esculenta zu ersehen ist, kann ich die
Angaben von Allen voUstandig bestatigen."

Another paper, appearing the same year (King, '08), gives an
account of the origin of the sex-cells in Bufo lentiginosus which
is in complete accord with the above, and states : "Allen's recent
account of the origin of the sex-cells in Rana pipiens agrees
essentially with what I have found in Bufo." Miss King finds
no evidence in the course of development of any transformation
of peritoneal cells into sex-cells as asserted by several writers
among whom may be mentioned Bouin and Dustin. This is
quite in accord with my observations upon Chrysemys ('06) in
which the sex-cells were traced to the period of sexual maturity
without finding any evidence of such transformation.

Miss May Jarvis ('08) in a paper upon ''The Segregation of the
Germ-Cells of Phrynosoma cornutum" (preliminary note) finds
the sex-cells to take their origin in the entoderm of the vascular
area on all sides of the embryo, even cranial to it, and notes a
few in the region of the brain. Her results are in their main
features confirmatory of my own work upon Chrysemys. The
following quotation from her paper is self-explanatory: ''Through
the courtesy of Dr. Allen, I have been enabled to examine the
more important stages in the migration of the germ-cells of
Chrysemys; they are similar to my own material, as my conclu-

30 BENNET M. ALLEN

sions, although differing from Dr. Allen's in details of early dis-
tribution and periods of migration, uphold his."

Rubaschkin ('08 and '09) in a couple of recent papers, has
shown that the sex-cells of the rabbit and guinea-pig are first to
be found in the entoderm at some distance on each side of the
hind-gut and that they follow a path almost identical with that
followed by the sex-cells of Chrysemys. These references to the
coincidence of the views of other recent writers with my own are
made to show that I do not stand alone in placing emphasis upon
the entodermal origin of the sex-cells in the vertebrates. At the
same time I wish, however, to disclaim any intention of making
at this time a sweeping claim that the sex-cells of all vetebrates
arise in the entoderm. Wheeler's work on Petromyzon ('99)
shows that they may be included in the mesoderm at the time
when that layer is split off from the entoderm. He has, however,
pointed out their similarity to the entoderm cells and their dis-
similarity to the mesodermal cells among which they lie.

I do not seek to discredit the work of Dustin upon the sex-
cells of Triton; although his statements about the origin of the
sex-cells in Rana and Bufo strike me as being very far from the
mark, because they are so radically at variance with not only
my own observations, but with those of King and Kuschake-
witsch as well. Dustin, in his attitude toward the work of others,
seems to consider that there must be a strict uniformity in all
forms in both the place of origin and in the movements of the sex-
cells. He has apparently studied this problem first in Triton
and at some length. His results, probably correct for that form,
he has attempted to apply to Rana and Bufo as well, undeterred
by the difficulties to which attention was called above. Dustin
is quite ready flippantly to dismiss my work upon Chrysemys,
because the results there expressed did not coincide with the
views that he had formed regarding the origin of the sex-cells in
Triton, Rana, and Bufo.^ The process of migration through the
entoderm is so clear in Chrysemys, that it is unmistakable.
The sex-cells are not only characterized by their larger size,

1 See postscript.

SEX-CELLS OF AMI A AND LEPIDOSTEUS 31

definite, rounded outlines and fine chromatin network, but by
their large yolk content and the fact that they do not divide
during the stages in dispute.

The sex-cells are migratory to a high degree. The path and
time of their migration may vary greatly within a given group
of animals, as illustrated by the case of Amia and Lepidosteus.
While in the forms that I have studied they are first to be ob-
served in the entoderm, I am quite open to conviction that in
other forms they may migrate from this layer into the potential
mesoderm before the two layers are separated, as shown by
Wheeler in Petromyzon. It is even conceivable that they may
lie, from the very beginning of development, in material destined
to form mesoderm — that they may never have existed among
cells actually or potentially entodermal. The more recent de-
velopment of our work along these lines, however, most cer-
tainly tends to show that it is usual among the vertebrates for
the sex-cells to first appear in the entoderm.

SUMMARY AND CONCLUSIONS

1. The sex-cells of both Amia and Lepidosteus have their
origin in the entoderm. In Amia they are first distinguishable
in the peripheral entoderm from the lateral angle of the subgermi-
nal cavity to the anlage of the blood cells.

In Lepidosteus they are first seen in the ventral and lateral
portions of the gut-entoderm, although analogy with Chryse-
mys leads us to assume that they may have migrated through
the entoderm to these regions from more lateral anlagen, similar
to those from which the sex-cells of Amia arise. In both forms,
the sex-cells arise only in the region of the hind-gut. None were
found at any considerable distance in front of it.

2. The path of sex-cell migration in Amia carries them out
of the peripheral entoderm directly into the overlying lateral
plates of mesoderm, along which they travel, to come to rest
near the medial edges of the latter. These portions are destined
to join above the intestine to form the mesentery. As the
splanchnic and somatic layers of the lateral plates of mesoderm

32 BENNET M. ALLEN

split to form the coelome, the sex-cells adhere to the somatic
layer at a point near the root of the developing mesentery — the
sex-gland anlage. They later sink into the peritoneum of this
region, which afterwards proliferates to form a long ridge — the
sex-gland. Very few sex-cells fall by the wayside in this migra-
tion, practically all reaching the sex-glands.

3. In Lepidosteus the sex-cells, first seen in the ventral and
lateral portions of the gut-entoderm, migrate to occupy a position
in the dorsal portion of it, from which they pass dorsally into
the loose mesenchyme that forms the substance of the developing
mesentery. As the mesentery becomes more narrow and com-
pact, owing to the increase in size of the body cavity, the sex-
cells migrate to its dorsal portion and laterally to the sex-gland
anlagen. Roughly speaking, one-half of the total number of
sex-cells reach the sex-gland anlagen, the remainder being dis-
tributed between the intestinal entoderm, the mesodermal layers
of the intestine, the mesentery and the tissues at and dorsal to
the root of the intestine.

4. The number of the sex-cells in Amia and Lepidosteus is a
matter of individual variation for those periods of development
during which they do not undergo division. The average number
in Amia, after the period when the migration from the entoderm
to the mesoderm has been completed, up to the latest stage in
which counts were made, was found to be 75. In Lepidosteus it
was 765, an average of 636 of these occurring in the mesoderm.

5. There is a close resemblance between the nuclei of the sex-
cells and of the yolk cells. This is especially true of certain
cells of the peripheral entoderm, although these grade by gradual
transition forms into the large nuclei of the vitelline entoderm.
This is probably due to the fact that both types of cells have
undergone but little differentiation in the course of development.

POSTSCRIPT

A few days before proof of this article came to hand, I received,
through the courtesy of the author, a reprint of an article by A. P.
Dustin, entitled, "L'Origine et I'Evolution des Gonocytes chez

SEX-CELLS OF AMIA AND LEPIDOSTEUS 33

les Reptiles," (Archives, de Biologie, 1910). This article deals
with the origin of the sex -cells in Chrysemys marginata, the form
which served as a subject for my own work of 1906. As noted
above, Dustin in his paper "Recherches sur Torigine des gono-
cytes chez les Amphibiens" 1907, exhibited scant respect for my
work on the sex-cells of Chrysemys. It was, no doubt, in large
part, this feeling that prompted him to repeat my work. While
he, no doubt, expected to find in this form a confirmation of his
previously expressed views, he is led to substantiate completely
my statements regarding the entodermal origin of the sex-cells.
He traces them along the same migration path that I demon-
strated four years before. For all this he now gives me full credit
and support; but takes issue with my statements regarding the
distribution of the sex-cells prior to their migration into the em-
bryo, and, furthermore, claims to have -evidence to show that
there is a new formation of sex-cells, due to a transformation of
ordinary peritoneal cells. These points of controversy and
certain other minor ones can not be considered here, but I promise
a full discussion of them in another place. I may say that I am
fully prepared to maintain my views upon all of the points at
issue.

On my part, the work that I have carried on upon Necturus
since this paper was written, has given me results quite similar
to those at which Dustin arrived in his work upon Triton. I
may say that preliminary studies have convinced me that the
sex-cells arise in an essentially similar manner in Amblystoma.
We then see that, in all three of these urodeles, the sex-cells
arise from the inner edges of the lateral plates of mesoderm. I
owe it to myself to call attention to the fact that I have at no
time disputed the accuracy of Dustin's work upon Triton. While
the evidence seems to me quite clear that this is the usual, if
not the universal, mode of origin of the sex-cells among the uro-
dele amphibians, I am ready to maintain with equal vigor the
entodermal origin of the sex-cells in the aruran amphibians,
at the same time admitting the possibility that exceptions to
this apparent rule may be discovered. I do not feel however,
that Dustin has proved his case in Rana fusca and Bufo vul-

34 BENNET M. ALLEN

garis. The discussion of his work above gives the reasons for
my position in this matter.

Not only does it seem probable that the sex-cells arise during
early stages in the mesoderm of the urodeles, but this seems to be
the case in the teleosts as well. The most recent and satisfac-
tory support of this view is contained in the excellent paper of
Dr. Gideon S. Dodds upon the ''Segregation of the Germ-Cells
of the Teleost, Lophius, " in the Journal of Morphology, 1910.
Here again, we must urge caution in forming a sweeping general-
ization from the facts thus far at hand. There is certainly a
wide field for work in the study of the origin of the sex-cells of
the vertebrates. It is a subject which should be approached in
a spirit of broad toleration for the views of others. The sex-
cells are cells that retain their early embryonic character after
the somatic cells have undergone specialization. It seems,
from a number of observations made by different authors, that
in most forms the sex-cells first make their appearance in the
entoderm — the germ layer whose cells appear to maintain their
primitive embryonic characters longer than do those of the other
germ layers. At the same time, unimpeachable evidence shows
that this apparently logical process is not universal, and I have
at no time claimed that it is. The sex-cells, as show^n by Nuss-
baum, Eigenmann, Beard and others, do not belong to any one
germ layer, but are, in a sense at least, independent of the som-
atic tissues. They are free to follow their own path in their
travels from the place of origin to the sex-gland anlagen, where
they finally come to rest. While this path is no doubt identi-
cal or similar in closely allied species and in more general divi-
sions of the vertebrates, I do not feel that we are justified in at-
tributing a high degree of phylogenetic importance to the different
steps in the migration paths through which they travel.

I wish to express my indebtedness for the work of our depart-
mental artists. Misses Hedge and Battey. I am indebted to Miss
Hedge for the execution of diagrams 1-6 and for figs. 9, 10, 14,
15, 21, 22, 25 and 26; and to Miss Battey for figs. 11, 12, 13, 23,
and 24. The remaining drawings are my own.

SEX-CELLS OF AMIA AND LEPIDOSTEUS 35

BIBLIOGRAPHY

Allen, Bennet M. 1906 Origin of the sex-cells of Chrysemys. Anat. Anz.
Bd. 29.

1907a A statistical study of the sex-cells of Chrysemys marginata,
Anat. Anz. Bd. 30.

19076 An important period in the history of the sex cells of Rana
pipiens. Anat. Anz., Bd. 31.

1909 The origin of the sex- cells of Amia and Lepidosteus. Anat.
Rec, Vol. 3.

DusTiN, A. p. 1907 Recherches sur I'origine des gonocytes chez les Amphi-
biens. Arch, de Biologie, tome 23.

Jarvis, Mat. 1908 The segregation of the sex-cells of Phrynosoma. Biol.
Bui., Vol. 15.

King, Helen Dean. 1908 The oogenesis of Bufo lentiginosus. Jour. Morph.,
Vol. 19.

KuscHAKEwiTSCH, S. 1908 Ueber den Ursprung der Urgeschlechtszellen bei Rana
pipiens. Stzber. math. phys. Klasse, k. bayer. Akad. Wiss., Bd. 38.

RuBASCHKiN, W. 1907 Zur Frage von der Entstehung der Keimzellen bei Sauge-
tierembryonen. Anat. Anz., Bd. 31.

1909 Ueber die Urgeschlechtszellen bei Saugetieren. Anat. Hefte, Bd.
39.

Wheeler, W. M. 1899 The development of the urogenital organs of the lamprey.
Zool. Jahrbuch., Anat. Abth., Bd. 13.

ABBREVIATIONS FOR ALL FIGURES

Arch., Archenteron

Coel., Coelomic cavity

Ect., Ectoderm

Gut End., Gut entoderm

Int., Intestine

Lat. Mes., Lateral plate of mesoderm

Mes., Mesentery

Meson., Mesonephros

Myo., Myotome

Nolo., Notochord

P. Card., Post cardinal vein

Periph. End., Peripheral entoderm
Roof End., Roof entoderm
S. €., Sex-cells
S. GL, Sex-gland

Sub-Germ. Cav., Sub-germinal cavity
Sub-Germ. End., Sub-germinal ento-
derm
Siv. Bl., Swim bladder
Vit. End., Vitelline entoderm

Wolff. D., \
W. D., J

Wolffian duel

(36)

PLATE 1

EXPLANATION OF FIGURES

1 Diagram to show the migration path of the sex-cells in Chrysemys mar-
ginata.

2 Diagram to show the migration path of the sex-cells in Rana pipiens.

PLATE 2

EXFLANAIION OF FIGURES

3 Diagram to show the migration path of the sex-cells in Lepidosteus osseus.

4 Diagram to show the last phase of the migration of the sex-cells in Lepi-
dosteus osseus.

5 Diagram to show the migration path of the sex-cells of Amia calva.

6 Diagram to show the last phase of the migration of the sex-cells in Amia
calva.

JOURNAL OF MORPHOLOGT, VOL. 22, NO. 1

PLATE 1

THE ORIGIN OF THE SEX-CELLS OF AML\ AND LEPIDOSTEUS

BENNET M. ALLEN

Rana

JOURNAL OF MORPHOLOGY, %'OL. 22, NO. 1

THE ORIGIN OF THE SEX-CELLS OF AMIA AND LEPIDOSTEUS

BENNET M. ALLEN

Lepidosteus

Peri ph. End

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1

it. End.

PLATE 3

EXPLANATION OF FIGURES

7 Transverse section through the hind-gut of an 8.6 mm. larva of Lepidos-
teus osseus. X 300.

8 Transverse section through the hind-gut of a 9.3 mm. larva of Lepidos-
teus osseus. X 300.

9 Transverse section through the hind-gut of a 10.7 mm. larva of Lepidosteus
osseus. X 300.

10 Transverse section through the hind-gut of a 14.1 mm. larva of Lepidosteus
osseus. X 300.

11 Transverse section through the hind-gut of a 17 mm. larva of Lepidosteus
osseus. X 300.

SEX CELLS OF AMLV AND LEPIDOSTEUS
EENNK.T M. A1.M:X

ERRATA

Gelatin plates 1, 2 and 3 should have been numbered 3, 4 and 5

JOURNAL OF MORPHOLOGY, VOL. 22. NO. 1

PLATE 3

EXPLANATION OF FIGURES

SEX CELLS OF AINHA AND LEPIDOSTEUS
F.ENNET M. ALLEN

JOURNAL OF MORPHOLOGY, VOL. 22. NO. 1

PLATE 4

EXPLANATION OF FIGURES

12 Transverse section of the rudimentary sex-glands of a 24 mm. larva of
Lepidosteus osseus. X 300.

13 Transverse section of a sex-gland of a 110 mm. specimen of Lepidosteus
osseus. X 300.

14 Part of a transverse section of a 17 mm. larva of Lepidosteus osseus, show-
ing the reduced vitelline mass.

15 Detail drawing of a portion of the vitelline mass of the above section.
X 300.

16 Transverse section through the region immediately lateral to the pos-
terior portion of the sub-germinal cavity of a 147 hr. embryo of Amia calva. X
300. This shows the place of origin of the sex-cells.

17 Section passing similarly through another specimen of the same stage of
Amia calva. X 300.

One sex-cell shown as it is pushing up into the mesoderm.

18 Transverse section through the hind-gut of a 6 mm. larva of Amia calva.
X 300.

19 Transverse section through the hind-gut of a 7 mm. larva of Amia calva.
X 300.

SEX-CELLS OF AML\ AND LEPIDOSTEUS

KENNF.T M. AI.I.EX

Periph. EncL.jg.^

«ir.-J?>Vsw B,, ^f:'^ ^1^ r^x

JOURNAL OF MORPHOLOGY, VOL.23, NO. 1

PLATE 5

EXPLANATION OF FIGURES

20 Transverse section through the hind-gut of a 9.1 mm. larva of Amia calva.
X 300.

21 Transverse section through the hind-gut and sex-gland anlage of an 11.4
mm. larva of Amia calva. X 300.

22 Transverse section through the young sex-glands of a 16 mm. larva of Amia
calva. X 300.

23 Sketch to show the orientation of the sex-glands in q 40 mm. specimen
of Amia calva as seen in transverse section.

24 Detail drawing of the sex-gland as seen in above sketch. X 300.

25 Drawing to show the orientation of the much reduced vitelline mass of
an 11.4 mm. larva of Amia calva.

26 Detail drawing of a portion of the vitelline mass indicated above. This
shows the resemblance that certain cells of the peripheral entoderm show to sex-
cells of this stage. X 300.

SEX-CELLS OF AMLV AND LEPIDOSTEUS
BENNET M. ALLEN

PLATE 3

JOURNAL OF MORPHOLOGY, VOL. 32, NO. 1

THE CYCLIC CHANGES IN THE OVARY OF THE
GUINEA PIG

LEO LOEB

From the Laboratory of Experimental Pathology of the University of Pennsylvania,
and from the Pathological Laboratory of the Barnard Skin and Cancer Hospital,
St. Loxds, Mo.

In the course of an experimental investigation into the causes
of the cychc changes taking place in the uterine mucosa and into
the factors underlying the formation of the maternal placenta in
mammals, we observed that cyclic changes in the structure of
the ovary correspond to the uterine cycle. It has of course been
known that at certain times ovulation takes place in the mamma-
lian ovary, and furthermore, changes have been described as occur-
ring in the ovarian follicles of certain mammals in connection
with copulation and during pregnancy; but the cyclic changes
taking place in the ovary 'quite independently of copulation
and of pregnancy and merely dependent upon ovulation have,
as far as we are aware, not yet been recognized. While
we know of no publication dealing with the cyclic changes
in the ovaries in general, a valuable study of the changes taking
place during pregnancy in two species of Insectivores and in one
species of Lemurid has been made by C. H. Stratz.i This author
comes to the conclusion that in the period following copulation
all the ovarian follicles become atretic; that during pregnancy
small follicles are formed but also become atretic before they can
develop; that only towards the end of pregnancy the follicles
begin to grow to a considerable size, and that they reach the stage
of maturity during the puerperium.

Stratz was not in a position to determine in an exact manner
the time elapsed since the last copulation of the animals the ova-

^ C. H. Stratz: der geschlechtsreife Saugethiereierstock. Haag. 1898.

38 LEO LOEB

ries of which he examined. He also seems to have examined a
relatively very limited number of ovaries of animals during the
different stages of pregnancy, and furthermore he studied only
certain parts of each ovary. A methodical study of ovaries of
non-pregnant animals was not undertaken. While his obser-
vation that after copulation all follicles become atretic is ap-
proximately, but not altogether correct, as far as its general
validity is concerned, in the guinea pig the processes taking place
in the ovaries during the subsequent stages differ from the con-
ditions described by Stratz in the case of Tupaja, Sorex and
Tarsius.

Furthermore Stratz does not recognize the essential factor
upon which the cyclic changes in the ovaries depend. The con-
clusions in the last chapter of his publication show this clearly.

He summarizes as follows: If we find all follicles atretic, the
animal has been pregnant. If at the same time a new corpus
luteum is present, we have to deal with an early stage of preg-
nancy. If we detect some normal follicles, besides numerous
atretic follicles and a new corpus luteum, we have to consider a
puerperal condition of the animal. A large number of atretic
besides a few normal follicles also* suggests a puerperal state.

These general conclusions are not justified; the changes of
the follicles do not, as Stratz assumes, depend upon pregnancy,
and if we should attempt to use the criteria given by Stratz in
the case of guinea pigs and mammals in general we would be
liable frequently to make mistaken diagnoses. Notwithstanding,
these necessary criticisms, the work of Stratz is very valuable
and it advanced to a considerable extent our knowledge of the
ovaries.

Since his publication no more detailed investigation into the
processes taking place in the ovaries under various conditions
has appeared, as far as we are aware. Within recent years, how-
ever, the question has been raised whether a new ovulation can
take place during pregnancy.

We limited our investigations to the study of the ovary of the
guinea pig. We examined several hundred pairs of ovaries of
animals in which the period of the sexual cycle at which the ovaries

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 39

were obtained had been ascertained. In each case the entire
ovary was cut into serial sections.

During the progress of our work new problems arose and an
accident made it impossible for us to re-examine all our material
in order to answer several questions which were raised at a
later stage of our investigation. We especially regret our in-
ability to determine the existence of follicles which were ready
to rupture, in certain cases in which these data would have been
of considerable interest. Our work is therefore incomplete in
some respects. We expect, however, very soon to be able to
supplement our present work, wherever necessary.

OVARIES OF GUINEA PIGS IN THE LAST STAGE OF PREGNANCY

The condition of the ovaries of a guinea pig in the last days
of pregnancy is as follows: there are small, medium sized and
large follicles without degeneration of granulosa cells. In other
large follicles various stages of granulosa degeneration are pre-
sent. Many follicles show further advanced stages of atresia,
in which connective tissue grows into the follicular cavity.
Especially numerous are the last stages of atresia in which the
zona pellucida is directly surrounded by very cellular connective
tissue. Mitoses are seen in the granulosa cells of the well pre-
served follicles. We also find here a few mature follicles which
are characterized by an increase in cytoplasm of the granulosa
cells. These follicles are large; their cavity is very wide. The
nuclei of the granulosa cells are not as densely packed in these
follicles as in the ordinary large follicles, this peculiarity being
due to the marked development of the cytoplasm. They can
be easily recognized in sections stained by haemotoxylin and eosin,
inasmuch as they appear stained more reddish, in contradistinc-
tion to the ordinary large follicles in which the blue color of the
nuclei predominates, while in the mature follicles the red stain
of the cytoplasm is a distinguishing feature. In these mature fol-
licles the number of mitoses is very much smaller than in the
ordinary large follicles. With the increase in the quantity of
cytoplasm and the relative decrease in the nuclear material,

40 LEO LOEB

the cell proliferation is diminished. The number of mitoses is
usually very small, or mitoses may be absent in such follicles.
Another characteristic feature is the relative lack of degeneration
of the granulosa in these follicles. While the ordinary large
follicles degenerate in the large majority of cases, the granu-
losa cells becoming karyorrhectic, as soon as the follicle attains
a certain size; the mature follicles are very much more resist-
ant. The changes in the granulosa cells described above and
which lead to the transformation of an ordinary large follicle
into a mature red-staining follicle, and simultaneously to a de-
crease in cell proliferation of the granulosa and to a diminished
karyorrhexis of the granulosa cells, probably produces a decrease
in cell metabolism, and this decrease in cell metabolism stands
perhaps in a causal relation to the decrease in cell multiplication
and to the greater resistance of the granulosa cell. A slight de-
gree of degeneration of the granulosa may even occur in the ma-
ture red-staining follicles; a few of the central granulosa cells
may degenerate; and in one case we observed even a fargoing
degeneration of the granulosa in a mature follicle. It becomes
therefore probable that these mature follicles also degenerate,
if ovulation does not take place. This transformation of an
ordinary large follicle into a mature follicle takes place only
to a limited extent; the large majority of the follicles degenerate
before they have reached the stage of full maturity. This holds
good even in the case of guinea pigs before delivery, in which a
rupture of follicles will soon take place.

The corpora lutea of pregnancy which, at the time at which we
examined the ovaries, were approximately fifty-six to sixty-four
days old and which had formed soon after copulation, show already
some retrogressive changes in the lutein cells. A considerable
number of the vessels entering the corpora lutea have a very thick
wall consisting of several rows of cells. A large number of the
vessels, however, have merely an endothelial lining. In many of
the vessels no lumen is visible, the circulation through the corpus
luteum being evidently not very active; some of the capillary
vessels have, however, a widely open lumen. The quantity of
the connective tissue in the centre of the corpus luteum is small.

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 41

on account of the previous proliferation of lutein cells which en-
croached more and more upon the space originally filled by the
connective tissue. The corpora lutea are large. The lutein cells
show signs of degeneration ; they are finely vacuolar and may have
a foamy appearance ; a certain number of cells take less eosin and
appear therefore pale. Many cells have a sharply defined, red-
staining outline. The nuclei also show changes; they are frequent-
quently deformed, indentated ; or they are round, vesicular, but
stain less with haematoxylin ; they appear somewhat karyolytic.
Mitoses could not be seen in the lutein cells. The degree of retro-
gressive changes may vary in different corpora lutea even in the
same ovary.

We see therefore that even before delivery and before a new
ovulation has taken place, degenerative changes set in in the cor-
pora lutea, and it accords with these retrogressive changes that
mitoses are absent or at least very rare in such corpora lutea.

Besides these corpora lutea of pregnancy we may find in such
ovaries 'yellow bodies' representing the last stage of retrogression
of corpora lutea. In the corpora lutea which were transformed into
such yellow bodies, degeneration must have set in approximately
sixty to sixty-five days ago. These 'yellow bodies' have the fol-
lowing structure: In their centre and periphery we find hyaline
connective tissue; between these two zones of hyaline connective
tissue a relatively small number of degenerated large lutein cells
is enclosed, in which, during the process of retrogression, a large
amount of yellow pigment was produced.

OVARIES OF GUINEA PIGS WITHIN TWO DAYS AFTER DELIVERY

In the period directly following delivery the condition of the
ovaries, as far as follicles and corpora lutea are concerned, is
approximately the same as in the period preceding it. The growth
and degeneration of the follicles still continue to take place, and
in follicles in which the granulosa has completely or almost com-
pletely degenerated an ingrowth of connective tissue and complete
atresia of the follicles occur. The retrogressive changes in the
corpora lutea also progress, but at a slow rate, and on the whole the

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1

42 LEO LOEB

corpora lutea are not very different from those found in the preced-
ing period. This description holds good for instance for ovaries of
a guinea pig extirpated ten minutes after complete delivery.

Soon after delivery (usually within a few hours) the guinea pig
is ready for a new copulation and ovulation, and after ovulation
changes take place in the follicles which will be described later.

The corpora lutea of the preceding pregnancy undergo no very
marked changes within the next two days after delivery, although
vacuolization of the lutein cells and degenerative changes in the
nuclei show probably a slight advance ; the lutein cells do not stain
as well with eosin and appear pale. If copulation take place soon
after delivery, a rupture of the mature follicles occurs within the
succeeding six or ten hours; but if copulation be prevented by isolat-
ing the female, ovulation frequently occurs, but does not need
to take place within thirty-six hours after delivery. In several
cases in which an actual copulation was prevented, in which how-
ever the male was in contact with the female for a short time after
delivery, the rupture of the follicles and the formation of new
corpora lutea took place in the usual way. The changes in the new
corpora lutea within the first two days after delivery are the same
as those described in a previous paper.^

In three cases the lower part of the uterus or the vagina of guinea
pigs were tied completely or incompletely towards the end of
pregnancy. This procedure led to the death of the fetuses, fol-
lowed by expulsion of the dead fetuses in a case in which the occlu-
sion had been incomplete. In another case the animal was killed
by chloroform six days after the application of the ligature, and
the fetuses were found dead; furthermore autolysis of the placenta
had set in. In these cases especially the periphery of the corpora
lutea of the preceding pregnancy showed vacuolization of the lute
tein cells. The nuclei were shrunken or somewhat chromatolytic.
Notwithstanding the degenerative changes visible in the corpora
lutea, no new ovulation had taken place. From these and other
observations it follows that delivery as such does not lead to far-

2 The formation of the corpus luteum in the guinea pig. Journal American
Medical Association, February 10, 1906.

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 43

going changes in the ovaries; that merely a slow progress takes
place in changes which had set in before delivery. We furthermore
see that without copulation a spontaneous ovulation does not need
to take place after delivery, notwithstanding the degenerative
changes in the corpora lutea; that ovulation can, however, occur
without copulation, and this seems to be the rule, if the male
had been in contact with the female for some time after delivery,
a copulation having been made impossible during this period of
contact.

OVARIES OF NON-PREGNANT GUINEA PIGS IN THE PERIOD
DIRECTLY PRECEDING OVULATION

This description applies to ovaries of guinea pigs which had
copulated a few hours previously, in which an ovulation had how-
ever not yet taken place — ovulation usually taking place approxi-
mately six to ten hours after copulation. In another case we
examined the ovaries of a guinea pig that was ready for copulation
('in heat') in which, however, an actual copulation had been pre-
vented by occluding the vagina by means of a strip of plaster.

The condition of the follicles in these ovaries was similar to the
condition found in ovaries preceding and immediately following
delivery; we find good follicles of small, medium and large size;
mitoses are present in the granulosa of such follicles. The majority
of the large follicles however show more or less degeneration of
the granulosa, with the exception of the few large follicles which
progressed to complete maturity; they showed the cytoplasmic
changes described above. In these as well as in some other well
preserved large follicles the theca interna appears somewhat
hyperemic. We also find the various stages of connective tissue
ingrowth and of the subsequent diminution in the size of the folli-
cles ('connective tissue atresia') which we described in the case
of the other ovaries. In this case we do not find corpora lutea of
a preceding pregnancy, but corpora lutea of an ordinary ovarian
period, not accompanied by pregnancy. These corpora lutea
are much smaller than those of pregnancy ; their lutein cells show
vacuolization, indicating the beginning of retrogressive changes.
Notwithstanding these retrogressive changes an occasional mitosis

44 LEO LOEB

can still be found in lutein cells. The corpora lutea of the second
last ovulation have in the meantime been transformed into yellow
bodies. Processes of degeneration have therefore set in in the
corpora lutea of non pregnant as well as of pregnant guinea pigs
before ovulation. These beginning degenerative changes do
however not prevent the occurrence of a few mitoses in the cor-
pora lutea of previously not copulated animals, while in the de-
generating corpora lutea of pregnancy we have so far not been
able to detect the presence of mitoses in lutein cells.

OVARIES OF GUINEA PIGS WITHIN THREE AND ONE HALF DAYS
AFTER OVULATION

In connection with ovulation certain far reaching changes take
place in the ovaries. All follicles, with exception of very small
ones, degenerate. These changes set in with ovulation, or they
may perhaps start somewhat earlier, namely, simultaneously
with those processes that bring about ovulation. As we have
pointed out above, the general degeneration of the follicular
granulosa which we find directly after ovulation cannot yet be
observed before ovulation. This sudden degenerative process is
quite independent of copulation; we found that it can be produced
through ovulation without a preceding copulation. We discovered
experimental means through which we can produce a spontaneous
ovulation without a preceding copulation. Such an ovulation is
followed or accompanied by the same degeneration of the granu-
losa. Moreover, if we keep a number of female guinea pigs separ-
ated from the males and if we examine their ovaries after various
periods of isolation, we find occasionally ovaries in which the
rupture of follicles had taken place a few days before. In this
case also the typical follicular degeneration takes place indepen-
dently of a preceding copulation.

Six and a half hours after a preceding copulation the ovaries
showed, besides the presence of newly ruptured follicles, the follow-
ing changes in the follicles : All, with the exception of very small
follicles, show granulosa degeneration; in the large majority of the
follicles almost the whole granulosa is found in a process of degen-
eration. We also find folHcles in the process of connective tissue

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 45

atresia. Similar conditions are found in other ovaries at the same
period.

Twenty-two hours after copulation some granulosa cells are
found degenerated even in small follicles, (follicles having a small
cavity) ; these degenerated granulosa cells are dissolved.

Similar changes take place in ovaries of guinea pigs in which
ovulation followed delivery. In a guinea pig in which copulation
took place two hours after delivery and in which the ovaries were
examined seventeen hours after copulation, only a few quite small
follicles without granulosa degeneration were found ; in the large
and also in the medium sized follicles much granulosa degenera-
tion had taken place, the central granulosa cells degenerating
first. Almost no entirely good follicles were left. As soon as the
interna becomes exposed, phagocytic cells (rounded off interna
cells) peneirate into the follicular cavity and these cells take up
debris of the granulosa. The degeneration of the granulosa cells
is as usual followed by ingrowth of connective tissue.

In other ovaries the granulosa may be degenerated to a great
extent, but some remnants may still be left. Especially the
granulosa cells of the discus proligerus survive usually the rest of
the granulosa. We find of course various stages of connective
tissue atresia besides the degeneration of the granulosa. From
these observations it follows that the onset of degeneration of the
granulosa must be extremely rapid.

If we extirpate the corpora lutea, from two to eight days after
copulation a new spontaneous rupture of follicles takes place in
most cases approximately from thirteen to fifteen days after the
previous copulation, even if the female had been kept entirely
isolated during the whole period following the extirpation of the
corpora lutea. This early spontaneous ovulation is accompanied
by the same follicular degeneration which we described above.

It is an interesting problem, whether an artificially produced
rupture of a follicle, with the subsequent development of a corpus
luteum, is accompanied by the same acute follicular degeneration.
Several years ago we made experiments in which we pricked or cut
the surface of ovaries of guinea pigs which were either 'in heat,'
without however having copulated, or which copulated a few hours
previously, or which had in some cases copulated from three to six

46 LEO LOEB

days previously. In only one case did we find a young corpus
luteum the origin of which could reasonably be attributed to the
cutting of the ovary and to the artificial ru^pture of a follicle.
In this case an animal had been used which showed the first symp-
toms characteristic for the period of heat. Three days after the
cuts had been made the ovaries were examined. One young
corpus luteum was found in the cortex of the ovary. Blood and
connective tissue were found in the center of the corpus luteum;
connective tissue and vessels grew into the corpus luteum, which
was very small. In this ovary we found good follicles of small
medium and large size; we also found large follicles with begin-
ning and wdth further advanced granulosa degeneration, and with
beginning ingrowth of connective tissue. In as much as in no
case of spontaneous rupture the follicles were found in a similar
condition at that period after the rupture, it is very probable that
we have in this case to deal with an artificial rupture of follicles and
that such an artificial rupture of follicles is not accompanied by the
rapid degeneration of the follicular granulosa.

On the basis of our previous results we can easily understand,
why in all probability we succeeded in one case only in causing an
artificial rupture of a follicle. Such an experiment does not promise
to be successful, unless we have the chance of opening a mature
follicle, and such an opportunity exists only at periods of very
short duration.

In these ovaries we find usually two or three generations of cor-
pora lutea; namely:

1. The young corpora lutea, developing in the recently rup-
tured follicles. These corpora lutea we have described elsewhere
in their development up to the sixth day.

2. Corpora lutea that had formed at the time of the preceding
ovulation, which had not been followed by pregnancy in female
guinea pigs which had been kept separated from males. These
corpora lutea are therefore in all probability approximately nine-
teen to twenty-eight days old. They show signs of beginning retro-
gression. Their lutein cells are more or less vacuolar, especially
in the periphery, where the vacuolization usually begins ; gradually
the vacuolization progresses to the central part. In the center of

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 47

the corpus luteum we find a relatively small amount of fibrous
tissue. We not only find capillary vessels but also vessels the wall
of which consists of two coats penetrating the corpus luteum.

These corpora lutea begin to shrink very soon, and three days
after the new rupture they are usually smaller than immediately
after the ovulation. Notwithstanding the degenerative processes
which are apparent in these corpora lutea, it is not uncommon to
to find still mitoses in the lutein cells of such corpora lutea within
the first twenty hours after the new rupture of follicles has taken
place. At a later period mitoses were not seen in this series. The
mitoses appear in the relatively well preserved, but they may be
present even in somewhat vacuolar lutein cells. It is possible
that occasionally mitoses occur also in endothelial cells of the capil-
laries.

3. The third generation is represented by yellow hyaline bodies.
They are the remnants of corpora lutea that formed forty or more
days ago.

If we examine ovaries of young guinea pigs, two and a half to
three months old, we may find only the first, or the first and second
generations of corpora lutea, but yellow bodies may be lacking.

We see therefore that preceding and following the rupture of
new follicles in non-pregnant animals, processes of degeneration
have begun in the corpora lutea of the preceding ovulation, and that
notwithstanding such processes of degeneration, mitoses may
occur in such corpora lutea for a short period following the new
ovulation. These corpora lutea which are not accompanied by
pregnancy are much smaller than the corpora lutea of pregnancy
and they shrink more rapidly. The absolute diminution in size
is more rapid than in the retrogressing corpora lutea of a preced-
ing pregnancy. Concerning the relative rapidity of retrogression
(the percentage decrease in size, the full size of the corpora lutea
being taken as the standard), we cannot make any definite state-
ment, not having carried out any measurements.

The mode of retrogression is the same in both ordinary corpora
lutea and in those of pregnancy. The vacuolization begins in
the periphery, where it becomes most marked, and from here it
proceeds into the interior of the corpus luteum.

48 LEO LOEB

OVARIES OF A GUINEA PIG APPROXIMATELY THREE TO FOUR
DAYS AFTER ABORTION

In one case the ovaries of a guinea pig were examined which on
examination had previously been found to be in a well developed
stage of pregnancy, but which had aborted about three to four
days previously. The four corpora lutea showed signs of degener-
ation. The lutein cells were vacuolar in the periphery, in the cen-
ter the cells stained pale red with eosin, the vesicular nuclei
showed a diminution in the amount of chromatin. The cell out-
lines were very sharp, staining red with eosin. In the center there
was dense connective tissue and many blood vessels had very
thick walls.

Follicles of small, medium and large size, with well preserved
granulosa, were present. A few mature, red staining follicles with-
out mitoses or degeneration in the granulosa were also found.
Many other large immature follicles showed various stages of
granulosa degeneration. There were of course also present
various stages of connective tissue atresia.

We see therefore that abortion is not followed by or associated
with marked changes in the follicles. Whether the mature follicles
which we found in these ovaries matured as a result of abortion, or
whether the mature follicles were present before the onset of abor-
tion, we cannot state with certaintj^, although it is more probable
that maturation of the follicles followed abortion. We also note
the beginning retrogressive changes in the corpora lutea. But in
this case also we cannot be sure that the degenerative processes
had not set in before the abortion had commenced.

OVARIES OF GUINEA PIGS FOUR TO SEVEN AND ONE HALF DAYS
AFTER OVULATION

Six days after an ovulation we find in the ovaries on the whole
the following condition of the follicles: There are well preserved
follicles of small and medium size, with mitoses in the granulosa
cells. A limited amount of granulosa degeneration is found only
in rare instances. In such follicles mitoses are absent or their

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 49

number is decreased. Follicles in an advanced state of connective
tissue atresia are frequent.

The character of the follicles at this period -of the sexual cycle
is the same in cases in which the last ovulation was preceded by
delivery, in which, therefore, in the previous period of the sexual
cycle a pregnancy was present, and in other cases in which the
previous period of the sexual cycle had not been comphcated by
pregnancy. We see therefore that within six days quite small
follicles, possessing only a very small follicular cavity, grow and
reach medium size. During this period the granulosa of medium
sized follicles did not degenerate, and no large follicles had as yet
developed. We find therefore principally, besides the follicles
with preserved granulosa, follicles in an advanced state of con-
nective ,^issue atresia.

Six days after ovulation we find the corpora lutea of the last
generation (corpora lutea six days old, as follows : The center of
the corpus luteum is filled by a more or less loose connective tissue.
Mitoses are present in the lutein cells as well as in the endothelial
cells of the capillaries. Almost all the vessels have a capillary
character. They penetrate into the central connective tissue. At
that period vessels with two coats (intima and muscle coat of the
media) can be observed for the first time, although they become
more frequent at a somewhat later period.

In guinea pigs in which a pregnancy and delivery preceded the
last ovulation, the corpus luteum of the preceding pregnancy
shows marked signs of degeneration. Especially the peripheral
cells are frequently coarsely, while the more centrally situated
cells are more finely vacuolar; but even in the latter the proto-
plasm stains less with eosin and the nuclei are slightly chromatoly-
tic ; the cells appear distinctly pale. The vessels are very thick and
at certain places in the periphery the connective tissue of the neigh-
borhood seems to begin to grow into the peripheral parts of the
corpus luteum.

The ordinary corpora lutea of the second generation (not accom-
panied by pregnancy) show marked vacuolization; they diminish
in size and in one case yellow pigment developed in a few of the
vacuolar cells. Therefore in the course of five to eight days

50 LEO LOEB

since the beginning of degeneration the retrogressive changes have
much advanced. The retrogressing corpora lutea of pregnancy of
the corresponding generation are much larger at this period than
the ordinary corpora lutea.

In a certain number of ovaries we also find a further (third)
generation of retrogressing corpora lutea, represented by yellow
bodies.

One corpus luteum deserves especial mention. In an ovary of
a guinea pig which had ovulated approximately four and a half
days before, five corpora lutea were found, four of which showing the
typical structure. In the fifth of these corpora lutea, however,
the lutein cells were arranged in the shape of glandular ducts.
This condition has perhaps been produced through a dissolu-
tion of the central cells. Otherwise the corpora lutea in this
ovary were normal.

The same typical changes in the folhcles noticed in ovaries of
this period after a preceding copulation and ovulation are also
found in ovaries in which a spontaneous ovulation took place inde-
pendently of a preceding copulation. As we stated above, such
a spontaneous ovulation can be produced through an early exci-
sion of the corpora lutea. The same follicular changes take place
also in pregnant animals in which, through an excision of the cor-
pora lutea about six to eight days after copulation, a spontaneous
ovulation is produced approximately thirteen to fifteen days
after the beginning of pregnancy, without the pregnancy being
interrupted.

We see therefore that these cyclical changes in the ovaries are
essentially independent of copulation and of pregnancy and are
directly connected only with ovulation.

OVARIES OF GUINEA PIGS SEVEN AND ONE HALF TO EIGHT AND
ONE HALF DAYS AFTER OVULATION

At this stage of the sexual cycle we find good follicles of small,
medium and large size with no, or only very little, granulosa
degeneration. We also find follicles in connective tissue atresia.
We see therefore that in approximately eight days follicles origin ally

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 51

very small have reached a large size. The new (eight days old)
corpora lutea grow actively during this period and show frequent
mitoses in lutein cells. The corpora lutea of the preceding ovula-
tion (second generation) continues to shrink and show marked
vacuolization of the lutein cells. If the second last ovulation were
accompanied by pregnancy, the retrogressing corpora lutea were
still larger.

The third generation of corpora lutea was represented by atre-
tic yellow bodies the age of which varied approximately between
forty-eight and ninety-five days.

OVARIES OF GUINEA PIGS TEN TO ELEVEN DAYS AFTER OVULATION

We fmd good follicles without granulosa degeneration of small,
medium and large size, besides various stages of granulosa degener-
ation and of connective tissue atresia, early stages with beginning
ingrowth of connective tissue included. In the granulosa of well
preserved follicles mitoses are present as usual.

At this stage — ten days after ovulation— the ovary presents
again its normal aspect. The follicles have grown to a large size
and undergo the ordinary retrogressive changes. The ten to eleven
days old corpora lutea are well developed; in the centre a relatively
small amount of connective tissue is present. Mitoses in the lu-
tein cells are usually frequent ; they occur perhaps also in endothe-
lial cells of capillaries. The large majority of the vessels have a
capillary character, but occasionally a vessel is seen with a double
coat of cells. Marked signs of degeneration are absent, but a few
slightly vacuolar lutein cells may occasionally be seen.

The second generation of corpora lutea, originating in the sec-
ond last ovulation, are small vacuolar bodies with much connective
tissue and thick vessels. If, however, this second last ovulation
had been followed by pregnancy, the retrogressing corpora lutea
of the previous pregnancy are as yet much larger; the lutein cells
have become very vacuolar; many thick vessels are present. In
some of the vacuolar lutein cells yellow pigment appears.

A third generation of corpora lutea is represented by yellow
bodies. They are, however, not found in all ovaries.

52 LEO LOEB

In this series of animals pregnancy had been prevented after
a preceding copulation, either by ligaturing the tubes within the
first two days after copulation, or by making long incisions into
the uterus approximately four to six days after copulation.

The ovaries were also examined in a certain number of other
guinea pigs of this period in which pregnancy existed. The accom-
panying pregnancy does not produce any marked change in the
ovaries and the preceding description applies on the whole equally
well to these ovaries.

OVARIES OF GUINEA PIGS THIRTEEN TO FIFTEEN DAYS AFTER
OVULATION

In this series of animals pregnancy was prevented in the same
manner as in the series of animals examined ten to eleven days
after ovulation. The follicles have approximately the same
character as in the previous period. We see the same varieties
of follicles. Small foUicles grow and become large and, after
having reached this stage, or even at a slightly earlier stage,
granulosa degeneration sets in with consecutive connective tissue
atresia. In the granulosa of well preserved follicles numerous
mitoses are present, and mitoses may even be found, if a slight
amount of granulosa degeneration has taken place. The corpora
lutea of the last ovulation (I generation) show more generally the
beginning of vacuolization, especially in the periphery of the
corpus luteum; but on the whole the corpus luteum is still well
preserved and usually mitoses are found in some of the lutein
and occasionally in cells belonging to blood vessels.

In the center we find connective tissue with thin spindle-shaped
nuclei, and a number of vessels with walls consisting of several
rows of cells penetrate into the central connective tissue. In
some of the lutein cells the protoplasm appears dense and stains
deeply with eosin. It appears probable that in such cells the nuc-
leus had started to divide by mitosis, but degenerative processes
seem to have set in and interrupted the process of the mitotic
division. We are however not certain that this interpretation,

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 53

which would perhaps agree with an opinion expressed by Regaud
and Dubreuil,^ is correct.

The second generation of corpora lutea is represented by small
vacuolar bodies with relatively much connective tissue and thick
vessels. These atretic corpora lutea originated at the time of the
second last ovulation and are therefore approximately thirty-three
to forty days old. If this second last ovulation had been followed
by pregnancy, the corpora lutea of this period are still much larger
than the corpora lutea of the corresponding generation without
an accompanying pregnancy; but a considerable shrinking of these
corpora lutea has also taken place. The vessels are to a great
extent collapsed. The lutein cells are finely or coarsely vacuolar,
take less stain, still possess nuclei and a distinct cell wall,
staining with eosin. The third generation of corpora lutea is
again represented by yellow bodies. They are not present in all
ovaries, but are found especially in the ovaries of older guinea
pigs. Occasionally the degenerating corpora lutea of the second
generation may also be absent.

In guinea pigs in which the last ovulation was followed by preg-
nancy, the condition of the follicles is very similar. The corpora
lutea of the first generation, however, are large and show frequent
mitoses in lutein cells, occasionally also in lutein cells the peri-
phery of which is vacuolar. There are possibly also mitoses pres-
ent in the endothelial cells. The retrogressing corpora lutea of
the second and third generations are in pregnant animals of a
similar character as those described in the ovaries of guinea pigs
of the same period without an accompanying pregnancy.

OVARIES OF GUINEA PIGS FIFTEEN TO NINETEEN DAYS AFTER
OVULATION

Pregnancy had in most cases been prevented by the same means
which were used in the preceding stages. In a few instances in
which pregnancy had occurred an early abortion followed. The
follicles exhibit on the whole the same character as in the preced-
ing stage ; we find good foUicles of small, medium and large size,

3 C. R. Soc. Biol., 54. 1908.

54 LEO LOEB

and follicles in various stages of granulosa degeneration and of
connective tissues atresia. We may also find large mature follicles.
In how many cases these latter are present, will still have to be
determined. In such animals a rupture of follicles is imminent.

In three guinea pigs a spontaneous ovulation had taken place
at this period, notwithstanding the absence of male guinea pigs.
In such cases young corpora lutea were found and, accordingly, a
condition of the follicles characteristic of a period directly follow-
ing ovulation. In the large majority of cases however a spontane-
ous ovulation did not take place in ovaries at this period of the
sexual cycle. In such cases the folUcles showed the character
described above.

The corpora lutea of the first generation, which originated as
a result of the last ovulation, show more or less signs of beginning
retrogressive changes as indicated by fine or coarse vacuoliza-
tion of the lutein cells. The intensity of this degenerative change
varies is different ovaries. On the whole the retrogressive changes
seem to be more marked in the nineteen days than in the sixteen
days old corpora lutea; but variations seem to occur, even in cor-
pora lutea of the same age. The vacuolization is usually most
marked in the periphery and progresses toward the center. Other
lutein cells are still more solid and mitoses in lutein cells can be
seen in the majority of the corpora lutea of this period. In cases
in which mature follicles are present and a spontaneous rupture
of follicles is therefore soon to be expected, the corpora lutea
show much vacuolization; but here also mitoses are still present in
lutein cells.

In some cases the retrogressive changes are still further ad-
vanced and a connective tissue capsule may appear in the peri-
phery of the corpus luteum. The marked vacuolization of peri-
pheral lutein cells may be accompanied by a diminution in the
lumen of blood vessels. Vessels with coats consisting of several
rows of cells are seen regularly in these corpora lutea. The con-
nective tissue in the center of the corpora lutea is usually dense and
relatively small in amount.

In those cases in which a new spontaneous ovulation had taken
place the vacuolization of the corpora lutea had still further pro-

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 55

gressed and under such circumstances mitoses were no longer
present in them.

The corpora lutea of the preceding (II) generation, originating
in an ovulation that took place at least thirty-seven days ago,
are sometimes represented by small bodies which are surrounded
by a thick connective tissue capsule; much fibrous tissue is found
in the center and the lutein cells between these two zones show
very large vacuoles. The vessels remaining in such structures
have very thick cellular walls. In other cases some yellow pig-
ment appears in such vacuolar cells and in still other cases we see
only yellow, atretic bodies. It is probable that the latter struc-
tures are found in cases in which a still longer time has elapsed
since the preceding (second last) ovulation. There may of course
have occurred a longer interval than twenty days between the
last and second last ovulation.

When the second generation was represented by a corpus luteum
of pregnancy, the retrogressive changes were also marked, shrink-
ing of the corpus luteum and vacuolization of the lutein cells are
pronounced, but such corpora lutea are still considerably larger
sixteen to nineteen days after the completion of pregnancy than
ordinary corpora lutea of the corresponding generation. Some of
the vacuolar cells may show a yellow pigmentation. In such
ovaries we may find a still older generation of retrogressing cor-
pora lutea present, represented by yellow atretic bodies which
owe their origin to an ovulation that took place more than a
hundred days ago; and if the last named (third last) ovulation were
followed by a pregnancy, this ovulation may have taken place
approximately one-hundred and fifty days ago. Not in all
animals are so many generations of corpora lutea found ; especially
in young animals (two to three months old only one generation
may be present.

If the last ovulation that took place fifteen and one half to
nineteen days ago were followed by pregnancy, the follicles in the
ovaries of pregnant animals of this period do not show any marked
difference from the follicles of non-pregnant animals at the corre-
sponding period after ovulation. In both cases we find good folli-
cles of various sizes and the different stages of retrogression of

56 LEO LOEB

follicles which we mentioned above. In the ovaries of pregnant
animals of this period we may also find mature follicles, the granu-
losa cells of which have more cytoplasm that stains red with eosin.
Such follicles show less granulosa degeneration and a decrease in
the number of mitoses is visible in the granulosa cells. Some degen-
eration of granulosa cells may however occur in these follicles and
their further fate will still have to be determined.

The corpora lutea of pregnancy (first generation) are well pre-
served. Fine vacuoles may however be present, especially in
the peripheral lutein cells. Mitoses are also present. They do
not show such pronounced signs of retrogression, as occur in cor-
pora lutea of non-pregnant animals of this period.

OVARIES OF GUINEA PIGS TWENTY TO TWENTY-SEVEN DAYS
AFTER OVULATION

At this period the proportion of animals in which a spontane-
ous ovulation had taken place, notwithstanding the separation of
females and males, is much greater than in the preceding period.
Among twenty-two guinea pigs a spontaneous ovulation had taken
place in eight, while in the fourteen other females no rupture of
follicles had as yet occurred. In at least one and possibly in more
of these fourteen guinea pigs a rupture was however imminent,
as indicated by the presence of mature, red-staining follicles. In
those animals in which ovulation had taken place within the last
few days the follicles were in the condition corresponding to that
stage after ovulation. The corpora lutea that originated as a
result of the ovulation twenty to twenty-six days previously
showed marked degeneration; the cells were vacuolar; in one case
the lutein cells formed a hyaline material in which the vesicular
nuclei were imbedded. Mitoses were present in only one case, in
which the rupture had taken place apparently within the last
twenty-four hours, but even vacuolar cells may divide mitotically.
Many blood vessels have thick cellular coats and the blood vessels
in general do not seem to be patent.

In all the other guinea pigs in which a new rupture of follicles
had not yet taken place the follicles behave approximately in the
same manner as in the previous stage; we see follicles of various

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 57

sizes without granulosa degeneration/ and follicles of large and
also of medium size in various stages of granulosa and connective
tissue atresia.

In the ovaries of the guinea pig in which a spontaneous rupture
of follicles was imminent, the twenty-two days old corpora lutea
also showed the signs of early degeneration; some of the cells
were still good, but the majority were vacuolar.

In the guinea pigs, in which a spontaneous ovulation had not yet
taken place, the corpora lutea of the last ovulation were also in a
process of degeneration, which was especially marked during the
later stages, twenty-four to twenty-six days after ovulation: here
the vacuolization was very pronounced, and occasionally con-
nective tissue began to grow into the periphery of the corpus
luteum. The vessels of these corpora lutea were very thick. In
some other ovaries, especially in those examined twenty and
twenty-one days after ovulation, the number of relatively well
preserved cells was still greater. On the whole the number of
mitoses found in lutein cells at this period is distinctly diminished.

The older generations of corpora lutea are represented by atre-
tic yellow bodies, which are however not present in all animals.
In one case a corpus luteum was present that originated as a result
of an ovulation that took place approximately ninety-three days
before and was accompanied by pregnancy. In this cases twenty-
seven days after delivery very little of the lutein tissue was left,
the blood vessels had very thick coats, and the fibrous tissue of
the remnant of the corpus luteum was very prominent.

If the ovulation which took place twenty to twenty-seven days
before were followed by a pregnancy, no new spontaneous ovula-
tion took place, 'the conditions of the follicles was the same as in
those guinea pigs in which the last ovulation was not followed by
pregnancy and in which no new spontaneous ovulation had as yet
taken place. The corpora lutea of pregnancy of this period showed
much less vacuolization, although a slight amount of it may have
been present, especially in the periphery of the corpus luteum.
Mitoses were more common in these corpora lutea of pregnancy
than in the ordinary corpora lutea of the same period. Their
size was also greater.

JOURNAI^ OF MORPHOLOGY, VOL. 22, NO. 1

58 LEO LOEB

In regard to the ordinary corpora lutea and the corpora lutea of
pregnancy of previous generations, the same retrogressive changes
which were described above in the ovaries of non-pregnant guinea
pigs of this period, were found in pregnant animals.

We see therefore that the condition of the corpora lutea indi-
cates the condition of the follicles, and conversely the condition
of the follicles indicates the history of the corpora lutea. At a
certain time (approximately ten days) after the ovulation a
certain equilibrium is reached between the growth and the degen-
eration of the follicles. Whether a quantitatively exact equili-
brium is reached, cannot yet be stated. In proportion to the
length of time which elapsed since the last ovulation, the probabil-
ity of a new spontaneous rupture, with the subsequent changes in
the follicles, becomes greater. At this and the preceding period
signs of degeneration are present in the ordinary corpora lutea,
which become the more marked the older the corpus luteum; the
number of mitoses in lutein cells decreases with advancing age ;
they may however still be present in corpora lutea immediately
following a new ovulation; the latter however is soon followed by
further progressing degeneration of the corpus luteum of the pre-
ceding ovulation. If the ovulation that took place twenty to
twenty-six days previously was accompanied by pregnancy,
no new spontaneous rupture of follicles took place, the prolifera-
tion of the lutein cells continued, and degenerative processes in
the corpora lutea were retarded.

Approximately twenty-five days after the completion of preg-
nancy the corpora lutea of pregnancy (second generation) have
become small vacuolar bodies with thick vessels and fibrous tissue,
while corresponding ordinary corpora lutea have at this time
apparently been transformed into yellow bodies.

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 59

OVARIES OF GUINEA PIGS TWENTY-SIX TO FORTY DAYS AFTER
OVULATION

In five animals in which, in order to prevent pregnancy, both
(and in one case one) of the Fallopian tubes had been ligated
within twenty-six hours after copulation, and in which at a later
operation incisions had been made into the uterus, no new ovu-
lation had taken place at the time of the examination, twenty-
six to thirty-four days after copulation. The corpora lutea
(twenty-six to thirty-four days old) showed very marked retro-
gression; they were very vacuolar; their size was always diminished
especially after thirty-two to thirty-four days, but differed some-
what in individual cases. Some corpora lutea formed small
bodies containing very dense fibrous tissue in the center and en-
closing in the periphery a relatively small number of very vacuolar
cells. Other corpora lutea were still somewhat larger and con-
tained a. few better preserved cells.

Besides the retrogressing vacuolar corpora lutea some atretic
yellow bodies could be found in some cases ; they were remnants
of corpora lutea at least fifty days old. In two other ovaries a
spontaneous ovulation had taken place recently and the condition
of the follicles was in accordance with the age of the new corpora
lutea. Here also the thirty to thirty-two days old corpora lutea
of the preceding ovulation were very vacuolar and contained blood
vessels with a thick coat and much dense fibrous tissue.

OVARIES OF A PREGNANT GUINEA PIG APPROXIMATELY THIRTY-
FIVE TO FORTY DAYS AFTER COPULATION

In these ovaries we found good follicles of small, medium and
large size without granulosa degeneration and with mitoses in
granulosa cells; other follicles showed various stages of granulosa
degeneration and of connective tissue atresia. Mitoses were
absent or diminished in number in follicles in which granulosa
degeneration existed.

In addition to the ordinary large follicles mature or almost
mature follicles were seen in which the cytoplasm of the cells was
well developed, and in which the granulosa contained only very

60 LEO LOEB

few mitoses which were found especially in the discus proligerus.
Some of the nuclei of the granulosa cells appeared somewhat con-
tracted in these follicles, but no marked degeneration of the gran-
ulosa cells was found.

The corpora lutea of pregnancy were large, the cytoplasm of
the lutein cells stained red yellow with eosin; the cell outlines
were quite distinct. The large majority of the lutein cells were
compact and did not show vacuoles; the nuclei were vesicular.
A few mitoses were found in lutein cells. Only very little con-
nective tissue was present in the center of the corpora lutea.
Some of the vessels had thick walls, while other vessels were of a
capillary character and had either a wide or narrow lumen. We
see therefore that also at later stages of pregnancy the follicles
continue to grow and to degenerate, and that even at this period
of pregnancy follicles may mature. The lutein cells of the corpora
lutea of pregnancy continue to show mitotic nuclear figures and
well preserved cytoplasm at a time when, in the ordinary corpora
lutea, retrogression is very far advanced.

OVARIES OF GUINEA PIGS IN WHICH COPULATION HAD BEEN
PREVENTED

A large number of ovaries were examined of female guinea pigs
which had been kept separated from males for various lengths of
time.

One set of guinea pigs was separated from males before sexual
maturity had been reached. The ovaries were examined, when the
animals were six and twelve months old. In every instance ovu-
lation had taken place repeatedly and we usually found the three
generations of corpora lutea which we described in the case of
guinea pigs which had copulated, namely relatively young cor-
pora lutea, retrogressing vacuolar corpora lutea and atretic
yellow bodies.

In another series guinea pigs were guarded against contact
with males after delivery, and were kept separated from males
for various periods of time. In this case a spontaneous ovulation
took place after delivery, at least in the majority of cases, even

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 61

without contact with males, and subsequently further ovulations
occurred. Under such conditions the successive ovulations do
however not occur in the same intervals in all animals; in some
cases a delay in ovulation may take place : this accords well with
our previous observations. The conditions of the follicles corre-
spond to the time elapsed since the last ovulation, as indicated
by the state of the corpora lutea.

Not in every case however does a spontaneous ovulation take
place without contact with male. In several cases neither new
nor retrogressing corpora lutea could be found in the ovaries of
guinea pigs which, according to their age, ought to have ovulated,
but in which no sign of heat had been noticed during an observa-
tion extending over a certain period of time. In other guinea pigs
which had been in heat recently, but in which copulation had
been prevented, no new ovulation corresponding to the period of
heat had taken place at the time of examination.

SOME OBSERVATIONS ON THE POSTFETAL DEVELOPMENT OF THE
OVARY OF THE GUINEA PIG

In connection with the cyclic changes in the adult ovary of the
guinea pig, just described, we thought it of interest to determine
the time at which these cyclic changes set in. For this purpose we
studied a series of ovaries at differents stages of the growing guinea
pig-

1. In the ovaries of a fetus near the time of birth many follicles
are present in the cortex. These follicles have not yet a cavity
and the largest follicles have a granulosa consisting of three, or four
rows of granulosa cells ; in the latter some mitoses can be seen.
No distinct differentiation appears in the connective tissue of the
different parts of the ovary.

2. In the ovaries of guinea pigs four, five and seven days old
we find a cavity in a certain number of the follicles ; no atretic
processes have as yet taken place. The theca interna cells are
distinguished from the surrounding connective tissue through the

62 LEO LOEB

increase in the size of their nuclei. The connective tissue around
the medullary canals is relatively dense. In the granulosa, theca
interna and in the ordinary connective tissue stroma mitoses are
frequent.

3. The ovaries of guinea pigs eighteen days old are larger; the
follicles also have increased in size. Small and medium sized and
in proportion to the as yet small size of the ovaries, relatively
large follicles are present. In some of the follicles degenerative
processes appear at this time, but the extent to which such changes
have taken place differs in the ovaries of different animals. In
the ovaries of some guinea pigs no degeneration of the granulosa
has as yet taken place. In the ovaries of another guinea pig a
few follicles showed a trace of granulosa degeneration, while in
another follicle the granulosa degeneration was pronounced.

In the follicles of some ovaries we find even a beginning ingrowth
of connective tissue into the follicular cavity, and in one case a
cavity of a follicle was filled with loose connective tissue. The
majority of the follicles are in a good condition; their cavity is
larger than at the preceding stage and the interna is better devel-
oped and consists of more rows of cells. Mitoses are present in the
theca interna and in the granulosa. The connective tissue be-
tween the follicles is a little more fibrous, and around certain
blood and lymph vessels it is somewhat edematous and rarefied.

4. In the ovaries of guinea pigs twenty-eight days old the major-
ity of follicles are in good condition and non-atretic; they are of
small and medium, but not yet very large size. In some ovaries
hardly any degeneration of follicles is visible ; in others we see some
follicles which have not yet attained their full size (corresponding
to the as yet small size of the ovaries) , presenting various stages of
granulosa degeneration. In some follicles the granulosa has been
entirely destroyed and connective tissue begins to grow into the
cavity.' In some cases we find quite atretic connective tissue
follicles. In some small and medium sized follicles the ova may
undergo (probably amitotic) nuclear division and a corresponding
segmentation of the cytoplasm, the granulosa being still intact.

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 63

In other cases, however, such ova are surrounded by connective
tissue.

The connective tissue of the ovaries shows more differentiation
at this period and is somewhat more fibrous.

5. In ovaries of guinea pigs one to two months old the size of
some of the folhcles, in correspondence with the growth of the
ovaries, enlarges. We see various stages of granulosa degenera-
tion and of connective tissue atresia. Granulosa degeneration may
take place in medium sized and in large follicles. In some ovaries
the large majority of follicles may show either granulosa degen-
eration or connective tissue atresia. Corpora lutea are not yet
visible.

6. Ovaries of guinea pigs three months old: Approximately at
this period the ovaries have become mature. We find various
stages of developing follicles and occasionally mature follicles
ready to rupture. We find the various stages of granulosa degen-
eration and of connective tissue atresia. We notice a greater differ-
entiation in the structure of the stroma in different parts of the
ovary.

Corpora lutea, which occasionally are already in the beginning
of degeneration, are present in some ovaries; in other animals
ovulation has not yet taken place.

It follows from these observations that degenerative processes
in follicles set in approximately fourteen to eighteen days after
birth, and ovulation and formation of corpoa lutea appear in
guinea pigs two to three and a half months old. The ovaries and
follicles must have reached a certain size, before ovulation sets
in. The time required for the development of small into large
follicles, with subsequent beginning of degenerative processes, is
somewhat longer in the young growing animal than in the mature
guinea pig, but in both the periods of time are of a similar order
(approximately nine and fourteen days respectively).

64 LEO LOEB

SUMMARY

The principal result of our investigations we can state as follows :
In the ovary of the guinea pig (and probably of mammals gener-
ally) cyclic changes take place independently of copulation and of
pregnancy.

A sexual period (the period between two ovulations) is accom-
panied by a series of changes in the follicles. As a result of the
conditions leading to or accompanying ovulation the granulosa
of all large and medium follicles undergoes a very rapid degenera-
tion, which is very marked within an hour or two after ovulation,
or perhaps even sooner. In the follicles in which the cavity is as
yet very small, the degenerative processes are very slight or absent.
These follicles do not seem to perish. These degenerative changes
affect equally both ovaries of one animal, even if a rupture of
follicles should have taken place in only one of the two ovaries.
The local effect of the rupture of the follicle can therefore not be
the cause of the follicular degeneration. Within the next few
days the small follicles grow and gradually attain a large size.
Eight days after ovulation large follicles are again noticeable.
As soon as good sized and medium sized follicles have been formed
they begin to undergo degenerative processes, the granulosa
degenerating and becoming dissolved and connective tissue grow-
ing into the follicular cavity. This process ends in an almost
complete disappearance of these follicles. In the meantime other
follicles grow and, having reached a large size, they also degenerate.
Thus after a first stage of general growth, comprising approxi-
mately ten days after ovulation, a certain equilibrium is reached in
which new follicles are growing to a certain size, and in which other
follicles of large or medium size degenerate. Whether certain quan-
titative differences in the proportion of the number of growing and
degenerating follicles exist at different periods of this second part of
the sexual cycle, will still have to be determined. This second
period of equilibrium begins approximately ten days after the last
ovulation, and it lasts until a new ovulation occurs. Gradually a
few large follicles undergo still further changes, the cytoplasm of
their granulosa cells enlarges, the number of mitoses in these

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 65

cells decreases and they become more resistant to those processes
which lead to degeneration in other follicles. The follicles in
which such changes have taken place are mature and ready to
rupture. In the meantime the follicles that ruptured during the
preceding ovulation developed into corpora lutea. The latter
represent principally the hypertrophic granulosa cells of the rup-
tured folUcles, which proliferate mitotically. After a certain
stage of development has been reached, degenerative processes
set in in the corpus luteum, which start in its periphery and pro-
ceed to the center. These degenerative processes set in very early,
are noticeable eighteen to twenty days and are usually marked
twenty to twenty-four days after the preceding ovulation.
Throughout this period of beginning degeneration, however, some
mitoses are still visible in certain lutein cells. At this period
usually a new ovulation takes place. The exact time at which
the new ovulation occurs varies however somewhat in different
animals, ovulation occurring earlier in some animals than in
others. In some cases it can be hastened through certain external
factors, especially copulation, but in the large majority of cases
it occurs sooner or later even without a preceding copulation.^
After the new ovulation has taken place, the degenerative processes
progress in the corpus luteum, although within the first twenty
hours after ovulation mitoses may still be found in certain lutein
cells. In the following period a considerable shrinking of the cor-
pus luteum takes place ; the connective tissue in the cortex and in
the periphery becomes hyaline and forms a relatively prominent
part enclosing a small number of very vacuolar cells. Gradually
yellow pigment is deposited in these vacuolar cells and thus the
corpora lutea become transformed into the atretic yellow bodies.
The new ovulation was of course again followed by the typical
changes in the follicles.

If the ovulation be followed by pregnancy, the principal changes
taking place in the ovaries are on the whole the same. The only

■* Whether or not in the guinea pig ovulation can take place independently of a
preceding copulation has been a subject of controversy. Concerning the litera-
ture of this question see William H. Kirkham, Biological Bulletin, vol. 18, no. 5,
April, 1910.

66 LEO LOEB

difference consists in a prolongation of the sexual cycle, which lasts
as long as the pregnancy continues. The changes in the follicles are
identical with those found in the ordinary sexual period not
accompanied by pregnancy.

After copulation the period of growth following the sudden
degeneration of the follicles is the same as in the ordinary sexual
period, but the period of follicular equilibrium is much pro-
longed.

During this period of follicular equilibrium certain follicles
can not only grow to a considerable size, but may even undergo
the additional changes which indicate the maturation of the folli-
cle. A rupture of follicles does not however take place during
pregnancy under ordinary circumstances.

The corpus luteum of pregnancy differs from the ordinary
corpus luteum mainly in its prolonged duration of growth and of
life. At a time when, in the ordinary corpus luteum not accom-
panied by pregnancy, mitoses have ceased to be present and the
retrogressive changes are very marked, mitoses are still seen in the
corpus luteum of pregnancy. In the corpus luteum of pregnancy
degenerative changes set in before the end of pregnancy has been
reached, and they continue after delivery. A short time after
delivery a new ovulation usually occurs, even if no copulation had
taken place after delivery. The retrogression of the corpora lutea
of pregnancy continues, but it requires much more time than the
retrogression of an ordinary corpus luteum.

The mechanism that governs the sexual cycle in the ovary can
be recognized only incompletely by observation and it has been
the subject of an experimental investigation, the results of which
we shall report in more detail elsewhere. We may however state
that our experiments have shown that through extirpation of the
corpora lutea the sexual cycle is shortened. The presence of well
functioning corpora lutea inhibits a new ovulation. Pregnancy as
such does not prevent ovulation. Ovulation can be made to
take place even in pregnancy, if the corpora lutea be extirpated
at an early period after copulation. And under such conditions
the typical follicular changes follow the ovulation during preg-
nancy. As soon therefore as degenerative processes have set

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 67

in in the corpora lutea, either during pregnancy or outside of
pregnancy, a new ovulation can take place. How far the presence
of ,the corpora lutea influences the transformation of ordinary large
follicles into mature follicles and how far its action merely con-
cerns the rupture of the mature follicles, remains still to be deter-
mined.

It follows from our observations that the time of ovulation
depends upon at least three different factors: (1) Changes taking
place in the ovaries. It is necessary that mature follicles have
been produced, before rupture can take place. Our experiments
indicate that cuts into an ovary causing an opening of a
follicle may possibly lead to the formation of a corpus luteum only
at a time when mature follicles are present. A certain time must
therefore have elapsed after ovulation before another ovulation
can take place. During this period small follicles reach their
full size. Thus a minimal time which must elapse between two
ovulations is required. (2) The time at which the influence of the
corpus luteum preventing ovulation ceases to be exerted. Our
observations make it very probable that the retrogressive changes
observed in the corpora lutea before ovulation indicate the neces-
sary cessation of functional activity. It is however noteworthy
that, notwithstanding such a cessation of activity, mitoses can
still be observed in the lutein cells at this period. Whether the
corpus luteum acts principally upon the last stage in the develop-
ment of follicles (maturation) or merely upon the rupture of folli-
cles will still have to be determined with certainly. We recall
however the fact that we observed the occurrence of mature
follicles during various stages of pregnancy, notwithstanding the
existence of corpora lutea. (3) Certain more or less accidental con-
ditions, as for instance copulation. It is probable that other cir-
cumstances also may accelerate or retard the rupture of the follicles.
Such factors act probably indirectly by causing changes in the
circulation in the ovaries. In the guinea pig these are not indis-
pensable, but their place can be taken by other factors; or even the
total absence of corpora lutea may in the guinea pig be sufficient
to allow a new ovulation.

68 LEO LOEB

In the guinea pig ovulation occurs in the large majority of cases
without any previous copulation. In many cases however copula-
tion is not without significance even in the guinea pig; it acceler-
ates ovulation. While, after delivery, a spontaneous rupture may
take place without copulation, in other cases it does not occur with-
out ovulation. Also in the ordinary period of heat ovulation does
not need to take place without copulation. Copulation is there-
fore not without importance; but in almost all of these cases ovu-
lation is only deferred and sooner or later it will take place with-
out the male. So far as the literature has been accessible to us it
appears that the role copulation plays had not been fully appre-
ciated by former investigators. Certain observations which we
made indicate that other factors besides a preceding copulation
may influence ovulation, and we intend to continue our investi-
gation in this direction.

Our observations enable us to give some data concerning the
time relations in the growth of various ovarian structures.

a Follicles. In about six days after ovulation small follicles
reach medium size. In approximately eight days large follicles
have developed and now degenerative processes set in. Mitotic
cell division is most pronounced in the granulosa before degenera-
tive processes have commenced; but mitoses may still be seen, if
a slight degree of degeneration exist.

h Ordinary corpora lutea. The development of corpora lutea
within the first six days after ovulation has been described in a
previous paper. At six days we see for the first time, besides the
capillary vessels, blood vessels with walls consisting of two rows
of cells penetrating into the corpus luteum; they become some-
what more frequent from the tenth day on. In the meantime
mitotic division of lutein cells continues and the increase in these
cells causes the central connective tissue to become smaller in
amount.

In corpora lutea ten to eleven days old a few vacuolar cells are
present in the periphery of the corpus luteum. From ten to fif-
teen days after ovulation vacuolization is still very slight in peri-
pheral luetein cells. From fifteen to eighteen days more fine or
coarse vacuolization may appear. Other lutein cells are still

CYCLIC CHANGES IN THE OVARY OF GUINEA PIG 69

more solid and mitoses are still present. If no new ovulation have
taken place, degeneration becomes more marked after twenty days;
twenty-four days after ovulation we noticed a small amount of
connective tissue growing into the periphery. At this period the
number of mitoses is already diminished. In cases in which,
between the eighteenth and twenty-sixth day after ovulation, a
new rupture of follicles sets in, the degenerative processes are
still more marked; mitoses may still be seen in the course of the
first day after rupture of the follicles, but they disappear after-
wards and the degenerative processes progress. The vacuoliza-
tion of the lutein cells increases, the corpora lutea shrink, the con-
nective tissue becomes gradually hyaline and is relatively pre-
ponderating in quantity over the lutein cells. About six days after
the new ovulation (in approximately twenty-six days old corpora
lutea) yellow pigment may be seen for the first time in the vacuo-
lar lutein cells. Eight days after the new ovulation the corpus
luteum is much shrunken, and ten to eleven days after the new
ovulation corpora lutea approximately thirty-one to thirty-two
days old have been reduced to small vacuolar bodies, around
which a strong connective tissue capsule may appear. Corpora
lutea thirty-three to forty days old (twelve to nineteen days
after new ovulation) still represent vacuolar bodies; but now
gradually the transformation into a yellow body sets in. Corpora
lutea about forty-five days old have the appearance of yellow
bodies and they may probably persist as such for a long time,
after the third ovulation has taken place. Thus three generations
of corpora lutea may be present side by side in the same ovary,
c. Corpus luteum of pregnancy. The corpus luteum of pregnancy
differs from the ordinary corpus luteum in the longer duration of
mitotic division, and the delay in retrogressive changes. Although
slight vacuolization may be noticeable at relatively early stages,
the corpora lutea of pregnancy are still in a good condition thirty-
five to forty days after ovulation and they may still show mitoses
at this period. Towards the latter part of pregnancy however
degenerative processes set in, vacuolization and loss in staining
power of the nuclei, and other changes, are noticeable. Mitoses
could not be seen at this stage, and they appeared to be absent

70 LEO LOEB

after delivery had taken place. From ten to twelve days after
delivery yellow pigment was seen in a few of the lutein cells in the
corpus luteum of the previous pregnancy.

Thirteen to twenty days after delivery the corpus luteum is still
much larger than an ordinary corpus luteum at the same period
after ovulation, but considerable shrinking has already taken
place. Twenty-seven days after delivery the corpus luteum is
very small and vacuolar, with much hyaline connective tissue,
but has not yet been transformed into a yellow body; but at a
later stage, approximately sixty days after delivery (or possibly
somewhat earlier) the corpus luteum appears as a yellow body,
and as such it may persist for some time.

d. In the developing ovaries degeneration of the granulosa and
connective tissue atresia of follicles are found as soon as the folli-
cles have reached a relatively large size; these retrogressive changes
first appear in guinea pigs approximately fourteen to eighteen
days old, while the first ovulation appears much later, namely
two to three and a half months after birth.

STUDIES ON CHROMOSOMES

VII. A REVIEW OF THE CHROMOSOMES OF NEZARA; WITH SOME
MORE GENERAL CONSIDERATIONS

EDMUND B. WILSON

From the Zoological Department, Columbia University

NINE FIGURES AND ONE PLATE

CONTENTS

Introduction 71

Descriptive 73

1 The second spermatocyte-division in Nezara 73

a The idiochromosomes 73

b The double chromosome 77

2 The first spermatocyte-division 78

3 The growth-period and spermatocyte-prophases 80

4 The diploid chromosome-groups 83

General 84

5 The idiochromosomes 84

a Composition and origin cf the XY-pair 85

b Modifications of the X-element 88

c Sex-limited heredity 94

d Secondary sexual characters 99

6 Modes in which the chromosome-number may change 99

Conclusion 105

INTRODUCTION

In the first of these 'Studies' ('05a) I described the idiochromo-
somes (X and Y-chromosomes) of Nezara hilaris as being of equal
size in the male, and reached the conclusion that in this species
no visible dimorphism appears in the spermatid-nuclei. In my
third 'Study' ('06), after examination of the female diploid groups,
this species was assigned a unique position as the single then

72 EDMUND B. WILSON

known representative of a type in which a pair of idiochromo-
somes can be identified in both sexes, but are of equal size in
both, and in which, accordingly, no visible sexual differences ap-
pear in the diploid nuclei. These conclusions, as is now appar-
ent, were based upon a wrong identification of the idiochromo-
some-pair, which is not the smallest pair, as I then believed, ^ but
one of the largest. When this fact was recognized, the true con-
ditions soon became evident.

I was led to re-examine Nezara hilaris by the fact (very sur-
prising to me) that in Nezara viridula, a southern species closely
similar to N. hilaris, the idiochromosomes of the male are ex-
tremely unequal in size, and the dimorphism of the spermatid-
nuclei is correspondingly marked. Upon returning to the study
of N. hilaris it soon became manifest that the dimorphism is
present in this species also, though in far less conspicuous form.
The size-difference between the X- and Y-chromosomes is here
often so slight that I did not at first distinguish it from an incon-
stant fluctuation of size, such as is sometimes seen between the
members of the other chromosome-pairs. When, however, the
identity of the XY-pair was correctly recognized, its constancy
of position and of size in the second division enabled me to make
an accurate comparison between it and the other bivalents; and
this fully established the constant inequality of its members,
which is constantly greater than that now and then seen in other
pairs. Both species also exhibit some other very interesting
features that I overlooked in my former studies.

Nezara can therefore no longer stand as a representative of
the third of the types distinguished in my third 'Study,' but
belongs with Euschistus, Lygaeus, etc., in the second type

1 This was in part because in most of the other forms known at the time the
idiochromosomes are in fact the smallest, or one of the smallest, pairs. In part,
also, I followed Montgomery ('01) who described in this species two small " chro-
matin nucleoli" in the spermatogonial groups, and believed them to be identical
with the chromatic nucleolus of the growth-period. In a later paper ('06) Mont-
gomery states these "chromatin nucleoli" to be "apparently not quite equal in
volume," and asserts that I was in error in describing them as equal. In my
material they are certainly equal in the great majority of cases. However, this
is not the idiochromosome-pair.

STUDIES ON CHROMOSOMES 73

DESCRIPTIVE

Since the two species agree very closely save in respect to the
idiochromosomes they may conveniently be considered together.
Before describing the divisions, attention may be called to a
striking difference between the two species in respect to the size
of the cells and karyokinetic figures. As a comparison of the
figures will show, the spermatocytes and maturation division-
figures of N. hilaris are much larger than those of N. viridula.
In the spermatogonia this difference is also apparent, though less
marked. In the ovaries, strange to say, it cannot certainly be
detected, either in the dividing cells or in the nuclei of the follicle-
cells or of the tip-cells at the upper end of the ovary. It would be
interesting to make a more accurate study of these relations;
but I will here only state that the differences between the two
species seem to arise mainly through greater growth of the
spermatocytes in N. hilaris. With this is correlated a greater
size of the testis as a whole; but the size of the entire body in this
species is but little larger, as far as I have observed, than in N.
viridula.

As regards the general features of the divisions, the diploid
groups of both sexes uniformly contain fourteen chromosomes,
the first spermatocyte-division eight and the second seven, the
idiochromosomes being, as is the rule in Hemiptera, separate and
univalent in the first division.

1. The second spermatocyte-division

a. The idiochromosomes. Polar views of the second division
always show 7 chromosomes which are usually grouped in an
irregular ring of six with the seventh near its center (fig. 3 j-m,
figs. 14, 15). In both species one chromosome of the outer ring
(s) can usually be distinguished as the smallest, though this is
not always evident owing to the apparent variations produced
by different degrees of elongation. This is the chromosome that
I formerly supposed to be the idiochromosome-bivalent, despite
its peripheral position, and despite the fact, which I had myself
described, that a similar small chromosome, also peripheral in posi-

JOUBNAL OF MORPHOLOGY, VOL. 22, NO. 1

EDMUND B. WILSON

EXPLANATION OF TEXT FIGURES

Figures 1 to 9 are from camera drawings, and are not schematized except
that in a few instances the chromosomes have been artificially spread out in a
series in order to facilitate comparison. Figs. 2 k-l are somewhat more enlarged
than the others. In all the figures d denotes the double chromosome or 'd-chro-
mosome,' s the small chromosome, X the large idiochromosome and Y the small.

*..»

5 /

Fig. 1 The second spermatocyte-division in Nezara viridula. a-d, metaphases
in side view; e-gr, anaphases; h, i, polar views of two sister-groups, middle ana-
phase, from the same spindle and in the same section.

tion, appears in several other pentatomids (e.g., in Euschistus, Coe-
nus and Mineiis). But Nezara forms no exception to the rule that
the central chromosome is the idiochromosome-bivalent. In N.
viridula this is immediately apparent in side views (often also in
polar views) where the central chromosome is seen to consist of two
very unequal components, the smaller being not more than one
fourth or one fifth the size of the larger (fig. 1 a-c). In the ana-
phases these separate and pass to opposite poles, while all the others
divide equally (fig. 1 e-g). Polar views of middle or rather late
anaphases, when both daughter-groups can be seen superposed
in the same section, clearly show the marked difference of the
two groups in respect to the idiochromosomes (fig. 1 h-i). All the
facts are here so nearly similar to those seen in Euschistus or
Lygaeus as to require no further description.

STUDIES ON CHROMOSOMES 75

In N. hilaris the conditions differ only in tiiat the two compo-
nents of the central chromosome are but sUghtly miequal; but in
the examination of at least two hundred of these divisions I have
never failed to detect the inequaUty. A series of side views are
shown in fig. 2 a-i, figs. 16-21, two of which show all the chromo-
somes. These figures illustrate practically all the variations
that have been seen in the idiochromosomes. The most charac-
teristic condition is that seen in 2 a, b, d, in which both idiochro-
mosomes (X and Y) are more or less elongated and united end to
end. Less often one of them assumes a more spheroidal form
(fig. 2 e, h, i, fig. 17). The size-difference, though always evident,
seems to vary slightly (perhaps because one or the other compo-
nent may be more or less compressed laterally), but is always dis-
tinctly greater than that now and then seen in other bivalents.

Fig. 2 j shows a mid-anaphase^ (cf. figs. 21-23) in which the
inequality would hardly be noticed without close study and the
comparison of other cases. Figs. 2 k and I are similar stages
showing all the chromosomes spread out in a series for the sake
of comparison. In both, the two idiochromosomes are easily
distinguishable,^ and the larger is seen to be 07ie of the three largest
chromosomes. Figs. 2 m-n, o-p, q-r and s-t are pairs of sister-
groups, in each case from the same spindle in anaphase. All of
these are selected from cases in which a distinct size-difference
appears between X and Y, but there are also many cases in which
this cannot be seen. Such a case was figured in fig. 4 e-f of my
first ' Study' the correctness of which is confirmed by re-examina-
tion of the original section. This condition is due simply to the
fact that the large idiochromosome is more elongated than the
small, so that the size-difference cannot be seen in polar view;
and for the same reason it is often not evident in polar views of
the metaphase.

2 This and the two following figures are a little more enlarged than the others .

' Fig. 2 I is the same group figured in fig. 4 d of my first 'Study,' carefully redrawn
and corrected. A comparison of the two drawings will show that in the latter a
distinct size-difference between X and Y is actually shown but is minimized by
the fact that the former is represented a trifle too small, the latter a little too
large. It is now also evident that they are connected by two connecting fibres
instead of by one.

EDMUND B. WILSON

%^ ill * ! ;

P b

• I'

4v "^

"^^ m ^
P n %^p

Fig. 2 The second spermatocyte-division in Nezara hilaris. a-i, metaphase
figures in side view, a and e showing all the chromosomes; j~l, mid-anaphases; in
A; and I all the chromosomes are shown artificially spread out in series; m-n, o-p,
q-r, s-t, four pairs of sister-groups from late anaphases, in polar view, in each case
from the same spindle.

*• r •^ t

STUDIES ON CHROMOSOMES 77

b. The double chromosome. A seco;iid interesting feature of the
second division that I formerly overlooked is the presence of a
remarkable double chromosome which in the metaphase has ex-
actly the appearance of a butterfly with widespread wings. This
chromosome (which may be called the d-chromosome) is shown in
profile view in 2 b~e and 1 a-d, 16, 17, 20, 24, 25. This is the only
chromosome in the second division that shows any approach to a
quadripartite form, audits characters are so marked as to constitute
the most striking single feature of the division. As the figures
show, it is one of the largest of all the chromosomes. It always
has an asymmetrical tetrad shape, giving exactly the appearance
of a smaller and a larger dyad in close union; and it always lies
in the outer ring, so placed as to undergo an equal division, and
with the larger wings of the butterfly turned towards the axis of
the spindle. In polar view (3 j-m) the duality is far less apparent
and sometimes invisible, even upon careful focussing. In N.
viridula the duality is always apparent in side view, but the but-
terfly shape is usually less evident than in N. hilaris.

In the initial anaphases the (i-chromosome divides symmetri-
cally, drawing apart into two bipartite chromosomes (2 j, k, I g);
but this is seldom evident save in profile view. Viewed from the
pole the duality does not now ordinarily appear, though it may
still sometimes be seen upon careful focussing. In the later ana-
phases the two components tend to fuse, and often can no longer
be distinguished. Not seldom, however, the duality is visible
even in the final anaphases; and sometimes this is so conspicuous
that the spermatid-group seems at first sight to comprise eight
instead of seven separate chromosomes (n, r, s, t).

Since the duality of this chromosome does not certainly appear
in the spermatogonial groups or in the first spermatocyte-division,
its peculiar form in the second division might be supposed to
result from some special mechanical relation to the spindle-fibers
in that division. This is, however, excluded by examination of
the interkinesis, in which the chromosomes are irregularly scat-
tered. • In these stages, even when the spindle is still very small
and the chromosomes lie in a quite irregular group, the butterfly
shape is already perfectly evident; and it shows no constancy of

78 EDMUND B. WILSON

relation to the spindle-axis^ often lying at right angles to the
latter. Apparently therefore its duality arises quite independ-
ently of the spindle or astral rays, and its constant position in
the fully formed spindle is the result of a later adjustment. In
this species, as in many others, each chromosome is connected with
the pole by a bundle of delicate fibers. In case of the d-chromo-
some this bundle is very broad, but I cannot be sure that it is
double.

At first sight any observer would, I think, take the c?-chromo-
some to be merely a result of the accidental superposition or close
adhesion of two separate dyads of unequal size ; but such an inter-
pretation is inadmissible. When all the chromosomes can be
unmistakably seen, the d-chromosome is found to constitute one
of the seven separate elements invariably present in this division;
and since the diploid number is 14 in both sexes this chromosome
must represent one chromosome, not two, of the original sperma-
togonial groups. It is certain, therefore, that the double appear-
ance does not result from close apposition of two separate chromo-
somes; it is therefore not a ''tetrad" in the ordinary sense of the
word — i.e., not one that results from the synapsis of two chromo-
somes that are originally separate in the diploid groups.

2. The first spermatocyte-divisiori

This division requires only brief mention. As stated, it shows
eight separate chromosomes, of which the only one that can be
positively identified is the Y-chromosome of N. viridula. This
chromosome, always immediately recognizable in this species
by its small size (3 c, d, f, g, i), figs. 12, 13), is usually central in
position, like the m-chromosome of the Coreidae, but this is not
invariable. Since it divides equally, and without association with
any other chromosome (3 g) it is evident that the two idiochro-
mosomes must be separate and univalent in this division. In N.
hilaris (3 a, b, figs. 10, 11) the eight chromosomes usually form an
irregular ring, there is no central chromosome, and neither idio-
chromosome can be certainly recognized. It nevertheless seems
a safe inference from what is seen in N. viridula that the two
idiochromosomes are here also separate and univalent.

STUDIES ON CHROMOSOMES

• •

•
f

« tifiXtl

xv-i

d ^

•c

^•%^ mid

Fig. 3 First and second spermatocyte-divisions in the two species of Nezara.
a, 6, first division, hilaris, polar views: c, d, corresponding views of viridula; first
division, hilaris, side view showing five of the chromosomes in position and the
other three to the right above;/, corresponding view of viridula; g, middle ana-
phase, viridula, showing division of Y; h, first division metaphase, hilaris, all the
chromosomes artificially spread out in series; i, corresponding view of viridula;
2, k, second division metaphase, hilaris, polar views; I, m, corresponding views
of viridula.

80 EDMUND B. WILSON

In this division the d-chromosome can not be identified in either
species. Figs. S e, f, h, i, show all the chromosomes of the two
species, in each case from a single spindle in side view. Most
of them have a simple bipartite form, but in each species two or
three of them often appear more or less distinctly quadripartite
as is, of course, often the case with the bivalents in this division.
In N. hilaris one of the largest chromosomes is usually more
elongated than the others, and each half shows a slight trans-
verse constriction. I suspect that this may be the d-chromosome,
but cannot establish the identification.

3. The groivth-period and spermatocyte-prophases

These stages fully bear out the conclusions based upon the
divisions and establish the identity of the idiochromosome-pair
with the chromatic nucleolus of the growth-period. Throughout
the growth-period each nucleus contains a single intensely stain-
ing spheroidal chromatic nucleolus and in addition a very large,
■clearly defined pale plasmosome, which is sometimes double.
Series of drawings of these two bodies (in each case from the same
nucleus, and in their relative position) are given in figs. 4 i-l and
m-p, from cells of the middle growth-period. They are also
shown in figs. 26-29. In these stages no sign of duahty is to be
seen in the chromatic nucleolus, even after long extraction or in
saffranin preparations. In later stages, as the chromosomes begin
to condense, this nucleolus becomes less regular in outline, and
gradually assumes a tetrad form, which becomes very clear as
the chromosomes assume their final shape. This transformation
may be traced without a break, successive stages being often seen
within the same cyst. Just before the nuclear wall breaks down
this tetrad is still clearly distinguishable from the others by its
asymmetrical quadripartite form, as seen in 4 y, z, which show all
the chromosomes (in each case from two successive sections).
Figs. 4 q-t show four views of this tetrad at this period in N.
hilaris, while u-x are corresponding views of N. viridula. These
figures (which might be indefinitely multiplied) show the marked
differences between the two species in respect to this tetrad,
obviously corresponding to that seen between the idiochromosome

Fig. 4 The diploid groups, nucleoli of the growth-period, and late prophase-
figures of the two species of Nezara. a, b, spermatogonial groups, hilaris ; c, d, the
same, viridula; e,f, ovarian groups, hilaris; g, h, the same, viridula; i-l, chromatic
nucleolus and plasmosome from the growth-period, in each case from the same
nucleolus in their relative position; m-p, corresponding views, viridula; q-t, the
idiochromosome-tetrad (chromatic nucleolus) from prophase nucleoli, hilaris;
u-x, corresponding views, viridula; y, late prophase nucleus, showing all the
chromosomes, hilaris (combination figure from two sections) ; z, corresponding
viridula, three of the chromosomes from adjoining section at the right.

82 EDMUND B. WILSON

bivalents of the two in the second division. ^ The two species
may in fact readily be distinguished by mere inspection of the
chromatic nucleolus at this period. Already at this time the two
components are here and there seen to be separating, but as a
rule they do not finally move apart until the nuclear wall has
dissolved. From this time forward they cannot be individually
identified with exception of the small idiochromosome of N.
viridula, which is obvious at every period.

As far as my material shows, the earlier stages of the idio-
chromosomes can not be so readily traced in Nezara as in some
other species, and the chromatic nucleolus can not actually be fol-
lowed backward to the spermatogonial telophases — as can be
done in such forms as Lygaeus or Oncopeltus, of which a detailed
account will be given in a later publication. The prophase -figures,
however, decisively establish its identity with an unequal pair of
chromosomes that divide separately in the first spermatocyte-
division; and in N. viridula, one of these is certainly the small
idiochromosome. It may therefore confidently be concluded
that the chromatic nucleolus is identical with the idiochromo-
some-pair, as in so many other cases. Comparison of the division-
figures proves that this pair can not be identical with the small
pair that I formerly supposed to be the idiochromosome-pair;
and this small pair is moreover usually recognizable in the pro-
phase groups (s, in 5 y, z) in addition to the unequal pair.

The foregoing facts make it clear that in Nezara the idiochromo-
somes undergo a process of synapsis at the same time with the
other chromosome-pairs, and that their separation before the first
division is a secendary process, to be followed by a second conju-
gation after this division is completed. A similar process often
takes place in many other Hemiptera. There are, however, some
forms, like Oncopeltus, in which the idiochromosomes are always
separate, from the last spermatogonial division through all the suc-
ceeding stages up to the end of the first division. In this case,
which I shall describe more fully hereafter, there can be no doubt
that the conjugation which follows the first division is a primary
synapsis, to be immediately followed by a disjunction.

^ CJ. the earlier figures of the corresponding tetrad in Brochymena in my first
'Studv,' fis. 7.

STUDIES ON CHROMOSOMES 83

4. The diploid chromosome-groups

In these groups the interest centers again in the identity of the
idiochromosomes and the d-chromosome. Of the 14 separate
chrosomomes present in the diploid nuclei of both sexes, none
shows any constant indication of duality (figs. 4 a-h). The d-
chromosome can not, therefore, be identified in these stages.
Secondly, in both species the diploid groups of the two sexes show
the same relation as in other Hemiptera of this type, though this
is, of course, more readily seen in N. viridula than in hilaris, owing
to the small size of the Y-chromosome. In the spermatogonia!
groups of this species (4 c, d) this chromosome is at once recog-
nizable while in the female groups {g, h) it is lacking, its place
being taken by one of larger size. In both sexes the small pair
(s, s) is also recognizable. In this species, accordingly, the Y-
chromosome is confined to the male line, and the Y-class of
spermatozoa must be male-producing, as in other forms.

In N. hilaris the Y-chromosome can not be identified (4 a, h),
but the relation of the spermatozoa to sex-production is shown in
another way, though less unmistakably than in N. viridula. As
already described, the large idiochromosome or X-chromosome is
one of the largest three chromosomes seen in the second division.
We should therefore expect to see five largest chromosomes in
the male diploid groups. This is clearly apparent in figs. 4 a, h,
and is also shown in the corresponding figures of N. viridula
(c, d) though not quite so clearly. One of these five in the male
should be the X-chromosome; and if the usual relation of the
spermatozoa to sex hold true, there should be six largest chro-
mosomes in the diploid groups of the female. This relation actu-
ally appears in nearly all cases, and is shown in figs. 4 e, f, g, h, in
each of which, again, the small pair (s, s) may be recognized.
Though this evidence is in itself less convincing than that afforded
by N. viridula (since the relation can not always be made out
with certainty) it is fully in harmony with the latter, and sustains
the same conclusion.^

^ This relation is shown in my original figures of N. hilaris, though not quite
as clearly as in the groups here figured. In my first 'Study' ('05) the five largest
chromosomes are very clearly shown in fig. 4 h, and are also evident in 4 q. In the
third 'Study' the relation is hardly evident in the male but fairly so in the
female (figs. 5 I, m).

84 EDMUND B. WILSON

GENERAL

5. The idiochromosomes

The case of Nezara shows how readily a morphological dimorph-
ism of the spermatid-nuclei may be overlooked when the X- and
Y-chromosomes do not differ markedly in size ; and it emphasizes
the necessity for the closest scrutiny of these chromosomes in the
study of this general question. In my fourth 'Study' I placed
with Nezara hilaris, as a second example of my original 'third
type/ the lygaeid species Oncopeltus fasciatus (Dall.), on the
strength of Montgomery's account of the conditions in the male
('01, '06) and my own unpublished observations on both sexes.
While I have carefully re-examined this case also, I am not yet
prepared to express an unqualified opinion in regard to it. Cer-
tainly, in very many of the cells of this species, at every period of
the spermatogenesis, the idiochromosomes (which are always sep-
arate up to the second division) seem to be perfectly equal. A
slight inequality may indeed be seen in some cases; but as far as I
can yet determine this seems to fall within the range of the size-
variation in other chromosomes and gives no positive ground for
the recognition of a morphological dimorphism in the spermatozoa.
A similar condition has been described in several other insects, not-
ably in some of the Lepidoptera (Stevens, '06; Dederer, '08; Cook,
'10), in the earwig Anisolaba (Randolph, '08) and apparently also
in the beetle Hydrophilus according to Arnold ('08). I see no rea-
son to question these observations; but the interpretation to be
placed on them is by no means clear at present. The experimental
evidence on the Lepidoptera seems to demonstrate that in at least
one case^that of Abraxas according to Doncaster and Raynor, —
it is the eggs and not the spermatozoa that are sexually dimorphic ;
that is, in the terms that I have recently suggested ('10a), in
this case it is the female that is sexually 'digametic' whUe the
male is 'homogametic' Baltzer's careful work on the sea-
urchins ('09) clearly demonstrates a cytological sexual dimorphism
in the mature eggs of these animals, and shows that the sperm-
nuclei are all alike. In cases, therefore, where no visible dimorph-
ism of the spermatid-nuclei is demonstrable, two possibilities

STUDIES ON CHROMOSOMES 85

are to be considered, namely, (1) that it may be the female which
(as in sea-urchins) is the digametic sex, and (2) that one sex or
the other may still be physiologically digametic even though this
condition is not visibly expressed in the chromosomes. The
first of these possibilities may readily be tested by cytological
examination of the female groups. The second can only be
examined by means of experiment, and especially by experiments
on sex-limited heredity. It is interesting that the work of Don-
caster and Raynor, cited above, and the more recent one of Morgan
on Drosophila ('10) have given exactly converse results, the former
demonstrating a sexual dimorphism of the eggs, the latter of the
spermatozoa. This agrees with the cytological data, as far as
they have been worked out. The researches of Stevens ('08, 10),
on the Diptera establish the cytological dimorphism of the sper-
matozoa in these animals, while all observers of the Lepidoptera
have thus far failed to find such dimorphism in this group. It
thus becomes a very interesting question whether a cytological
dimorphism of the mature eggs may be demonstrable in the
Lepidoptera; though a failure to find it would in no wise lessen
the force of the experimental data. Physiological differences be-
tween the chromosomes are of course not necessarily accompanied
by corresponding morphological ones — indeed such a correlation
is probably exceptional.

(1) (a) Composition and origin of the XY~pair. The facts
seen in Nezara again force upon our attention the puzzle of the
Y-chromosome or 'small idiochromosome.' It is remarkable
that two species so nearly akin as N. hilaris and N. viridula should
differ so widely in respect to this chromosome; though this is
hardly so surprising as the fact that in Metapodius this chromo-
some, as I have shown ('09, '10) may actually either be present or
absent in different individuals of the same species. These facts
show, as I have urged, that although the Y-chromosome shows a
constant relation to sex when it is present, it can not be an essen-
tial factor in sex-production. As the case now stands this might
be taken as a direct piece of evidence against the view that the
idiochromosomes are concerned with sex-heredity. Further, as
I have pointed out ('10) in Metapodius the introduction of super-

86 EDMUND B. "WILSON

numerary Y-chromosomes into the female has no visible effect
upon any of the characters of the animal, sexual or otherwise;
and this might be urged against the whole conception of qualita-
tive differences among the chromosomes and of their determina-
tive action in development. It is especially in view of these
and certain other general questions that I wish to indicate some
of the many possibilities that must be taken into account in the
consideration of this problem. My discussion is throughout
based upon the assumption that the chromosomes do in fact
play some definite role in determination, and that they exhibit
qualitative differences in this respect. I do not hold that they
are the exclusive factors of determination; though it is often con-
venient, for the sake of brevity, to speak of them as if they were
such.

(2) Cytologically considered, the morphological dimorphism
of the spermatozoa seems to have arisen by the transformation
of what was originally a single pair of chromosomes comparable
to the other synaptic pairs. We have at present no information
as to whether the members of this pair were equal or unequal in
size; but in either case there are grounds for the assumption that
its two members differed in some definite way in respect to the
quality of the chromatin of which they were composed. This
pair, which may be called the priixiitive XY-pair, has undergone
many modifications in different species, but without altering its
essential relation to sex. In the insects (Hemiptera, Coleoptera,
Diptera) its most frequent condition is that of an unequal pair, con-
sisting of a 'large idiochromosome' or 'X-chromosome,' and a
''small idiochromosome" or ' Y-chromosome,' the latter being con-
fined to the male line, while the former appears in both sexes —
single in the male and paired in the female. That all gradations
exist between cases where X and Y are very unequal (as in many
Coleoptera and Diptera and in some Hemiptera) and those in which
they are nearly or quite equal (Mineus, Nezara, Oncopeltus) gives
some ground for the conclusion that in the original type the
XY-pair was but slightly if at all unequal.

By disappearance of the free Y-member of this pair has arisen
the unpaired odd or 'accessory' chromosome, which accordingly

STUDIES ON CHROMOSOMES 87

has no synaptic mate. This condition seems to have arisen in
more than one way. It is almost certain that in many cases the
Y-chromosome has disappeared by a process of gradual and pro-
gressive reduction (as indicated by the graded series observed
in the Hemiptera (Wilson, '056, '06). In some cases (of which
Metapodius is an example) the same result may have been pro-
duced suddenly by a failure of the idiochromosomes to separate
in the second spermatocyte-division (Wilson, '096). A third pos-
sibility, first suggested by Stevens ('06), is that the X-element
may have separated from a YY-pair with which it was originally
united. This possibility seems to be supported by recent obser-
vations on Ascaris megalocephala, where the X-chromosome is
sometimes fused with one of the other pairs, sometimes free
(Edwards, '10).

(3) We have as yet no positive knowledge as to how the X-
niember of the XY-pair originally differed, or now differs, from
the Y, or as to how this difference arose— a definite answer to
these questions would probably give the solution of the essential
problem of sex. There are, however, pretty definite grounds for
the hypothesis that the X-member contains a specific ' X-chroma-
tin' that is not present in the Y-member, and that the XY-pair
is heterozygous in this respect. If this be so, the primary sexual
differentiation is therefore traceable to a condition of plus or
minus in this pair, accompanied by a corresponding difference
between the nuclear constitution of the two sexes. (Cf. Wilson,
'10a.) Further, there is also reason for regarding the heterozy-
gous condition of this pair as due to the presence of the X-chroma-
tin in one member of a pair which is (or originally was) homozy-
gous in respect to its other constituents. The latter may be
called collectively the 'Y-chromatin'; and we may, accordingly,
think of the XY-pair as being essentially a YY-pair with one
member of which the X-chromatin is associated.^ Both the X-

^ This suggestion is in principle the same as one earlier made by Stevens ('06,
p. 54) that the Y-chromosome represents "some character or characters which
are correlated with the sex-character in some species but not in others," with one
member of which the X-chromosome is fused; and that "a pair of small chromo-
somes might be subtracted from the unequal pair, leaving an odd chromosome."

55 EDMUND B. WILSON

chromatin and the Y may themselves be composite, thus giving
the possibility of many secondary modifications. The point of
view thus afforded opens many possibilities for an understanding
of sex-limited heredity, as indicated beyond.

(6) Modifications of the X-element. This view of the XY-pair
is based upon two series of facts, of which the first includes the
various modifications of the X-member of the pair seen in dif-
ferent species. It is, perhaps, most directly suggested by a study
of the pentatomid species Thyanta custator. ' In this common and
widely distributed species I have found two races, which thus far
can not be distinguished by competent systematists, ^ but which
differ in a remarkable way in respect to both the total number
of chromosomes and the XY-pair. In one of these races (which
I will call the 'A form'), widely distributed throughout the south
and west, the total number in both sexes is 16, and the XY-pair
of the male is a typical unequal pair of idiochromosomes, exactly
like that seen in many other pentatomids {e.g., Euschistus,
Coenus or Banasa). These are shown in fig. 5 a, b, their mode
of distribution being the usual one. The second race (the 'B
form') is thus far known from only a single locality in northern
New Jersey. It differs so remarkably from the A fonn that I
could not believe the observations to be trustworthy until repeated
study of material, collected in four successive years, established the
perfect constancy of the cytological conditions and the apparent
external identity of the two forms. In this race the XY-pair is
represented by three small chromosomes of equal size, which are
always separate in the diploid groups and in the first spermato-
cyte-di vision (fig. 5i), but in the second division are united to
form a linear triad series (5 c, d) . This group so divides that one
component passes to one pole and two to the other {oe, h), the

^ I am indebted to Mr. E. P. Van Duzee for a careful study of my whole series
of specimens of both races. He could find no constant differential between them.
Additional studies of this material are now being made by Mr. H. G. Barber.

Addendum. Since this paper was sent to press Mr. Barber, after prolonged
study, has reported his conclusion that the 'A form' is Thyanta custator of
Fabricius, while the 'B form' is probably Thyanta calceata of Say, which has
long been regarded as a synonym of former species.

STUDIES ON CHROMOSOMES

» •

•••••», ••;.•#

III *i'it t

f X 9 p T 'g

• •

Fig. 5 Comparison of the XY-group in various Hemiptera. (a-i are orig-
inal; the others from Payne.) a, b, Thyanta custator, 'A form,' second division
in side view; c, d, corresponding views of the 'B form'; e-h, anaphases of same;
i, polar view of first division of same; j, k, metaphase chromosomes, second divi-
sion, Diplocodus exsanguis; i, similar view of Rocconot^ anrtailicornis; m, similar
view of Conorhinus sanguisugus; n, Sinea diadema; o, Prionidus cristatus; p,
Gelastocoris oculatus; q, anaphase chromosomes of the same species; r, the XY-
group, from the second division, AchoUamultispinosa; s, diagram, slightly modified
from Payne, to show the distribution of the XY-components in the second divi-
sion of the same species.

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1

90 EDMUND B. WILSON

latter being usually in close contact and in later anaphases some-
times hardly separable (5^), though now and then all three compo-
nents are for a time strung separately along the spindle in the
early anaphases, so that no doubt of their distinctness can exist
(5/). Comparison of the diploid groups of the two sexes shows
that those of the male contain but three of these small chromo-
somes and those of the female four, the total respective num-
bers being 27 and 28 (instead of 16 in both sexes, as in the A
form).

These facts make it perfectly clear that one of the small chromo-
somes in the male passes to the male-producing pole, and therefore
corresponds to the Y-chromosome ; while the other two, taken to-
gether, represent the large idiochromosome, or X-chromosome, of
the A form — precisely as in the reduvioids the single X-chromosome
of Diplocodus is represented by a double element in Fitchia,
Rocconota or Conorhinus (Payne). Had we no other evidence
on this point we might assume simply that the original X-chro-
mosome has divided into two equivalent X-chromosomes. But
there are other facts that give reason for the conclusion that the
breaking up of a single X-chromosome into separate components
means something more than this. In the B form, as in' Fitchia or
Rocconota (fig. 5 I), the X-element consists of two equal compo-
nents, but in Conorhinus the two components are always of un-
equal size (5 m). In Prionidus and in Sinea there are three equal
components (5 n, o), in Gelastocoris four equal ones (5 p, and
in A choUa multispinosa five, of which two are relatively large
and equal and three very small (5 r, s). In every case these com-
ponents, though quite separate in the diploid groups (and usually
also in the first spermatocyte-division) act as a unit in the second
division, though not fused, and pass together to the female-produc-
ing pole (Payne, '09, '10).

In the foregoing examples the X-element is accompanied by a
synaptic mate or Y-chromosome. The following are examples of
a similar breaking up of the X-element into separate components
when such a synaptic mate is missing. In Phylloxera (Morgan)
the X-element consists of two unequal components, sometimes
separate, sometimes fused together. In Syromastes (Gross,

STUDIES ON CHROMOSOMES 91

Wilson) it consists of two unequal components, always separate,
in the diploid groups but closely in contact (not fused) in both
spermatocyte-divisions. The recent work of Guyer ('10) indi-
cates a similar condition in the X-element of man. In Agalena
(Wallace) there are two equal components, always separate.
Finally, in Ascaris lumbricoides (Edwards, '10) there are five com-
ponents, separate, and scattered in the diploid groups but closely
associated in the spermatocyte-divisions.

In all these cases the significant fact is that not only the number
but also the size-relations of these components are constant ; and
in many of these forms this fact may be seen in such multitudes
of cells, and with such schematic clearness, as to leave no manner
of doubt. It seems impossible to understand this series of phe-
nomena unless we assume that the single X-chromosome is essen-
tially a compound body — i.e., one that consists of different con-
stituents that tend to segregate out into separate chromosomes.
We are led to suspect, further, that the composition of the X-
element, even when it is a single chromosome, may differ widely
in different species because of its great variations of size as between
different species. For instance, in the family of Coreidae it is
in some cases very large (Protenor), in others of middle size (Che-
linidea, Narnia, Anasa), in others one of the smallest of the chromo-
somes (Alydus). Similar examples might be given from other
groups.

In the case of Thyanta, therefore, it seems a fair assumption
that the double X-element of the B form likewise represents at
least a partial segregation of the X-chromatin from other con-
stituents ; and the latter may plausibly be regarded as represent-
ing the 'Y-chromatin' of the original X-member of the pair.
In other words, we may think of the triad element as a YY-pair,
one member of which is accompanied by a separate X-chromosome.
In accordance with this its formula should be X.Y.Y, while that
of the A form is XY.Y; and this may also be extended to other
forms of similar type. If this be admissible, the male formula, as
regards essential chromatin-content, becomes in general XY.Y
and the female XY.XY, both sexes being homozygous for the
Y-constituents, while in respect to X the male is heterozygous,

92 EDMUND B. WILSON

the female homozygous. The puzzle of the Y-chromosome
would thus be solved; for although a separate Y-chromosome,
when present, is confined to the male line, its disappearance
only reduces the male from a homozygote to a heterozygote in
respect to the Y-chromatin, and the introduction of supernumer-
ary Y-chromosomes into the female (as in Metapodius) brings in
no new element.

I * c

l^x ^x X^^

X X-

Fig. 6 Compound groups formed by union of the X-chromosome with other
chromosomes in the Orthoptera. (a and b, from Sinety, the others from McClung. )
a, triad group; first division of Leptynia, metaphase; b, division of similar triad in
Dixippus; c, triad group formed by union of the l^-chromosome with one of the
bivalents, first spermatocyte-prophase, Hesperotettix; d, the same element from
a metaphase group; e, the same element in the ensuing interkinesis; /, the com-
pound element of Mermiria, from a first spermatocyte prophase; g, the same ele-
ment in the metaphase (now, according to McClung, united to a second bivalent
to form a pentad) ; h, the same element after its division, in the ensuing telophase.

The same general view as that outlined above is suggested by the
constant relation known to exist in some cases between the X-
chromosome and a particular pair of the 'ordinary chromosomes.'
The first observed case of this was recorded by Sinety ('01) in
the phasmid genera Leptynia and Dixippus (fig. 6a,h), where the
X-chromosome is always attached to one of the bivalents in the

STUDIES ON CHROMOSOMES 93

first spermatocyte-division, and passes with one half of the bivalent
to one pole. Since the spermatogonial number in Leptynia (36)
is an even one and twice that of the separate chromosomes present
in the first spennatocyte-division, it may be inferred that the
X-element is already united with one of the ordinary chromo-
somes in the spermatogonia, though Sinety does not state this.
Somewhat later McClung ('05) discovered essentially similar rela-
tions in the grasshoppers Hesperotettix and Anabrus (fig. 6,
c-e) , and in case of the first named form was able to establish the
important fact that it is always the same particular bivalent with
which the X-chromosome is thus associated. In respect to the
intimacy of this association, a progressive series seems to exist,
since in Leptynia it seems to take place in the spermatogonia, in
Hesperotettix only in the prophases of the first spermatocyte-
division, while in Thyanta the union is only effected after the
first division is completed.

Finally, the recent observations of Boring ('09), Boveri ('09)
and Edwards ('10) seem to establish the fact that in Ascaris megalo-
cephala the X-element, whether in the diploid groups or in the
maturation-divisions, may either appear as a separate chromo-
some (which has the usual behavior of an accessory chromosome)
or may be indistinguishably fused with one of the ordinary chromo-
somes.

These relations may, of course, be the result of a secondary
coupling; and I m^yself formerly so interpreted them ('09c). But
in view of what is seen in Thyanta or the reduvioids we may
well keep in mind the possibility that they are expressions or
remnants of a more primitive association, like that which I have
assumed for an original XY-pair. Whatever be their origin, the
effect is the same — a definite linking of the X-chromatin with
that of one of the other pairs.

Fig. 7 shows, in purely schematic form, the general conception
of these relations that has been suggested above, the X-chromatin
being everywhere represented in black. A is the primitive XY-
pair from which all the other types may have been derived. By
simple reduction of such a pair arises the ordinary or typical
idiochromosome-pair (B) ; and from either A or B may be derived

EDMUND B. WILSON

the other types (C-G)/ or the more compHcated ones shown in
fig. 5. I represents the possible mode of separation of the
X-element from a YY-pair, as suggested by Stevens; and this
may be reaUzed in Ascaris megalocephala (H). J and K are,
schemes of the relations seen in Hesperotettix, Anabrus and
Mermiria (cf. fig. 6). These may be direct derivatives of a
primitive XY-pair, as the diagram suggests, or may be a result

^_^ 0^ C. T/iyanfo

.il^ D. Fitchla

. dl^ E. Cotiorhiniis

F. Protenor'

-^^)X G. Syro/nastes

HAscarcs

' /{e^perotetUx

Fig. 7 Diagram illustrating the possible relation of the various types of idio-
chromosomes to a primitive XY-pair. Explanation in text.

of secondary coupling of X with other elements. In either case
X may itself have such a composition as is indicated in F (Prote-
nor).

(c) Sex-limited heredity. (1) The foregoing considerations have
an important bearing on the problem of sex-limited heredity,
for they give us a very definite view of how such heredity may be
effected. It is not my intention to consider this subject in ex-

* These figures are not intended to indicate the precise mode of segregation of
the X- and Y-chromatins of the X-element, but only illustrate possible modes.

STUDIES ON CHROMOSOMES 95

tenso; but I wish to indicate some of the possibiUties that have
been opened by the cytological results, even at the risk of offering
what may be regarded as too speculative a treatment of the matter.
It is obvious that atiy recessive mutation should exhibit sex-limited
heredity when crossed with the normal or dominant form, if it be due
to a factor contained in {or omitted from) the X-element. For in-
stance, in the remarkable Drosophila mutants discovered by
Morgan ('10) the experimental data establish the fact that white
eye-color (which seems to follow the same type of heredity as
color-blindness in man) is linked with a sex-determining factor in
such a way that when the white-eyed male is crossed with the
normal red-eyed female, the former character is never transmitted
from father to son, but through the daughters to some of the
grandsons (theoretically to 50 per cent), though the daughters are
not themselves white-eyed ; that is, after such an initial cross, white
eyes fail to appear in the Fi generation in either sex and in the
F2 generation appear only in some of the males. As Morgan
points out, this follows as a matter of course if the factor for white
eye be identical with, or linked with, a sex-determining factor in
respect to which the male is heterozygous or simplex, the female
homozygous or duplex. The X-element exactly corresponds in
mode of distribution to such a sex-determining factor; for this
chromosome, too, is simplex in the male, duplex in the female
and its introduction into the egg by the spermatozoon produces
the female condition, its absence the male. This chromosome
therefore, as I have shown ('06), is never transmitted from father
to son, but always from father to daughter. Conversely, the
male zygote always receives this chromosome from the mother.
So precise is the correspondence of all this with the course of sex-
limited heredity of this type that it is difficult to resist the con-
clusion that we have before us the actual mechanism of such
heredity — in other words, that some factor essential for sex is
associated in the X-element with one that is responsible for the
sex-limited character.

This will be made clearer by the accompanying diagram (fig.
8) where the X-element assumed to be responsible for a recessive
sex-limited character is underscored (X) . This character may

EDMUND B. WILSON

be regarded as due to the absence of some particular constituent
that is present in the normal X-element.

9 Line

Zygotes (xX)

Line.

Gametes (x^ (^ (^ (y) 2

Zygotes (XX

Gamete (X) (^ (^ (^ 4

^y^otes(xx)(pi) (xf){Tf)5

R(^ R^
[DD] [RM

r6 W^

Fig. 8 Diagram of the distribution of the X- and Y-elements in successive
generations, illustrating sex-limited heredity. The underscored X-element (X)
is assumed to bear a factor for a recessive character {e.g., white eye-color), while
X represents the normal or dominant character {e.g., red eye-color). Y (being the
absence of X) likewise represents the recessive character.

Upon pairing the affected male (XY) with the normal female
(XX) there are in the Fj generation but two possible combina-
tions, XX and XY. The affected X-chromosome here passes

STUDIES ON CHROMOSOMES 97

into the female, and the male is normal; but the female of course
likewise shows only the normal (dominant) character. In the
following F2 generation (5) there are four possible combinations
XX, XX, XY and XY, two of each sex. Though X is present in
half of each sex, the character appears only in the males, XY,
again because of its recessive nature. By crossing together males
of the composition XY and females of composition XX, some of
the resulting females will have the composition XX , and the sex-
limited character is thus made to appear in the female.

When the female is the heterozygous or digametic sex — as in
sea-urchins, in Abraxas, the Plymouth Rock fowls, etc. — exactly
the converse assumption has to be made. Here, as Spillman
('08) and Castle ('09) have pointed out, the observed results
follow if the sex-limited character {e.g., lacticolor color-pattern
in Abraxas) be allelomorphic to, or the synaptic mate of, a sex-
determining factor, X, that is present as a single element in the
female but absent in the male. The formulas now become'^
(as Spillman has indicated) XG (9 grossulariata), GG (d^ gross.)
XG (9 lacticolor) and GG (cf lact). XG X GG then gives
in Fi XG and GG (gross. 9 and d^), G having passed from the
female to the male. The following cross, XG X GG gives in F2
the four types XG, XG, GG and GG, — i.e., grossulariata appear-
ing in both sexes but lacticolor only in the female. By crossing
XG with GG some of the progeny will have the composition GG
(cT lacticolor). The other combinations follow as a matter of
course.

This interpretation is in all respects the exact converse of
that made in the case of Drosophila, which is also the case with

^ These formulas are in substance the same as those stated by Mr. SpiHman in
a private letter to the writer, and are a simplified form of those suggested by Castle
('09). The interpretation thus given seems both the simplest and the most satis-
factory from the cytological point of view of all those that have been offered. It
obviates the cytological difficulties that I urged ('09) against Castle's formulas,
and renders unnecessary the secondary couplings that I suggested. All these
ways of formulating the matter conform, of course, to the same principle and
differ only in details of statement. Whether the synaptic mate of X is directly
comparable to the Y-chromosome of other insects (in which case the female formula
becomes XY and the male YY) is an open question.

98 EDMUND B. WILSON

the experimental results, as Morgan has pointed out. It seems
probable that all the observed phenomena may be reduced in
principle to one or the other of these schemes, though many modi-
fications or complexities of detail probably exist. A possible basis
for many such modifications seems to be provided by the cyto-
logical facts already known.

(2) We might assume that in cases of the first type {e.g., Droso-
phila) both sex and the characters associated with it are deter-
mined by the same chromatin; and such a possibility should
certainly be borne in mind. But in view of the widely different
nature of the characters already known to exhibit sex-limited hered-
ity it seems improbable that we can regard them as all alike due
to the same chromatin. In the light of the conclusions that have
been indicated in regard to the composition of the X-element, it
seems more probable that such characters may be determined by
the various other forms of chromatin (' Y-chromatin') associated
with the X-chromatin. If these constituents be identical with
those contained in the free Y-chromosome (the synaptic mate of
X) sex-limited heredity would of course not appear, since the two
members of the pair would be homozygous in this respect. It
should make its appearance as a result of the dropping out, or
other modification, of certain Y-constituents of the X-element, and
such a mutation might arise in either sex.

We may perceive here the possibility of understanding many
different kinds of sex-limited heredity, perhaps of complex types
that have not yet been made known. Such a possibility is sug-
gested, for example, by the remarkable relation discovered by
McClung ('05) in Mermiria (fig. 6/-/i, fig. 7 in diagram), where
the X-chromosome is in the first spermatocyte-division attached
at one end to a linear chain of four other elements to form a
pentad complex, to which may be given the formula XA . ABB.
This so divides as to separate XA from ABB. The interpretation
to be placed upon this is a puzzling question under any view, and
apparently must await more extended studies on both sexes, per-
haps also on other forms, before it can be fully cleared up. Even
here the possibility exists, I think, that the entire complex may
have arisen by the differentiation of a single original XY-pair;

STUDIES ON CHKOMOSOMES 99

but this question is clearly not yet ready for discussion. How-
ever such associations have arisen, the result is equally appli-
cable to the explanation of sex-limited heredity.

(d) Secondary sexual characters. Castle ('09) has offered the
interesting suggestion that the free Y-chromosome may be re-
sponsible for the determination of secondary sexual characters
in the male. Though I have criticized this view ('09c) I now
believe it may be true for certain cases. It is obviously excluded
when the Y-chromosome is missing; and since nearly related
species — in Metapodius even different individuals of the same spe-
cies — show the same or similar secondary male characters whether
this chromosome be present or absent, it seems probable that
these characters are in general determined in some other way.
But if, as I have suggested, sex-limited heredity may arise through
a modification of the Y-constituents of the X-element, it follows
that the YY-pair thereby becomes heterozygous. In such case,
the free Y-chromosome, being confined to the male line, should
continue to represent characters that are no longer present in
the female, and hence would be indistinguishable from secondary
male characters otherwise determined. It has further become
evident (as is indicated below) that the chromosome-groups are
so plastic that their specific composition may vary widely from
species to species. It may very well be, therefore, that Castle's
suggestion may apply to some forms.

6. Modes in which the chromosorne-number may change

The constant and characteristic duality of the 'd-chromosome'
in the second division suggests a series of questions regarding the
mode in which the chromosome-number may change that have
an important bearing on those already considered. The appear-
ance of this chromosome must suggest to any observer that it is
a compound body, consisting of two closely united components
that are invariably associated in a definite way ; but it is especially
noteworthy that its duality does not certainly appear before the
last division.^ This case must be added to the steadily increasing
evidence that chromosomes which appear single and homoge-

100 EDMUND B. WILSON

neous to the eye may nevertheless be compound bodies. An
important part of it is derived from the modifications of the X-
element reviewed above; but the evidence is now being extended
to the 'autosomes' or ordinary chromosomes as well. The double
chromosome of Nezara suggests either the initial stages of a sep-
aration of one chromosome into two or the reverse process — in
either case that we have before us one way in which the number,
and the composition, of the chromosomes may change from species
to species. This is supported by the recent observations of
Miss E. N. Browne ('10) on Notonecta. In N. undulata there
are always, in addition to a typical unequal XY pair, two small
chromosomes that appear in all the divisions as separate elements.
In N. irrorata there is always but one such chromosome, the total
number in each division being accordingly one less than in N.
irrorata. N. insulata presents a condition exactly intermediate,
there being sometimes one and sometimes two such small chromo-
somes. When, however, but one seems to be present, the second
may often be seen closely adherent to one of the larger chromo-
somes; and the latter may positively be identified, by its size, as
always the same one. It can hardly be doubted, as the author
points out, that we here see three stages in a process by which
the chromosome-number is changing, either by the fusion of two
originally separate chromosomes, or by the separation of one into
two. It is of the utmost importance that this process affects
a chromosome that can be positively identified as the same in
each case; for this proves that the change is not an indefinite
fluctation but the expression of a perfectly orderly process.
While there is here (as in the case of the d-chromosome of Nezara)
no way of knowing in which direction the change is taking
place, either alternative involves the conception that the indivi-
dual chromosomes may be compound bodies, whether as a re-
sult of previous fusion or as possible starting points for a subse-
quent segregation.

The facts now known indicate at least four possible ways in
which the chromosome-number (and in three of these also the
composition of the individual chromosomes, may change from
species to species.

STUDIES ON CHROMOSOMES 101

One is that suggested by the foregoing phenomena, i.e., the
gradual fusion of separate chromosomes into one or the reverse
process.

A second mode may be the gradual reduction and final disap-
pearance of particular chromosome-pairs, as was suggested by
Paulmier ('99), and afterwards by Montgomery and myself, in
case of the microchromosomes, or ' m-chromosomes' of the co-
reid Hemiptera. In respect to the size of these chromosomes, a
graded series may be traced from forms in which they are very
large (as in Protenor) through those where they are of intermediate
size down to cases where they are very small (as in Archimerus)
and finally to such a condition as that seen in Pachylis (fig.
9 j-l) where they are almost as minute as centrioles and may
almost be regarded as vestigial. Four of these stages are shown
in fig. 9. In Protenor {a-c) the wi-chromosomes are so nearly
of the same size as the next smallest pair that they often can not
be positively identified in the spermatogonia! groups. In Lepto-
glossus phyllopus {d-f) they are always recognizable, though not
much smaller than the next pair. In L. oppositus or L. corcu-
lus they are a little smaller. In Anasa (the form in which they
were first discovered by Paulmier) they are of middle size {g-i) ,
representing perhaps a fair average of the group. Several other
genera {e.g., Metapodius) show intermediate stages between this
condition and that seen in Archimerus (figured in my second
'Study,' and more recently by Morrill) where the ??i-chromo-
somes are almost as small as in Pachylis. It is most remarkable
that throughout this whole series the m-chromosomes exhibit
essentially the same behavior (Wilson, '056, '06), usually remain-
ing separate throughout the entire growth-period and only con-
jugating in the final prophases of the first spermatocyte-division,
to form a bivalent which with rare exceptions occupies the center
of the metaphase group; in some forms, also (e.g., Protenor, Aly-
dus) they show a marked tendency to condense at a much earlier
period than the other chromosomes. The m-chromosomes of
Pachylis, excessively minute though they are, exhibit a behavior
in all respects as constant and characteristic as even the largest
of the series. In the Lygaeidae they seem to be present in some

102

EDMUND B. WILSON

%•

\'l •!»•

% #

Fig. 9 Comparison of the m-chromosomes in Hemiptera. (In each horizon-
tal row are shown at the left a spermatogonial group, in the middle a polar view
of the first spermatocyte-division, at the right a side-view of the same division.)
a-c, Protenor belfragei; d-^, Leptoglossus phyllopus; g-i, Anasa tristis; j-l,
Pachylis gigas.

STUDIES ON CHROMOSOMES 103

species (Oedancala, t. Montgomery), in others absent (Lygaeus).
In the Pyrrhocoridae (Pyrrhocoris, Largus) they are absent as
far as known. So characteristic is the behavior of these chromo-
somes as to leave not the least doubt of their essential identity
throughout the whole series ; and this series may be regarded as a
progressive one, in one direction or the other, with the same reason
as in case of any other graded series of morphological characters.
The series thus shown in case of the m-chromosomes is as gradual
and complete as in case of the Y-chromosome, and may with
the same degree of probability be regarded as a descending one.
Thirdly, it is probable that the chromosome-number may
change by sudden mutations that produce extensive redistribu-
tions of the chromatin-substance without involving any commen-
surate change in its essential content. Were gradual changes,
chromosome by chromosome, the usual mode of modification,
we should certainly expect to find such conditions as are seen in
Nezara, in Notonecta, or in the Coreidae, more frequently. In
some groups, however, we find wide differences of chromosome-
number between species even of the same genus, and even be-
tween those that are very nearly related, without any accompany-
ing evidence of a gradual process of transition — for instance,
among the aphids and phylloxerans (Stevens, Morgan) or in the
heteropterous genera Banasa and Thyanta. (Wilson, '09d.) In
Banasa dimidiata the diploid number is 16 in both sexes, in the
nearly related B. calva 26. Of the two races of Thyanta custa-
tor described above, apparently identical in other visible char-
acters, one has in both sexes the diploid number 16, with a
simple X-chromosome, while in the other the diploid number of
the male is 27 and that of the female 28, and the X-chromosome
consists of two components. It is improbable that the dif-
ferentiation of these two forms has been accomplished by grad-
ual modifications, chromosome by chromosome. It seems far
more likely that the change took place by sudden mutation, invol-
ving a redistribution of the nuclear material which changed its
form but not in an appreciable degree its substance. In the well
known case of Oenothera gigas, derived by sudden mutation from
Oe. Lamarckiana, a change by sudden mutation is known to be

104 EDMUND B. WILSON

PLATE 1

EXPLANATION OF FIGURES

All the figures from photographs of sections. Enlargement 1500 diameters.

10, 11 First spermatocyte-division (N. hilaris)

12, 13 The same (N. viridula)

14, 15 Second spermatocyte-division (hilaris)

16-25 Side views of second division (hilaris). The XY-pair shown in 16-23, the

d-chromosome in 16, 17, 20, 24, 25; the small chromosome is evident in

10, 12, 13, 14, 15, 17, 18.

22 Initial separation of X and Y

23 Early anaphase, X and Y separating near the center (hilaris)

26-28 Nuclei from the growth-period, showing chromosome-nucleolus and plas-

mosome (hilaris)
29 Corresponding stage (viridula)

STUDIES ON CHHOMOSOMES

EDMUND B. WILSON

* K

*•*

•^

26
JOURNAL OF MORPHOLOGY,

VOL. 22, NO. 1

-'y

STUDIES ON CHROMOSOMES 105

a fact (Lutz, '07; Gates, '08), though it may be due in this instance
to a simple doubling of the whole group. Such cases led me sev-
eral years ago to the conclusion ''that the nucleus consists of
many different materials that segregate in a particular pattern
. . . and that the particular form of segregation may readily
change from species to species" (Wilson, '09d, p. 2).

Such changes must involve corresponding ones in the morpho-
logical and physiological value of the individual chromosomes;
and we must accordingly recognize the probability that these
individual values, though constant for the species, may change
from species to species as readily as does the number. Despite
the conformity to a -general type often exhibited by particular
genera or even by higher groups, the individual chromosomes are
therefore per se of subordinate significance; and it may often be
practically impossible to establish exact homologies between those
of different species. How closely this bears on the origin of the
diverse conditions seen in the composition of the XY-pair is
obvious.

Lastly, it is almost certain that changes of number may some-
times arise as a result of abnormalities in the process of karyoki-
nesis, such as the passage of both daughter-chromosomes, or of
both members of a bivalent, to one pole. In Metapodius I found
('096) direct evidence of this in case of the XY-pair itself, and
endeavored to trace to this initial cause the remarkable variations
of number that occur in this genus. Many other observers have
recorded anomalies of this kind, in both plants and animals. It
seems entirely possible, as has been suggested by McClung ('05)
and by Gates ('08) that definite mutations may be traceable to this
cause; though probably such abnormalities may in general be
expected to lead to pathological conditions.

CONCLUSION

Some of the suggestions offered in the foregoing discussion are
admittedly of a somewhat speculative character; but they are not,
as I venture to think, mere a priori constructions, but are forced
upon our attention by the observed facts. The time has come

JOURNAL OF MORPHOLOGT, VOL. 22, NO. 1

106 EDMUND B. WILSON

when we must take into account more fully than has yet been done
the new complexities and possibilities that have continually been
unfolded as we have made better acquaintance with the chromo-
somes. In this respect the advance of cytology has quite kept
pace with that of the experimental study of heredity; and it has
established so close and detailed a parallelism between the two
orders of phenomena with which these studies are respectively
engaged as to compel our closest attention.

Studies on the chromosomes have steadily accumulated evi-
dence that in the distribution of these bodies we see a mechanism
that may be competent to explain some of the most complicated
of the phenomena that are being brought to light by the study
of heredity. New and direct evidence that the chromosomes
are in fact concerned with determination has been produced by
recent experimental studies, notably by those of Herbst ('09)
and Baltzer ('10) on hybrid sea-urchin eggs. But the interest
of the chromosomes for the study of heredity is not lessened, as
some writers have seemed to imply, if we take the view — it is in
one sense almost self-evident — that they are not the exclusive
factors of determination. Through their study we may gain an
insight into the operation of heredity that is none the less
real if the chromosomes be no more than one necessary link in a
complicated chain of factors. From any point of view it is
indeed remarkable that so complex a series of phenomena as is
displayed, for example, in sex-limited heredity can be shown to
run parallel to the distribution of definite structural elements,
whose combinations and recombinations can in some measure
actually be followed with the microscope. Until a better expla-
nation of this parallelism is forthcoming we may be allowed to hold
fast to the hypothesis, directly supported by so many other data,
that it is due to a direct causal relation between these structural
elements and the process of development.

A second point that may be emphasized is the remarkable con-
stancy of the chromosome-relations in the species, and their no
less remarkable plasticity in the higher groups. The scepticism
that has been expressed in regard to constancy in the species finds,
I think, no real justification in the facts. It is perfectly true that

STUDIES ON CHROMOSOMES 107

individual fluctuations occasionally are seen in the number of the
chromosomes, in the process of synapsis, in the distribution of the
daughter-chromosomes, and in all other cytological phenomena.
It is, however, also true that most observers who have made pro-
longed, detailed and comparative studies of any particular group,
have sooner or later reached the conviction that in each species
all the essential relations in the distribution of the chromosomes
conform with wonderful fidelity to the specific type. So true is
this that the species "may often at once be identified by an expe-
rienced observer from a single chromosome-group at any stage of
the maturation-process. No one, I believe, who has engaged for
a series of years in the detailed study of such a group, for instance,
as the Hemiptera or the Orthoptera, returning again and again to
the scrutiny of the same material, can be shaken in the convic-
tion that the distribution of the chromosomes follows a perfectly
definite order, even though disturbances of that order now and
then occur. But it is equally important to recognize the fact
that this order has undergone many definite modifications of
detail from species to species, and that while all cases exhibit cer-
tain fundamental common features, we cannot without actual
observation predict the particular conditions in any given case.
It is now evident that the larger groups vary materially in respect
to specific conditions. For instance, in the orthopteran family of
Acrididae (McClung) the relations seem to be far more
uniform than such a group as the Hemiptera, where great spe-
cific diversity is exhibited, the details often changing from species
to species in a surprising manner — witness the species of Aphis
or Phylloxera (Stevens, Morgan), those of AchoUa (Payne) or
of Thyanta (Wilson). In these respects, too, the cytologist finds
his experience running parallel to that of the experimenter on
heredity; and here, once more, we find it difficult not to believe
that both are studying, from different sides, essentially the same
problem.

December 13, 1910.

108 EDMUND B. WILSON

LITERATURE CITED

Arnold, G. 1908 The nucleolus and microchromosomes in the spermato-
genesis of Hydrophilus piceus. Arch. Zellforsch., vol. 2,

Baltzer 1909 Die Chromosomen von Strongylocentrotus lividus und Echinus
microtuberculatus. Arch. f. Zellforsch., Bd. 2.

1910 Ueber die Beziehung zwischen dem Chromatin und der Ent-
wicklung und Vererbungsrichtung bei Echinodermenbastarden. Habi-
litationsschrift, Wurzburg. Engelmann, Leipzig.

Boring 1909 A small chromosome in Ascaris megalocephala. Arch., f. Zell-
forsch., vol. 4.

BovERi, Th. 1909 " Geschlechtschromosomen" bei Nematoden. Arch. f. Zell-
forsch., Bd. 4.

Browne, E. N. 1910 The relation between chromosome-number and species
in Notonecta. Biol. Bull., vol. 20,1.

Castle, W. E. 1909 A Mendelian view of sex-heredity. Science, n. s., March 5.

Cook, M. H. 1910 Spermatogenesis in Lepidoptera. Proc. Acad. Nat. Sci.,
Philadelphia, April.

Dbderer, p. 1908 Spermatogenesis in Phyllosamia. Biol. Bull., vol. 13.

Edwards, C. L. 1910 The idiochromosomesin Ascaris megalocephala and Ascaris
lumbricoides. Arch. f. Zellforsch., vol. 5.

Gates, R. R. 1908a The chromosomes of Oenothera. Science, n. s., vol. 27,
Aug. 2.

1908b A study of reduction in Oenothera rubrinervis. Bot. Gazette,
vol. 46,

1909 The behavior of the chromosomes in Oenothera lata x O. gigas.
Ibid., vol. 48.

Gross, J. 1904 Die Spermatogenese von Syromastes marginatus. Zool. Jahrb.
Anat. u. Ontog., vol. 20.

GuTER, M. 1910 Accessory chromosomes in man. Biol. Bull., vol. 19.

Herbst, C. 1909 Vererbupgsstudien, VI. Die cytologischen Grundlagen der
Verschiebung der Vererbungsrichtung nach der miitterlichen Seite.
Arch. Entwicklungsm., Bd., 27.

LuTZ, A. M. 1907 A preliminary note on the chromosomes of Oenothera La.
marckiana and one of its mutants. Sci., n. s. 26.

McClung, C. E. 1905 The chromosome complex of orthopteran spermatocytes.
Biol. Bull., vol. 9.

STUDIES ON CHROMOSOMES 109

Montgomery, T. H. 1901 A study of the chromosomes ofMetazoa. Trans.
Am. Phil. Soc, vol. 20.

1906 Chromosomes in the spermatogenesis of the Hemiptera Heterop-
tera. Trans. Am. Phil. Soc, vol. 21.

Morgan, T. H. 190Ga A biological and cytological study of sex-determina-
tion in phylloxerans and aphids. Jour. Exp. Zool., vol. 7,

1910 Sex-limited inheritance in Drosophila. Science, n. s. 32, July 22.

Morrill, C. V. 1910 The chromosomes in the oogenesis, fertilization and
cleavage of coreid Hemiptera. Biol. Bull., vol. 19.

Patjlmier, F. C. 1899 The spermatogenesis of Anasa tristis. Jour. Morph.,
vol. 15, Suppl.

Payne, F. 1909 Some new types of chromosome distribution and their rela-
tion to sex. Biol. Bull., vol. 16.

1910 The chromosomes of Acholla multispinosa. Biol. Bull., vol. 18.

Randolph, Harriet. 1908 On the spermatogenesis of the earwig, Anisolaba
maritima. Biol. Bull., vol. 15.

Sinety, R. de 1901 Recherches sur la biologie et I'anatomie des phasmes. La
Cellule, t. 19.

Spillman, W. J. 1908 Spurious allemorphism. Results of some recent investi-
gations. Am. Naturalist, vol. 42.

Stevens, N. M. 1906 Studies in spermatogenesis, II. A comparative study of
the heterochromosomes in certain species of Coleoptera, Hemiptera
and Lepidoptera, etc. Carnegie Inst. Pub. 36.

1908 A study of the germ-cells of certain Diptera, etc. Jour. Exp.
Zool., 5, 3.

1910 The chromosomes in the germ-cells of Culex. Jour. Exp. Zool.,
vols.

Wallace, L. B. 1909 The spermatogenesis of Agalena nsevia. Biol. Bull.,
vol. 17.

Wilson, E. B. 1905a Studieson chromosomes, I. The behavior of the idiochro-
mosomes in Hemiptera. Jour. Exp. Zool., vol. 2.

1905b Studies on chromosomes, II. The paired microchromosomes,
idiochromosomes, and heterotropic chromosomes in Hemiptera. Jour.
Exp. Zool., vol. 2.

1906 Studies on chromosomes, III. The sexual differences of the
chromosomes in Hemiptera. Jour. Exp. Zool., vol. 3.

1909a Studies on chromosomes, IV. The accessory chromosome in
Syromastes and Pyrrhocoris. Jour. Exp. Zool., vol. 6.

no EDMUND B. WILSON

1909b Studies on chromosomes, V. The chromosomes of Metapodius,
etc. Jour. Exp. ZooL, vol. 6.

1909c Secondary chromosome-couplings and the sexual relations in
Abraxas. Science, n. s. 29, p. 748.

1909d Differences in the chromosome-groups of closely related
species and varieties, etc. Proc. Seventh Internat. Zool. Congress,
Aug. 1907.

1910a The chromosomes in relation to the determination of sex.
Science Progress, no. 16, April.

1910b Studies on chromosomes, VI. A new type of chromosome-com-
bination in Metapodius. . Jour. Exp. Zool., vol. 9.

1910c Note on the chromosomes of Nezara. Science, n. s. 803, May 20.

THE TRANSPLANTATION OF OVARIES IN CHICKENS^

C. B. DAVENPORT

From Carnegie Institution of Washington: Station for Experimental Evolution

Dr. C. C. Guthrie ('08) has reported the results of transplant-
ing ovaries from black to white hens and vice versa. A black-
plumaged hen furnished by transplantation with 'white' eggs and
mated to a white cock gave ''about equal numbers of white and
spotted" chicks. Guthrie thinks that these black spots indicate
that the black-plumaged foster-mother infected the engrafted
'white' eggs. So far Guthrie. But a person familiar with the
results of hybridizing v/ill appreciate that Guthrie's result is bet-
ter explained on the assumption that the engrafted ovary was
absorbed and that the white sperm fertilized the regenerated
'black' eggs of the black hen. For the white by black cross gives
white offspring with black spots in the female chicks only, i.e.,
half of all, as Guthrie found.

In a second set of experiments, Guthrie found that when a
white hen carrying a 'black' ovary was mated to a White Leg-
horn male, the offspring were either white or black or spotted.
Guthrie says : "The black, therefore, must have come through the
black ovary." But the student of hybridization on poultry will
recognize at once that, if the white-plumaged cock produced only
' white ' germ cells, none of his offspring would be black even if
the eggs were 'black.' Hence, the cock must have had 'black'
germ cells and, very likely, the hen also, since 'White Leghorn'
hens that carry 'black' germ cells are very common and fre-
quently show, in adult life, a pure white plumage.

If two 'White Leghorns' with 'black' germ cells be mated expec-
tation is that in four chicks one shall be black; one spotted, and

' A preliminary paper covering these results was read before the Society for
Experimental Biology and Medicine, June 1910.

Ill

112 C. B. DAVENPORT

two white; Guthrie got five chicks, one black, one spotted and
three white.

Guthrie found that a black hen containing a 'white' ovary,
mated with a black cock gave black-plumaged chicks, of which
two out of six had white feet. He concludes that the white condi-
tion of the feet must have come from the engrafted eggs of the
White Leghorn. In criticism it must be pointed out that the
cross, white egg X black sperm, normally gives offspring whose
plumage color is white, either pure or with black specks. The
fact that all the offspring had black plumage proves that the eggs
were the normal ' black ' eggs regenerated by the black hen. The
white toes are frequently found in the offspring of two black birds.
Thus in my pen 1041 two extracted blacks (Sumatras) mated
give ten black chicks in six of which white toes are recorded. The
results of this cross of Guthrie's confirm the conclusion that the
transplanted ovaries were not functional and that the normal
ovaries had regenerated.

To -test the possibility of such regeneration of ovaries I removed
the ovaries of some hens in the autumn of 1909 and transplanted
into them eggs from dissimilar hens. The operated birds were
then mated to cocks resembling the soma of the so-called 'foster-
mother.' Were there regeneration of the ovary the offspring
should be of the straight breed; but if the 'grafts' persisted and
became functional the chicks should be hybrids.

Experiments 1 and 2, operations: The protocol of the grafting
operations is as follows :

No. 11379, pure-bred Dark Brahma bantam, hatched February,
1909; made to fast two days. On September 29, 1909, injected
with 0.005 grain of atropin in 1 cc. of water, etherized in about
twelve minutes and opened up between two left intercostals.
Large ovary, badly torn in removal, removal tolerably complete.
One piece of ovary from no. 11605 fastened by cotton thread to
inesentery near attachment of ovary. Sewed up.

No. 11605, hatched March, 1909, from White Leghorn,-Houdan
ancestry. Clean-footed, with five toes on each foot, V-comb,
modified high nostril, plumage color white (with black recessive).
On September 29, 1909, injected 0.005 grain atropin in 1 cc. of

THE TRANSPLANTATION OF OVARIES IN CHICKENS 113

water. Etherized in twenty minutes. Plucked feathers and
opened body wall between last two ribs. Large ovary completely
removed or nearly so, in three or four pieces. Hemorrhage slight.
Stitched in small piece of ovary of no. 11379 to peritoneum near
attachment of old ovary. Sewed up. Bird recovered rapidly.
Some Dark Brahma in ancestry, but its characters had become
eliminated.

Results, Experiment 1. Mated in pen 1027, no. 11605 9 (with
engrafted ovary from no. 11379, Dark Brahma) and 11291°",
straight Dark Brahma. Table 1 gives the juvenile character-
istics of 1, the male; 2, the White Leghorn-Houdan, so-called
foster-mother; 3, the hen from which the ovaries were transplanted ;
4, expectation on the hypothesis that the graft succeeded; 5,
expectation on the hypothesis that the graft failed and the proper
ovary was regenerated ; and 6, the observed characteristics of the
young offspring.

An examination of table 1 shows at once that it cannot be true
that the engrafted ovary replaced the hen's proper ovary, for if it
had, columns six and four should agree. On the contrary, col-
umn six agrees essentially with column five and supports the
hypothesis that the engrafted eggs did not become functional.

One discordant fact there is, however, namely, the occurrence in
column six of three cases of cinnamon offspring. Such offspring
are to be expected on the hypothesis that some eggs of the graft
became functional. If that hypothesis be true, then the other
characters of the same individuals should be like those of the
pure Dark Brahma. Of the three the first has extra toes, split
comb and a boot of one row; it is no Dark Brahma; the second
has extra toes, wide nostril and a two rowed boot; it is not a
Dark Brahma; and the third has really black down with some red
at the tips, five toes on the right foot, a split comb and one row
of feathers on the shank; so it is not a Dark Brahma. These
therefore, are not from the engrafted Dark Brahma eggs. They
represent cases of imperfect dominance of the black down over
cinnamon. The conclusion to be drawn from this experiment is
that the engrafted eggs did not mature in the foster-mother.

s ^

white; white + specks (or smoked) 21
black, white below ISl
cinnamon above, white below 3J

3 rows, 12; 2 rows, 15;
1 row, 1; row, 1;
unrecorded, 3

pea, 16;

split, 23; unrecorded, 3

4 toes, 18; 5 toes, 24

narrow, 12; Intermediate and wide, 12;
unrecorded, 18

50 per cent white
or white +
smoky; 50 per
cent black back

intermediate,

3 to rows
pea, 50 per cent;

split pea, 50 per

cent

4, 50 per cent

5, 50 per cent
narrow, 50 per cent
Interm. and wide,

50 per cent

cinnamon; light
below

heavy
pea

4
narrow, gr. 2

light cinnamon;
gray below

heavy, 5 to 7 rows
pea

narrow, gr. 2

a
1

white (+ smoky)
[black back]

absent

Y [V, 1]

5 [4]
wide, grade 7

cinnamon; gray
below

heavy, 5 to 7 rows
pea

4
narrow, grade 2

3 ,

a ^

THE TRANSPLANTATION OF OVARIES IN CHICKENS 115

Experi7nent No. 2. No. 11379, Dark Brahma with engrafted
ovary from no. 11605 (White Leghorn-Houdan) was mated in
pen 1050 with 14122^, a single-comb Black Minorca. Table 2
gives the juvenile characters of 1, the male parent; 2, the Dark
Brahma, so-called foster-mother; 3, the hen from which the ova-
ries were transplanted; 4, the expectation of offspring on the
assumption that the graft succeeded; and 5, that the graft failed
and the proper ovary was regenerated; also, 6, the observed char-
acters in the offspring.

Without exception the characters of the offspring are clearly
those of the Dark Brahma X Minorca cross and none of the White
Leghorn or Houdan differential characters enter into their com-
position. The grafted ovary produced no eggs that developed,
the extirpated ovary was regenerated.

Experiments 3 and 4, operations. The protocol of the grafting
operations is as follows:

No. 11541 9 is a white-plumaged hen derived from a cross
between 8681 9 ,aWhiteLeghorn-Minorca-Polishbird, and 7811 cf,
a Houdan cross hatched (in pen 905) in February, 1909; fasted
two days. On October 2, 1909, injected with 0.005 grain atropin
in 1 cc. of water; etherized and opened. Ovary very large, two
large pieces (60 per cent) of ovary removed. Strong hemor-
rhage. Two small pieces of ovary from no. 11383 9, Dark
Brahma, sewed with peritoneum close to ovarial artery. Sewed
up. Bird slow in recovery.

No. 11383 9, straight Dark Brahma, hatched February, 1909,
from mating 907: 7549. On October 2, 1909, injected with atro-
pin, etherized and opened, ovaries small, incompletely removed.
Two large pieces of ovary of no. 11541 sewed into peritoneum.
Sewed up. Bird recovered rapidly.

Results, experiment 3. No. 11541, White Leghorn-Black
Minorca-PoHsh-Houdan hybrid, with engrafted ovary from No.
11383, Dark Brahma, was mated in pen 1027 with 11291 o^, a pure
bred Dark Brahma. The results of this mating are given in
table 3.

Experiment 4- No. 11383 9, pure-bred Dark Brahma with
engrafted ovary from no. 11541 (White Leghorn-Black Minorca-

;

1
1

i ^

m
\ "*

1 i

1 1" i i

°? . £ S &a 1 i

If 1 ss ss 1

>
O

IM f li

i i

is "^ -i 5

F 1 i i:

1 !

ill i:

S £

■SI"

S 8

s S

I i
■Is

„• t

S « "^

s fe g

g ft

a -^

S. S

a a ftS

I - « 1 1

1 1

- 1

2.
"foster-mother"

« i

. > 1 1 ,

1 1 1 1

i
5

ill " I .

1" 1 1 § ^

: 2

i
i

' .

5
1

118 C. B. DAVENPORT

Polish — Houdan hybrid) was mated with a Black Minorca
14122 d". The results of this mating are given in table 4.

Experiment 5. No. 11693 9 , used in this experiment, is a white
bird that had ' smoke ' on down when hatched. It is of somewhat
complex origin. Its mother was an Fi hybrid between a Black
Spanish cock and a White Leghorn; its father had the same ele-
ments and also white Silkie in its ancestry. No. 11693 has, con-
sequently, black recessive. It has a single comb, is free of the
skin pigment of the Silkie, is clean-shanked and has four toes on
the right foot and five on the left.

On September 19, 1909, this pullet (which was hatched March,
1910) was treated with atropin, etherized during half an hour and
opened as usual between the last two ribs. All of the ovary, as
far as could be seen, was removed. Pieces of ovary from no.
11280 ? (a straight-bred Dark Brahma bantam) were placed in
contact with the peritoneum, near the removed ovary, but not
stitched in, as the bird showed signs of succumbing. The cut
was sewed up and the bird set aside where it lay quiet for half an
hour.2 The Dark Brahma from which the ovary (whose eggs
measured 0.5 mm. in diameter) was removed died in consequences
of hemorrhage.

Later No. 11693 was mated with 11291 o^ (in mating 1027:
11693). He is a straight-bred dark Brahma bantam cock, used
also in experiments 1 and 3. The results are shown in table 5.

Experiment 6. No. 11826 9, hatched March, 1909, a pure bred
Dark Brahma was opened October 2, 1909, and ovary imperfectly
removed. Ovary of no. 12550 (a White Leghorn-Minorca-Polish-
Houdan hybrid) sewed on to peritoneum at point of removal.
The ovary had been kept out of body of hen about ten minutes,
but covered and moist.

In the late winter of 1910 no. 11826 9 was mated in pen 1050
with 14122 cf, a single-comb Black Minorca. The results are
given in table 6.

See postscript.

i Is

"!i

I I
e o

e »

^ i

3 I

il I

I -I ^ I

J3 S

¥ i °

a go

s §
6 I

^"1

S3
<§ I

IJ
II

I I

l-g

as
•o

2 ca

■< ^

S S

3 b
■< o

(»
fc

■\

§2

1 K5

3 "

^ 73

£ g

_a _.___!

l|s

§ s.

1
1

S
« 1

z
o

Ill

8 °

■3

i !P

-c fe

E >,

. (B

5-0

><
<

j^'

-1

a g

III

s-^ ^

a « o

3i§

lit

J
1

~ tc _

s::2

■*

111

i<i"

Isl

■^^

:=!-§

-g

§2

•-^2

s >.

lil

_l 1

s - 5

i!i

•7;

1«

m
<

8 :

E'"

y. "

^ 1

fe §

^ -s

1 1
11

III

43 -s S I
"^ IS s

1
3

■€

d
3

3 rows. 13

4 rows, 1

pea, 18

•6

i i

'St

All black; white
below

intermediate, to
3 rows

pea, modified

low, grade 1 or 2

white, 25 per cent;
white, black
specks, 25 per
cent; black and
white 50 per cent

absent

Intermediate grade
5 to 2

I 5

white [black and
white]

absent

wide, grade 8

. s

cinnamon above;
gray below

heavy, 5 to 7 rows

pea

low, grade 1 or 2

FATHER

black; white below

absent

single

narrow, grade 1

i
i

i
1

JOURNAI, OF MORPHOLOOy, VOL. 22, NO. 1

122 C. B. DAVENPORT

CONCLUSIONS

In the six experiments described above there is no evidence that
the engrafted ovary ever became functional but all results are in
accord with the conclusion that the more or less completely extir-
pated ovary regenerated and produced an abundance of eggs.
With the results the data of Dr. Guthrie's paper are not in dis-
accord. His data, like ours, furnish no evidence for the survival
of the engrafted ovaries, far less of an effect of the soma of the
foster-mother on the introduced germ plasm.

Cold Spring Harbor, N. Y.
September 26, 1910.

POSTSCRIPT

On January 4, 1911, No. 11693 ? was killed and opened on the
left side. An ovary of fairly typical size for a hen entering her
second year of laying was found. It contained numerous eggs,
4 to 5 mm. in diameter. Slightly ventrad of the main artery of
the ovary is an irregular mass 5x4x2 mm. of cheesy consistency,
imbedded in and covered by peritoneum. Its general appearance
is that of a dried, hardened ovary, with clear traces of follicles.
It doubtless represents the engrafted ovary, entirely encysted
in the peritoneum.

January 30, 1911.

LITERATURE CITED

Davenport, C. B. 1906 Inheritance in poultry. Publication no. 52, Carnegie
Institution of Washington.

1910 Inheritance of plumage color in poultry. Proc. Soc. Exper.
Biol, and Med., vol. 7, p. 168.

Guthrie C. C. 1908 Further results of transplantation of ovaries in chickens.
Jour. Exp. Zool., 5, pp. 563-576.

THE EFFECTS OF INBREEDING AND SELECTION

ON THE FERTILITY, VIGOR AND SEX RATIO

OF DROSOPHILA AMPELOPHILA

W. J. MOENKHAUS

Indiana University, Bloomington, Indiana

CONTENTS

Introductory 124

Material and methods 124

Inbreeding and selection on fertility and vigor 126

1 Introductory 126

2 Sterility 127

a Character of the sterility 127

b Degrees of sterility 128

3 Inbreeding and vigor 131

4 Sterility and selection 134

6 Discussion of results 138

Sex-ratio and selection 141

1 Introductory 141

2 The normal sex-ratio 141

3 Control of sex-ratio by selection 142

a History of strain 206 143

b History of strain 207 147

c Discussion 147

4 Influence of male and female in determining the sex-ratio 148

5 Discussion of results on sex-ratio 151

Summary 153

Literature cited 154

123

124 W. J. MOENKHAUS

INTRODUCTORY

The present report includes the results of two series of experi-
ments on the fruit fly — -Drosophila ampelophila. One set con-
cerns itself primarily with the effects of inbreeding and the other
with sex-ratios. The experiments on inbreeding grew out of work
I had been carrying on on hybridization. In these hybridiza-
tion experiments the effects on the developmental processes of
hybrids between species too remotely related were especially
emphasized. The converse of these experiments was, naturally, to
study the effect upon the young between individuals too closely
related. Fishes, upon which all my experiments in hybridization
were made, do not lend themselves for purposes of inbreeding
without elaborate breeding facilities. Mice seemed suitable for
this purpose but, both at the outset of these experiments and since,
these creatures have proven miserable failures in my hands.
Among the insects, I tried the common willow beetle but this
proved to throw only one generation annually in this latitude.
It was desirable to have an animal with a brief life history, whose
food could be easily obtained at all seasons and in which the sexes
could be readily distinguished. In these respects the fruit fly is
almost ideal. The facts herein considered confine themselves to
this species.

The experiments on sex-ratio suggested themselves in connec-
tion with the inbreeding experiments and so were carried out
along with the latter and after they were completed.

MATERIAL AND METHODS

The strain which is mostly under discussion in my inbreeding
experiments came from a well-filled female that was taken from the
window of my residence in Bloomington. Other strains were
started at the onset. Some of these came from the banana bunches
at the various groceries and others came from fruit which I had
laid out for this purpose. None of these were carried further than
two or three generations excepting two, called 6 and 7 in my
records. The latter was discontinued after the tenth generation

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 125

since it had been from the beginning apparently less prolific.
The strain 6 was carried for over seventy-five generations and is
the one on which the experiments in inbreeding of this report are
based.

For vivaria, tall stender dishes, tumblers, quinine bottles and
lamp chimneys were given a trial. They were discarded in favor
of 8-dram shell vials. These were compact, so that a large number
of matings could be kept in a small space, and they were most con-
venient in manipulating the pairs during the frequent changes to
new cages that was necessary all along. The open end of the
shell vial was closed with a plug of absorbent cotton, not too
compact, so as to afford some ventilation. The flies are strongly
positive to light, so that the vials could be laid with their bottom
toward the light and the cotton plug removed with safety for the
introduction of food etc. Small trays holding fifteen of these
vials were used and in this way the experiments could be readily
and compactly stored in the incubator, or they could be packed
into a valise to be taken along wherever I went. The food was
exclusively well-ripened bananas. To prevent the larvae ;rom
pupating in the food, narrow strips of blotter or filter paper were
introduced in which they seemed to be especially fond of pupating.
It is, of course, apparent that the greatest care had to be taken
to avoid contamination from flies without. The stock food had
to be scrupulously watched and the instruments kept clean to
avoid the introduction of eggs laid on them by extraneous females.
The bananas, especially, as they come from the stores, are likely
to be infected with eggs and larvae if the skin be in any way bruised
or split.

The brothers and sisters were paired off, always within the
first ten or twelve hours of their life as imagos. Up to this time
mating has not occurred. In fact I have never found a pair that
copulated during the first twenty-four hours or, if so, that pro-
duced fertile eggs.

126 W. J. MOENKHAUS

INBREEDING AND SELECTION ON PERTILITY AND VIGOR
1. Introductory

That continued inbreeding acts deleteriously on the fertility
and vitahty of a race is a beUef so firmly and generally established
that it is seldom questioned. This has its origin largely in the
common experience of breeders whose observations, unfortun-
ately, are too often unreliable. There are not wanting experi-
ments such as those of Van Guaita ('98) and Bos ('94) and others,
scientifically conducted, which bear out this conclusion.

On the other hand, it is refreshing to encounter in the literature
such reports as that of Gentry ('05) who believes not only that
inbreeding is not necessarily harmful, but also that it maybe
beneficial to conserve and intensify the good points in his breed.
Gentry's experiments were made on Berkshires. The most pro-
longed tests of close inbreeding that have been recorded were
made by Castle ('06) on the same species with which the present
paper deals. He inbred (brothers with sisters) for fifty-nine gener-
ations. He concludes that such close inbreeding does not neces-
sarily result in a loss of productiveness and of vigor; at least that
inbreeding cannot be regarded as a causal factor. Some of his
results so nearly parallel those of the present writer that further
reference to his results will be made in the body of the paper.

During the early part of October, 1903, a number of pairs were
started breeding. These came from various sources in Bloomington.
These different pairs were reared for the most part only a few
generations, excepting pair No. 6 which was continued for about
four and one-half years. During this time over seventy-five
generations were produced. Toward the close of this period no
exact count was kept of the generations so that only an approxi-
mate figure can be given. Five pairs of brothers and sisters were
mated in each generation to insure against accidents that might
terminate the strain if but one mating were made.

Along at the fifth and sixth generation it became more and
more difficult to keep the strain alive with the five pairs of brothers
and sisters that were mated each generation. The failure of an

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 127

occasional pair to produce young had hitherto been attributed to
accidental conditions of food, etc., but this no longer seemed a satis-
factory explanation of all the failures to produce young. This con-
dition, was, therefore looked into more thoroughly. This was done
by laying out instead of five pairs a much larger number from thf:
offspring of a given productive pair. The greatest care was taken
with the food, temperature etc. and it soon developed that a
variable per cent of the pairs were sterile. These sterile pairs were
to all appearances normal. It was clear now that, while inbreed-
ing had not reduced the general vitality of the strain thus far,
there had appeared a high degree of sterility.

2. Sterility

* a. Character of the sterility. Examination of all the matings
brought out the fact that in all cases eggs were present in large
numbers. This seemed to suggest that the difficulty lay in the
larvae either failing to emerge from the egg envelope or, succeeding
in this, failing to carry themselves through the feeding stage or the
transformation.

By a careful search of the food of the sterile pairs, after suffi-
cient time for the larvae to mature had been allowed, it became
evident that the difficulty lay at a time earlier than the pupal
stage for none of the latter could ever be found. The food sup-
plied these sterile pairs was the same as that of the fertile ones since
it could not be foretold which pairs were going to prove infertile.
Furthermore, the infertile pairs were usually kept for from
twenty to thirty days, the best of food being supplied them from
time to time. The same search showed that no larvae were pres-
ent, at least so far as direct inspection of the food under a dis-
secting microscope could be depended upon.

It was always possible, of course, that the larvae failed to carry
their development very far, and, thus, being small when they first
emerge from the egg, might have been overlooked. It became
necessary, consequently, to take the eggs as they were laid from
time to time and keep them under observation to see whether the
larvae ever emerged. This was done by placing a piece of banana

128 W. J. MOENKHAUS

in the vial with a sterile pair and from time to time removing the
eggs one by one with the point of a needle and placing them on a
piece of moist filter paper in a separate vial. Usually twenty were
placed in each vial and some food added for the larvae, should they
emerge. Inspection of the eggs after twenty-four, forty-eight and
seventy-two hours would readily reveal the number of eggs that
had produced larvae. I have laid out thus at a great expense of
time literally thousands of eggs from many infertile pairs, in
many cases all the eggs that a given pair produced during the first
twenty-five days of its life, but I have never seen a single egg that
had hatched. Eggs of fertile pairs thus laid out will readily hatch
so that all the larvae will have taken to the food twenty-four hours
after the eggs are deposited.

Such infertile pairs copulate frequently and it would seem that
impregnation should follow. I have never sectioned the eggs
to see whether spermatozoa enter the eggs or whether they con-
tain partially developed larvae which fail to hatch. I have,
however, been able to determine in this strain which of the sexes
is at fault. This was done in the following manner. After a
pair by sufficient trial had proven itself infertile, the male was
mated to a virgin female of a fresh strain that had not been inbred
and possessed a high degree of fertility, and the female was simi-
larly mated with a male, usually one whose fertility had been estab-
lished. Sixty-four such cases were tried and in no case did the
females fail to produce young and in no case did the males pro-
duce any although repeated copulations took place. It is evident
from the foregoing, that, in this strain, the sterility lies exclusively
in the male and that the female has lost, apparently, nothing in
fertility. Castle (p. 735) reports, on the other hand, that either
sex may be sterile. However, Castle took no account of the eggs
and larvae but merely the production of pupae, so that his steril-
ity cannot be with certainty compared to mine. It would seem,
however, that in some strains infertility may be strictly confined
to the males and in others to both sexes. That sterility is com-
plete for all males, when it occurs, is shown by both our results.

h. Degrees of sterility. The foregoing experiments concerned
themselves with such pairs as were completely sterile. Other pairs

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 129

of brothers and sisters from the same parents, however, were fer-
tile. Judging from the productiveness of these, there was often
a wide divergence. It would seem that, as a result of inbreeding,
we had a condition of fertility ranging from absolute infertility
to comparatively high fertility among the different pairs of brothers
and sisters from any given pair of parents. To test this the follow-
ing experiment was carried out: About two-hundred eggs from
each of fifteen pairs of flies were laid out after the fashion indicated
above. Ten of these pairs had been inbred for seventeen genera-
tions while five belonged to fresh stock that had not been inbred.
Of the ten pairs of the inbred strain, five belonged to a strain
which had arrived at a very low degree of fertility, namely only
36 per cent of the forty- two pairs tested were fertile (table 3,
seventeenth generation, strain, A) . These five pairs were brothers
and sisters to many of the sterile pairs considered in the preceding
section.

The other five pairs (of the ten inbred) were from a strain which
had been held by selection to a high degree of fertility, namely
97 per cent of the thirty-four pairs tested were fertile. Both of
these strains were descended from common great grandparents
(table 3, seventeenth generation, strain B).

We have, thus, for comparison three conditions, namely, (1)
eggs from a highly infertile inbred strain; (2) eggs from a highly
fertile inbred strain; and (3) eggs from a presumably norma
strain that had not been inbred. It should be added that the
five pairs were taken at random and were not selected. Approxi-
mately the first two-hundred eggs of each pair were laid out in
batches of about twenty to twenty-five to the vial. The number
of eggs that hatched was noted in each case and also the number
that emerged as imagos. Table 1 gives the summary of results.

From this table it appears that from the eggs which were taken
from the inbred pairs with low fertility practically as large a per
cent (97.27) hatched as from the eggs that came from the inbred
pairs that showed a high fertility (98.2). The same is true in
regard to the number that produced imagoes, 86.8 per cent and
85.1 per cent respectively. The fact clearly brought out here is
that when infertility arises in this strain it arises suddenly and

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1.

130

W. J. MOENKHAUS

does not present all intergradations. In other words, one does not
find that among a large number of brothers and sisters some pairs
whose eggs only partially hatch and other pairs that range in this
respect, on the one hand, to perfect fertility and, on the other, to
complete sterility. The fertility is either completely lost or it is
of a high degree. Furthermore, when we compare the inbreds with
the normals (not inbred) in regard to the percentage of eggs hatched
no essential difference is observable. It would seem, therefore,

TABLE 1

Inbred {low fertility)

■ PAIRS

NUMBER OF NUMBER OP ^^,^,^^n^^ PER CENT OF
EGGS PLACED EGGS HATCHED 1 EMERGED '' ^°^^ HATCHED

PER CENT OF
IMAGOS
EMERGED

193 184 160

95.3
94.0
98.0
100.0
100.0

82.9

200 188

201 197
198 198
123 123

169
182
180
104

84.5

90.5

90.9

84.5

Total

915

890

795

97.27

86.8

Inbred (high

fertility)

201
173
204

197
175

198
172
200
193
169

182
156
161
165
145

98.5
99.4
98.0
97.9
96.5

90.5

90.1

78.9

83.7

82.8

Total

950

932

809

98.2

85.1

Normals (not inbred)

215
70
153
224
158
146
223

211
70
152
218
155
127
222

193
48
132
144
144
109
205

98.1
100.0
99.9
97.3
98.1
87.7
99.9

89.7

68.5

86.2
64.2

e". .:

91.1

74.6

91.9

Total

1189

1155

975

97.2

82.0

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 131

that the pairs that had not completely lost then- fertility, in so
far as hatching their eggs is concerned, had suffered no deteriora-
tion whatever as a result of seventeen generations of closest in-
breeding.

A fact of further importance brought out by table 1 is that of
the percentage of eggs that successfully produced imagos. This
does not differ essentially in the two groups of inbreds nor do
these differ essentially from the normals. Castle used as his
measure 'productiveness,' meaning thereby the number of pupae
that were successfully produced. Making allowance for some
pupae which do not emerge, the imagos produced in my experi-
ments were an approximation to his 'productiveness.' Inbreed-
ing, consequently, does not affect adversely the productiveness of
pairs that show any fertility at all.

Castle found that his strains showed an annual fluctuation in
productiveness, the period of least productiveness falling in the
late autumn and early winter. My own experiments extended
over about four and one half years and, although I have been on
the lookout for this, I have never observed it. As Castle himself
suggests, this fluctuation was probably a function of the tempera-
ture of the room. My flies were kept in a room which varied from
60 to 80 degrees and, when this was not possible, they were placed
in an incubator kept at about the same range of temperature. It
may also be that the productiveness of his strain ran low at this
time of the year because they were placed in new hands at the open-
ing of the college year. My observation has been that it takes
some time for a new man to learn all the conditions that make for
a favorable rearing of these creatures so that Castle's low produc-
tive periods may be merely a measure of the training period of the
experimentor.

3. Inbreeding and vigor

At the outset of the experiments it was the expectation of the
writer that such rigorous inbreeding would early and violently
show itself in the vigor and fertility of the animals. In this, how-
ever, he was largely disappointed. In the strain that is here under
consideration no untoward results could be detected during the

132 W. J. MOENKHAUS

first five or six generations. As previously stated, up to this time
the method consisted in placing pairs of brothers and sisters in
each of five vials to insure against mishaps. These mishaps con-
sisted of drying up of the food, attacks of fungus and in some cases
the escape of the flies themselves during the process of feeding etc.
Those pairs that produced young were regarded as having es-
caped these various possible mishaps and were taken as indica-
tions of the vitality and productiveness of the strain. The expecta-
tion at that time was that any deleterious effect of the inbreeding
would show itself in the offspring of any of the pairs. Conse-
quently, when a given pair would produce offspring that was num-
erous, all well formed^ vigorous, and in no apparent way differing
from normal offspring, to see whether some slight influence might
not be present that could not be detected by ordinarj^ observation a
definite measure was taken of (1) their rate of reaction to light and
gravity, (2) the total number of eggs produced and (3) the percen-
tage of eggs which hatched and emerged. An attempt was made to
determine their length of life but this proved too prolonged to
allow one to carry it out together with all the other incidents of
the already too laborious experiments.

The reaction of this animal toward light and against gravity
is well known. To get a measure of the rate of reaction the ani-
mals were made to travel through a glass tube that had been
blackened for 16 cm. on the inside. This tube had a light placed
at one end and was inclined about twenty-five degrees. From a
glass vial the flies were admitted, one at a time, into the tube and
the time from the moment of entrance into the blackened portion
of the tube to their emergence was recorded. It was found essen-
tial that the two batches of flies (inbreds and normals) should be of
the same age, be reared under the same conditions and that the
temperature of the room be the same for the two batches. The
results are as foflows: at a temperatureof 27.2° C. 133 normals
took 16 seconds, average, to travel the distance, and 140 inbreds
took 15.4 seconds. The two sexes in these two groups were about
equal in number. In both groups the males travel the distance
on an average in three seconds less time. It is clear from this
that the normals and inbreeds are equally responsive to these two

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA

133

agents and that the latter have not suffered in this regard as a
result of inbreeding.

In order to determine the total number of eggs produced it was
necessary to isolate the pairs and twice each day pick off all the
eggs that had been deposited in and around the food provided.
This proved to be a most laborious process, for the eggs are too small
to be followed safely with the naked eye and had to be removed
individually with the point of a needle. Too much value must not
be attached to this measure for the reason that the rate and,
therefore, probably the number of eggs deposited seems to depend
somewhat, at least, on the condition of the food present, and for the

TABLE 2

Strain 6

Number of generations inbred 2

Number of days eggs were counted 27

Total number of eggs laid 433

617

480

6
34

724

455

516

Strain 7

Number of generations inbred ....

26
654

33
662

539

6
23

498

33
907

Number of days eggs were counted

Total number of eggs laid

28
429

reason that only the eggs deposited during the first twenty-five or
thirty days were counted. These creatures live to he very much
older. We have kept females alive 153 days, but after the first
twenty-five or thirty days the eggs come only in small numbers.
Table 2 gives the actual counts of several females of both strains
6 and 7.

We see from the above counts that no material reduction has
occurred in egg production during nine and ten generations of
inbreeding. Such variations as occur may, of course, represent
individual differences in the females.

The data given in table 1 of the relative hatching and emerging
qualities of the young of normals and of pairs inbred for seventeen
generations shows that there is no difference in this respect.

134 W. J. MOENKHAUS

In so far as the above determination may be taken as a measure
of the vitality of this species we are justified in concluding that
from six to seventeen generations of inbreeding no appreciable
deterioration has resulted. No such exact determinations were
made in later generations, and it is possible that eventually the
effects of inbreeding would manifest themselves, but my observa-
tions during seventy-five or more generations does not lead me
to believe this.

Jf.. Sterility and selection

Along at the thirteenth and fourteenth generations the sterility
had become very pronounced. Of the offspring of some of the
pairs, more than 50 per cent of the males were sterile. On the
other hand, while practically all pairs showed at least some degree
of sterility this varied very much in the different brothers and sis-
ters of the same brood. That this sterility was a direct physiolog-
ical result of the inbreeding seemed to me very doubtful. To
find the effects of inbreeding showing itself in such a specific way
upon the males only, did not, to say the least, meet expectations.
Furthermore, sterility was not wholly wanting in forms that had
not been inbred.

It was highly desirable to continue the experiments on inbreed-
ing, and yet to keep the strain alive, it was necessary to find some
way to eliminate this high degree of sterility. The process that
was most effective was selection. By continuing the strain of
those pairs whose offspring showed the highest degree of fertility
but at the same time continuing the rigorous inbreeding, it was
possible almost completely to eliminate the sterility. This at the
same time gave one of the severest tests as to whether inbreeding
was the responsible factor, for if the sterility could be eliminated
by continuing the very process of inbreeding the latter could not
well be held to be the cause of it.

This was done as follows: In the fourteenth generation three
fertile pairs of brothers and sisters from the same brood were iso-
lated and mated. The offspring of each of these were mated in
pairs to determine the degree of sterility. By reference to table

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 135

^-. o

^68

3!^

:ig

136

W. J. MOENKHAUS

3, it will be seen that the pair marked A produced offspring out of
which nine of twelve pairs tested were infertile; pair' 5 produced
offspring of which four pairs out of fourteen tested were infertile
and pair C threw offspring with five pairs out of fifteen infertile.
We have here, then, three pairs showing a wide variation in the
degree of fertihty of their offspring. Pair A showed 75 per cent of
the pairs infertile and pairs B and C approximately the reverse
ratio. In the further progress of the experiment pair C was dis-

TABLE 4
Strain A

NUMBER
PAIRS TESTED

NUMBER
PAIRS FERTILE

NUMBER PAIRS
INFERTILE

PER CENT PAIRSPE
FERTILE

^ CENT PAIRS
INFERTILE

18 (1)

52
51
52
66

27
37
37
46
19

25
14
15
11
9

51
72
71
80
69

49
28
29
20
31

18 (2)

18 (3)

18 (4) . . .

18 (5) . .

Average for 238 pairs 69 per cent.

Strain B

18(1)...
18(2)...
18(3).^.

18(4)...
18(5)...

100
100
100
68
100

Average for 93 pairs 92.5 per cent.

continued so that only pairs A and B were used. I shall in the
further description of the experiment refer to the descendants of
A as strain A and of B as strain B.

Before entering upon the experiment of selection it was neces-
sary to ascertain whether, without selection, the descendants of
pairs A and B continued to show a low and high fertility respec-
tively. Accordingly, a single one of the fertile pairs of the 15th
inbred generation of strain A and B was tested. Reference to
the table shows that in strain A 27 pairs or 57 per cent of the forty
seven pairs tested were infertile, while in strain 5 none of the thirty-

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 137

seven pairs tested were infertile. The same process was repeated
with a pair of the sixteenth generation of the two strains. Strain
A showed twenty-seven or sixty-four per cent of the forty-two
pairs tested infertile and strain B one or three per cent of the
thirty-six pairs tested.

Up to this point in the experiment only a single pair in each
generation was tested as to the fertility of its offspring. It might
well be that by chance in each case a pair of low fertility was taken
in strain A and a pair of high fertility in strain B. To eliminate
this possible error five pairs w^ere taken in each strain and the
fertility of their offspring determined. It was further desirable to
obtain an estimate of the variability in the fertility of the pairs in
the two strains as well as to get a more correct estimate of the
average fertility of both. In the diagram these five pairs are
designated as 18 (1), 18 (2), etc. Table 4 shows the number of
pairs of offspring tested for each pair and the number and per-
centage of pairs fertile and infertile.

The fertility thus varied in strain A from 51 per cent in
18 (1) to 80 per cent in 18 (4), with an average fertility of 69
per cent. In strain B the fertility was much less variable in the
different pairs, the only exceptions being 18 (4), the average fertil-
ity being 92.5 per cent.

We now have definitely established two strains, one of low
and another of high fertility. The important part to be empha-
sized here is that this was produced by the process of selection from
among the variable offspring of generation fourteen of the inbred
strain. To make the experiment more complete it was now neces-
sary to obtain a highly fertile strain out of the one with low fertility.
Accordingly strain B was discontinued at this point and attention
restricted to strain A. Five pairs, 19 (1), 19 (2), 19 (3), etc., were
taken from among the offspring of 18 (4) because this showed the
highest percentage of fertility. These were tested in the same way
as in the preceding generation. Table 5 gives the details.

By selection it will be seen that the average fertility was raised
from 69 per cent in the 18th generation to 75 per cent in the 19th
generation. Among the five pairs used one 19 (2) showed an
unusually high fertility (96 per cent). This pair was accordingly

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1

138

W. J. MOENKHAUS
TABLE 5

19(1).
19(2)
19(3)
19(4)
19(5)

Average fertility of 239 pairs 75 per cent.

taken to select from. Five pairs were again taken as before. The
results appear in table 6.

Thus it will be seen that all five pairs showed a uniformly high
degree of fertility. The average fertility of all the pairs was raised
to 93. 8 per cent.

5. Discussion

From the above series of experiments a number of important
facts are birought out. 1. Sterility, as it appeared in the strain
under consideration, is strongly transmissible through inheritance.
2. It is readily controlled by selection. 3. Inbreeding is probably
not the physiological cause of it.

That this sterility is transmissible cannot be doubted. The
faithfulness with which this occurs appears in the strains A and B.
Both were derived from a common pair that showed a variability
with respect to this character in the three pairs of its offspring

20(1).
20 (2) .
20 (3) .
20(4).
20 (5) .

Average for 211 pairs 93.8 per cent.

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 139

tested. One of these possessed a high degree of sterility, while the
two other pairs showed a low degree. The descendants of the latter
constituting strain 5, retained this low degree of infertility through-
out. Similarly the descendants of the former, constituting strain
A, retained their high degree of infertility up to the time when
selection away from this condition was introduced. In the latter
process the transmissibility of the character is again emphatically
revealed. In the eighteenth generation, pair 4 showed a lower
degree of sterility than any of the remaining four pairs of brothers
and sisters. Breeding from this pair at once showed offspring
with a decided decrease in sterility, compared with the eighteenth
generation, the average of the nineteenth generation being 75 per
cent of the pairs fertile as compared to 69 per cent of the latter.
Again, in the nineteenth generation, pair 19 (2) showed a much
lower degree of infertility than the other pairs. Continuing the
strain from this pair, this character is faithfully reproduced in
the offspring in that they average fertility of the latter is raised
to 93.8 per cent.

It is important to note in this connection that Castle, in his
experiments upon Drosophila, found that productiveness (which
as previously noted is quite a different thing from the sterility
here considered) was similarly transmissible and amenable to
selection. Furthermore, Castle's experiments would seem to indi-
cate that this character of productiveness behaves, in inheritance,
after the Mendelian fashion, low productiveness acting as the
recessive character. We have evidently to do here, both in the pro-
ductiveness in Castle's experiments and in the sterility in my own,
with characters that are germinal for they behave as such. In
the strain upon which my experiments Vv^ere made we have the
further remarkable condition that the infertility is inherited
only by the males.

It is clear that whatever the causal factor or factors to which the
sterility may be attributed, it is relatively insignificant compared
to the effect of selection upon it. Furthermore, the modification
is a germinal one. That inbreeding may be responsible for its
prevalance in the strain seems probable, but that it is responsible

140 W. J. MOENKHAUS

for its origin is not believed. We have seen that the general vital-
ity of the strain, as measured by its productiveness and its reaction
to light and gravity, did not suffer as a result of seventeen gener-
ations of closest inbreeding. Failing in this, it is not probable that
its effect would show itelf in so specific a way as the sudden and
complete sterility in certain males of the strain. The improba-
bility is further supported by the fact that the inbreeding may be
continued unabated if only care be exercised in the selection of
the brothers and sisters to be mated, thereby even eliminating
practically what sterility may have existed.

It is much more probable that the sterility arose spontaneously
in this strain or that it is present to a varying degree in this
species. With the character present and highly transmissible
and subject to selection it is only necessary to carry on indiscrim-
inate breeding to have the character appear in varying intensi-
ties depending upon the chance combinations. The rule of
inbreeding would be only to intensify the chance combination of
the character and to insure the more or less continued presence
in the successive generations.

That this character of sterility is not unique to this inbred
strain is evident from its rather frequent presence in pairs not
inbred. In my own experience this sterility nearly always showed
itself in the males. In one instance I found among a brood, besides
a sterile male, two females that failed to deposit eggs although
eggs were evidently present in the oviducts. Similarly Castle
found in his strain a considerable amount of sterility, and this
in some cases among the females. We see, therefore, that sterility
is not altogether rare even in broods that were not inbred.

The same facts doubtless hold for the character of productive-
ness. Castle has shown this to be transmissible and amenable
to selection. Inbreeding does not produce it but is instrumental,
with indiscriminate mating, in intensifying it, or, if the strain be
not eliminated thereby, of preserving it in the strain.

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 141
SEX-RATIO AND SELECTION
1. Introductory

The once rather generally accepted notion that nutrition was
an influential factor in the control of sex, based on the experiments
of Yung ('85), Born ('81), and others, has given place to the now
as commonly accepted idea that sex is determined prior to or
at the time of fertilization and is independent of the food. The
experimental work of Cuenot ('99) King ('07) and others, and the
splendid cytological researches of Wilson and his students are
largely responsible for this change of view and have been so fre-
quently reviewed in the various recent discussions of the problem
of sex that they need not be further detailed here.

The writer tried some starvation experiments on Drosophila
in 1904. During the past year more extensive experiments were
carried on under his direction by Mr. Claude D. Holmes, on the
effects of starvation during successive generations upon the sex-
ratio. These are published under a separate title ('10). It will
suffice in this connection, to state that the results coincide with
those of recent workers, namely that nutrition does not affect the
sex-ratio.

2. The normal sex-ratio

One fact was very apparent in these earlier tests and in all sub-
sequent experiments, that, under the varying conditions in these
creatures were reared, there was the same persistance of the pre-
dominance of females over males. Below (table 7) is given the

FOOD

TOTAL.
NUMBER
BEARED

NUMBER OP
MALES

NUMBER OF
FEMALES

RATIO

10506
2161
4048

10218

4972

995

1943

4757

5534
1166
2105
5461

1:1 113

Grapes

1:1 171

Tomatoes and grapes

Bananas . . . .

1:1.083
1*1 14

Total

26933

12667

14266

1-1 126

142 W. J. MOENKHAUS

summary of four determinations on a large scale to obtain the
normal sex-ratio. The flies were reared in the following manner.
Mason jars containing a large quantity of food were exposed to
flies in nature. The jars were left open until the larvae began to
pupate when all flies were excluded by tying a guaze over the top.
As the imagos emerged from time to time they were preserved
and the sex-ratios determined. For 26933 individuals, the ratio
was one male to 1.126 females.

In regard to these determinations only one question, so far as I
can see, can be raised. This is the academic one of the greater
mortality of the males during development or, to push the matter
back a little further and to make it applicable to recent develop-
ments in our idea of sex, the greater mortality of the male deter-
mining sex cells. In reference to this it may be pointed out that
the developmental conditions were as nearly normal as one can
imagine. There was an abundance of food, air, light and mois-
ture, and the larvae pupated in the remnants of the food in much
the same manner as one finds them doing in nature. In this con-
nection the experiments of Miss King ('07) on the influence of
food on the sex ratio of Bufo are of importance. In this she finds
that the mortality among the males is not greater than among the
females. From these facts and from the knowledge that has come
to me from the extensive rearing of Drosophilas for six years I am
convinced that the sex-ratio in this species is not one of equality.

3. Control of sex-ratio by selection

If the sex-ratio of this species, then, is that of 1 male to every
1.126 females, this should be regarded as specific just as any other
of the specific characters of the species. It should, therefore, be
subject to fluctuations and to control like other specific characters.

Starting with this conception of sex-ratio, I wished to see
whether it were possible to control this, within limits, of course,
by the process of selection. The results of these experiments
I propose to detail below.

To apply the selective process on the sex-ratio, the following
simple method was employed. Two pairs were selected from

INBREEDING AND SELECTION IN DROSOPHII,A AMPELOPHILA 143

nature, the one showing a high, the other a low female ratio.
These were selected as the parents of the two strains to be developed.
From among the offspring of each of these two pairs a number of
single matings were made. From among these the pair that
showed the most favorable ratio in the desired direction was
selected to continue the strain. The same process was repeated
as often as desired.

From a number of pairs taken from a banana bunch in Bloom-
ington June 12, 1907, two such pairs were obtained. These two
pairs go by the numbers 206 and 207, showing the following ratio :

206— 52 cf: 135 9 or 1:2.59
207—84 d^ : 75 9 or 1 : 0.89

A. Strain 206 {high female ratio). The 206 strain will, for
convenience, be called the female strain and the 207 strain the
male strain, although, as will appear, the latter never developed
into a predominantly male strain. In tables 8 and 9 are given in
diagramatic form the results of selection for five generations in the
former and six generations in the latter. At the margin the genera-
tions are numbered 1, 2, 3 etc., and the sex-ratios are indicated.

The sex-ratio of the eleven pairs of brothers and sisters mated
from the first generation of the female strain (206) varied from 1 :93
(76 d'; 71 9) to l:-7.00 (8cf:56 9).

The unusually high female ratio in the latter is probably attri-
butable to the small number of individuals obtained from this
pair. Two of the pairs threw a predominance of males (table
8 nos. 4 and 8). With the exception of no. 5, all the remaining
pairs threw a high female ratio. The ratio for all the pairs was
1:1.67 (578 &: 969 9). We have here a female ratio very much
higher than that characteristic of the species (1:1.14) and yet
considerably below that of the parent pair (1 :2.59). This may be
regarded as a regression toward the normal ratio. It should be
pointed out here that too much emphasis should not be placed
upon the exact figures representing the ratios in the different pairs,
since the number of individuals at best are rather small. In most
cases, however, when the number of offspring obtained is fairly
large, the ratio approximates the true one, so that in any given

144

W. J. MOENKHAUS

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 145

1-1 02

146 W. J. MOENKHAUS

pair from which a fairly large number of offspring has been
obtained shows a high female ratio for instance, this may be
taken as a pretty safe indication that the female ratio would be
high if all or a much larger number had been obtained.

For the next generation ten pairs were taken from brood 9 with
a ratio of 1.1.94. Brood 3, with a ratio of 1 :2.31, would have been
a more favorable one to select from, but this is not always possible
since the matings must be made before all the offspring have
emerged and therefore all the data for the complete ratio is ob-
tained. Onh^ four pairs of this series of matings came through
safely, due purely to the lack of time to give them the attention
they should have had. The four pairs threw the following sex-
ratios: 1: 2.33; 1:2,29; 1:1.27; 1:1.81. The ratio for the entire
brood was 1 :1.82 (215 d" : 391 9 ). This ratio was somew^hat more
predominantly female.

Pairs were now selected from the brood 8 with a ratio of 1 :2.29.
Of the seven pairs mated the offspring of only four was obtained
and the number of young in each case was quite small. The
ratio for all the offspring of the generation was 1:2.17 (93 cf to
201 ?). The total number here involved is so small that not
too much importance should be attached to the increased female
ratio.

For the matings of the next generation there is little doubt
that an unfortunate selection was made. The brood from which
the matings were taken showed a ratio of 1 :2.46 but this ratio was
based on numbers so small (52) that it probably did not represent
the true ratio of the pair. This may account for the drop in the
ratio for all the broods of the 4th generation to 1:1.36 (354 d'
483 9).

Two sets of matings were now made from as many broods of the
fourth generation. One of these series was again taken from the
brood showing the most favorable female ratio 1:1.90 (85 cf
162 9), but the other series was taken from a brood showing a
relatively low female ratio, 1:1.04 (64 & 67 9). From the
former the ratio of five pairs was obtained showing a ratio of 1 :1.39
(372 cT — 518 9) and from the latter the ratio of 7 pairs, show-
ing a ratio of 1 :1.07 (496 c^ : 535 9 ).

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 147

b. Strain 207 {low female ratio). From pair 207 with a ratio
of 1 :0.89 (84 o" : 75 9 ) it was hoped to develop by selection a
strain showing a low female ratio. Seven matings from the first
generation produced 536 d" and 579 $, or a ratio of 1:1.08.
The range of ratios of the individual pairs was from 1 :1.22 (99 d" :
121 9) to 1:0.86 (79 cT: 68 9). This selection was continued
for four generations, the matings being made from broods with a
low female ratio. The ratios of all the offspring in the successive
generations were 1:1.06 (220 cf : 223 9) 1:1.10 (581 c^ : 640
9); 1: 1.04 (142 o^ : 147 9); 1:1.17 (518 d' : 607 9) for the
second, third, fourth and fifth generations respectively (See Table
9) . This low female ratio showed itself rather uniformly in all the
individual matings, a notable exception occurring in the fifth gener-
tion (see Table 9, pair 3.) with a ratio of 1:2.53 (45 c^r 144 9).
On the other hand no pairs threw a great preponderance of males,
the most notable among those from which a large number of
progeny was obtained being pair 2 in the third generation in
which the ratio was 1:0.87 (115 o-: 101 9). For the sixth
generation two sets of matings were made as in the fifth genera-
tion of the strain 206. One of these was made from a brood with a
ratio of 1:2.53 (45 d" : 144 9) and the other from a brood with a
relatively low female ratio, 1:1.36 (72 c?: 98 9). From the
former the total progeny of eight matings gave a ratio of 1:1.42
(461 d : 654 9 ) and from the latter the ratio of eleven matings
was 1.1.05 (944 d^:997 9).

c. Discussion. It seems from the above experiment that the sex-
ratio in this creature is a strongly transmissible character. Start-
ing with a pair that throws an offspring showing either high or a
low female ratio it was possible to maintain, by selection, a strain
maintaining the respective ratios. The offspring from a given
pair, when mated in pairs, show^ a considerable variation in the
sex-ratio of their children. It is thus possible to develop a strain
with a low female ratio from one with a high female ratio, or the
reverse, as is shown in the fifth and sixth generation of experi-
ment 206 and 207 respectively (tables 8 and 9). The sex-ratio is
clearly amenable to selection like any other character.

148 W. J. MOENKHAUS

It is an interesting fact that it is possible to develop a strain with
a high female ratio much more easily and pronouncedly than a
male strain. I have repeatedly tried to hold the sex-ratio to or
below that of unity but without success. Not infrequently pairs
will throw a predominance of males but it has not been possible
to hold them there. The best I have ever been able to do is to
hold it considerably below that of the normal, but never as low as
unity. On the other hand, it is relatively easy to select in the
direction of females even to the extent of 1 to 2.

It should be observed that in the breeding of these strains the
most rigorous inbreeding was practiced. It might, therefore,
be that the difficulty of selecting for a low female ratio results
from the possibility that inbreeding tends toward the elimination
of the males. My extensive experience in inbreeding these crea-
tures, however, does not bear out this explanation. Furthermore,
in the sixth generation of the high female strain it, was possible
in two generations to reduce this ratio to near unity notwith-
standing that the same rigorous inbreeding was continued.

4. Relative influence of male and female in determining the sex-ratio

Having thus produced two strains showing a decided difference
in the sex-ratio of their offspring I wished to determine two further
points. First, whether the maternal or the paternal elements had
an equal share in the control of this ratio, and second, whether this
ratio was determined in the process of fertilization. To this end
reciprocal crosses were made between the two strains and the pro-
portion of the sexes in the offspring ascertained. Three experi-
ments were performed in the following manner. From among
a brood of each of the two strains a large number of individuals
were taken. Before sexual maturity a number of males and females
were isolated, while the remainder were allowed to reproduce. The
latter gave a control for each of the strains. The isolated virgin
females of one strain were mated with the males of the other.
Each experiment thus consisted of four multiple matings. (1) A
number of brothers and sisters belonging to the male strain. This
furnished a control for the male strain. (2) A number of brothers
and sisters belonging to the female strain. This furnished a control

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 149

for the female strain. (3) Females from the male strain mated
with males from the male strain, and (4) the reciprocal of '(3)'.

In crossing two strains as in the above experiment three possi-
bilities might obtain. First, that the two sexes have an equal
influence in determining the sex-ratio; second, that either sex
have a predominant influence and third, that a ratio result unlike
that obtaining in either of the parental strains. While the first
is probably the expected result, the experiments show in a
most decided way that the male has little or no influence in deter-
mining the sex-ratio in this species (tables 10, 11 and 12). In
most of the cases the ratio of the offspring falls pretty closely
around that of the strain from which the females were taken. In
two instances the ratios exceeded 100 per cent influence. The re-
maining ones, with the exception of strain 244 in which the male in-
fluence amounted to 35 per cent show the female influence almost
near enough to 100 per cent to justify one in regarding the differences
merely as fluctuations incident to the small number of individuals
involved. The unusually great influence of the male in strain
244 might be accounted for in two ways. First the number of
individuals involved in this experiment are relatively small so
that the ratios of both the control and the crossed broods are not
as reliable as in the other experiments. Secondly, the flies used
for this experiment were taken from the earlier generations of
the two strains, before, we may believe, any considerable selec-
tion had been appUed to fix the character of the respective strains.
Indeed, this seems to be borne out in the other experiments.

The materials of the three experiments were not all taken from
the same generation but were taken from different generations in
the development of the strain. Thus, in experiment 1 the broods
were taken from the first generation of strain 206 and 207. In
experiment 2 the broods came from the second generation of
strain 206 and the third of 207. The third experiment was made
from the fourth and fifth generations of strains 206 and 207
respectively. Arranging these experiments in a series, based on
the length of time that selection had been practiced on the broods
used, we see that the male influence decreases as the selective time
increases.

150

W. J. MOENKHAUS

TABLE 10

Experiment 1

No. of strain mated |

No. 242
2122 X 2122

No. 245
2122 9 X2149d'

No. 243
2149 X 2149

No. 244
214g9 X212CC?

• 9

Number of individuals

Sex-ratio (actual)

208
1.00

194
0.98

463
1.00
1.00

475
1.03

1.288

171
1.00

273
1.60

225
1.00
1.00

311
1.38

Theoretical ratio

1.288

Influence of male parents....
Influence of female parents..

7 . 3 per cent
92.7 per cent

35 per cent
65 per cent

TABLE 11
Experiment 2

No. of the strains mated. . |

No. 271
252io X 252io

No. 274
252,0 9 X2558cf

No. 272
255s X 2558

No. 273

2558 9 X252iocf

Number of individuals

332
1.00

545
1.69

589
1.00
1.00

919
1.56
1.365

739
1.00

818
1.106

680
1.00

698
1.026

Theoretical sex-ratio

1.00

1.365

Influence of male parents... .
Influence of female parents. .

22 per
78 per

cent
cent

100

per cent (—13)
per cent (1 . 13)

TABLE 12

Experiment 3

No. of strain mated [

No. 279
2758 X 2758

No. 281
2758 9 X2787cf

No. 280 No. 282

278r X 2787 i 278; 9 X 2758c:f

9 &

9 ' C

Number of individuals

Sex-ratio (actual)

289
1.00

42-,
1.477

382
1.00
1.00

551 1022
1.50 1.00
1.249:

1044 752

1.021 1.00

1.00

825
1.083

Theoretical ratio

1.249

Influence of male parents....
Influence of female parents..

per cent 13 per cent
100 per cent 87 per cent

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 151
TABLE 13

PER CENT
OF FEMALE
INFLUENCE

PER CENT

OF MALE

INFLUENCE

Experiment 1-from broods selected for one/ 212 214' 92.7 i 7.3

generation \ 214 212; 65 [ 35

Experiment 2-from broods selected for 2 and 3/ 252 255' 78 22

generations \ 255 252' 100 I

Experiment 3-from broods selected for 4 and 5/ 275 278 100

generations \ 278 275j 87 13

This fact of the prevaihng or exclusive influence of the female in
determining the sex-ratio occurs in some other species of animals.
Phylloxerans (Morgan '09) and Dinophilus apatris (Korschelt '82) .
On the other hand, Whitney ('09) seems to have shown that in
rotifers certain eggs which will produce males if unfertilized
are changed to females, if impregnated. In the case of Droso-
phila, we can not be certain that the sex-ratio is established before
fertilization since the experiments do not with certainty entirely
exclude the male influence.

5. Discussion of sex-ratio

It is not the intention to enter into an elaborate discussion of
the problem of sex control. The literature is certainly already
sufficiently burdened with such. The writer wishes merely to
point out briefly a few conclusions about sex in this species which
his results seem to warrant.

The property of sexuality possessed by this species expresses
itself not in the equal production, numerically, of its two states,
male and female, but in an unequal production. Studies in normal
sex-ratios involving a sufficiently large number of individuals are
not numerous. The unequal production of the two sexes in the
human species is well established. Montgomery ('08) has given
the data of a large number of individuals of Theridium and finds
a marked inequality in the sexes. The general assumption seems
to be that an equal sex-ratio is the rule. It is not improbable,

152 W. J. MOENKHAUS

however, that, as careful determinations upon different species
multiply, the condition of unequal ratios will be found increasingly
common. Any theory of sex must take into consideration this
normal inequality in the sex-ratios.

Sex-ratio like color, size etc., is a character belonging to a
species. Sexuality of course is not, for it is common to all species
reproducing by the sexual method. The particular form of sexu-
ality, however, the proportion of the two sexual persons to which
it gives expression in the process of differentiation, this is specific.
For Drosophila ampelophila, the ratio of one male to 1.126 females
is a specific character. This is not a ratio of merely the present
generations but has been transmitted from generations remote.
It is inherited. It is the expression of the physiological condition
to which the species has been developed by its environmental
demands.

Like other specific characters this ratio should be subject to
modification, but this should not be more easily done or by other
methods, in general, than those used in the modification of other
characters. From this view point it should not be expected that
the sex-ratio in an animal could be materially changed by such
agents as food, temperature,etc. A change in the proportion of the
sexes involves a much more fundamental modification than simple
starvation or the reverse is likely to induce. In regard to other
characters, we have long ago ceased to regard them as modifiable
by such methods, but in the case of sex, it is only recently that
their futility is being entertained. The most potent factor and
the one most generally used to modify a character is selection.
If the experiments herein recorded prove what they are held to
prove, this process of selection is a potent factor in the modifica-
tion of the sex-ratio also. It would be interesting to try to line
this fact up with the chromosomal conception of sex. However,
the writer regards this as the task of those who are engaged in
these interesting and important investigations.

INBREEDING AND SELECTION IN DROSOPHILA AMPELOPHILA 153
SUMMARY

1. Drosophila ampelophila may be inbred (brothers and sis-
ters) for seventy-five or more generations.

2. Inbreeding in itself is not deleterious to the fertility or
vigor of this species.

3. ' Infertility normally occurs to a varying degree among the
offspring of any pair. Promiscuous inbreeding among such off-
spring may perpetuate and even intensify this character. When
sterility appeared in the strain experimented with, it was always
complete, appeared suddenly and was confined to the male.

4. By the judicious selection of the brothers and sisters
to be mated from a brood that shows a high degree of infertility,
this infertility can be eliminated by selection although continuing
the inbreeding in the closest possible way.

5. There is a wide divergence in the fertility and productive-
ness among the different pairs taken in nature, but by the proper
selection and closest inbreeding these may be readily brought
to either a high or low state with respect to these characters.

6. Many generations of closest inbreeding does not neces-
sarily cause any loss in size, perfection of form, rate of reaction to
light and gravity, egg production or length of life and sex-ratio.

7. The normal sex-ratio of this species in nature when reared
under diverse conditions of food is one male to 1.126 females.

8. Different pairs in nature show a wide divergence in the sex-
ratio of their offspring.

9. When the offspring from a pair with a given ratio are mated
in pairs their offspring will show a wide range in the sex-ratio but
in the aggregate will tend to reproduce the ratio of the brood to
which they belong.

10. Sex-ratio is therefore a character that is strongly trans-
missible. By the proper selection of pairs tending to throw a
high female ratio on the one hand or a low female ratio on the
other it is possible to develop strains characterized by high or
low female ratios.

11. In this species it is comparatively easy to develop a strain
with a female ratio considerably higher than the normal but very

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 1

154 W. J. MOENKHAUS

difficult to develop a strain with a female ratio much lower than
the normal or even one in which the sexes are equal in number.

12. Sex-ratio is one of the qualities that is, like color, an inher-
ent characteristic of this creature, strongly transmissible and
amenable to the process of selection.

13. The female is almost wholly responsible in the transmis-
sion of the sex-ratio. For, if females from a strain possessing a
high female ratio be mated with males from a strain possessing a
low female ratio or vice versa, the offspring will show a sex-ratio
which is wholly or very near that of the strains from which the
females were taken.

14. Sex is probably very little, if at all, influenced at fertiliza-
tion in this species, but is probably determined much earlier and
by the female, but there seems no reason why this may not be
influenced by various factors and in some species at fertilization.

LITERATURE CITED

Born, G. 1881 Experimentelle Untersuchungen iiber die Entstehung der Ge-
schlechtsunterschiede. Breslauer iirtzliche Zeitschrift. Bd. 3.

Bos, J. R. 1894 Untersuchungen fiber die Folgen der Zucht in engster Blutsver-
wandtschaft. Biol. Centralbl. pi. Bd. 14.

Castle, W. E. and others. 1906 The effects of inbreeding, cross-breeding and
selecLion upon the fertility and variability of Drosophila. Proc. Amer.
Acad. Arts and Sciences, vol. 41 .

CuENOT, L. 1899 Sur la determination du sexe chez les animaux. Bull, scientif.
de la France et de la Belgique, t. 32.

Gentry, N. W. 1905 Inbreeding Berkshires. Proc. Amer. Breeders Association
vol. 1.

Guaita, G. VON 1898 Versuche mit Kreuzungen von verschiedenen Rassen des
Hausmaus. Ber, ub. d. Verhandl. d. Naturforsch. Gesellsch. zu Frei-
burg, Bd. 10.

Holmes, Claude D. 1910 The effect of starvation for five successive genera-
tions on the sex-ratio in Drosophila ampeloph'.la. Indiana University
Studies No. 2.

King, Helen D. 1907 Food as a factor in the determination of se> in Amphi-
bians. Biol. Bull., vol. 13.

KoRSCHELT, E. 1882 tJber Bau und Entwickelung des Dinophilus apatris. Zeit-
schrift f. wiss. Zool. Bd. 37.

Morgan, T. H. 1909 A biological and cytological study of sex determination
in Phylloxerans and Aphids. Jour. Exp. Zool. vol. 7.

Whitney, D. D. 1909. Observations on the maturation stages of the partheno-
genetic and sexual eggs of Hydatina senta. Jour. Exp. Zool. vol. 6.

Yung, E. 1885 De 1' influence des \ariations du millieu physico-chimique sur
le development des animaux. Arch, des Sci. phys. et naturelles, t. 14.

CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF
COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. E. L. MARK, Director. No. 23L

THE MECHANISM OF LOCOMOTION IN GASTROPODS

G. H. PARKER
INTRODUCTION

The snail's foot in locomotion is so striking and so easily ob-
served that it has excited the interest of naturalists for a long
time and yet a complete solution of even the mechanical prob-
lems connected with its action seems not to have been attained.
Within recent times a number of investigators have attacked
the problem of locomotion in snails, but their efforts have been
directed chiefly toward the elucidation of the action of the neuro-
muscular mechanism rather than toward an understanding of
the external mechanical conditions that accompany locomotion.
It is the object of this paper to consider, in the light of the more
recent investigations and from the standpoint of renewed observ-
ation, the external mechanical factors involved in the movements
of the gastropod foot.

The observations recorded in this paper were made partl}^ at
the Bermuda Biological Laboratory, at the Harvard Zoological
Laboratory, and at the Biological Laboratory of the United States
Bureau of Fisheries at Woods Hole. I am under obligations to
the directors of the laboratories mentioned for the materials and
opportunities for carrying on these studies.

TYPES OF MOVEMENT

When the locomotor movements of the foot in many species of
gastropods are compared, a surprising diversity is found. These
different types of movement have been well classified by Vies

155

156 G. H. PARKER

('07) and are apparently characteristic not only for species but
for larger groups of gastropods. In the majority of species thus
far examined, the pedal waves course forward over the foot, thus
agreeing in direction with the animal's locomotion. Vies has
appropriately designated this type of movement as the direct type
and has given the following gastropods as examples; the pulmo-
nates (including Onchidium), Aplysia, Aeolis, Doris, Haliotis,
Trochus, Cyclostoma, and certain small species of Littorina. I
can confirm this statement for such of these molluscs as I have
examined, namely, many pulmonates, including Onchidium, and
I can add to this list Crepidula fornicata. In other gastropods
the waves pass over the foot from anterior to posterior and this
type has been designated by Vies as retrograde. As examples he
has given Acanthochites fascicularis, Littorina littorea, and L.
rudis. Besides confirming Vies' observation on Littorina lit-
torea, I can add to this list Dolabrifera virens Verrill, Tectarius
nodulosus Gmel., Nerita tessellata Gmel., and Chiton tuber-
culatus Linn. According to the observations of Jordan ('01,
p. 99) Aplysia belongs under this head and not under that of the
direct type as given by Vies.

In both chief types of movement several subtypes can be dis-
tinguished as determined by the lateral extent of the pedal waves.
In some gastropods each wave extends over the functional width
of the foot and thus the foot is occupied by only a single series
of waves. This subtype has been termed by Vies monotaxic,
and is exemplified by the pulmonates and chitons. In addition
to these gastropods, Dolabrifera virens also has a monotaxic
wave. In other gastropods the foot is functionally or even struc-
turally divided along the median plane and exhibits a double sys-
tem of waves, one right and the other left. This subtype has been
designated ditaxic by Vies and is exemplified by Haliotis, Tro-
chus, and Cyclostoma among the direct types, and by Littorina
littorea among the retrograde types. Besides confirming Vies'
statement as to Littorina littorea, I can add Tectarius nodulosus
and Nerita tessellata as ditaxic gastropods. In Tectarius the
waves on the two sides of the foot usually alternate and they are

LOCOMOTION IN GASTROPODS 157

SO extensive that never more than two waves can be seen on one
side of the foot at once. The foot, therefore, moves forward in
alternate steps, first on the right side and then on the left, the
motion resembUng that of a person in a sack walk. In Nerita
the wave begins anteriorly as a single wave whereupon it breaks
and passes down the right and left sides of the foot to unite as
one wave again at the posterior margin. These two conditions
of alternate waves, as in Tectarius, and opposite waves, as in Nerita.
will probably be found exemplified in other ditaxic gastropods.
In certain small species of Littorina with direct movements, Vies
has described four parallel sets of waves, fulfilling the require-
ments of a tetrataxic subtype. This occurs, according to Vies,
only in connection with the retrograde type of movement. I
have seen no example of it.

Among those snails that I have examined, one species, Ilyan-
assa obsoleta (Say), seems to find no place in Vies' classification.
This snail is a vigorous, active creeper. Its foot covers a large
area compared with the size of its body. Anteriorly the foot is
truncated and auriculate; posteriorly it is bluntly rounded. Its
ventral surface is whitish, flecked over with irregular grayish
splotches. In resting, the snail uses chiefly the posterior part of
the foot, the anterior part being sometimes more or less with-
drawn into the shell. In locomotion the anterior part seems to
be the more active. Notwithstanding the fact that this snail
is very easily observed in active creeping and that its foot is
marked in a most favorable way for exhibiting wave-like move-
ments, I have never been able to discover any evidence of such
movements. When in locoixiotion, the whole foot seems to glide
at a uniform rate over the surface of attachment such as that of a
glass plate. Only along the anterior edge and over a small por-
tion of the anterior ventral surface of the foot, can slight varia-
tions in the rate of movement be discovered and these variations
are so local and scattered that they can in no sense be regarded
as forming a wave. The movement of the foot of Ilyanassa has
a most striking resemblance to that of the foot of a planarian in
which cilia may be the chief motor organs, but on testing the foot

158 G. H. PARKER

of Ilyanassa with carmine suspended in seawater. not the least evi-
dence of cilia could be discovered. I therefore believe that Ily-
anassa moves by a form of muscular activity that does not appear
as pedal waves and it is not improbable that other gastropods
will be found that have the same peculiarity. That Vies recog-
nized something of this kind may be inferred from his statement
that no changes in color can be seen in the creeping foot of Na-ssa,
Buccinum, Aeolis, etc., and that the direction of the waves in
these instances can be judged only by the deformations produced
at the edge of the foot. As Vies makes no further mention of
Nassa in his subsequent account, I suspect that it is more or less
like Ilyanassa and is capable of little or no pedal-wave movement.
The locomotion of such gastropods I should designate as due to
arhythmic pedal movements as contrasted with rhythmic pedal
movements, such as have been fully classified by Vies.

It is a significant fact that all gastropods, irrespective of their
type of movement (direct or retrograde), are restricted to
forward locomotion. None, so far as I am aware, can reverse
and move backward as, for instance, an earthworm can. What-
ever differences these various types of pedal movements possess,
they still lead to but one result, the forward locomotion of the
snail.

THE GASTROPOD FOOT AS A HOLDFAST

The snail's foot subserves the double function of attachment
and locomotion. As means of attachment snails secrete a bed
of mucus, and use the foot as a sucker. Both methods are com-
monly employed by the same species, but in a given form one
method is usually developed much in excess of the other. For
instance, in Helix pomatia, Limax maximus, and other allied
species, the moist surface of the expanded foot will stick with
some tenacity to glass. But if such an animal be allowed to creep
its length over a glass surface and thus spread a bed of mucus on
which it can rest, it will be found to have multiplied the strength
of its attachment many times. The mucus adheres to the glass
and the surface of the foot to the mucus very much more power-

LOCOMOTION IN GASTROPODS 159

fully than the foot alone can adhere to the glass. That this
attachment is due chiefly to the adhesive properties of the mucus
and not to the sucking action of the foot, is seen from the fact that
the attachment can be completely accomplished over a minute
hole in a plate of glass. When a snail in such a position is seized
and drawn off, air is sucked in through the hole in the glass as the
middle of the foot rises, showing that under these extreme cir-
cumstances, the foot does act as a sucker, but in the ordinary
resting state of the snail no such suction is exerted. All snails
with which I am acquainted deposit more or less mucus and
though this is sometimes so small in amount that it can be demon-
strated only by means of powdered carmine, it serves, I believe,
in so far as it is present, as a means of attachment. This produc-
tion of mucus is highly developed in the pulmonates. Its rela-
tion to creeping on the surface-film of water, as exhibited by many
fresh-water snails, has long been recognized.

In some snails the foot serves as an organ of attachment chiefly
through its power of suction. The general surface of the foot is
applied closely to the substrate after which the central portion
is lifted thus converting the foot into a sucker. This kind of
attachment is well exemplified in Patella, Crepidula, etc. Crep-
idula fornicata can be made to creep over a surface of glass and
can move with ease and security over a minute hole in the glass.
If, however, the snail is disturbed by being touched several times
when its foot is over the hole, it will actually dislodge itself by
endeavoring to suck firmly to the glass, for in so doing it will
fill to repletion the forming concavity on the underside of its foot
by sucking water through the underlying hole. When one con-
trasts the difficulty with which Crepidula is dislodged from its
natural surface of attachment, particularly after it has been in-
duced to exert full suction, with the ease with which it can be
made to dislodge itself when over a small hole, the magnitude of
its power of suction becomes apparent. The action of the foot
of Aplysia as a suction apparatus has already been demonstrated
by Jordan ('01). These two methods -of attachment, suction,
and adhesion through mucus are the chief means by which snails
hold to the surfaces on which they creep.

160 . G. H. PARKER

THE GASTROPOD FOOT AS A LOCOMOTOR ORGAN

Locomotion by the gastropod foot, is not dependent upon
ciliary action but is a muscular operation as shown by Dubois
and Vies ('07) . The precise way in which the movements of loco-
motion are accomplished can best be made out by examining good
examples of direct and retrograde movement. The first is well
exemplified in Helix pomatia and Limax maximus; the second in
Chiton tuberculatus and Dolabrifera virens.

In an expanded and actively creeping Helix pomatia, the foot
may measure as much as seven to eight centimeters in length
by two and a half in width. Over this a succession of transverse,
dark-brownish waves run from posterior to anterior. At any
instant there may be as many as ten or a dozen such waves on the
foot. Each wave is separated from its neighbor by a space equal
to about three-times its own thickness. The waves travel over
the foot in about thirty seconds, or at a rate of a centimeter in
seven to eight seconds. These records, taken from a normal in-
dividual, agree fairly well with those given by Bohn ('02) and by
Biedermann ('05).

As the snail creeps, it spreads from the mucous gland at the an-
terior edge of its foot a broad path of slime over which it makes
its way. An active snail marks its course in this manner by a
long track of slime. A somewhat exhausted snail, when placed
upon an appropriate substrate, will almost always creep far enough
to lay a mucous path that will subtend the whole of its foot, after
which it will cease creeping. If it is removed to another position,
it will usually repeat this operation, but it will seldom creep far-
ther. This habit is doubtless connected with the effectual at-
tachment of its foot to the substrate.

Locomotion in Helix, like that in other pulmonates (Kiinkel,
'03), is apparently inseparable from the wave movement of its
foot. When a snail is placed upon a glass plate preparatory to
creeping, it lengthens and expands its foot; almost immediately
thereafter pedal waves appear and the animal begins to move
forward. Such a snail will creep over a perforation in a glass plate

LOCOMOTION IN GASTROPODS 161

without isuc'king air through the perforation, thus demonstrating
that its attachment in locomotion/ as in rest, is due to adhesion
and not to suction. In fact in a creeping Helix the foot not only
does not suck but actually presses on the substrate. If, as the
snail creeps, a bubble of air is introduced under it by a capillary
tube or other means, this air will usually escape at the edge of the
foot in such a way as to show that it was under considerable pres-
sure. The action of such bubbles demonstrates that the foot as
a whole is firmly attached to the mucous substrate, in fact presses
against it. Locomotion in Helix pomatia, then, has to overcome
under ordinary circumstances only the adhesion of the foot and
this is accomplished apparently by the pedal waves. In snails
in which the attachment is due to suction as well as to adhesion,
locomotion requires that both attractive forces shall have been
overcome, but, as suction is muscular, it seems likely that this
would be relaxed somewhat, as seems to be the case in Crepidula,
before locomotion begins.

How the pedal waves accomplish locomotion is still a disputed
question. According to von Uexkiill ('09, p. 181), who has fol-
lowed Jordan ('01) and Biedermann ('05) in many particulars,
each pedal wave is formed by the contraction of the longitudinal
muscles of the foot and takes the form of a slight swelling on the
underside of the organ. Such a wave, as von Uexkiill rightly re-
marks, would effect nothing by way of locomotion unless some
portion of the foot were fixed. Von Uexkiill ('09, p. 187) believes
that the foot is provided with some such mechanical device as the
setae of the earthworm, which, resist backward movement while
they allow forward motion and that, therefore, the region in front
of each wave may be regarded as a fixed region. Hence the con-
traction waves would always draw that portion of the foot where
they temporarily were forward over the substrate toward the
fixed point in front and as a result forward locomotion would be
accomplished.

Although this explanation is free from mechanical objections,
it is doubtful whether it really applies to the case in hand. Von
Uexkiill has maintained in support of this view, that a snail can

162 G. H. PARKER

be slipped over a glass plate more easily forward than backward,
just as an earthworm can be drawn over an appropriate surface
more easily headward than tailward. I must confess that I have
not been able to convince myself that there is any difference
in this respect in Helix pomatia or Limax maximus; both
seem to slip over the glass forward and backward with equal
ease.

Moreover, the view advanced by von Uexklill is based upon
what I believe to be a somewhat erroneous conception of the pedal
wave. Biedermann ('05, p. 11) pointed out that the foot of Helix
pomatia has great advantages over that of many other gastro-
pods for studies of this kind because of the numerous small specks
contained in its outer layer. These specks can be discerned clearly
by means of a hand lens and they give a true picture of the move-
ments of the foot. As watched through a plate of glass over which
the anuTial is creeping, they can be seen, as Biedermann has de-
scribed, to move momentarily forward, then come to rest, and then
again to move forward. This is best demonstrated on a sheet of
glass on which there are numerous scratches. Such scratches
serve as landmarks and by them it can be seen that the minute
specks in the foot do remain essentially fixed in position and then
momentarily move forward to assume again for a brief period a
position of rest. When this motion is examined in relation to the
foot as a whole, it is evident that the forward motion takes place
in the dark waves and that quiescence is characteristic of the
intermediate lighter portions of the foot. Each wave, then, is a
pulse of forward motion and the rest of the foot is momentarily
quiescent. The area covered by the waves is probably a fourth
or a fifth of the total area of the foot. At any moment, therefore,
about three-fourths to four-fifths of the surface of the foot is
stationary and about one-fourth to one-fifth is moving forward.
In other words the snail stands on the greater part of its foot while
it moves forward with a much lesser part.

Essentially the same conditions as have been described for Helix
pomatia can be demonstrated in Limax maximus. If particles
of carmine be driven into the substance of the median, active
band of the foot of this slug, they can be seen to exhibit exactly

LOCOMOTION IN GASTROPODS 163

the same type of movement as has been described for the specks
in the foot of HeUx. In Limax the waves, however, are light in
color, instead of being dark as in Helix, and their surfaces, .as
seen in the air, are marked with fine wrinkles transverse to the
longitudinal axis of the animal. These wi'inkles show that the
waves are regions of longidudinal contraction, as has been main-
tained by most recent writers on this subject.

The chief error in most previous accounts of the locomotion of
the gastropod foot is found in the physical configuration ascribed
to the underside of this organ. Biedermann ('05, pp. 10, 17)
states that the waves are convexities on the surface of the foot
and that they press more firmly against the substrate than does
the rest of the foot. This view was adopted by von Uexkiill
('09, p. 187) in his discussion of gastropod locomotion. In Helix
pomatia it is by no means easy to determine whether the waves
are convexities or not, for the reason that they are at most only
very slightly different in level from the general surface of the
foot. On inspecting by reflected light the free ventral surface of
a part of a Helix foot over which waves were running, I was un-
able to tell with certainty whether the surfaces of the waves were
convex, concave, or flat. If, however, the creeping foot be closely
studied through glass, evidence of a conclusive kind can be found.
If, under these circumstances, a very minute air bubble entangled
in the mucus under the snail is watched, it will be seen to change
its form and position slightly as each wave passes over it. As
the wave approaches it, it will elongate slightly on its face
next the wave and at times move a little towards the wave, and
as the wave leaves it, it will elongate slightly in the opposite
direction and at times follow slightly the retreating wave. The
motions of the bubble are exactly those that should be expected
provided the wave exerted a slight suction in its passage and the
reverse of what would occur supposing the wave pressed upon the
bubble. The evidence, though slight, is clear and I, therefore,
believe that each wave on the underside of the foot of Helix
pomatia is a slight concavity.

Although the configuration of the surface of the wave in Helix
pomatia could be determined only indirectly, in Limax maximus

164 G. H. PARKER

it can be seen with distinctness. If the anterior part of the
foot of this slug be appHed to a glass surface, the pedal waves
appear quickly over the whole foot. On inspecting the portion
of the foot not yet in contact with the glass, the waves can be
identified as dark bands alternating with light areas. On examin-
ing from the side the portion of the foot not yet in contact with
the glass, it can be clearly seen that the waves are concavities
in the foot as compared with the areas between the waves. I
am, therefore, entirely convinced that, contrary to the opinion
expressed by Biedermann and others, the pedal waves of the gas-
tropods are concavities and not convexities on the foot. In these
concavities, which are probably filled with the more fluid portion
of the mucus, the foot moves forward, the rest of this organ being
temporarily at a standstill.

The mechanical advantage of this arrangement must be obvious.
The snail is attached to the substrate chiefly by adhesion to the
denser mucus. This attractive force is overcome by drawing
certain parts of the foot, the region of the waves, away from the
substrate. These parts are then in a position to move with re-
duced resistance and are momentarily shifted forward while the
snail supports itself on the rest of its foot. As this release from
adhesion is propagated as a wave over the whole of the foot, this
whole organ, together with the rest of the snail, is eventually
moved forward. At first thought it might seem that such a
wave movement could not produce so uniform a motion as snails
show, but it must be remembered that the uniformity of this
movement is seen only in parts of the animal some distance from
the foot. On the foot itself the operation is alternate movement
an4 rest, which becomes more and more continuous motion as
points on the body more and more distant from the foot are reached.
The locomotion is in many fundamental respects like that of the
human being. In our locomotion each foot is alternately at rest
and in motion and yet distant parts of our body, like the head,
show a motion which in comparison with that of our feet is almost
continuously uniform. In fact, a ditaxic gastropod with alter-
nate, direct, single waves on the foot would almost exactly re-
produce the method of locomotion found in the human being.

LOCOMOTION IN GASTROPODS 165

Tectarius, as already noted, practicalh^ fulfills these conditions
except that its waves are retrograde. This general theory of
the mechanics of gastropod locomotion is an elaboration of the
views already set forth by Jordan ('01).

It is not my purpose in this paper to enter into an account of
the musculature by which the movements already described are
carried out, for I have made no observations on this part of the
subject. It is, however, pertinent to show that the elements of
motion implied in the preceding description are not inconsistent
with the general structure of the snail's foot. The work of Jor-
dan ('01), Biedermann ('05), and others shows conclusively, I
believe, that the musculature of the snail's foot works against
the elastic-walled, fluid-filled cavities of the animal's interior
and that these cavities are often temporarily closed from one
another. It is these spaces which, acting collectively as a vacuo-
lated, erectile tissue, give rise to such rigidity as is possessed by
the expanded foot of the snail. In this tissue two sets of mus-
cles, longitudinal and dorso-ventral, have been identified. The
dorso-ventral muscles lift the foot locally from the substrate.
They are imbedded in the vacuolated tissue already mentioned
and when they contract, their dorsal ends, being more firmly
set than their ventral ones, serve as relatively fixed points and
the ventral ones, therefore, move. The mechanical support that
these muscles receive comes primarily from the tissue adjacent
to their dorsal ends which in turn gets its support from other
tissues reaching to the parts of the foot fixed on the substrate in
front and behind the region of elevation. The "action of the ven-
tral end lifts the foot locally and overcomes adhesion in the given
region. When the muscle relaxes, the portion of the foot that was
elevated is returned to its former level chiefly by the elastic action
of the vacuolated tissue and the muscle recovers its original length
and position. This action of the dorso-ventral muscles takes
place in sequence from behind forward and thus a concave wave
runs on the surface of the foot from tail to head.

The second element in the pedal wave is the forward movement
of that portion of the foot which is temporarily lifted from the
substrate. This must be accomplished by the contraction of the

166 G. H. PARKER

longitudinal muscles and can be best pictured by reference to the
accompanying diagrams. These diagrams represent steps in the
passage of a concave wave over the foot of a snail from an anterior
position to a posterior one (left to right in the diagram) whereby
the pointlc is temporarily released from full adhesion to the mu-
cous surface, moved forward, and brought to full adhesion again.
The point x is supposed to be associated with a particular longi-
tudinal muscle fiber, number 2, through whose action it is moved.
In A, this fiber is shown in its relaxed condition with the wave
approaching. In B, the wave has released the point x from full

I 2 3

A ' ' — '

B i 1

C ' i

D ' '

K ' ' '—~

?-^- ^-^-^ ' ^<^

adhesion. In C, fiber 2 has contracted and since the posterior
end of it is over a released part of the foot and the anterior end
over a fixed part, the posterior end with the underlying point x
has been moved anteriorly. In D, the fiber remains contracted
and the point x has come again to adhere to the substrate. In E,
the wave has reached the next longitudinal fibre anterior, number
3, which has contracted and drawn out the relaxing fiber, number 2,
to its original length and position in reference to point x. The
contraction of each longitudinal fibre then serves two purposes:
it moves the foot forward as the releasing wave passes over the
region and it extends the relaxing posterior fiber. In this way each

LOCOMOTION IN GASTROPODS 167

point on the foot is lifted, moved forward, and set down again
and thus the foot, and with it the animal as a whole moves for-
ward. From this theoretic consideration, it is evident that the
theory of pedal-wave action advanced in the preceding paragraphs
is entirely consistent with such an arrangement of muscles as has
long been known to occur in the gastropod foot.

Vies ('07) has called attention to the fact that the majority of
theories as to the locomotor action of the gastropod foot apply
only to the direct type of movement and do not take into account
the retrograde type. The theory put forward in this paper is
believed to apply equally well to both types. Among retrograde
gastropods. Chiton tuberculatus is an excellent example. This
mollusc uses its foot as a sucker, but nevertheless can creep with
considerable rapidity. It exhibits, as a rule, not more than two
waves on the foot at a time; these course posteriorly at the rate
of about a centimeter in five seconds. In a Chiton creeping over
a glass plate, the wave when viewed from the side can be seen to be
an area lifted well off the substrate. This feature is much more
conspicuous in Chiton than in any other mollusc that I have ex-
amined. As in the pulmonates, the surface of the Chiton foot in
direct contact with the substrate is motionless ; that in the wave
area moves forward. At any moment about one quarter of the
Chiton foot is moving forward while the animal supports itself on
the remaining three quarters.

In Dolabrifera the foot is pear-shaped in outline with the
rounded end posterior. It is about 8 mm. in length. In creeping,
one to two waves can be seen on its surface at once; each wave
sweeps the length of the foot in about seven seconds. As in Chiton,
the waves can be clearly seen to be areas in which the foot is lifted
completely from the substrate to which the rest of the foot is
firmly applied. The pedal surface is mottled and in the wave
area it can be seen to be moving forward, whereas on the rest of
the foot it is motionless. The total wave area is about one-half
the total area of the foot.

The conditions in Chiton and in Dolabrifera are essentially
similar to those in the pulmonates, except that the pedal waves
progress posteriorly instead of anteriorly, i.e., the dorso-ventral

168 G. H. PARKER

muscles contract in sequence from the anterior to the posterior
end instead of the reverse and the longitudinal muscles follow
the same sequence; otherwise they act as they do in the direct
type. It is evident from this brief discussion of the nature of the
waves in the retrograde type that the theory developed in connec-
tion with the direct type applies perfectly to this second type.

It remains still to point out that what I have called the arhyth-
mic form of pedal locomotion, a form well exemplified in Ilyan-
assa, may be explained on the same general basis as that which
has just been given for the two types of arhyhmic locomotion. If
the foot of such a snail as Ilyanassa be thought of as composed of
a multitude of small areas, each one of which can be lifted from
the substrate, moved forward, and set down again separately,
and that this action takes place irregularly and without reference
to any sequence, it can easily be seen how the animal could move
forward but without the formation of pedal waves. It is my be-
lief that this is the condition in the foot of the arhythmic gastro-
pods, but because of the small size of Ilyanassa, I have not been
able to subject this opinion to experimental test.

Before closing this paper, I wish to add a word concerning the
very remarkable method of locomotion observed by Carlson ('05)
in Helix dupetithouarsi. The movement carried out by this
snail is appropriately described as a gallop, both from its rate and
configuration. The snail on strong provocation lifts the head and
projects it forward, and eventually brings it to the ground, thus
initiating a giant wave which proceeds backward over the length
of the body. Several such waves may be present at once. Carl-
son suggests that this movement is only an exaggerated form
of the ordinary locomotion, but I am inclined to agree with Jordan
('05, p. 104) that this is probably an entirely different type of
locomotion and I suspect that this snail also possesses the typical
pedal wave. In fact it seems to me likely that the gallop was,
so to speak, superimposed on the pedal wave system and, had the
snail when in gallop been examined from below, the pedal waves
would have been seen in operation in conjunction with the body
waves. I am the more inclined to the view that the gallop is an
independent form of locomotion as compared with the pedal

LOCOMOTION IN GASTROPODS 169

waves, because in the gallop the body waves of this species, as
reported by Carlson, were retrograde whereas the pedal waves in
all Helices thus far reported are direct.

SUMMARY

Ordinary gastropod locomotion is accomplished either without
pedal waves (arhythmic) or with pedal waves (rhythmic). In
rhythmic locomotion the waves may run from posterior to ante-
rior (direct) or the reverse (retrograde). The foot may exhibit
one (monotaxic), two (ditaxic), or four (tetrataxic; series of
waves. In the ditaxic foot the waves may be alternate or oppo-
site.

The gastropod foot is an organ of attachment through adhesion
(mucus) or suction, or both.

The pedal wave is an area of the foot that is lifted off the sub-
strate as compared with the rest of the foot and thereby freed
more or less from aahesion. It is also the region of the foot that
moves forward, the rest of the foot remaining temporarily station-
ary. Locomotion is the cumulative result of local forward motion
on the part of one section of the foot after another till the whole
foot has been moved. The same type of muscular movement as
that seen in rhythmic locomotion can be present in a diffuse form
(not wave-like) in a gastropod foot and will result in locomotion.

170 G. H. PARKER

BIBLIOGRAPHY

BiEDERMANN, W. 1905 Studien zur vergleichenden Ph}^siologie der peristalt-
ischen Bewegungen. II. Die locomotorischen Wellen der Schnecken-
sohle. Arch. f. ges. Physiol., Bd. 107, pp. 1-56, Taf. 1-2.

BoHN, G. 1902 Des ondes musculaires, respiratoires et-locomotrices, chez les
Annelides et les Mollusques. Bull. Mus. Hist. Nat., Paris, tome 8,pp.
96-102.

Carlson, A. J. 1905 The physiology of locomotion in gastropods. Biol. Bull. ,
vol. 8, pp. 85-92.

Dubois, R., et Vles, F. 1907 Locomotion des Gasteropodes. Compt. rend.
Acad. Sci., Paris, tome 144, pp. 658-659.

Jordan, H. 1901 Die Physiologic der Locomotion bei Aplysia limacina. Zeit. f.
Biol., Bd. 41, pp. 19&-238, Taf. 2.

1905 The physiology of locomotion in gastropods. Biol. Bull., vol. 9,
pp. 138-140.

KtJNKEL, K. 1903 Zur Locomotion unserer Nacktschnecken. Zool. Anz., Bd.
26, pp. 560-566.

Uexkull, J. V. 1909 Umwelt und Innenwelt der Tiere. Berlin, 8vo, 261 pp.

Vles, F. 1907 Sur les ondes pedieuses des Mollusques reptateurs. Compt.
rend. Acad. Sci., Paris, tome 145, pp. 276-278.

THE REGULATORY PROCESSES IN ORGANISMS

C. M. CHILD

Rull Zoological Laboratory, University of Chicago

Introduction 171

The organism as a physico-chemical system 173

1. The relation between metabolism, and structure 173

2. Physiological correlation and the physiological system or individual . . 179

3. The basis and nature of physiological correlation ISl

The nature of regulation 182

1. Organic or physiological equilibrium and equilibration 182

2. Regulation as equilibration 188

The regulatory processes 199

1. The relation between form regulation and functional regulation 199

2. The inducing conditions and the results 199

3. The provisional classification of the regulatory processes 200

a. The two methods of regulation 200

h. Regulatory compensation 202

c. Regulatory transformation 205

The nature of reconstitution 207

1. Restitution or reconstitution? 207

2. The initiating factor in reconstitution 211

3. The process of equilibration in reconstitution . 212

4. The complexity of reconstitution .215

5. The limits of reconstitution 217

Reproduction in general as a form of reconstitution 218

Conclusion 221

Bibliography 222

INTRODUCTION

Of late years the term 'regulation* has come into such general
use and has been applied to so wide a range of organic phenom-
ena, that it seems desirable to attempt a general consideration
and analysis of the regulatory processes from the present view-
point of physiology. Tne biologist who takes the position that
there is at the present time, when the investigation and analysis
of the physics and chemistry of the organic processes is still only

JOURN.\L OK MORPHOLOGY, VOL. 22. NO. 2

172 C. M. CHILD

at its beginning, no adequate basis for 'vitalistic' interpretations
of regulatory phenomena finds but little satisfaction or enlighten-
ment in Driesch's 'entelechy' or in other assumptions of the neo-
vitalistic school.

In the present state of our knowledge these views are and must
remain expressions of personal opinion. Driesch's first two ''Be-
weise der Autonomie der Lebensvorgange" (Driesch, '01, '03 etc.),
which are based on certain phenomena of form regulation, con-
stitute proofs only when we accept Driesch's premises, and as I
have pointed out (Child, '08b) these premises are pure assump-
tions. Neither Driesch nor anyone else has placed them on a
foundation of fact. The existence of the 'harmonious-equipo-
tentiai system,' for example, which is of so great importance to
Driesch, is a matter of assumption, not of fact. So far as the sys-
tems, which according to Driesch belong in this category, have
been thoroughly examined, they have shown themselves to be
neither harmonious nor equipotential in Driesch's sense, and to
the extent which he has assumed. It is of course easy to assume,
as Driesch has done, that the harmony of these systems is due to
entelechy and their limitations to physico-chemical factors, but
such assumptions, since they are so manifestly invented ad hoc,
do not carry conviction to the minds of most biologists, what
ever, their effect upon their author.

Much the same is true of other modern vitalistic hypotheses:
as expressions of personal opinion, they are of great interest in
the histor}^ of scientific thought, but none of them thus far has
presented any convincing arguments in its own support.

Furthermore, with the exception of Driesch's analytical con-
sideration of the regulatory phenomena in organisms, most of
the recent published works of general character, which concern
themselves primarily with the regulations which involve the
visible morphological features of the organism, e. g., the books of
Morgan ('07), Korschelt ('07) and Przibram ('09), have been de-
voted chiefly to the descriotive, rather than the analytical and in-
terpretative aspects of the subject.

In view of these facts, an attempt at physiological analysis of
the regulatory processes or of some of them can scarcely be re-

REGULATORY PROCESSES IN ORGANISMS 173

garded as superfluous. The further my own investigations in
this field proceed, the more completely I am convinced that those
phenomena, which we are accustomed to call regulations are
among the most characteristic, perhaps it is not too much to sa}-,
the most characteristic phenomena of life.

The views of different authors concerning the relation between
regulatory and ' normal' or ' typical' phenomena are very different.
Roux ('95, II, pp. S43-4), for example, makes a sharp distinction
between typical and regulatory development, though he admits
that the distinction is analytical rather than practical. For
Driesch regulation is areturn approach to the normal condition, after
this condition has been disturbed by some external factor. Many
physiologists, on the other hand, have used the term 'regulation '
in a much broader sense, as applying not only to the extra-normal
or extra-typical but at least to many of the most typical phe-
nomena of life.

Because there is no general agreement concerning the real basis
and nature of these processes, and because the regulations are of
great importance for any interpretation of life, it seems worth
while to undertake a brief analysis of them and particularly of the
regulations which involve form and structure to a large extent.
The present paper is concerned with such an analysis.

THE ORGANISM AS A PHYSICO-CHEMICAL SYSTEM

1. The relation between jnetdbolism and structure

The structural basis of living organisms consists primarily of
colloids. These colloids, together with water, make up the greater
portion of what we are accustomed to call protoplasm and in this
protoplasm the various reactions and processes which character-
ize life occur. The universal association of colloids with life
suggests that these substances play an important part in some
manner in determining some of the characteristic features of
life.

Metabolism has been commonly conceived in the past as con-
sisting, on the one hand, of the synthesis of an exceedingly complex

174 C. M. CHILD

and highly labile molecule, and on the other, of the breaking down
of this molecule in functional activity. According to this view the
colloid structure built up is used in function and must be continu-
ally replaced.

During recent years, however, many facts have been discovered
which seem to make necessary some modification of this view of
the relation between metabolism and colloid structure. In the
first place, most proteids, and indeed most organic colloids, are
relatively inactive chemically. The common interpretation of this
fact has been that death involved, or perhaps consisted in a change
from lability to relative stability of the proteid molecule. But the
recent work of Fischer and others upon proteids makes it highly
probable that the proteid molecule, although very large, is not so
complex as has been supposed, but may be polymeric in high
degree. Thus the assumption of extreme lability of this molecule
in the living organism becomes even more difficult than before.

Work along other lines has demonstrated that the nitrogen me-
tabolism is only a fraction of the total metabolism of the organism
and that it does not necessarily increase in proportion to functional
activity. Moreover, the nitrogen requirement for maintenance in
animals is apparently much smaller than had been supposed. All
of these facts seem to indicate that in the living organism, as in
vitro, the proteids, or many of them, are relatively inactive chemi-
cally; that after they are formed, they are excluded to a large
extent from metabolism simply because of their relative inactivity.
If these conclusions be correct, the accumulation of proteids
in the organism is a process not very different from the deposition
of other forms of inactive substance in or about the cell, e. g.,
chitin, insoluble salts, etc. Indeed it is highly probable that the
accumulation of most or all structural substances in the organism
is due to the fact that they are relatively inactive under the exist-
ing conditions. After they have arisen in metabolism they per-
sist or disappear much more slowly than other substances, simply
because under the exi.sting conditions they do not enter chemical
reactions as readily as other substances.

But the proteids and other substances deposited in the cell are
not absolutely inactive and undoubtedh' do enter metabolism

REGULATORY PROCESSES IN ORGANISMS 175

to some extent at all times. Under certain conditions, however,
i. e., in the absence of certain other substances, they, or some of
them, may reenter metabolism to a much larger extent and fur-
nish energy. It is a familiar fact that in the absence of nutritive
material from without the organism uses up its own substance,
^. e., the relatively inactive substances which under other conditions
had accumulated in it. In certain of the lower organisms this
process may continue until the organism is reduced to a minute
fraction of its original size. These facts do not, however, conflict
with the suggestions made above as to the relative inactivity
of structural substance, but serve rather to confirm the idea that
the accumulation of these substances, when other nutritive mate-
rial is present, is due to their relative inactivity.

Various authors have attempted to distinguish between a mor-
phological and a functional metabolism, but it is doubtful whether
such a distinction is valid, except as an expression of the fact that
substances of different degrees of chemical activity arise in the
course of metabolism and that certain of the less active substances
constitute the structural basis of the organism, while others
undergo chemical transformation and elimination.

It is of course not merely the nature of the substances them-
selves, but the existing conditions as well, which determine the
degree of activity or inactivity. Under certain conditions a cell
or an organ may accumulate certain substances and so acquire
a certain characteristic structure, while under altered conditions
these substances may rapidly disappear and others be accumu-
lated. Thus, for example, the oocyte, during its growth period,
accumulates yolk, which under the existing conditions is almost
wholly inactive chemically and so appears as a structure-building
substance. But when fertilization occurs the conditions within
the cell are so altered that the accumulated yolk rapidly reenters
metabolism and serves as nutritive material. In fact we may say
that the egg does not produce yolk because it is to develop into
a new organism, but that it develops as it does because it has
accumulated yolk. In the periodic changes in the cell connected
with growth and division there is also abundant evidence for the
occurrence of changes of this character.

176 C. M. CHILD

If we may accept this view of structure in the organism, and all
the facts are in its favor, then it is actually very similar in its rela-
tion to the energy current to the morphological characteristics
of a river system except of course that the latter are mechanically
produced. The constructed islands and bars, the depositions of
the river, represent those particles or masses which have, under
the conditions existing at a given time and place, been left behind
by the current. Under certain conditions the river may produce
structure of a certain kind at a certain point in its course, while
under different conditions this structure may disappear and give
place to structure of a different kind.

But the most important fact for present purposes is that in the
organisms, as in the river, structure, as soon as it appears, begins
to influence the metabolism, the energy current. From this time
on the metabolic processes, like the flow of the river, occur in a
certain structure and here the mutual interactions begin. Of the
character of these interactions in organisms we are only beginning
to obtain some vague conceptions, but that they occur, it is
impossible to doubt.

Perhaps a few words will not be out of place concerning the
bearing of these facts and suggestions upon the theory of * forma-
tive substances ' which has played a considerable role in embryo-
logical investigation during the last few years. Most of the sup-
porters of this theory have attempted to identify the so-called
formative substances with visible granules or other accumula-
tions in the cytoplasm, without considering the fact that the
appearance of these substances in visible structural form indicates
that they are, at least for the time being, relatively inactive, and
that they are first of all products or incidents of metabolism
(Child, '06b). Of course some or all of these substances might
reenter metabolism under altered conditions and so play a part
in determining its character, but the important point is that they
are indications of a difference in metabolism already existing in
the different regions where they are formed. We might expect
that the differences in metabolism, which are certainly more im-
portant as formative factors than these accumulations of granules,
would persist in the different regions, even if the granules could

REGULATORY PROCESSES IN ORGANISMS 177

be removed. Fortunately the embryologists themselves have now
a method of removing these granules from their usual position,
i. e., the method of centrifuging the egg, and the results of recent
experiments along this line indicate, as was to be expected, that
the granules are quite unessential to regional localization and dif-
ferentiation of the embryonic structures.

It is evident from the above suggestions that our fundamental
conceptions of the relation between structure and function in
organisms must be intimately connected with our ideas concerning
the nature of colloid substances and their significance as a sub-
stratum or medium for chemical reactions. Within recent years
it has been pointed out repeatedly that these substances afford
various means for the partial or total isolation, of different chemi-
cal reactions in organisms and that their mere presence may bring
about such isolation, e.g., by the formation of semipermeable mem-
branes. Here then we have a physico-chemical basis for localiza-
tion and differentiation. Moreover, the changes in the physical
aggregate condition of colloids, together with the possibility of
the simultaneous existence of different phases of high and low
water-content, must play a part in determining the degree and
place of dissociation of various substances, and therefore in de-
termining the speed of reactions in different regions, as well as
the occurrence or non-occurence of certain reactions at partic-
ular points. It is then, to say the least, highly probable that
the possibilities of localization, physiological specification and
the accompanying possibilities of physiological correlation of
parts and of regulation are very closely connected with the fact
that the formation of colloids is a component of the reaction
complex known as metabolism. Moreover, as I have attempted
to show in another paper (Child, 'lib), the accumulation of rela-
tively inactive substances, particularly colloids, in the cell is
undoubtedly a factor in senescence, in that it constitutes an
obstacle to the metabolic interchange and so brings about a de-
crease in the rate of metabolism.

If the conclusion be correct, that the visible structural elements
of the cell are, at least for the time being, relatively inactive chem-
ically, then it follows that these elements do not represent the

178 C. M. CHILD

'living substance' in the stricter sense, but are really the least
'alive' of any part. The reactions which furnish the energy of
life undoubtedly occur, at least in large measure, in the more fluid
parts of the cell, the parts which present the least characteristic
structure. The so-called living substance is actually then, so far
as it presents a visible structure, chiefly a substratum or m€dium
in which the reactions occur, and is itself the product of past reac-
tions. That these structural elements, as they accumulate, must
modify the rate and character of the reactions to an increasing
extent, cannot be doubted. The advancing specialization of
metabolism in different organs and cells is probably closely con-
nected with the fact that these parts produce different structural
elements, either in consequence of an original specification or
in consequence of different correlative or external conditions which
induce specification.

If we accept this view of the relation between function and
structure in the organism, we must give up the idea of a definite
'living' substance in the chemical sense, and the basis of life
becomes, not a specific substance, but a series of reactions in a
field or medium of a certain complex constitution, which is itself
the product of past reactions. We can agree with Driesch ('01,
p. 140), as regards the absence of a specific living substance, though
we cannot follow him in his further conclusions along this line.
The life process has become individualized, not because of entele-
chy, but because it forms its own field or medium of action, as the
river forms its channel, particularly in the later stages of its course,
where deposition exceeds erosion. To put the matter briefly,
life as we know it consists not in metabolism alone nor in a speci-
fic substance or structure alone, but in the physiological correla-
tion of processes in a structural medium or substratum of a cer-
tain constitution, which makes possible localization and correla-
tion of processes.

It is then the existing relation between the processes and the
structural substratum, the mutual interaction and dependence of
both, that forms the basis for the phenomena of regulation or
equilibration which occur in the organism.

KEGULATORY PROCESSES IN ORGANISMS 179

2. Physiological correlation and the physiological system or
individual

Organisms in general appear in the form of more or less sharply
defined physiological systems or individuals and in the more com-
plex organisms we can distinguish systems or individuals of vari-
ous kind and degree. What is the basis of this unity?

The experimental investigation of orgaaisms has led those
who are not yet ready to accept vitalistic hypotheses to the con-
clusion that two factors are chiefly involved in the formation of a
living system or individual, viz. constitution and physiological
correlation. In its grosser aspects the first of these is the morpho-
logical, the second the physiological factor. Most of us believe,
however, that the morphological features of organisms are
essentially visible expressions of dynamic processes past or pres-
ent and that sooner or later we must interpret constitution in
dynamic terms. The factor of physiological correlation in the
organism is essentially the problem of physiology, for in the final
analysis function is impossible without such correlation.

Wherever in the universe unity can be recognized, there some
sort and some degree of correlation must exist, either conceptually
or as a datum of nature, between the elements which compose the
unit, and vice versa, wherever correlation between conceptual or
phenomenal elements is recognized or established, there a unity
of some sort and some degree exists. On the other hand, the char-
acter of the unity is determined by the nature or constitution,
however this may have arisen, of its elements.

There is at present no adequate ground for believing that organ-
isms differ from other phenomena in these respects. We cannot
conceive an organic individual without correlation of some sort
between the parts which compose it, nor can we conceive it with-
out elements or parts of a certain more or less characteristic con-
stitution.

The development of morphology and its separation from other
fields of biology during the latter half of the nineteetith century has
led, particularly in the field of zoology, to the consideration of the
problem of constitution apart from that of correlation. But the

180 C. M. CHILD

introduction of the experimental method into zoology has already-
demonstrated the limited scope and value of pure morphology
for the interpretation of life. In the organism as we find it, the
two factors, constitution and correlation are mutually determining.
We cannot alter the constitution without altering the correlation
of parts, neither can we alter correlation without changing con-
stitution to a greater or less extent. The cases of so-called self-
differentiation constitute no real exception to this statement,^
which has the value and significance of a law of nature. Morphol-
ogy and physiology are inseparable except analytically and their
artificial separation can lead only to the formulation of many
pseudo -problems and to uncertain or false conclusions and hypo-
theses.

In so far as the organism is a physiological, i. e., a physico-
chemical individual or unity, in so far must physiological correla-
tion exist between its parts in the form of actual physical and chem-
ical processes, conditions and substances. Until it is proven by the
profoundest investigation and the strictest analysis that physiolog-
ical correlation does not suffice to account for the organic indi-
vidual, there is no need of turning to the vitalistic hypotheses for
an interpretation.

Indeed our knowledge of physiological correlation is in its
earliest stages. One need only refer to the work on the conduction
of stimuli in plants and through protoplasm in general and to the
investigations of recent years on the thyreoid, the adrenals, the
reproductive organs, the pancreas, etc., as organs of chemical
correlation and to the work on hormones, to become aware of the
advances in knowledge along this line within the last few years.
At present we are willing to believe, in fact we find it difficult
not to believe, that every metabolically active organ in the body
is an organ of chemical correlation. And we also know that many

1 As Roux ('95, p. 822 etc.) has pointed out development depends primarily
upon correlation and absolute self-differentiation cannot occur. In cases where
parts differentiate in a relatively high degree of independence from each other, we
must believe, and in some cases, e. g., the nemertean egg, we know that this condi-
tion is preceded by, and is the result of an earlier condition in which the parts are
in much closer correlation.

REGULATORY PROCESSES IN ORGANISMS 181

of these correlative factors show a high degree of specificity.
Moreover, the chemical factors are by no means the only factors
in correlation: mechanical and other physical factors also play a
part. And finally, the experimental investigations themselves
have demonstrated the importance of physiological correlation
in morphogenesis.

In short, there is at present every reason to believe that the
existence and continuity in time and space of organic individuality
are essentially dependent upon physiological correlation, i.e., upon
processes and conditions which are accessible to scientific investi-
gation and analysis.

3. The basis and nature of physiological correlation

That Dhysiological correlation is in general dependent upon the
physical and chemical processes and conditions in the various
parts which make up the individual cannot be doubted. These in
turn are deDendent upon the constitution of the parts, which itself
depends in part upon preexisting correlation and to a greater or
less extent upon conditions and processes in the extra-individual
environment. At every step in our consideration we recognize
the mutual interdependence of constitution and correlation.

But if we consider the organic individual only as it exists at
the present time, then we may say that the existing physiological
correlation between parts is dependent upon the conditions and
processes in the parts, however these may have been brought
about.

In general we can recognize at present three main groups of
correlative factors : first, mechanical or mass correlation (Roux,
'95, II, p. 240), which results merely from the existence of mass
without respect to constitution; second, substantial or material
correlation, which consists in the actual transference or transpor-
tation of substance possessing a certain physical or chemical con-
stitution, e. g., chemical correlation; and third, dynamic correla-
tion, of which the essential feature is the transmission of energy
rather than the actual transportation of material over any appre-
ciable distance. None of these forms of correlation can be sharply

182 C. M. CHILD

separated from the others in the final analysis, but in its extreme
forms each type is readily distinguishable.

The physiological correlative effect of a part upon others is
then the result of all that that part is and has been in the past, of
its physical and chemical constitution, its position, its relation to
external factors and of the changes which are occurring in it. It
is apparent that there exists in physiological correlation the possi-
bility of an almost infinite variety and specificity. Driesch ('09)
has recently maintained that the specificity of the 'Restitutions-
reiz' together with the specificity of the reaction to it constitute'
an 'Individualitat der Zuordnung' which is inexplicable on a
physico-chemical basis and which therefore constitutes a new and
independent 'proof of the 'Autonomic der Lebensvorgange.'
Comment seems scarcely necessary. One sees here merely an
assertion, a jump at conclusions, but no proof, where proof of
the most convincing character is absolutely essential. If vitalism
can present no more convincing arguments than this its future
prospects in science are not bright.

THE NATURE OF REGULATION
1. Organic or physiological equilibrium and equilibration

One of the most characteristic features of organisms is, as Roux
('95, I, pp. 145, 154, 392, etc.) has said, their continued existence
as individuals, their 'Dauerfahigkeit' amid changing internal and
external conditions. On the other hand this 'Dauerfahigkeit'
is only relative, not absolute, i. e., it is limited. The organism is
constantly changing, and- so far as our knowledge goes, never twice
the same, yet the continuity of individuality is obvious.

Nevertheless the continuity of the existence of individuality
must not be emphasized to the exclusion of the fact that under
certain conditions this individuality may disappear, at least in the
simpler organisms, and be replaced by other individualities in
larger or smaller number. Certain factors concerned in this physi-
ological disintegration will be discussed below, but for the pres-
ent we are concerned with the individual, the system as we see it

REGULATORY PROCESSES IN ORGANISMS 183

in the organism or part which constitutes a unity cHstinct to a
certain extent from others.

When we investigate the processes in the organism, we find that
they are very intimately connected with one another: a change in
one conditions changes in others. Moreover, and this is an impor-
tant point, the physiological specification of different parts is not
in most cases absolute. In the highest, most complex forms abso-
lute specification is doubtless approximated more or less closely
in certain organs, but in general we find that the processes in differ-
ent parts of the organism are not fixed in character. The character-
istic series of reactions in a part does not represent the only pos-
sible series, but rather the particular series determined by a par-
ticular complex of conditions. A certain process may occur at one
time in a certain part, at another time in others. In short there
is more or less possibility of substitution among the different
parts.

Let us suppose, for example, that a certain correlative factor x
originates in a certain part. Under certain conditions this factor
may influence various other parts, a b c — n,of which one, a, let us
say, reacts with greater speed or intensit}^ than others. The
reaction of this part may itself produce new correlative factors and
so alter conditions in the others that their reactions are changed.
But if we suppose that the receptivity of the part a to the cor-
relative factor X is decreased, or that the part a is itself removed or
rendered incapable of reaction, then the reactions of a b — n or
of some of them are not altered or prevented by the effect of the
reaction of a, and these parts may take the place of a in the sys-
tem, though perhaps at first reacting more slowly or less intensely
than a, until their constitution has become altered by repeated
or continued reactions.

Through such a series of reactions the individuality of the organ-
ism is maintained, or restored, even though it may have lost a part.
We see exactly such reactions in various organisms, and we can
devise physico-chemical systems which show similar correlation
between parts and a similar method of maintenance or restora-
tion of something approaching the preexisting condition. In the
system which we devise such processes are processes of equili-

184 C. M. CHILD

bration. We find that the system is capable within certain limits
of attaining or approaching a condition of equilibrium after a
disturbance of a previously existing equilibrium.

And again in the law of mass action and the general principles of
chemical equilibrium, together with what we know of katalysis, we
have the possibility of accounting for a great variety of processes
of equilibration in the organism. Until we have exhausted these
and other physico-chemical possibilities and found them inade-
quate, we have no adequate reason for oelieving that organic indi-
viduality and its maintenance are anything unique.

The fact that physiological correlation exists between different
parts of an organism must necessarily determine a certain relation,
a certain proportionality in the activities of the different parts.
It is this relation, this proportion in activity determined by corre-
lation which constitutes what we call organic or physiological
equilibrium in the organism. This equilibrium is dynamic, not
static, it is an equilibrium of processes, not of masses and it must
be dependent either upon physiological correlation, or upon some-
thing else which controls the supply of energy to this or that part
in very much the way in which the man in charge controls the
workings of a complex machine, e.g., a steam-shovel, turning the
steam into this or that cylinder as required for the harmonious
working of the whole. Driesch's entelechy is comparable to the
man in charge of the engine.

But it is not the mere existence of an organic equilibrium which
constitutes the real problem; it is the apparent power of adjust-
ment, of equilibration, the harmony of action of the parts, as in the
engine, which has been regarded as the strongest argument for
vitalism. How, the vitalist asks, is it conceivable that a machine
with such capacity of adjustment, of equilibration as the organism,
which can even repair itself, can be constructed and continue to
exist and work unless there is something comparable to the man in
charge concerned in these processes.

As a matter of fact this question is based on a wrong conception
of the organism. The organism as we see it, i. e., morphologically,
is not the machine whose action constitutes life, but rather simply
a part of the products of that machine, which accumulate during

REGULATORY PROCESSES IN ORGANISMS 185

its action and as they accumulate, alter and determine the character
of its activity. In other words, as the products are formed they
become a part of the machine. Starting with the egg, the organism
is not, as Driesch asserts that it must be according to the 'machine
theory,' a machine developed /or function (Driesch, '05, p. 790),
but rather a machine developed by function. The result at any
stage represents morphologically the products of a preexisting
machine and physiologically the action of the machine as altered
from the preceding stage by the products of its own activity.
Each stage of development is the result of the machine plus the
product of the preceding stage. Our experiments have shown
that physiological correlation, not predetermined harmony is the
basis of development, and that where a predetermined harmony
appears to exist it is certainly in some cases, probably in others,
the result of an earlier condition of correlation.

On the basis of this conception of the organism it is inconceiva-
ble that processes of adjustment of the parts to each other, i. e.,
processes of equilibration, should not occur, both in development
in nature and under experimental conditions. The parts are what
they are, not simply because of their original constitution, but
because they have been acting in correlation with each other.
From the moment the organic machine began to work in the first
organism 'adjustment' of the parts to each other began and it has
continued ever since. Could we but read it completely, every
part is a record, an epitome, more or less complete according to
circumstances, of what has been going on, not merely in itself,
but in the whole organism. Moreover, in different parts this
record is written in different characters, in different languages,
according to the constitution of the part.

The distinction which Roux makes between the formative and
the functional periods of development (Roux, '95, II, p. 281), is,
according to this view, not fundamental in character. The for-
mative period is functional and the functional period is formative.
But this distinction is based upon the fact that at a certain more or
less sharply defined stage of development the accumulated products
of the activity of the machine begin to play a more or less definite
role in its further physical and chemical activity. The adult

186 C. M. CHILD

organism is not then to be compared with a machine constructed
of certain definite Darts, which have been put together in some
way, and which, after completed construction, begin to function.
It is much more nearly comparable to a river, which molds its
banks and bottom, forming here a bar, there an islaad, here a
bay, there a point of land, but still flowing on, though its course,
its speed, its depth, the character of the substances which it
carries in suspension and in solution all are altered by the structural
conditions which it has built up by its own past activity. In
such a system a wide range of equilibration exists and we see both
the adjustment of function to form and of form to function. The
relation between structure and function in the organism is similar
in character to the relation between the river as an energetic proc-
ess and its banks and channel. From the moment that the river
began to flow it began to produce structural configurations in its
environment, the products of its activit}^ accumulated in certain
places and modified its flow% but just so long as the flow continues
the process of equilibration goes on.-

If we consider merely a certain region of the river with the water
containing certain substances in suspension and in solution enter-
ing at one end, depositing some of these substances and taking up
others as conditions determine in the course of its passage, and
finally passing out at the other end bearing certain substances
more or less different from those which it brought in, the analogy
becomes even more complete. In fact this region of the river,
together with its bed, shows a real, though chiefly a mechanical
rather than a chemical metabolism.

I believe that this comparison between a river with its channel
and the organism is far more than a fanciful analogy. Theindivid-
dual organism is merely a section from that current of energy which
constitutes the essence of life, and in the individual we see the
mutual correlation and interaction between the current and the
conditions under which it finds itself, between the energetic process

- Rignano ('07) has referred briefly to this analogy between the river and the
organism, using the case of a river equilibrating itself in connection with the piers
of a bridge to illustrate the process of equilibration in organisms. See also
Delage, L'Heredite, etc., 1903.

REGULATORY PROCESSES IN ORGANISMS 187

and the structural features which its activity has produced. As the
banks and the channel are 'adjusted' to the activity of the current,
and the current to the morphological characteristics of the banks
and bed, so, and in no otherwise are structure and function in the
organism, correlated with each other. It is absolutely inconceiva-
ble that 'adjustment,' equilibration should not occur. So long
as the current flows, equilibration must take place in one way or
another.

The organism has often been compared to a flame. Roux par-
ticularly has carried out this comparison in detail (Roux, '05, p.
109, et seq.). Although this analogy contains much that is valua-
ble and on the chemical side is much closer than that of the river,
yet on the other hand the morphological features of the river are
more nearly comparable to those of the organism, In their locali-
zation, their often complex structure and their modifying effect
upon the activity of the current. For these reasons I have chosen
the river rather than the flame as a physico-chemical system with
which the organism may be compared.

When we take the view of the organism suggested above, I
believe that Driesch's first two proofs of the autonomy of vital
processes (Driesch, '03, p. 74, etc. Cf . also p. 197 below) appear in
their proper light. They apply only to the morphological con-
ception of the organism as a machine constructed for function,
i. e., to the banks and channel without the river. In the organism
the current is working from the beginning, the organism is func-
tioning in one way or another, and the real machine is the process,
the function, plus the existing structure which past processes have
produced, just as in the case of the river the real machine is the
current plus the banks and channel. The process of development
in the organism is comparable, not to the digging of a channel into
which, after its completion, the water is turned, but to the forma-
tion of a channel with certain characteristics determined by a
variety of conditions, by the activity of the current itself. From
the moment the current begins to flow, structure and function
become mutually interdependent and mutually determining,
but there can be no river -structure without the current. Machines
like the steam engine, constructed by man and considered without

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

188 . C. M. CHILD

their motive power, are comparable rather to the dead than to the
living organism. They are merely the conditions under which the
energy acts, but the living organism consists from the beginning
of these conditions plus the energy. Development is not com-
parable to the construction of such a machine by man, but rather
to its action after the steam is turned on. Every steam engine
possesses a certain power of equilibration dependent upon its
constitution, and the only reason its powers in this direction are so
narrowly limited is because the energy current and the structure
have not been working together from the beginning.

The only possible basis for a scientific, as opposed to a philo-
sophical vitalistic hypothesis is the proof that the energy of or-
ganic life is something essentially different from the energj^ of the
physico-chemical world. When the vitalists shall succeed in prov-
ing this or even in making it probable, then their views will be given
more general consideration. But even the most extreme among
this school at the present day do not attempt such proof. If
we admit that the energy of the organism is not different from that
in the physico-chemical world, then I believe we are forced to
regard the organism as a physico-chemical system, for as I have
shown above, physico-chemical systems exist in which the relation
between structure and function, between the conditions of action
and the energy itself, are of the same character as in the organism
itself and give rise to a power of equilibration of the same character.

2. Regulation as equilibration

From what has been said it will be at once apparent that the proc-
esses which we commonly call regulatory are processes of equili-
bration in the organism (Holmes, '04, '07, Child, '06, '08a). They
enable the organism to persist and to maintain its individuality
under changing conditions, although it cannot be supposed that
the condition of dynamic equilibrium is the same for different con-
ditions, and indeed we have evidence that it is not. But within
certain Umits, and for certain factors, the organism is capable of
a greater or less degree of equilibration, when a change in external
conditions occurs.

REGULATORY PROCESSES IN ORGANISMS 189

The question at once arises as to whether all processes of equili-
bration are to be regarded as regulations, or only certain of them.
By zoologists the term 'regulation' has been applied mostly to
processes occurring under experimental conditions outside the
usual range of conditions in nature and the regulations of form and
structure have been the chief, though not the only objects of inves-
tigation. Jennings ('06) has used the term with reference to
phenomena of behavior which are characteristic features of life and
not of abnormal or pathological conditions. Among the physiolo-
gists also we find the term often used as referring to various
changes in metabohsm and reactions of different kinds in response
to conditions to which every individual is subjected repeatedly.

If we define regulation as a return or approach to a condition of
dynamic equilil^rium in a living organism after a previously exist-
ing condition has been disturbed by some external factor (Child , '06) ,
we shall include all the above phenomena as well as many others.
According to this definition, the simplest reflex as well as the
restoration of a missing part is a regulation, the simplest correla-
tive compensation in metabolism, as well as the development of
a whole from an isolated blastomere of an egg.

Moreover, when a complex part of an organism undergoes an
equilibrating change in reaction in response to altered correlation
with another part or other parts, a regulation occurs as truly as
when the whole organism responds to some change in conditions
outside of it. In short, regulations are equilibrating reactions to
changes external to the reacting system, whether this system be a
part or a whole of an organism.

And finally, regulation is not limited to the return or approach
to the preexisting condition, but may be an approach to a condi-
tion very different from that, i.e., the organism or the part may
become something more or less widely different from what it
was originally. In every case of regeneration of lost parts some of
the cells become something different from what they were before
the part was removed, and their change is a reaction to altered
conditions and specifically to altered correlation.

But that the regulatory process is always and necessarily of
advantage to the organism does not follow from the definition.

190 C. M. CHILD

So far as it enables the organism to persist, it may be of advantage,
and I see no escape from Roux's argmnent (Roux, '95, 1, p. 145,
154, etc.), that systems possessing such reactions will persist longer
than others. But not all regulatory processes are of advantage
to the organism and many of them, e. g., the so-called axial heter-
morphoses, lead to its destruction or its disruption, but they are
no less regulations because of this result.

According to Driesch ('01, p. 92), ''Regulation ist ein am leben-
den Organismus geschender Vorgang oder die Anderung eines sol-
chen Vorgangs, durch welchen oder durch welche eine irgendwie
gesetzte Storung seines vorher bestandenen 'normalen' Zustands
ganz oder teilweise, direkt oder indirekt, kompensirt und so der
'normale' Zustand oder wenigstens eine Anniiherung an ihn wie-
der herbeigefiihrt wird."

If we accept this definition, then the processes which do not
constitute a return or approach to the previously existing 'normal'
condition are not regulations. This normal condition is nothing
but the condition which corresponds to a certain complex of exter-
nal factors or to changes within certain limits. Under changed
conditions a new equilibrium, not the old, is established. In
short, if we accept such a definition, we not only exclude many
processes which are as truly regulatory as any, but we are forced
to assume the existence of an entelechy or other similar principle
to account for the 'normal' condition and its maintenance.

Regulatory processes are determined in character and direction
by the nature of the organism, on the one hand, and the nature and
amount of the external change, on the other. Under the given
conditions, the organism or part is capable of doing only the one
thing; under other conditions, or with a different constitution, the
regulation may occur in a different manner and may often lead
to a different result. In Planaria, for example, the course and result
of regulation differ according to the size of the piece, the region
of the body from which it is taken, the temper iture, the nutritive
conditions and other factors. To say that the pieces always
produce a whole under all these conditions means but little, for
the wholes which they produce are not alike. In plants the charac-
ter of the external change often plays a very large part in deter-

REGULATORY PROCESSES IN ORGANISMS 191

mining the character of the regulatory processes. In many
cases, however, in both plants and animals, the action of the exter-
nal factor is so indeterminate, or the external conditions are so
complex that equilibration may occur in various ways under what
seem to be, but are not actually similar conditions. Thus, as
Jennings has pointed out, in the regulation of behavior the dis-
turbance, the stimulus, may merely bring about reactions of an
indeterminate character, which sooner or later, in one way or
another lead to equilibration. Evidently then the relation between
the character of the external change and the character of the regu-
latory process differs very widely in different cases.

The initiating factor in regulation is the external change, the
disturbance of the preexisting condition. This change brings about
changes within the organism or the part and these in turn lead to
changes in the correlative factors, and so to equilibration or to
disruption and death, in case the external change is such that
equilibration of the system as a whole is impossible. But so long
as the energetic processes of life continue in the system, equilibra-
tion of some sort must occur. To return to the analogy of the
river, so long as the water flows, equilibration of some sort occurs,
whatever the changes and whatever the obstacles. The river may
alter its course, it may transform its banks and its channel so that
they bear little or no resemblance to those existing before the
change, it may divide into a number of streams, each of which
pursues its own course, according to the conditions under which it
finds itself, and builds up its own structural characteristics.
In all cases, however, unless the conditions are such as to stop the
flow of the water, equilibration takes place in some manner.

The range of regulatory capacity in the organism represents then
merely the range of possibilities within which the flow of the cur-
rent of energy which constitutes metabolism and which is the
essential feature of life, is possible. Within these limits it is abso-
lutely inconceivable that regulation or equilibration should not
occur. The nature of the process depends upon the nature of the
orgaaism and the conditions which it meets.

Equally inconceivable is the occurrence of regulation as a
process of life under conditions which stop the metabolic current.

192 C. M. CHILD

Organisms which meet such conditions are simply eliminated.
The survival and elimination of organisms is determined primarily,
not by their morphological characteristics, but by their capacity
for regulation of one kind or another under the conditions in
which they find themselves. To attempt to understand the course
of evolution from morphological characteristics alone can only
lead to confusion and failure. Only a knowledge of the nature of
the metabolic current in organisms and the possibilities of its
equilibration under different conditions can lead to a theory of
evolution and heredity which will stand.

For example, the evolution of animals and plants, like fevery
other evolution, is based primarily upon differences in the meta-
bolic processes. These undoubtedly originated as regulations and
as soon as they had arisen, gave different possibilities ot further
regulation : in the course of the realization of these different possi-
bilities in accordance with the conditions of existence, animals and
plants with their different morphological characteristics have
arisen. In each case the visible structure represents merely a
partial record of the realized possibilities. All the structural
'adaptations' in both animals and plants are based upon the proc-
esses of equilibration of the energy current and must sooner or
later be expressed in terms of this current and its environment.
They are not the primary and essential features of the organisms,
they give us merely an outline, a diagram of the most character-
istic activities of the energy current. As the banks and channel
of the river, even after the water has ceased to flow, enable us to
gain some conception, though a very incomplete one, of what the
river has done in the past, so the structure of the organism is
merely a rough sketch of what the current of life has done in the
"way of deposition, arrangement and removal of materials along
its course. Many of the past activities of the current are not dis-
tinguishable in the structure because their effects were slight or
transitory, or because they have been masked or altered by later
activity of a different character. As the river in some process of
equilibration, e. g., in a flood, a period of increased energy, may
sweep away many of the records of its previous activity, so the

REGULATORY PROCESSES IN ORGANISMS 193

organism, in a period of increased metabolism may remove the
structural evidences of past metabolism.

In the present state of our knowledge we should think it absurd
to attempt to account for the configuration of the banks and bed
of the river without taking into account the action of the current.
It would remain a miracle, which we could ascribe to the caprice
or other quality of a personal creator, or to some other mysterious
natural force. In the same way, when we attempt to interpret the
structure of organisms without direct reference at every step to the
current of energy of which the structure is evidence, we must neces-'
sarily go astray or end in confusion or in the most bizarre hypothe-
ses. We can do as Driesch has done and shift the burden to the
shoulders of entelechy, to which we can ascribe such qualities as
may please us. Or we can speak of biophores and determinants,
pangens, or whatever we please to call them, or we may pin our
faith to the visible chromosomes, but these are nothing but
creators of a type which appeals to certain minds.

On the other hand, when we take as our starting point the process
of metaboHsm, we are proceeding as the physiographer has learned
to proceed in his study of rivers. As we learn how metabolism
produces structure we shall be able more and more completely to
interpret the nature and the past history of the organism from its
structure, but at every step we must return to the process, the
current, in order to understand, and we can never hope to under-
stand all through structure, simply because structure is an incom-
plete record. Life is first of all an energy process, a flowing current.
All that is relatively stable, all that persists as visible form and
structure, represents merely some past action of the current
occurring under certain conditions. Almost sixty years ago Huxley
said concerning the cells: ''They are no more the producers of the
vital phenomena than the shells scattered along the sea-beach are
the instruments by which the gravitative force of the moon acts
upon the ocean. Like these, the cells mark only where the vital'
tides have been, and how they have acted. "-^ And even yet the
truth of these words is not recognized as it should be by biologists.

* British and Foreign Medico-chirurgical Review, vol. 12, p. 314, Oct. 1853.
Cited from Whitman, the inadequacy of the cell-theory, Jour. Morph., vol. 8, 1893.

194 C. M. CHILD

Only when we take into consideration the motive power and
the method of its action under the given conditions, can we hope
really to advance in our knowledge of how things come to be as
they are in the organism, or to determine and predict what they
shall be. Man has attained his present position by acquiring
knowledge and control of energy in nature. Can he hope to ad-
vance in his insight into the problems and his control of the proc-
esses of life in any other way?

According to this point of view, life, like every other continu-
ous energetic process, is essentially a series of equilibrations, of
regulations. When regulation shall cease, evolution and li'e will
also cease. The power of regulation in organisms is nothing uni-
que, but is something which they possess in common with all
energetic processes in nature,which continue for any appreciable
time. In fact, strictly speaking, all energetic processes in nature
are equilibrations.

As was suggested above, the range of regulatory capacity in
organisms is undoubtedly due in large measure to the fact that
the process of metabolism produces certain colloid substances,
among which the proteids and lipoids are the most characteristic.
With the first proteid synthesis under certain conditions in nature
the processes of regulation of the type which we find in organisms
began. Perhaps we may say that life began here also. The reaction
which was concerned in the first synthesis must of course have
preceded the completed synthesis, but as water apart from the
channel which it forms for itself in its environment is not a river, so
a given chemical reaction, or a series of reactions, apart from the
conditions which it produces where it takes place, is not life. We
may say if we please that life began as a chemical reaction, but we
must recognize the fact that the occurrence of that reaction pro-
duced certain characteristic conditions, which played a part in
determining the course and character of further reactions : in short,
the reaction determined the existence of structure and the mu-
tual interrelations between structure and function: and finally,
with the existence of structure of colloid nature, the possibility
of regulation of the organic type also appeared, and regulation
began.

REGULATORY PROCESSES IN ORGANISMS 195

My purpose in laying special emphasis upon the point that regu-
lation is an essential characteristic of life and that life must cease
when regulation ceases, is merely to show that the extreme forms
of regulation, which occur under experimental or accidental con-
ditions, are in no way different from the processes of life apart
from experimental or accidental interference. The capacity for
regulation is not something secondary or something acquired in
the course of evolution, but it is as inseparable from life itself as
the power of equilibration from the flow of the river. Not only life
but the universe is an unceasing series of regulations. Every
experimental investigation performed with living organisms is,
so far as it does not lead to the death of the organism, an investiga-
tion of organic regulation, and death itself is an equilibration,
though of another type.

To set the regulations ofT as a special category of phenomena,
occurring only in organisms and of secondary or incidental signi-
ficance in these, must of necessity lead to conclusions of the same
character and value as those which would be reached by one who
should attempt to investigate the phenomena of equilibration in
the river, without considering the flow or the resistance of its
banks and bed. Such a one would doubtless marvel at the won-
derful harmony of action displayed by the simultaneous disap-
pearance of a part of the bank and the encroachment of the water
upon it, or by the appearance of an island and the division of the
river into two channels. He would doubtless call attention to the
remarkable fact that both the channel and the river were narrow
and deep at some points and broad and shallow at others. He
might wonder why stones moved along where the bed was steep
and only fine particles where it was nearly horizontal. If he were
of an investigating turn of mind, he might throw stones into the
river and observe the consequences, or he might dig a ditch and
turn part of the water into it. Thus he would observe further
remarkable harmonies of action. If he were inclined to look for
causes, he would probably conclude that the complex of phenom-
ena was determined and controlled by some mysterious being
or principle, which, judging from his own ability to bring about
harmony of action between different things in his world, he would

196 C. M. CHILD

conceive to be more or less like himself, though greater, more per-
fect and more powerful. Doubtless also he would give it a name.

But when once the idea of the flow of the river as a motive power
has entered his mind, his whole attitude toward what he has seen
is altered. He sees that it is the current which carries partic.es
away from the bank or stones and mud along the channel. On
the other hand he sees that the banks confine the river, that the
island, which it has formed divides it, that it accommodates its
form to the ditch which he has dug and at the same time begins
to change it. He begins to realize that the remarkable harmonies
which he has observed are the result, on the one hand, of the flow
of the river, i. e., in i further analysis, of the characteristics of
water, and on the other, of the nature of its banks and bed. He
will also realize in time, that just as long as the flow continues
these harmonies of action will continue to occur. Then he may
begin to investigate the characteristics of currents and of water in
general, and later we find him devising water-wheels, dams, pumps
etc., i. e., bringing about the most various harmonies of action
between the flow of water and other phenomena.

His conception of what he saw was at first more or less similar
to that of the vitalist concerning organisms and all his investiga-
tion could only end in speculation, which did not advance his real
knowledge. But when he once began to reahze the action of the
current as an energetic and a constructive process, then he saw
that the harmonies of action were only apparent, not real, be-
cause he was dealing with mutually dependent phenomena
rather than with those which were independent and predeter-
mined.

Driesch, for example has maintained in criticism of some of my
own earlier statements, that development is for function (Driesch,
'95, p. 790) and the same view is apparent in his repeated compari-
son of the organism to a machine constructed by man. This is as
if our hypothetical man should maintain that because he could
dig a ditch and turn water into it, therefore the channel of the
river must have been constructed by some 'entelechy,' or other
principle for the water, and then the water turned in. And more
specifically, Driesch's 'proofs' of the autonomy of vital proc-

REGULATORY PROCESSES IN ORGANISMS 197

esses, ^ which are based on the phenomena of regulation are not
proofs at all, because the 'machine' which he has in mind is com-
parable to the dead, rather than to the living organism, to the
river frozen solid, rather than to the river flowing. Tf we could
separate a portion of this frozen river with its channel from the
rest it would of course remain what it was, i. e., a part, so long as it
remained frozen. But if we divert any sufficient quantity of water
from the flowing river it is capable of forming a whole which shows
all the essential characteristics of the original river, though not
identical with it. In short each flowing river, with its banks and
bed is a 'machine' according to Driesch's definition, ''eine typische
chemisch-physikalische Spezifitatskombination" (Driesch, '01, p.
187), and it may become whole when parts are taken from it or
when their relative position is changed; moreover, when it is
divided, each part may form a whole essentially similar in its
processes and structure to the original whole. The existence of
such a 'machine' is therefore a sufficient refutation of these 'proofs'
of Driesch's. So long as the current flows such regulatory proc-
esses are not only possible but necessary, when the conditions
arise. Neither the organism nor the river 'remain w^iole' when
parts are taken from them, but they become new wholes, which
under similar conditions, may become more or less like the original
whole (Child, '08b), but which under other conditions, may be
different.

Driesch's error is two-fold : although his general definition of a
'machine' is sufficiently broad, his argument in the 'proofs' is
based only on a certain type of machine, viz., that constructed by
man for function, a type which is wholly passive during its con-

* A brief statement of the first two ' proofs' is as follows :

"Erstens: Eine Maschine bleibt nicht dieselbe, wenn man ihr beliebige Telle
nimmt oder ihre Telle beliebig verlagert; deshalb kann das sich auf Basis harmon-
isch-aquipotentieller Systeme abspielende Formbildungsgeschehen kein maschi-
nelles chemisch-physikalisches Geschehen sein.

"Zweitens: Eine nach den drei Dimensionen typisch spezifisch verschiedene
Maschine bleibt nicht ganz, wenn sie geteilt wird, deshalb liegt der Geneseaquipo-
tentieller Systeme mit komplexen Potenzen im Bereiche des Formbildungsges-
chehen kein maschinelles chemisch-physikalisches Geschehen zu Grunde."
Driesch, '03, p. 74.)

198 C. M. CHILD

struction, rather than on a type in which, as in the organism, the
structure at each stage is determined by the function and the
structure in the preceding stage. The organism is comparable not
to the constructed 'machine' alone but to the machine plus the
constructing activity, and since Driesch has confined his argument
to the type of machine constructed by man for a definite purpose,
he is very naturally and logically led to the assumption of a
constructor. His 'proofs' are equivalent to the argument that
because a ditch built by man for a particular purpose and possess-
ing a specific structure but containing no water does not remain
whole or the same when we take away parts of its banks or bot-
tom, therefore the river, as we see it in nature cannot be a physico-
chemical system.

Similarly in his consideration of the organism he has failed to
take account of the constructive activity of the continuous flow
of energy in a given environment. The organism is, he says, con-
structed for function. His position is identical with that of our
hypothetical man who concluded that the channel of the river
must have been constructed for the water, and like him, Driesch
has given his imagined constructor a name, or rather has adopted
an old one for it, viz., entelechy. Most of us have concluded from
our observations and experiments that the channel of the river is
constructed by the activity of the current and we have some rather
conclusive evidence upon that point. Before he can hope to see
his views accepted, our man must actually prove or make it at
least probable that this is not so. The burden of proof lies wholly
upon him. And similarly, until Driesch can make it at least
probable that the organism is constructed for and not by function,
instead of merely assuming this to be the fact, he cannot expect
to find wide acceptance for his views. Nowhere in Driesch's work
do we find any convincing evidence upon this point: Driesch has
simply chosen to assume that it is so. I am of course aware that
Driesch regards entelechy as in constant connection with physico-
chemical factors and as working with these as means. But I
see no reason why, if we postulate an entelechy for the organism,
we should not at least be consistent and postulate another for the

REGULATORY PROCESSES IN ORGANISMS 199

THE Rl<](iULAT()RV PROC^KSSES

1. The relation between form regulation and functional
regulation

A distinction between regulations of form and regulations of
function has \'ery commonly been made. As might be expected
from his conception of the organism, Driesch ('01) has attempted
to draw the line very sharply. But if we adopt the point of view
suggested above the distinction becomes apparent rather than
real. First of all every regulation in organisms is primarily an
energetic process and secondly, it occurs in a certain structure
and must affect that structure to a greater or less extent. On the
other hand, every change in structure must lead to a regulation
of function. Structural and functional regulation are in fact insep-
arable in organisms. If we go further and interpret structure in
terms of the constructive energy, we may say that all regulations
are essentially functional, i. e., energetic.

It is sufhciently evident from what has been said, that the flow
of energy in the organism is essential for regulation, and that the
structure must play a part in determining its character. It should
be possible, therefore, to interpret the regulatory processes in
terms of the energy current, i. e., metabolism, and the preexisting
structure. To refer again to the analogy of the river, both the
channel and the current are involved to a greater or less extent in
each equilibration in the system. The distinctioa between form
regulation and functional regulation is then in part conventional
and connected with the separation of morphobgy and physiology
from each other, and in part a matter of convenience, since some
regulatory processes involve the visible structure to a much greater
extent than others. As in the classification of other natural phe-
nomena, we separate for convenience of thought or reference a
graded series into a number of (in this case two) different classes.

2. The inducing conditions and the results

As already noted, the first factor in regulation is a change of
some sort in the external conditions affecting the system. In the

200 C. M. CHILD

case of a part of an organism this change may be a change in
physiological correlation resulting from changes in other parts,
however produced. This change, external to the system concerned,
produces an internal change of some sort in some part or parts, and
this in turn alters the physiological correlation between the com-
ponents of the system affected. So long as life continues, these
correlative changes must result in equilibration in one way or
another. The processes of equilibration may be very different in
different cases : they may bring about sooner or later a return or
approximation to the preexisting condition — according to Driesch,
this alone constitutes regulation and only when the preexisting
condition was the 'normal' condition. On the other hand, the
correlative changes may result in the establishment of, or ap-
proach to a condition of equilibrium more or less widely different
from the preexisting, and I believe that most, if not all regulations
which we usually regard as an approach or return to the preexist-
ing condition actually represent an approach to a new equilibrium,
often only slightlj^ different from the old ; it seems at least doubt-
ful whether the organism ever really returns to a preexisting
condition in the strict sense.

The new ecjuilibrium may diffei- (quantitatively or qualitatively
from the old, or it may even result in the separation of the system
into a larger or smaller number of systems, more or less completely
isolated from each other. In all these cases, as the rate or character
of the metabolic processes are changed, changes in structure as well
as in function occur to a greater or less degree. The following
suggestions foi- a classification of the regulatory processes are
based primarily upon the metabolic processes concerned.

3. A provisional classification of the regulatory processes

a. The two methods of regulation. It is evident that any really
analytical classification which is based upon the conception of
regulation suggested above must take account, not merely of
the visible features, but of the character of the different energetic
processes, since regulatioji is, according to this view, essentially
a complex of energetic processes in a substratum of a certain

KK(aJLAT<)RY I'UOCESSES IN OlUiANISMK 201

constitution. Such a classification must also be available for both
the so-called functional and form regulations, since every regula-
tion probably involves both to some extent.

Driesch's classification of the regulations (Driesch, '01, p. 95
et seq.) is based upon a conception of regulation so widely differ-
ent from the one developed in this paper that it does not assist us
in distinguishing the processes involved. From Driesch's point
of view, the physico-chemical processes in regulation are to a
large extent of secondary importance and therefore cannot serve
as a basis for classification.

At present, however, we are practically unable to attain the
proper basis for classification, since our knowledge of the processes
involved is incomplete. Nevertheless we can distinguish with
more or less certainty the resemblances and differences between
different equilibria, and the following suggestions are based upon
the character of the equilibria.

We may distinguish two chief type of regulatory processes,
first, quantitative equilibrations or compem^alions, and second,
qualitative changes in equilibrium or frans/orma/zons. In compen-
sation the rate or intensity of the processes, their continuation in
time or their extension in space are concerned: in the transforma-
tions their character as energetic processes, i. e., the nature of the
chemical reactions and the physical changes. In the compensation
the system remains much like that previously existhig, as regards
its character and the processes of equilibration are quantitative.
In the transformations a new system, qualitatively different
from that previously existing, arises as the result of equilibration.

Most regulations, as they occur in nature and experiment, involve
both compensations and transformation in various degrees. This
is especially true of the regulatory processes which follow the
removal of a part. Here some parts of the organism undergo trans-
formation in consequence of altered correlation, while compensa-
tory processes of various kinds are evident, both in the increase
in size of the new part and often also in a decrease in old parts.

Moreover our use of these terms will depend upon the particular
processes to which we have reference in a given case. The process
of compensatory growth, for example, is highly complex in charac-

202 C. M. CHILD

ter aad coasists in a variety of both compensatory and trans-
formatory processes, but when an increment in all of these proc-
esses occurs, the change is quantitative in character and when it
constitutes a process or part of a process of equihbration we are
justified in calling it a compensation.

It is of course evident that a classification of regulatory processes
must finally become identical with the classification of processes
occurring in the organism, for, as I have pointed out above, the
regulations are not a peculiar form of organic activity; they repre-
sent merely the equilibrations resulting from the existence of physi-
ological correlation between parts. But we shall probably always
have occasion to refer to the organism, the system, as a whole
undergoing equilibration or, relatively speaking, in equilibrium,
consequeatly some means of distinguishing between the different
methods of equilibration is useful. This is the chief significance
which any classification of the regulations can possess.

h. Regulatory compensation. Several different types of compen-
sation can be distinguished, though they do not of course in most
cases exist in nature or even in experiment apart from other
processes. The following divisions under this head are suggested :

Incremental compensation: The system shows an increment
as compared with that previously existing.

Decremental compensation: the system shows a decrement as
compared with that previously existing.

Reversional compensation: an increment or a decremeat in
some part of the system, indaced by some external factor is
correlatively more or less completely elimiaated and the system
approaches its previous condition.

Alterative compensation: an increment or decrement in one
part produces change in the opposite direction in another or in
others, so that the proportional relations in the system differ from
those previously existing.

The first step in all compensations is of course a change in some
part {a) induced by some factor external to the system. What
particular form of compensation shall occur depends upon the
degree of the change in the part and upon the character of the
correlation existing between it and other parts (6, c, d-7i). If for

REGULATORY PROCESSES IN ORGANISMS 203

example the part a dominates the other parts, or if the change
in a is so great that the correlative factors resulting from it
become dominant, then an increment in a may bring about an
incremental compensation, a decrement in a a decremental com-
pensation. On the other hand, if the parts b, c, d-n, or certain of
them, dominate a, then they may inhibit or reverse the incremen-
tal or decremental change in a and reversional compensation
results. A.nd finally, alterative compensations occur whenever
changes in one part induce correlatively changes in the opposite
direction in others.

An incremental compensation occurs when increased metabo-
hsm and growth follow the ingestion of food, a decremental com-
pensation, when decreased activity of a sense organ or a muscle
induces a correlative decrease in activity and perhaps atrophy in
parts with which it is connected, e. g., the center in the case of the
sense organ, the tendon, or even the bone in the case of a muscle.
In various temperature regulations in warm-blooded animals we
have reversional compensations, and finally, in many cases of
regeneration and probably also often in normal development, al-
terative compensations occur, e. g., when increased growth of one
part retards correlatively the growth of another or perhaps induces
reduction in it.

The chemical substances which arise in the course of metabo-
lism in certain parts very often produce compensations of various
kinds in other parts. A good example is the correlative effect of
increase in the carbon dioxide in the blood through the nervous
system upon the rate of respiration. The recent work of Bayliss
and Starling and others on 'hormones' gives us some insight into
various other cases of compensation and other regulatory reactions ;
the distribution of nutritive substances in the starving animal and
under various other conditions also constitutes compensations of
various kinds; the correlative changes in so-called functional
structure are in many cases very characteristic compensations.

Incidentally it may, be pointed out that the view of the relation
between metabolism and structure suggested above affords a
basis for interpretation to a certain extent of the processes of
functional hypertrophy and atrophy from disuse.

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

204 C. M. CHILD

We have seen that structure-formation of some sort, i. e., the
accumulation of relatively inactive substances in or about the cell
is a characteristic feature of metabolism. The close relation be-
tween the syntheses and the oxidation processes has been pointed
out repeatedly by Loeb as well as by others. The structural
substances, when once formed, play only a relatively small part in
further metabolism, provided other more active substances, i. e.,
nutritive materials, are at hand, and provided the general char-
acter of metabolism is not changed. Any condition, e. g., the
'functional stimulus' which leads to increased metabolic activity
of the particular kind which constitutes what we call the special
function of the cell or part leads, when nutritive material is pres-
ent, to increased accumulation of the inactive substances and
hypertrophy is the result. On the other hand, in the absence of the
functional stimulus, or when its frequency or intensity is decreased,
the use of nutritive material aad the accumulation of structural
substance do not occur or are less rapid, and the result is that
below a certain level of functional activity the gradual breaking
down of the accumulated substance, which is not immediately
connected with the special functional activity of the part, exceeds
the constructive processes and decrease in size and atrophy occur.
The constructive processes continue only, or very largely, in
connection with the functional stimulus and, for the addition of
new structure nutritive material must be taken in from without,
but this functional activity does not under these conditions,
increase proportionally the rate of reentrance of the structural
substances into metabolism; in fact, if other more active sub-
stances are present in sufficient quantity, the structural substances
may be spared to a large extent.

Hypertrophy and atrophy are then the result of two different
kinds of processes, the one connected with the specialized function
of the part in its relation to other parts, the other to a considerable
degree independent of this except in starving animals. In its 'func-
tional activity' the part builds structure, but does not destroy it to
so great an extent. The destructive process is largely independent
of function and goes on more or less continuously. Whether hy-
pertrophy or atrophy shall occur in a given case depends merely

REGULATORY PROCESSES IN ORGANISMS 205

on whether the one or the other of these processes is the more
rapid. Hypertrophy is then in no sense a 'regeneration in excess' ;
it is merely a direct result of increased metabolic reactions of the
kind which constitute or accompany the 'function' of the part
concerned. Nevertheless, it is without doubt a compensation.
The occurrence of one series of metabolic reactions determines the
occurrence of another series according to chemical laws; the one
series furnishes energy, the other forms relatively inactive sub-
stances, which persist as structure. Closely related to functional
hypertrophy is the growth in size of regenerating parts after their
formation: here the 'functional stimulus' is the quantitative
factor in the correlative influences from other parts. This factor
induces a certain rate or frequency of reaction in the small new
part, which leads to rapid accumulation of material, i. e., to hyper-
trophy (Child, '06a, p. 407). But as the structural substance
accumulates, the structure itself constitutes an obstacle to metab-
olism (Child, '116) the rate of hypertrophy decreases and finally
equilibrium is attained.

c. Regulatory transformation. The character of many metabolic
reactions is more or less definitely known, but the exact relation
of the reactions to the production of a particular kind of visible
structure is a much more difficult matter to determine. The
visible characteristics of organic structure are by no means ade-
quate criteria of the character of the processes involved in its
formation. We are not always justified in concluding from the
differences in the visible appearance of structures that the proc-
esses concerned in their formation are actually different in nature.
Great differences in appearance may arise in the same colloid
substance in consequence of differences in aggregate condition
or phase. But when we find substances of different constitution
in different cells or parts, it is evident that processes of different
character were concerned in their formation. Consequently we
can often determine that a transformation has occurred by the
change in the character of the structural substance. Many fea-
tures of correlative differentiation, whether in ontogeny in nature
or under experimental conditions are undoubtedly transforma-
tions, e. g., the formation of a bud from a differentiated cell in

206 C. M. CHILD

the plant, in consequence of the removal of other vegetative tips,
the formation of a hydranth from cells of the stem of Tubularia,
etc.

The difficulty lies in distinguishing qualitative from quantita-
tive regulations. In living organisms the two are evidently very
closely associated, and probably in every regulation which we
can observe directly both are concerned. And in the final analy-
sis the question of the relation between quality and quantity in
general is involved, though this is scarcely a biological problem.

As regards the further classification of the regulatory transfor-
mations, I think that at present the most satisfactory basis for
classification is the comparison of the new system with that exist-
ing before regulation. The following division of this group of
regulations is therefore suggested:

Progressive transformation: the regulatory formation of a sys-
tem possessing a greater degree of complexity, more varied locali-
zation of structure and function and consequently more varied
correlation than the system existing before regulation.

Regressive transformation: the regulatory formation of a sys-
tem of simpler character than the preexisting.

Transgressive transformation: the regulatory formation of a
system which cannot be distinguished as more or less complex,
but merely as different from the preexisting.

Suh a classification serves merely to suggest the various pos-
sibilities. As our knowledge of the processes concerned in the
changes of the organic system increases, the basis of classification
will change. Without doubt many progressive transformations
occur in normal development. The adult organism is certainly a
more complex and qualitatively different system from the blas-
tula, and we know that correlative factors have been concerned
in the changes in many parts. A regressive transformation occurs
when a part undergoes dedifferentiation in consequence of altered
correlation, as in various cases where cells which give rise to new
parts first lose their old differentiation.

The group of transgressive transformations possesses little
more than a conventional significance, since it is based upon
difference from the normal which is essentially merely the usual

REGULATORY PROCESSES IN ORGANISMS 207

Very probably various 'sports' and mutations can be placed under
this head, perhaps also certain of the neoplasms, more specific-
ally the malignant tumors, though this is by no means certain.
But whatever the categories which we may establish for the dif-
ferent regulatory processes, the important point is that we should
at least make the attempt to find a physico-chemical basis for our
analysis. If we do this we cannot separate structure and function
since both are merely different aspects of the same process-com-
plex and are dependent upon and determine each other. Doubt-
less we shall still find it convenient to speak of form regulation as
distinguished from functional regulation, but we must remember
that the distinction is not a real one and that every regulation in
the organism is undoubtedly a regulation of both form and func-
tion, of both structure and reactions. Furthermore, we must
regard our experiments on regulation as means of analyzing the
factors of the process. With the proper care in experiment we can
do much toward determining the nature and action of various
correlative factors in regulation, and every step in this direction is
a step in advance in our knowledge of the system which constitutes
the organism,

THE NATURE OF RECONSTITUTION '

1. Restitution or reconstitution f

When an organism 'restores' a missing part or in general when a
part of an organism forms a whole, the process seems at first
glance to be so obviously a restoration in which something re-
moved is replaced, that the term 'restitution' has found very
general favor. Although I have used this term to a large extent,
it has always seemed inadequate, for the reason that the proc-
ess is not simply one of restoration but something more. There
is no case of so-called restitution known in which the changes
following the removal of a part are limited to the formation of a
new similar part. In every case changes of one kind or another,
quantitative or qualitative, or both, occur in other parts, some-
times limited chiefly to parts adjoining the part removed, some-
times extending throughout the system. The removal of a part

208 C. M. CHILD

brings about not simply its restitution, but an equilibration
extending more or less widely, and strictly speaking, probably
throughout the system. The system reconstitutes itself and the
whole formed is different from the original quantitatively or
qualitatively^ The whole process is a complex of equilibrations,
of compensations and transformations, resulting in that which
we call a whole, but no two wholes are alike.

If for example, we consider this process in a piece of the Plan-
arian body, we find that it differs in rate and character according
to size of the piece, region of the body from which it is taken, tem-
perature and other factors which influence metabolism. The ani-
mal formed as the result resembles the original in its general shape
and activity, but it is far from being identical with it. It is usually
smaller than the original, the pharynx may be in quite different
position in the body, and the arrangement, number and form of the
intestinal branches differs more or less widely, according to con-
ditions. Moreover, under various conditions, various degrees
and kinds of incompleteness appear. Some pieces develop only a
single eye, or the eyes are partially fused or otherwise different
from those in the original animal, some pieces develop no pharynx
and no posterior end, others no head or an imperfect' one, some
develop the postpharyngeal intestinal branches more rapidly
and more completely, others the prepharyngeal branches, some
produce a larger, others a smaller head, some show more 'regen-
eration,' others more 'redifferentiation' and so on. If we place
the different sorts of wholes under closely similar conditions and
give them food they become more or less like each other because
these conditions bring about further regulations but these regu-
lations do not properly belong to the regulatory process which re-
sulted from the isolation of the part, but are independent of it and
are such as were occurring in the original animal during its life. It
is probably not too much to say that no two pieces of the
Planarian body attain the same condition in the process of regu-
lation. When we say that because some or all of them produce
wholes they are all potentially alike, we are simplj- assuming that
all wholes must be alike, which is obviously untrue. As each piece
is different at the start from the others, so it attains a di^erent

REGULATORY PROCESSES IN ORGANISMS 209

result. Driesch's assumption of 'equipotentiality' of different
parts is shown by the facts themselves to be incorrect, and I
believe that his 'harmonious-equipotential systems' do not exist
in nature, as systems capable of development but only as ab-
stractions of the human mind.

For these reasons the term 'restitution' seems to me to carry with
it implications which, when we analyze them, we cannot, in the
light of the facts, accept. The piece does most certainly not
restore what it lost : it reconstitutes itself into something more or
less widely different from that of which it formed a part, and this
something often possesses visible structural characteristics which
we have come to regard as characteristic of a whole. Seeing these
resemblances, we abstract from the differences and say that it is
the same as that of which it formed a part.

Closer examination shows us that even visibly it is not the same,
but different; moreover, visible characteristics are not the sole
criterion of resemblance and difference in organisms. The processes
occurring are just as characteristic as the visible structure. I
have shown elsewhere (Child, 'lib) that in Planaria the process
of form regulation results in a rejuvenation, the pieces after
undergoing regulation are physiologically younger than the ani-
mals from which they were taken, and the degree of rejuvenation
varies with the degree of reconstitutional change. Manifestly
then these pieces are not the same after regulation as the wholes
of which they formed parts. To say that they are is simply to
deny the facts as they stand before us.

In many cases, moreover, the pieces do not produce anything
that can be called a whole. Pieces of Planaria may produce
double heads or double tails, tailless heads or headless forms. For
those who with Driesch regard the formation of a 'whole' as the
uniform result of so-called restitution, these cases are difficult to
interpret. The process of regulation has apparently followed the
wrong track, it has gone astray and so failed of the correct result.
But when we take the position that the part, when isolated, under-
goes a reconstitution, which differs in its results according to the
existing conditions, internal and external, we see that these 'abnor-
malities' differ from 'normal' results simply because different

210 C. M. CHILD

conditions existed at the beginning or elsewhere in the course of
the regulation, and in many cases we can determine what those
conditions are. For example, pieces of Planaria which give rise to
'wholes' at a certain temperature and under certain other conditions,
can be made to produce headless forms or tailless heads, accord-
ing to the region of the body from which they are taken, by sub-
jecting them to lower temperatures, by starving the animals before
beginning the experiment, by placing the pieces in dilute alcohol
or ether, etc. Nothing has gone astray in t'lese cases, there is no
error, the same laws have been followed as when 'wholes' are pro-
duced, different conditions simply lead to different results. Else-
where (Child, '10c) I have attempted to analyze some of the con-
ditions which bring about so-called heteromorphosis in Tubu-
laria and other forms, and have shown that they are similar in
character to those which bring about asexual reproduction in
nature.

There are many cases in which the occurrence of reconstitution
as opposed to restitution is so obvious that there can be no ques-
tioning it. A piece from the body of Hydra, for example, does not
restore the missing parts, but reconstitutes itself into an organism,
smaller, simpler, possessing fewer tentacles and undoubtedly
physiologically younger than the original animal. In Clavellina
also, as Driesch himself has shown (Driesch, '02), the isolated
branchial region or a part of it does not replace the missing parts,
but undergoes a process of reconstitution. Tn these cases the physi-
ological effect upon the parts remaining of the removal of cer-
tain parts is so great that these parts do not retain their original
structure, and a dedifferentiation and redifferentiatioQ occurs.
But the effects so apparent in these cases are simply more extreme
than in cases where only a small part is removed. Przibram ('07)
has called attention to a number of cases which show verj" clearly
that the removal of a part results, not in the restoration of the
original, but in the establishment of a new equilibrium, differing
more or less widely from that.

It is obvious that the process of reconstitution is an equilibra-
tion and just as obvious that it leads to different results under dif-
ferent internal and external conditions. As different rivers differ

REGULATORY PROCESSES IN ORGANISMS 211

from each other, so may the results of reconstitution, even in differ-
ent pieces of the same individual, differ from each other. More-
over, as a certain amount of water does not, except under certain
conditions, form a river, so does a piece of an organism reconstitute
itself to a whole only under certain conditions.

2. The initiating factor in reconstitution

As Driesch has pointed out (Driesch, '01), it is evident that
reconstitution occurs as the result of the absence CNichtmehr-
vorhandensein') of something. What is this something? Is it
the structure, the form, or is it activity? In plants it is possible
to bring about reconstitution, i. e., the formation of new buds, roots,
etc., by inhibiting the metabohc activity of the existing growing
regions without their removal, e. g., by enclosing them in plaster or
in an atmosphere of hydrogen, or even by applying anesthetics
locally between them and the regions concerned. In another
paper I have considered numerous cases of this sort and have dis-
cussed their significance at length (Child, 'Ua). These facts show
very clearly that it is not the form or structure which is involved
but a process, an activity, whose effect is transmitted in some way
from one part to another. If we stop the metabolism of the one
part for a time the effect on the other is the same as if the first
part had been removed. These facts alone should be sufficient to
prevent us from regarding form regulation as a process distinct
from functional regulation. It is the absence of the effect of cer-
tain processes in a certain part or in certain parts, which initiates
reconstitution. In short, it is the absence or decrease below a cer-
tain point of certain physiological correlative factors which were
previously present, that initiates reconstitution. In the absence
or decreased effectiveness of these correlative factors, the remain-
ing parts, still being subjected to other correlative factors, which
may themselves gradually change in consequence of the removal
or decreased activity of the part, react in a manner different from
their previous reactions, simply because they are under different
physiological conditions. Reconstitution is then initiated by a
change in physiological correlation. Recently Driesch ('09) has

212 C. M. CHILD

expressed himself in somewhat similar terms, but he finds neverthe-
less, as already noted above, that the 'Individuahtat der Zuord-
nung' between agent and effect cannot be accounted for on a
physico-chemical basis and therefore regards it as a new 'proof
of the autonomy of vital processes.

3. The process of equilibration in reconstitution

The process of equilibration in reconstitution differs in many
details in different cases, but it possesses certain more or less char-
acteristic features, and it is desired to call attention briefly to
some of these. The change in ph37siological correlation is the inter-
nal factor which has disturbed the preexisting condition, whatever
that may have been. This change may or may not lead to equili-
bration of the living organism. If the change be great, if the other
parts possess but little capacity for altering their reactions, it
may lead to death. On the other hand, it may lead to reconstitu-
tion in various ways according to conditions.

Let us consider first the case where a part is removed and is
formed again without any great changes in other parts, e. g., the
'regeneration' of the posterior end of Planaria.

In the absence of the correlative factors which originated in the
part removed (a), certain regions (6) of the remaining parts {b,c

d n), which were before prevented by these correlative factors

from reacting as their own constitution and the correlative factors
from other parts would determine, now begin to react in this man-
ner. In the region adjoining the part removed, i. e., in the cells b,

the correlative factors originating in the parts cd n are more or

less siniilar in their effect to those which affected the part removed
(a). So far as they are and remain similar, and so far as the con-
stitution of b permits, this region will be forced by the correlative
factors to react more or less in the manner of a, which is no longer
present, and b will replace a more or less completely and more or
less rapidly, according to conditions in the particular case^ If

' The formation of a new head in Planaria or a new hydranth in Tubularia is a
somewhat different process from the formation of the proximal or posterior end.
In these forms the anterior or distal region is physiologically dominant over parts

REGULATORY PROCESSES IN ORGANISMS 213

the cells of the region b adjoining the part removed, be chiefly
affected correlatively by the removal of a or if they react much
more rapidly than others further away, the process of formation of
a will be a 'regeneration.' But if parts further away from a are
also affected to a considerable extent by the change and are cap-
able of reacting as rapidly or almost as rapidly as b, then they may
also take part in the process of replacing a, which then takes on
more or less completely the character of a 'redifferentiation.' In
some cases we can determine experimentally whether a part shall
be formed chiefly by regeneration of redifferentiation. In Planaria,
for example, the amount of regeneration, as opposed to rediffer-
entiation, in the formation of a new posterior end increases with
increasing distance of the cut surface from the old posterior end.
The farther the level of the cut from the old posterior end, the
more completely is the development of the new part confined to
cells near the cut, and vice versa. The cells near the cut are those
which are most affected by the removal of the part a. Even when
this part does not develop anew, they react by healing the wound.
That is to say, they change their reactions most rapidly of all cells,
they lose their old specification, they become capable of a more
rapid metabolism (Child, 'lib) and being subjected to the correla-
tive factors of the parts cd n, they begin to develop into some-
thing more or less like a in advance of other parts. The processes
in these cells establish certain correlative factors which determine
that the cells farther away from the cut shall remain as they are or
take other forms of reaction.

But if we decrease the rate of metabolism in Planaria by extreme
starvation or by the use of anesthetics, then parts which under the

posterior or proximal to it and controls their development directly or indirectly.
Briefly stated, the regulatory formation of a dominant part is a reconstitution re-
sulting primarily from isolation, while for the formation of a subordinate part
correlation with other parts of the original system is necessary. For example, a
piece of the tubularian stem may reconstitute itself into a hydranth without any
other parts (Child, '07a, b, c), or a piece of Planaria into a head without other parts
but a tubularian stem or stolon or a planarian tail is never formed except in con-
nection with a more distal or anterior region of the original organism. The ques-
tion of the dominance and subordination of parts and its significance will be dis-
cussed more fully elsewhere.

214 C. M. CHILD

usual conditions are formed chiefly by regeneration, e. g., the head,
may be formed largely by redifferentiation (Child, '10a). T.iis
means simply that the cells near the cut do not react so rapidly
s under the usual conditions, so that other cells further away have
ime to change their reactions and take part ia the process, while
ordinarily they would be prevented from doing this by the cor-
relative factors arising from the activity' in the cells near the cut.
These instances are merely special cases under the general rule
that the less rapidly the missing part is replaced, the more ex-
tensive are the changes in the remaining parts, so far as their con-
stitution permits change.

Regeneration and redifferentiation in their extreme forms repre-
sent the extreme terms of a graded series, of which all terms are
essentially of the same physiological character, i. e., all consist
in a change in reaction in consequence of a change in physiological
correlation. The designations 'regeneration' and 'redifferentia-
tion' serve merely as convenient descriptions of the visible phe-
nomena.

Where all the cells of the remaining parts are so sensitive to the
absence of the correlative factors originating in the part removed
that they cannot maintain themselves after its removal, the old
structure of all the remaining parts may disappear to a greater or
less extent, i. e., a 'dedifferentiation' occurs, as, for example in the
isolated pieces of the branchial region of Clavellina. In the differ-
ent cells of the mass the metabolic processes become less specified
and in this respect it approaches the 'embryonic' condition, and
the correlative factors in the mass approach those existing in the
embryo. But during this process some of the cells have been sub-
jected to correlative factors more or less similar to those to which
the part removed was subjected at some stage of development,
consequently these become in some degree the physiological repre-
sentatives of that part. In short the system becomes physiologi-
cally a whole, but in consequence of the rapid dediffereatiation
of the old parts it corresponds to a whole in a relatively early
stage of development. From this condition renewed differentia-
tion as a whole results necessarily from continued metaboUsm,

REGULATORY PROCESSES IN ORGANISMS 215

i. e., continued life. The new whole is, however, different from the
old in size, physiological conditions, number and proportion of
various parts, etc.

In the process of regeneration in the stricter sense the new part
is usually at first smaU and increases rapidly in size. I believe
that this growth in size is essentially similar to the functional hy-
pertrophy of organs. The part which was removed possesses a
certain size in relation to other parts, because its size was deter-
mined chiefly by correlative factors. Just so far as the new develop-
ing part is subjected to similar correlative factors, it will
tend to attain the same size as the part removed. Consequently
it does not always attain the same size with respect to other parts.
In Planaria the relative size of the new head differs according to the
region of the body from which the piece is taken, to the nutritive
condition and various other factors. The process of reconstitu-
tion ceases when a certain stage, differing under different condi-
tions, is attained. This stage represents an equilibrium of physi-
ological correlation, i. e., of interaction between the parts; it is
primarily a dynamic equilibrium, a proportionalit}' of processes,
not of form. We can alter this condition of equilibrium experi-
mentally by food, by starvation, by temperature, and in short by
all factors which affect the processes.

In various papers Driesch has distinguished a number of differ-
ent forms of reconstitution (restitution). His distmctions are
based primarily upon differences in the visible phenomena of
development or dedifferentiation and for him the chief interest
lies in the recognition of the different forms, rather than in the
attempt to determine how they differ from each other physiologi-
cally, since from his point of view the physiological factors are in
many cases only 'means' which the enfelechy employs. It is
impossible to consider here these various forms of reconstitution,
and since my point of view is so widely different from that of
Driesch such a consideration would show merely that his basis
of distinction could not be accepted for purposes of a physiological
analysis. While it is convenient to distinguish different forms or
methods of reconstitution, I believe that it is much more impor-
tant to resolve the phenomena into processes.

216 C. M. CHILD

Ij.. The complexity of reconstitution

The process of reconstitution is not a simple process which can-
not be analyzed, but rather an exceedingly complex one; in fact
its complexity is of the same order and character as that of devel-
opment. It consists of a series of compensations and transforma-
tions in different parts of the system. It is only when we take into
account the complexity of physiological correlation between parts
and the almost infinite possibility of change and variety in this
correlation that we have any hope of gaining an insight into the
complex series of events. The specificity of correlation and reac-
tion does not, as Driesch apparently beheves' (Driesch, '09) con-
stitute a physico-chemically insoluble problem except when we
follow Driesch in ignoring the energy current as an equilibrating
factor in the organism and as the efficient factor in construction of
the visible and tangible characteristics. The energy current
performs its work under specific conditions in each case and leads
to a specific result. As soon as a specific condition arises in any
part of the system, from whatever cause, it determines other speci-
fic conditions in at least certain other parts. From the experi-
ments on Planaria it is perfectly apparent that the cells at every
level of the body posterior to the ganglia, are capable under cer-
tain conditions of developing into a head, but under the usual
conditions they are prevented from doing this because the cor-
relative factors arising from the presence, ^. e., the activities, of a
head determine their activities in another direction. As soon as
the old head is eliminated from the system, those cells, which in
consequence of their past correlation are most similar to it, be-
gin at once to form a new head, provided the piece is not too small
and as soon as this occurs it determines correlatively a variety
of reactions in other cells. The same may be said of the reconsti-
stitution of any part. The place where a particular part shall
arise is determined by constitutional and correlative factors in
the existing system — so far of course as external factors are not
concerned — and as soon as one such place is determined, it deter-
mines others and so on. 'Ihus any case of reconstitution consists
of a series of regulations, each of which determines others. This

REGULATORY PROCESSES IN ORGANISMS 217

is shown very clearly by the fact that isolation of a part from
certain others during the course of reconstitution may alter the
course of the process in the part, according to the degree and
character of the isolation, i. e., according to the correlative
factors which are eliminated. By means of experiments of this
kind it is possible even now to determine the action of various
correlative factors in different stages of the process of reconsti-
tution.

5. The limits of reconstitution

In every case the reconstitutional processes are limited and
determined by existing conditions as the river is limited and
defined by its banks and channel which its own activity has
constructed in the environment through which it flows. It is not
true without qualification that any part of certain organisms is
capable of giving rise to any part. Driesch's often repeated state-
ment to this effect requires modification and limitation. The part
is at most only provisionally capable of giving rise to any part;
in other words, only when it constitutes a component of a system
possessing certain characteristics, i. e., only when it is subjected
to or isolated from certain correlative factors. This is apparent
from every recorded series of observations on reconstitution, except
perhaps the most superficial. In some cases the system may con-
sist of a single cell, in others of a large number of pells, but the
fact remains the same. The power of reconstitution is limited,
not unlimited. As I hope to show elsewhere for Planaria, and
as I have shown for Tubularia (Child, '07a, '07b, '07c), the
investigation of these limitations is of the greatest importance
in throwing light upon the nature of the reconstitutional
processes. When we find that the removal of a certain part,
or even a certain amount of material, determines a different
result from the removal of another part or a larger or smaller
amount of material, ■ we are forced to the conclusion that the
part or the material removed has some connection with the
character, place or other factors in the result, and furthermore,
when w^e find that inhibition of the metabolic processes or
certain of them in the part or the material is as effective in certain

218 C. M. CHILD

respects as the removal of the part or the material, we have at-
tained a basis for investigation and analysis which is proof against
such assumptions as those which Driesch has made, e. g., concern-
ing the nature of the 'harmonious-equipotential system.' For
Driesch the limitations of the reconstitutional processes appear
to be of secondary importance, but I believe that any one who
will investigate and analyze these limitations at all thoroughly
will find that they are not only essential features of the regulatory
processes, but that they afford us one of the best means of gaining
some insight into their nature. As water does Qot constitute a
river, except under certain limiting conditions, so certain sub-
stances or processes do not constitute an organism or even life
except under certain limitations. The water contains the poten-
tialities for giving rise to any kiad of a river, as well as other
specific 'machines,' but none of these exist until the specific hmita-
tions are present. The case is essentially similar as regards the
organism. The specific 'machine' exists only so far as the limita-
tions exist. And the investigation of the limitations of reconsti-
tution affords at present one of the best methods, if not the best,
for demonstrating this to be a fact.

REPRODUCTION IN GENERAL A8 A FORM OF RECONSTITUTION

In another paper (Child, '11a) I have discussed at length the
significance of physiological isolation of parts as a factor in repro-
duction. I have shown that certain degrees and kinds of physiolog-
ical isolation of parts may arise as the result, first, of an increase in
size; second, of decrease in correlative control or physiological
dominance of a part in consequence of decreased activity in it;
third, of decreased conductivity^ or transmissibility of correlative
processes, agents or conditions; fourth, of decreased receptivity,
sensitiveness or irritability of certain parts to the correlative factors
originating in other parts. Furthermore, we know from experi-
ment, as I have shown, that in a considerable number of cases
physiological isolation of parts serves as well as physical isolation
by section to bring about reconstitution; and if it were possible
to perform the experiment, it is practically certain that we should
find the same to be true for many other cases.

REGULATORY PROCESSES IN ORGANISMS 219

In the paper just referred to I have also attempted to show that
at least a great variety of natural and experimental forms of
reproduction, reduplication of parts, etc. are essentially processes
of reconstitution following physiological isolation of parts. The
chief difference between them and the cases of reconstitution
following experimental section is, first, that the isolation of the
part or parts is brought about within the organism physiologic ally
and not by the crude method of cutting the organism into pieces;
and second, that this isolation is usually partial at first and differs
in degree and kind in different cases.

And finally, I have called attention to certain evidence in sup-
port of the view that the formation of sex cells and the develop-
ment of organisms from them are processes not fundamentally
different from other forms of reproduction, i. e., that the sex cells
are first physiologically parts of the organism like other organs,
and that the development of a new organism from them is initiated
by changes similar in character to those which occur in other parts
capable of reconstitution, when they are physiologically or physi-
cally isolated (Child, '10b, '10c).

The evidence bearing upon the first point is briefly as follows;
first, the sex cells always arise in, or attain by migration particular
regions of the body in a particular organism, therefore, their physio-
logical correlation with other parts cannot be purely nutritive
in character, for if it were, there is no reason why they should
not take the most various positions in the same species. Second,
they undergo characteristic differentiations during the life of the
individual, as do other organs and these differentiations begin
at a certain stage of development of the organism, i. e., at or near
the end of the period of vegetative growth. This cannot be ac-
counted for by quantitative differences in nutrition at different
stages, because the growth of the primitive germ cells in earlier
stages often requires very large amounts of nutritive material. If
this development is predetermined, then physiological correla-
tion between the germ cells and other parts must have existed
at some earlier stage, or else we are forced to a hypothesis of pre-
established harmony, which amounts to some form of vitalism.

Moreover, in organisms, which show both asexual and sexual

220 C. M. CHILD

reproduction in the same individual, the asexual reproduction
occurs earlier in the life cycle than the sexual, and in organisms
which produce naturally both parthenogenetic and non-partheno-
genetic eggs, the parthenogenetic eggs appear earlier than the
non-parthenogenetic. In both these cases the earlier product is
usually capable of a greater degree of regulation when it is iso-
lated from the parent body, than the later; in other words, both
the non-parthenogenetic egg and the sperm appear from their
behavior to be more Highly specified or differentiated than the
asexual or parthenogenetic reproductive elements. There is then
considerable evidence in support of the view that the history of
the germ cells, like that of other organs, is in part the result of
physiological correlation.

As regards the 'stimulus to development,' I have shown by
experiment (Child, 'lib) that in Planaria the process of reconstitu-
tution after physical isolation as well as extreme starvation fol-
lowed by feeding, accomplish rejuvenation and that it is highly
probable that various other factors bring about similar changes.
And finally, I have considered the facts which indicate that the
process of fertilization and the conditions inducing artificial par-
thenogenesis produce changes in the egg similar in character to the
rejuvenation occurring in the other cases.

From this point of view, the stimulus to development of the
egg is essentially a process or the beginning of a process of recon-
stitution and so is similar in its physiological effect to the factors
initiating the various processes of asexual reproduction. Experi-
mental reconstitution following section is then merely a special
case of reproduction occurring under certain conditions, or we
may say just as correctly that each form of reproduction in nature
or experiment is a special case of reconstitution occurring under
certain special conditions.

If this view be correct, then the fundamental problems of devel-
opment and heredity are before us in every case where a physically
or physiologically isolated part of an organism produces a new
organism, just as truly as they are in sexual reproduction. In fact
sexual reproduction constitutes the most complex case of all, but
I am convinced that a recognition of its essential similarity to the

REGULATORY PROCESSES IN ORGANISMS 221

processes following experimental section and the physiological
solation of parts is of the greatest significance for our conception
and solution of the problems of inheritance and development.

CONCLUSION

It is sufficiently evident from what has been said that I consider
the phenomena of regulation in organisms as congtitating the es-
sential chacteristic of life as a coatinuing process. I agree with
Jennings that the problem of regulation is the fundamental prob-
lem of life. All of our experimental investigations on living organ-
isms are directly concerned with the problem of regulation in
one way or another. Tn fact there are only two possible methods
of investigating and analyzing the phenomena of life : one is con-
cerned with regulation in the living organism, the other with the
observation and analysis of the results of stopping the life-proc-
esses at this or that particular point, under these or those parti-
cular conditions. In the one case we observe and control the proc-
ess in its action, in the other we seek to determine the effects of
its past action. As we can watch the river at work and investi-
gate the processes of equilibration resulting from alteration of its
flow in one way or another, so we can investigate the living organ-
ism. And as we can stop the flow of the stream or divert it into
other channels and determine something of what it has done along
its course up to a certain time by examination of its channel, so
from the dead organism, we can determine something of its past
activity.

But the conclusions drawn from the examination of the channel
of the 'dead' river are only fragmentary at best. Only by observ-
ing and controlling the river in action is it possible to acquire any
adequate conception of what it really is. And so, I believe, with
regard to the organism : the living organism will teach us more than
the dead one though we must work with both. And when
we work with the living organism we come at once face to face
with the problem of equilibration, of regulation. And finally, I
believe that the further our knowledge of the processes of equili-
bration, in the organism advances, the greater will be the difficulty
of finding an adequate foundation in biolog}'^ for vitalistic or
dualistic hypotheses.

222

C. M. CHILD

BIBLIOGRAPHY

Child, C. M. 1906a Contributions toward a theory of regulation. I. The
significance of the different methods of regulation in Turbellaria. Arch,
f. Entwickelungsmech. Bd. 20, H. 3,

19066 Some considerations regarding so-called formative substances.
Biol. Bull. vol. 11, no. 4 .

1907a An analysis of form regulation in Tubularia. I. Stolon formation
and polarity. Arch, f . Entwickelungsmech. Bd. 23, H. 3.
1907b An analysis etc. IV. Regional and polar differences in the time
of hydranth-formation as a special case of regulation in a complex
system. Arch. f. Entwickelungsmech. Bd. 24, H. 1.
1907c An analysis etc. V. Regulation in short pieces. Arch f. En-
twickelungsmech. Bd. 24. H. 2.

1908a The physiological basis of restitution of lost parts. Jour. Exp.
Zool., vol. 5, no. 4.

1908b Driesch's harmonic-equipotential systems in form regulation.
Biol. Centralbl. Bd. 28, nos. 18 und 19.

1910a Analysis of form regulation with the aid of anesthetics. Biol.
Bull., vol. 18, no. 4.

1911b A study of senescence and rejuvenescence, based on experiments
with Planaria. Arch. f. Entwickelungsmech. Bd. 31, H. 4.
1911a Die physiologische Isolation von Teilen des Organismus,
Vortr. u. Aufs. ti. Entwickelungsmech. H. 11.

Driesch, H. 1901 Die organischen Regulationen. Leipzig.

1902 tJber ein neues harmonisch-aquipotentielles System und iiber
solche Systeme iiberhaupt. Studien uber das Regulationsvermogen der
Organismen. 6. Die Restitutionen der Clavellina lepadiformis. Arch,
f. Entwickelungsmech. Bd. 14, H. 1 u. 2.

1903 Die 'Seele' als elementarer Naturfaktor. Leipzig.

1905 Die Entwickelungsphysiologie von 1902-1905. Ergebnisse der
Anat u. Entwickelungsgesch. Bd. 14 (1904).

1908 The science and philosophy of the organism. Vol.1. London.

1909 Der Restitutionsreiz. Vortr. u. Aufs. ii. Entwickelungsmech.
H vii.

Holmes, S. J. 1904 The problem of form regulation. Arch. f. Entwickelungs-
mech. Bd. 17, H. 2 u. 3.

1907 Regeneration as functional adjustment. Jour. Exp. Zool. vol.
4, no. 3.

Jennings, H. S. 1906 Behavior of the lower organisms. New York.

KoRSCHELT, E. 1907 Regeneration und Transplantation. Jena.

Morgan, T. H. 1907 Regeneration. Ubersetzt von M. Moszkowski. Leipzig.

Przibram, H. 07 Equilibrium of animal form. Jour. Exp. Zool., vol. 5, no. 2.
1909. Experimental-Zoologie. 2. Regeneration. Leipzig u. Wien.

RiGNANO, E. 1907. Die funktionelle Anpassung und Paulys psychophysische
Teleologie. Riv. di Scienza, vol. 2.

Roux, W. 1895. Gesammelte Abhandlungen fiber Entwickelungsmechanik der
Organismen. Bd. 1, u. 11. Leipzig.

PARAMAECIUM AURELIA AND PARAMAECIUM
CAUDATUM

LORANDE LOSS WOODRUFF

Fruyn the Sheffield Biological Laboratory, Yale University

ONE FIGURE

Leeuwenhoek in 1677 described^ some '' little animals longer
than an oval" which he had discovered two years previously,
and there is some reason to believe that this is the first published
record of an organism belonging to the genus Paramaecium.
The name Paramaecium, however, was first employed by HilP
to designate certain small organisms which were more or less
oblong, in contrast to others which were round or decidedly
vermiform, and either the present species aurelia or caudatum
is probably the animal which he designated as 'Paramaecium
species 3.'

Although Hill was the first to attempt to apply scientific names
to microscopic animals, it remained for O. F. Miiller to give
a general classification of these forms, and to apply the Linnean
nomenclature. He began this work on the infusoria as a sec-
tion of a treatise entitled, Vermium terrestrium et fluviatilium His-
toria, which appeared in two volumes in 1773. Unfortunately
he did not live to see the publication of his special work, Animal-
cula Infusoria fluviatilia et marina, 1786, which was edited by his
friend, O. Fabricius. Miiller described a Paramaecium and ap-
plied the specific name aurelia in the former of these works. In
the latter work he described and figured^ Paramaecium aurelia

^Philosophical transactions, London, 11, 133. 1677.
^History of animals, 3, 1751.
'Plate 12, figs. 1-14.

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

223

224 LORANDE LOSS WOODRUFF

together with several other forms, which at the present time are
assigned to other genera. Miiller's description is as follows*: —

Paramaecium. Vermis inconspicuus, simplex, pellucidus, mem-
branaceus, oblongus.

Paramaecium aurelia. Paramaecium compressum, versus ant-
ica plicatum, postice acutum.

Thus the organism described by Baker^ as ''Animalcules in
pepper water, first sort/' by Joblot^ as Chausson, by Ellis^ as
Volvox terebella, etc., received the name which, in spite of var-
ious vicissitudes, has come down to the present time.

The next great student of the lower organisms, C. G. Ehren-
berg, in the first two of his treatises,* described several species of
Paramaecium, and one of these is Paramaecium aurelia. In his
third treatise^ he described still another species which he named
Paramaecium caudatum.^" Five years later, in 1838, Ehrenberg
brought out his monumental monograph. Die Infusionsthierchen
als vollkommene Organismen, and in this work he described these
two species as follows :^^

Paramecium Aurelia, Pantoffelthierchen.

P. corpore cylindrico, subclavato, antica parte pauUo
tenuiore, plica longitudinali obliqua in os multum
recedens exeunte, utrinque obtuso.

Paramecium caudatum, geschwanztes Pantoffelthierchen.

P. corpore fusiformi, antica parte obtusiore, postica magis
attenuata.

Thus Ehrenberg described, on the basis of shape and size, the
two common forms of colorless paramaecia which appear in

*Page 86.

*The microscope made easy, London, 1742. 3rd ed., 1744, p. 72, PI. 7, fig. 1.

''Observat. fait, avec le microscope, Paris, 1754.

''Observations on a particular manner of increase in the Animalcula of vegetable
infusions, etc. Phil. Trans., London, 1769.

^Abhandl. der Akademie d. Wissensch. zu Berlin, 1830, 1831.

"Ibid, 1833.

lOEhrenberg notes that Herrmann (Naturforscher, 1784) applied the name cau-
datum to a form which was probably a species of Amphileptus; also Schrank
(Fauna boica, 1803) used the same name.

iiPp. 350-352. PI. 39, figs. 6, 7.

PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 225

modern systematic works as P. aurelia O. F. M. and P. cau-
datum Ehrbg.

Dujardin, in 1841, in his treatise on the Infusoria '^ recognized
but two species of Paramaecium as follows :

Paramecie Aurelie. — Paramecium aurelia.

Corps ovale oblong, arrondi ou obtus aux deux extremit^s,
plus large en arriere. — Long de 0, 18 a 0, 25.

Paramecie a queue. — Paramecium caudatum.

Corps fusiforme, obtus ou arrondi en avant, aminci en
arriere.— Long de 0, 22.

His figures of the two species show clearly the characteristic
form which he considered diagnostic.

Various investigators, including Stein, and Claparede and
Lachmann, questioned the justification of considering these two
forms as distinct species, basing their opinions, as had Ehrenberg
and Dujardin, solely on external characters, and they united
these two forms under one species, and applied Miiller's original
name, P. aurelia. This union of aurelia and caudatum into one
species was accepted by all the subsequent students of Paramae-
cium, e.g., Balbiani, Blitschli, Engelmann, Gruber and Kolliker
and consequently all the early literature on the conjugation of
this infusorian, refers to the organism as P. aurelia, although it
had but a single micronucleus.

Maupas, in 1883, in his studies on the ciliates,i^ wrote:- —

Tous les auteurs jusqu'ici ont decrit Paramecium aurelia comme ne
poss^dant jamais qu'un nucleole d'assez grande taille et mesurant de
0mm,005 h. 0mm,008. C'est en effet la forme que Von rencontre la
plus frequemment. Mais j'ai observe aussi de nombreux individus
pourvus de deux nucl^oles plus petit s et de structure diff^rente de la
piecedente. lis etaient de forme spherique et composes d'un corpus-
cule central opaque vivement color^ par les teintures et ne mesurant que
0mm,003; enveloppe d'une couche corticale mesurant en diametre
0mm,005, claire et ne se colorant pas.

i^Histoire naturelle des Zoophytes. Infusoires, etc. Paris, 1841. Pp. 481-483,
PI. 8, figs. 5, 6, 7.

"Contributions a I'etude morphologique et anatomique des Infusoires cilies,
Arch, de zool. exp. et gen., (2), I, 1883, p. 660.

226 LOKANDE LOSS WOODRUFF

Thus Maupas tacitly accepted the current view that there was
one large species of Paramaecium, but observed, for the first
time, that certain paramaecia have a different nuclear apparatus
from that previously described. This author, however, in 1888,
stated that in his earlier work he, as all his immediate predeces-
sors, had confused two species, and he wrote^^ as follows :

Ces deux formes de micronucleus constituent le caractere distinctif
le plus important entre les deux esp^ces de Parameeies. La premiere
forme appartient toujours et uniquement au P. caudatum, la seconde,
egalement toujours et uniquement, au P. aurelia.

Pour Ehrenberg et Dujardin, P. caudatum se distingue par un corps
allonge, fusiforme, obtus en avant, aminci en arriere: P. aurelia par un
corps plus large, presque ovale, obtus aux deux extremites. Ces differ-
ences de contour general, tout en ^tant reelles, ne sont pas absolument
rigoureuses; car, si on ne trouve jamais de Paramecie k un seul micro-
nucleus affectant la formed trapue obtuse, il n'est pas tr^s rare d'en
rencontrer a deux micronucleus, ayant pris la forme allongee k queue.
Dans ce dernier cas, il est impossible de savoir a quelle esp^ce on a
affaire, sans une preparation permettant de voir les micronucleus.
Ce charact^re distinctif, bas^ sur le contour general, n'a done qu'une
valeur relative. II est cependant bon d'en tenir compte; car lorsqu'on
s'est exerce k bien distinguer les deux especes, il suffit presque toujours
et trompe rarement.

Le P. caudatum paralt avoir une taille un peu plus grande que celle
du P. aurelia. Ainsi, j'ai mesur^ des premiers depuis 120 jusqu'a
325 /x, tandis que les seconds ont varie seulement entre 70 et 290 /i.
En outre, P. caudatum se conjugue avec une taille variant entre 125 k
220 n, et P. aurelia entre 75 a 145 fi. Pendant la conjugaison, le deroule-
ment rubanaire, preparant la fragmentation du nucleus, s'effectue chez
le P. aurelia, des le stade D, tandis que chez le P. caudatum il ne com-
mence que vers le milieu du stade G. Chez cette derniere espece, le
nucleus mixte de copulation donne naissance finalem'ent a huit corpus-
cules, chez P. aurelia il n'en produit que quatre; il en resulte que chez
celle-ci I'^tat normal se trouve r^tabh des la premiere bipartition qui
suit la conjugaison, et chez P. caudatum seulement apres la seconde.

Toutes ces differences anatomiques et physiologiques me paraissent
plus que suffisantes pour justifier la distinction des deux especes. II

"Sur la multiplication des Infusoires cilies, Arch, de zool. exp. et gen., (2), 4,
1888, pp. 231-235.

PARAMAECIUM AURELIA AND PARAMAECIUM CAU DATUM 227

est fort possible que Claparede et Lachmann aieiit eu raison, en consider-
ant la forme caudatwn comme plus typique que la forme aurelia. Si,
en effet, on examine avec soin les dessins de O. — F. Miiller, on penche
k croire que le vieux micrographe a vu et figure la premiere seulement.
En se conformant strictement au principe de la loi de priorite, ce serait
done le nom aurelia, donne par Miiller, qui devrait etre conserve a la
forme fuselee. Mais, d'un autre cote, Ehrenberg et Dujardin ont dis-
tingue ce type et I'ont decomme caudatum. Si nous lui conservons la
vieille denomination aurelia, il devient impossible de transmettre le
qualificatif caudatum a la forme qui, le plus souvent, est obtuse k
ses deux extremites. II faudrait alors creer un nouveau nom. Je
crois plus simple de conserver les denominations d'Ehrenberg.

Since 1889, when Maupas^^ and Hertwig^'', in studies on conju-
gation added further evidence for the distinction of the two forms,
they have been generally accepted as 'good' species. Calkins,
however, again raised the question in 1906: ''I personally believe
that the slight differences that distinguish the two types of Para-
mecium are not of specific value, and hold that P. caudatum
should be regarded as a mere variant of P. aurelia. "^^ He based
this view chiefly on the following observations. One of a pair
of ex-conjugants of P. caudatum, which he was studying by his
well-known accurate culture methods, reorganized as P. caudatum
and the other as P. aurelia, i.e., the latter had two small micrp-
nuclei, instead of one, and remained in this condition for about
forty-five generations in pedigree culture, and then reverted to
the caudatum type with one large micronucleus. While the
aurelia phase existed, the rate of«division was comparatively
slow, and when the caudatum phase was reassumed the rate of
division immediately increased considerably. Calkins also con-
sidered the relative size of the two forms, and the conjugation
phenomena as described by Maupas and Hertwig, and concluded
that these are not of such a character as to warrant their being
considered diagnostic.

i*Le rajeunissement karyogamique chez les cilies, Arch, de zool. exp. et gen., (2),
7, 1889.

i^Ueber die Konjugation der Infusorien, Abh. kgl. bayr. Akad. d. Wiss. Miinchen,
2, CI. 17, 1889.

"Paramecium aurelia and Paramecium caudatum. Studies by the pupils of
W. T. Sedgwick, 1906.

228 LORANDE LOSS WOODRUFF

Jennings, in his studies on heredity in Paramaecium/^ showed
that he could readily isolate a considerable number of pure lines
from a wild culture, and that these pure lines breed true, i.e.,
there exist inherent hereditary differences in size, persisting when
all other conditions remain the same. These different lines fall
usually into two main groups, one group having a mean length
greater than 170^, and the other having a mean length less than
140ai. But he was able finally to isolate a line intermediate in
size, and thus to bridge over the gap. As Jennings points out,
even if it were not possible to isolate a strain of intermediate
size between the two large groups, this would not give a basis
for distinguishing two species. However, he states: ''I may
be permitted to add to the precise data thus far given a personal
impression or surmise. Though, as I -have shown, intermediate
lines occur, I believe that it will be found that most Paramecia
can be placed in one of the two groups that we have called ' cau-
datum' and 'aurelia'. In other words, if my impression is cor-
rect, most lines will have a mean length either below 145 microns
or above 170 microns; rarely will lines be found whose mean falls
between these values. Such at least has been my experience in
a large amount of work. Furthermore, I am inclined to believe
that those belonging to the smaller group (mean length below
145 microns) will be found to have as a rule two micronuclei;
those belonging to the large group but one micronucleus. This
matter is worthy of special examination."

Jennings and Hargitt in 1909 made this examination and in a
preliminary communicatioif stated^* that ''two sets of races
could be distinguished, one set having two micronuclei, the other
but one. The races with two micronuclei were all smaller than
those with one. The larger races together thus correspond with
what had before been described as P. caudatum, the smaller races
with P. aureha. The two differ also in the size, position and

i^Heredity, variation and evolution in Protozoa. II. Heredity and variation of
size and form in Paramecium, with studies of growth, environmental action
and selection, Proc. Amer. Philosophical Society, 47, no. 190, 1908.

''Characteristics of the diverse races of Paramecium, Proc. Amer. Soc. Zool-
ogists, 1909 meeting, in Science, March 25, 1910.

PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 229

staining relations of the micronuclei, in ways that correspond
to the descriptions of Hertwig and Maupas. But in rare cases
specimens of the caudatum races have two micronuclei, those
of aurelia races but one, thus confirming the observation of
Calkins on this point."

In accordance with the conclusions of Calkins, I have used the
specific name aurelia to include both the aurelia and caudaturn
forms; but my extended study of Paramaecia cultures has demon-
strated that these two forms are remarkably constant, and I am
inclined to the view that they are distinct species, in the sense
in which this term is generally used in biological work. The data
on which I base this conclusion are chiefly as follows: the pedi-
gree culture of P. aurelia which I have had under daily observa-
tion for (so far) more than three and one half years, during which
time more than 2100 generations have been attained, has bred
practically true to the aurelia type as described by Maupas in
the passage quoted. The pedigree culture of P. caudatum which
I have carried for nearly seven months, and which has attained
more then 300 generations up to the present time, has bred prac-
tically true to the caudatum type as described by that author.

The pedigree culture of P. aurelia was started on May 1, 1907,
with a 'wild' individual which was found in a laboratory aquar-
ium, and was carried on at Williams College during May and June,
1907; at the Woods Hole Marine Biological Laboratory during
parts of the summers of 1907 through 1910; and at Yale Univer-
sity during the academic years from 1907 to the present time,
November 30, 1910. The pedigree culture of P. caudatum was
started on May 14, 1910, with a 'wild' individual collected from
a pond at New Haven, Conn., and was carried on at Yale Univer-
sity except for a period of a few weeks in the summer when it
was taken to the Woods Hole Laboratory.

The original specimen of each culture was placed in about five
drops of culture fluid on a glass slide having a central ground
concavity, and when the animal had divided twice, producing
four individuals, each of these was isolated on a separate slide to
form the four lines of the respective cultures. The pedigree
cultures have been maintained by the isolation of a specimen from

230 LORANDE LOSS WOODRUFF

each of these lines practically every day up to the present time, thus
precluding the possibihty of conjugation taking place between
sister cells. The number of divisions of each line has been recorded
daily at the time of isolation and the average rate of these four
lines has been again averaged for ten-day periods (cf. fig. 1).
The culture medium has consisted of materials collected prac-
tically at random from laboratory aquaria, hay infusions, ponds,
etc. The infusions were thoroughly boiled to prevent the con-
tamination of the pure lines of the pedigree cultures by 'wild'
individuals. Permanent preparations have been preserved from
time to time for the study of the cytological changes during the
life history.

In the light of this experience with cultures I shall consider
each of the characters emphasized by Maupas.

Shape. The general shape of the aurelia and caudatum forms
is, in nearly all specimens, quite distinctive; aurelia is slightly
more broad at the posterior than at the anterior end, while cau-
datum, as the name implies, is quite pointed at the posterior
end as compared with the anterior end. The posterior end, in
the specimens in my pure culture, is markedly pointed, and
being free from endoplasmic inclusions, appears transparent and
clearly delineated even under a lens with a magnification of ten
diameters. I have been accustomed to allow stock material
from my pedigree aurelia culture to multiply in large flasks of
hay infusion, for various experiments on conjugation, etc.
Frequently I have used this material for ni}^ elementary class in
biology and I have found that even the novice has called attention
to the fact that the shape of the ends was reversed as compared
with the figure of caudatum in the text-book. McClendon, how-
ever, stated that in his study of aurelia and caudatum he found
''no characters of outward form" which were diagnostic.

Changes in the vitality of my pedigree lines never have been
very marked, and consequently I have not had organisms, in
the direct lines of my pedigree cultures, representing physiolog-
ical extremes to compare. Numerous experiments, however,
have been made with 'stock' material left over after the daily
isolations of the pure lines, which have clearly shown that, for

PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 231

example, even when the aurelia and caudatum cultures are sub-
jected to unfavorable environmental conditions, as, for example,
scarcity of food, the very great majority of individuals retain
the shape which is characteristic of the race.

Size. As has frequently been pointed out, size alone is an
entirely inadequate character on which to base species. It is
significant, however, I believe, that during the long life of my
pure strains, I have never observed the relative size of the indi-
viduals of the aurelia and caudatum forms, when bred under
identical conditions, to change greatly during any single period.
Experiments have shown that even when the two forms have
been bred under diverse conditions, for example, aurelia in a
medium rich in food and caudatum in a medium with a very
small amount of bacterial growth, the size of the caudatum form
always has remained sufficiently great to render it distinguish-
able from the aurelia form. On the basis of size alone, then, it
has been possible, with great accuracy, to separate the two forms
when mingled together. It is probable, of course, that I began
my pedigree cultures with very typical specimens of the aurelia^"
and caudatum groups as described by Jennings. If such be the
case, then my cultures add considerable evidence in favor of
the different strains which Jennings has isolated. It appears
to me, however, that what that author has done for Paramaecium,
can probably be done for many closely related species of infusoria,
and the very fact that he did find it difficult to secure an inter-
mediate race between the aurelia an-d the caudatum groups is a
strong point in favor of the. distinctness of the forms.

Vitality. It has been customary to regard the rate of repro-
duction of infusoria in culture as a just criterion of vitality.
Maupas wrote i^^ "Cette faculte de reproduction (aurelia) resem-
ble beaucoup a celle de la precedente espece (caudatum)." My
cultures completely corroborate this statement, for during the
six and one half months of the life of the caudatum culture, 324
generations have been attained, while during the same period,

^"For further details of the culture see: L. L. Woodruff, Two thousand genera-
tions of Paramaecium; Archiv fiir Protistenkunde, 21, 3, 1911.
'^'Sur la multiplication des Infusoires cilies, loc. cit., p. 234.

232

LORANDE LOSS WOODRUFF

under identical conditions, the aurelia culture has advanced from
the 1785th generation to the 2117th generation, or 332 genera-
tions. This gives a difference of only eight generations in the rate
of reproduction of the two forms during seven months (cf. fig. 1).
These cultures obviously do not support the statement, frequently
made, that aurelia is a weaker form than caudatum.

Maupas remarked that P. aurelia was one of the most common
infusoria, and Jennings found that a typical wild culture could

2.5

2.0

1.5

0.0

Fig. 1 Diagram showing the comparative rate of division of the pedigree
cultures of Paramaecium aurelia and Paramaecium caudatum, when bred under
identical conditions, from May 14, 1910, to November 30, 1910. During this pe-
riod P. aurelia (designated by continuous line) advanced from 1785 to 2117 genera-
tions, while P. caudatum (designated by broken line) advanced from 1 to 324
generations. The rate of division is averaged for ten-day periods. The ordin-
ates represent the average daily rate of division of the four lines of the cultures.

be resolved into caudatum and aurelia groups. It has been my
experience that it is as easy to procure one form as the other in
the wild state. Certainly my aurelia culture, which theo-
retically would provide individuals to the number represented
by 2 to the 2117th power, gives more evidence of vitality and
reproductive power than has been demonstrated for any other
animal.

Conjugation. I have no data in regard to the conjugation of
either of these forms, for, so far, in all experiments with stock

PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 233

material left over after the daily isolations from my pure lines,
I have failed to observe a single syzj-^gy, either between aurelia
lines or caudatmn lines, or between aurelia and caudatum lines.
Jenning's" experiments on conjugation in Paramaecium bring
out data which add further evidence that in certain strains at
least a predisposition to conjugation does not exist. Maupas
wrote: ''C'est bien certainement une des especes (aurelia) qui
se recontrent les plus frequemment a I'etat conjugu^."

Maupas, as we have seen, pointed out a difference in the nuclear
phenomena during conjugation which he held to be of diagnostic
value, and Hertwig apparently showed that aurelia has two micro-
nuclei at the reorganization after conjugation. Calkins, on the
other hand, has shown that P. caudatum, in one case, reorgan-
ized with two micronuclei and later reverted to the uninucleate
type. Such a case can readily be considered a ' sport ' which has
arisen possibly by the persistence of the stage with two micro-
nuclei immediately following the separation of the conjugants,
or by the precocious division of a single micronucleus previous to
the first regular vegetative division after conjugation. Although,
as Calkins stated also, forty-five generations is a long time for
an abnormality, if it be such, to persist; nevertheless, I believe
it is very significant that, whereas during the presence of two
micronuclei the division rate averaged only 0.8 of a division per
day, after the loss of one of the micronuclei the division rate
increased to the remarkable rate of 2.2 divisions per day, on the
average for a period af four months. It is also of interest that
the other exconjugant which reorganized 'normally' as caudatum
failed to live.

So far as I am aware, the following statement^'* by Simpson is
the only record of a possible case of conjugation between aurelia
and caudatum: ''Out of twenty-one attempts I had but two par-
tial successes. Conjugation took place on two slides: the period
was normal. After separation each of the ex-conjugates divided
once : on the third day they died off. In anticipation of something

"What induces conjugation in Paramecium? Jour. Exp. Zool. 9, 2, 1910.
^^Observations on binary fission in the life-history of Cihata, Proc. Royal
Soc. Edinburgh, 1901, pp. 407-408.

234 LORANDE LOSS WOODRUFF

of this sort from analogy in higher forms, I intended to let the
two pairs run their natural course, foregoing the desire to examine
their nuclear condition. In view, therefore, of the incomplete-
ness of the experiment, it is perhaps unwarrantable to draw any
results regarding hybridization and infertility, or even the 'fixity
of species' so far down in the animal scale." Simpson gives no
data to prove that these were actually syzygies between the two
forms, but if they were, it is obvious that they were not fertile.
Jennings and Hargitt stated that they had been unable to induce
the two forms to conjugate.

In view of the fact that, for example, Maupas studied conjuga-
tion of both P. aurelia and P. caudatum, and Hertwig studied
conjugation of P. aurelia, and also that Jennings observed con-
jugation in both his aurelia races and in his caudatum races, it
is clear that aurelia forms conjugate and caudatum forms con-
jugate, but there is no positive evidence that conjugation takes
place between individuals of aurelia and caudatum.

Macronucleus. The normal macronucleus of aurelia was de-
scribed by Hertwig and Maupas and that of caudatum agrees
very closely. It is an ellipsoidal body with a smooth contour,
except for a slight depression, in which the micronucleus is usually
located. But the form of the macronucleus of both aurelia and
caudatum frequently departs very greatly from the 'normal'
condition. It is not unusual to find paramaecia of my aurelia
cultures with the macronucleus resolved into several parts. These
parts apparently may be gathered together into a typical nucleus
for division, or the cytoplasm and micronuclei may divide, the
macronuclear fragments which are in the posterior part form-
ing the macronucleus of one daughter cell and those in the
anterior part forming the macronucleus of the other daughter
cell. I shall reserve the full discussion of these interesting
changes for a special paper. It is - important to emphasize
the fact that these are not pathological conditions, since the
general vitality, as indicated by the rate of division, is not
appreciably affected.

PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 235

Calkins,-^ however, found nuclear fragmentation in degenerat-
ing individuals of caudatum, Wallengren-^ and Kasanzeff^^^
showed that various changes including fragmentation of the mac-
ronucleus occur when paramaecia are starved, and Popoff-'
described a large increase in size and fragmentation of the macro-
nucleus in degenerating caudatum which paralleled the conditions
observed in specimens ripe for conjugation. He also obtained
similar changes by subjecting the animals to various reagents. ^^
Mitrophanow^* emphasized the fact that the structure of the
macronucleus varied considerably under the influence of diverse
conditions, and he described fragmentation and figured spherical
pieces which very closely resembled micronuclei.

It is evident, then, from my cultures that the macronucleus of
both aurelia and caudatum is subject to great morphological vari-
ation without appreciably affecting the rate of reproduction, i.e.,
it is entirely normal. It is also apparent from the work of the
other authors cited that degeneration changes become manifest
in the fragmentation of the macronucleus. Consequently the
macronucleus presents no character which is of permanent diag-
nostic value.

Micronucleus. Maupas, as we have seen, regarded the micro-
nucleus as the chief distinguishing character of aurelia and cau-
datum, and my cultures substantiate his view. Fixed and stained
individuals show that the micronuclei of the aurelia culture for
over two thousand generations have conformed in a remarkable
degree to the aurelia type as described by the French investi-
gator, and the micronuclei of the caudatum culture have conformed
to his caudatum type.

2*Studies on the life history of Protozoa. IV. Death of the A Series, Jour.
Exp. Zool., 1, 3, 1904.

-^Inanitionserscheinungen der Zelle, Zeit. f. allg. Physiologie, I, 1, 1901.

-"Experimentelle Untersuchungen ueber Paramecium caudatum. Inaug.' — Diss.,
Zurich, 1901.

-^Depression der Protozoenzelle und der Geschlechtszellen der Metazoen,
Archiv fur Protistenkunde, R. Hertwig Festband, 1907.

2«Experimentelle Zellstudien III. Ueber einige XJrsachen der physiologischen
Depression der Zelle. Archiv fiir Zellforschung. 4, 1909.

-"L'appareil nucleaire des Param6cies, Arch. Zool. Exp. et Gen., (4), I, 1903.

236 LORANDE LOSS WOODRUFF

It is not only the presence of two micronuclei, but their pecul-
iar morphology, as emphasized by Maupas, which is character-
istic of the aurelia type. I have found one individual of the au-
relia culture with three micronuclei, and a few specimens in
which I have been unable, in total mounts, to distinguish a single
micronucleus or more than one micronucleus. ,But when only
one micronucleus could be seen it has been of the aurelia type, and
other individuals of the culture at the same period have had the two
characteristic micronuclei. I have observed a variation in the
number of micronuclei in various pedigree cultures of hypotrichs,^"
Popoff has found reduplication in Stylonychia mytilus and Para-
maecium caudatum during degeneration, and Kasanzeff has
observed the same in starved P. caudatum. Thus, while my
cultures of Paramaecium and various hypotrichous species sub-
stantiate Wallengren's and Calkins' statement that the micro-
nuclei are the most stable elements in the cell, and the last to be
visibly affected by environmental changes, nevertheless it is
apparent that they are subject to variations under certain
unknown conditions. Temporary variation, therefore, cannot
be considered as having weight in determining species. The essen-
tial fact is, however, that throughout the existence of my aurelia
and caudatum cultures, the morphology of the micronuclei has
conformed to Maupas' description for the respective species.
It must be borne in mind also that P. caudatum has been the
subject of more extended study by exact culture methods than
any protozoon except P. aurelia, and in all these long pedigree
cultures it has bred true to the caudatum type, at least with respect
to the single micronucleus. Calkins, for example, in his impor-
tant investigations on the life history of this form, carried three
distinct cultures, by the aid of artificial stimuli during periods
of physiological depression, through 379, 570, and 742 generations
respectively. McClendon, also, studied mass cultures of Para-
maecium for considerable periods and stated that he never
found individuals ''with different numbers of micronuclei in the
same culture. "''^

^"An experimental study on the life history of hypotrichous Infusoria, Jour.
Exp. Zool., 2, 4, 1905.
"Protozoan studies, Jour. Exp. Zool.. 6, 2, 1909.

PARAMAECIUM AURELIA AND PARAMAECIUM CAUDATUM 237
SUMMARY

Briefly stated, I am convinced from my study of paramaecia
that —

1 . A very great maj ority of individuals of aurelia and caudatum
can be distinguished on the basis of shape alone;

2. A very great majority of individuals of aurelia and cau-
datum can be distinguished on the basis of size alone;

3. The power of reproduction, or general vitality, of aurelia
and caudatum is practically identical;

4. The macronucleus of aurelia and caudatum is subject to
such great variation that it affords no diagnostic feature;

5. The micronuclear apparatus of aurelia and caudatum affords
crucial diagnostic characters.

I have summarized the various characters of the two forms as
they have shown themselves in my long pedigree cultures, and it
is evident that they have conformed practically identically to
the Maupasian types — such variations as have appeared not being
so great as have been observed to occur in undisputed species,
or as one would expect to find when the intimate relation of the
unicellular organism to the environment is considered. Therefore,
I believe, that since one of the crucial tests of a species is
its ability to breed true to its type indefinitely, aurelia and cau-
datum have adequately met this test during more generations
than any other animal under observation, and accordingly
Paramaecium aurelia O. F. M. and Paramaecium caudatum Ehrbg.
should be regarded as distinct species. ^-^ ^^

*2In this paper I have followed the spelling of the name of the genus as given
by its founder, except in direct quotations from other authors.

'*I have the satisfaction to note that my conclusions are in accord with the
final results published by Jennings and Hargitt in the last number of this
journal, which was received when this paper was in press. Hargitt says,
"There is cytological warrant for distinguishing caudatum races from aurelia
races, and it seems probable that it will continue to be convenient to distinguish-
these as two species."

MALE ORGANS FOR SPERM-TRANSFER IN THE

CRAY-FISH, CAMBARUS AFFINIS: THEIR

STRUCTURE AND USE

E. A. ANDREWS

From the Zoological Laboratory, Johns Hopkins University

THIRTY -ONE TEXT FIGURES AND FOUR PLATES

INTRODUCTION

The present paper is a contribution to our knowledge of the
means that lead to the fertilization of the egg. It is part of the
history of the sperm outside the body of the animal.

Sexual reproduction in most complex animals involves the trans-
fer of the sperm from one animal to another, before the eggs can
be fertilized.

Among animals the various methods by which the sperm is
transferred may be grouped under the three heads, diffuse, direct,
indirect. By diffuse sperm transfer we mean the discharge of the
sperm into the water, where it may meet the eggs outside of the
female, as in certain coelenterates, echinoderms and annelids,
or may be drawn into the body of the female, as in certain lamel-
libranchs. By direct sperm transfer we mean the method found
in the majority of complex animals, in which there is more or less
direct application of the terminal parts of the passages leading
the sperm to the exterior to the passages leading from the exter-
ior direct to the eggs. In this group there is commonly a true
copulation.

By indirect sperm transfer we mean those peculiar complex
methods of getting the sperm from the testis to the eggs that are
found in a few cases amongst the great groups of animals, as in

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

239

240 E. A. ANDREWS

earthworms, spiders, some cephalopods and leeches. The es-
sence of these cases of indirect sperm transfer Hes in the fact that
while the sperm is transferred by organs, and not by floating,
yet these organs either do not put the sperm into the egg passages,
or else if they do they are not organs directly concerned with the
discharge of sperm, or both may be true. In indirect sperm trans-
fer there is no true copulation, or intromission, but at most con-
jugation or clasping.

The three methods are not always sharply separable, and may
be regarded as only convenient groupings of physiological processes
that occur here ^nd^here among animals without reference to
their systematic positions. The diffuse method is obviously the
one open to the most hazard in the sperm and eggs meeting; the
direct method by intromission the best assured method; the in-
direct method most perculiar and needing special explanation in
each case.

In the crayfishes and lobsters most interesting cases of indirect
sperm transfer occur, and it is the purpose of the present paper to
describe the organs of the males that are used to transfer the sperm
to the receiving organs that have been described in previous com-
munications (1, 2, 3). In these animals the male transfers the
sperm to the outside surface of the female where it remains till
the eggs are laid, when fertilization takes place outside the body.
In the American lobster and the sixtj^ and more species of cray-
fishes of the genus Cambarus that are found in all but the most
western parts of North America, the sperm on the shell of the
female is stored in a special pocket, or receptacle, but in the other
genera of crayfishes, all the world over, there is no such receptacle
and the sperm is believed to be distributed over the shell of the
female in separate spermatophores. While the sperm pocket has
been described (1,2, 3) the organs of the male that fill the pocket
have had only such consideration as was necessary for the sys-
tematist, who found them to be of the greatest value in distin-
guishing species and in forming subgenera.

In the present paper the anatomy of the male organs is exam-
ined and their use as organs of sperm- transfer is explained.

ORGANS FOR SPERM-TRANSFER 241

CAMBARUS AFFINIS

While the sexual habits of all the species of Cambarus, agree
in the main, the species afFmis has been more studied, and as in
describing the female organs concerned in sperm transfer we first
considered this species, we will also give chief attention to the
male organs of sperm transfer in this species.

As elsewhere described (4, 5) conjugation is here a long series of
activities of the male accomplishing the accurate adjustment of
the essential transfer organs of the male to the receptacle of the
female. The receptacle of the female is a single pouch in the shell,
but the transfer organs of the male are three pairs of outgrowths.
On each side of the body there is a papilla, or special termination
of the sperm duct, and two limbs, those of the first and second
segments of the abdomen, which we may call the stylets.

To introduce the sperm into the receptacle the papilla must be
adjusted both to the first and to the second stylet and both sides of
the body play a necessary part in the process of sperm transfer.

THE PAPILLAE

One external instrument concerned in the process of sperm
transfer is the modified end of the sperm duct that emerges from
the base of the fifth leg, on each side of the animal. These organs
are the papillae.

Since systematic work has been done largely upon preserved
specimens it is not so generally known that the sperm duct ends in
life, in a soft, turgid protuberance, which may be so collapsed after
death as to leave only the rounded hole in the firmer shell as the
apparent ending of the sperm duct. These papillae lie concealed
by the stylets, at rest, but on raising the stylets the papillae are
seen as conspicuous, clear, tubes about 3 mm. long and 1^ mm.
wide, jutting out from the base of each fifth leg.

At the time of conjugation the papillae are also concealed from
view since the necks of the first and the second stylets form a
nicely adjusted frame about the papillae and this frame is fitted
in between the bases of the fifth legs. Certain in and out move-

242 E. A. ANDREWS

ments of these bases seem to adjust the papillae so that they fit
accurately into the orifices of the first stylets (figs. 30, 31) and by
that means the sperm discharged from each papilla is passed into
the cavity of the stylet.

The papilla (P. fig. 1) is on the under side of the large first seg-
ment of the leg and projects downward and toward the median
plane; but its tip turns away from the middle line of the body.
The papilla is a cone with bent apex. It is translucent and dis-
tended with colorless blood. When directly injured, or upon les-
sening of the blood pressure from injury elsewhere, the papilla
collapses, being but a thin uncalcified protrusion of the skin, kept
turgid, or erected, by blood pressure. Within the papilla one can
see a large central tube passing toward the tip and also chalky
white masses suspended between the central tube and the thin
outer wal!s.

On the shell at the base of the papilla there is, anteriorly, a
single row of very long setae (fig.l ) that form a sort of protective
screen over the anterior face of the papilla.

Sections show that the papilla is a continuation of the deferent
duct, blood cavity and skin, so constructed that the bent, conical
apex, with its soft walls can be adjusted to the hard opening of
the stylet so as to fit hermetically, as a tense rubber bag might.
Moreover the bent tip can be opened to discharge the sperm,
when special muscles remove the obstructing valve that holds the
tube closed.

A lengthwise section through this delicate papilla (fig. 2) shows
that the central tube is a direct continuation of the deferent duct
that leads the sperm from the testes to the tip of the papilla.
Between this duct and the outer cuticle there is a large space
fall of blood, traversed by little connective tissue and in it are the
white bodies just mentioned, now seen to be small tubular glands,
opening into the central duct. The central duct presents two
strikingly distinct parts; the one continued from within the leg
has the thick muscular wall and peculiar secreting lining of the
deferent ducts, the other is lined by the thin cuticle inflected at
the orifice at the tip of the papilla, and lacks muscle. In place
of muscle the wall has only epidermis, which extends irregular!}'

ORGANS FOR SPERM-TRANSFER

243

Fig. 1 Posterior face of the left fifth leg of a living male 95 mm. long to show
the translucent papilla. (P.) 2. Qq.

Fig. 2 Longitudinal section through the papilla and the base of the fifth leg,
showing the orifice of the sperm duct, the valve, the muscles and the glands.
2. 90 mm. A.

244 E. A. ANDREWS

into the blood space as the tubular glands alluded to above. The
orifice at the tip is small and is not closed by any muscle, but
apparently by blood pressure only. The part of the tube lined
by cuticle has its lumen much reduced by a valve, or great longi-
tudinal ridge, which extends out as far as the abrupt bend at the
orifice. In a cross section (fig. 3) this ridge is well seen, as is also
the fact that some muscle fibres run into it and that the glands are
chiefly on the side opposite the ridge. The ridge appears to act
like a valve to hold this part of the tube closed, while contractions
of the muscle would tend to open the tube wide and let the sperm
pass to the orifice, which would then be forced open by the internal
pressure of the sperm squeezed by the muscles of the wall all along
the length of the duct, or some extent of it at all events.

The upper part of the duct, as seen in the cross section fig. 4,
has its thick muscles arranged chiefly in transverse fibres and is
lined by an epithelium that evidently in large measure breaks
down to furnish a great mass of secretion about the sperms. It
is probably this secretion that envelops the sperm in the form of
macaroni-like tubes, when they pass out in a slow stream.

THE STYLETS

The most complex of the organs concerned in sperm transfer
are the modified limbs of the abdomen which we will call the sty-
lets. In the male the sixth pair of abdominal appendages form
the large side parts of the tail fan while the third to the fifth in-
clusive are the simple and apparently rather useless swimmerets.
The first and second pairs are specially constructed to serve as
transfer organs for the sperm.

These appendages of the first and second somites are much
stouter and longer than the following swimmerets and have a very
firm attachment to the abdominal sterna. The calcified ridge
across the middle of the sterna is much more developed in the
first and second somites, and where the appendages are fastened
it rises up as a decided elevation which remains as a stump when
the appendage is cut off. On the second somite these stumps are
far apart, (some 10 mm. in a male of 100 mm.) while on the first

ORGANS FOR SPERM-TRANSFER

245

Ex. m. ',

Fig. 3 Cross section of the sperm duct and valve along the line 3 of fig. 2,
showing the duct closed by valve ridge. 2. A.

Fig. 4 Cross section of the sperm duct along the line of 4- of fig. 1, showing the
muscular wall and the lining epithelium disintegrating in secretion. 2. A.

Fig. 5 Extreme tip of right first stylet, showing the groove bottom coming to
the surface, posterior face. 2. A. Ex. m.— the external mass. M.m.— the in-
ternal mass. IS' — the level of the section, fig. 13.

Fig. 6 Diagram of stylet as in plate i, fig. i, to show location of glands in the
interior, and the location of the sections, 7 to 13, shown in figs. 7 to 13.

24G E. A. ANDREWS

somite they are in contact at the median Une of the abdomen.
The eUiptical transversely elongated stumps of the first append-
ages are 5 mm. long and those of the second about 3 mm.

Commonly these appendages are carried forward horizontally
under the thorax between the thoracic legs in a deep depression
of the thoracic sterna. The first pair lie close side by side with
their median faces in contact. The second pair lie over and largely
conceal the first, since their form enables them to come to the
middle of the body beneath the first pair in spite of the fact that
their bases are attached to the sterna, so far from the middle line.

In a dead male one may move the appendages upon their at-
tached bases as follows:

The first may be moved upon its base from the horizontal up
toward the vertical only about 45°. The membrane on the anter-
ior face of the joint at the base of the appendage is stretched to
its limit when the appendage is pulled up a little beyond sixty
degrees, so that this appendage is never vertical and cannot swing
back and forth through a wide arc as do the ordinary swimmerets.
The distance traversed by its tip is some two cm. The appendage
may also be rotated a very little at its base and moved from side
to side a little so that ts tip travels some 5 or 6 mm.

The apex of the second may be drawn from the horizontal up
a little beyond the vertical; but neither the basal protopodite
nor the endopodite travels more than 90°. They are set together
at a large angle, so that while the main length of the appendage is
horizontal the basal part never is, and when the base goes back
some 90 degrees the horizontal part is swung past the vertical
line. The tip traverses some 3 cm. The base may be rotated
a little and moved from side to side so that the apex travels 6 or
7 mm.

STRUCTURE AND ANATOMY OF THE FIRST STYLET

The first abdominal appendage of the male is a very stiff cal-
cified mass of the general shape of an awl, some 17 mm. long, but
having two tips. There is a groove along more than half its length
and the base is articulated to the ventral shell of the anima' so

ORGANS FOR SPERM-TRANSFER 247

that the appendage has very Httle mobihty back and forth
through some 45°. The normal position of the stylet is pointed
forward under the thorax, where it lies horizontally in a deep
groove, but in use it is dropped down and backward toward a
vertical position. It has an anterior face, which is usually carried
as the dorsal side, a posterior face which is usually the ventral
aspect, and an outer and an inner or median face.

The general appearance of the stylet is seen in the photographs,
figs. I, II, III, IV, which represent respectively the posterior, median
(or rather median and posterior somewhat diagonally), the anter-
ior and the outer faces of the same left stylet. Fig. i, the posterior
face, is the view got by looking at the under side of the crayfish,
after lifting up the second stylet, which lies over the first and
largely conceals it.

The first pair of stylets do not spring from the sternal surface
far apart as is the case with the common, unmodified swimmeret,
but they arise very close together; in fact the median faces, (fig.
II,) of the two come into contact so that these two appendages really
form one mass. If looked at from the dorsal side, the two are
seen to lie in contact at the base and all along the distal half,
leaving between the constricted parts of the two a square opening
that is occupied, in rest, by part of the second stylet.

In describing the stylet we will distinguish the base, the neck,
and the scroll or spiral that contains the groove. The scroll ends
in two tips, the more slender, side outgrowth, or spatula, and the
real end bearing the groove, the canula

The base is some 6 mm. wide and long and only 2 thick, being
flattened from before back. The posterior or ventral face of the
base, fig. I, presents a wide groove bounded on the median side by
a rounded knob and on the outer side by a long ridge which, as it
passes on to the neck, bears a tuft of long, finely plumose setae,
that are seen again in profile in figs, ii, iv. In this deep groove
the second stylet lies when not in use, so that the two appendages
are firmly packed together under the thorax of the male.

The part of the base joined to the sternum of the animal is an
oblique eUiptical area, around the edge of which the hard shell
gives place to the soft articular membrane that makes it possible

248 E. A. ANDREWS

to cut the whole appendage away from the sternum. In this
membrane there is an articular, whitish plate that is seen in figs.
Ill and IV. The whole base is pyramidal and except the posterior
all its faces (figs, ii, iii, iv) are convex and rounded.

The neck is the narrowest part, before the sudden enlargement
of the spiral part; it is the smallest of the three regions; and is best
seen in figs, ii, iii, iv. The neck passes gradually into the base and
ends abruptly at the spiral. It is some 3 mm. long and 2 wide
and thick. It has an angle along the ventral face that continues
the ridge of the base up to the outer part of the spiral.

The spiral or scroll may be likened to a long triangular plate
with its edges rolled in together so as to leave a groove between
them, but it is a plate some 8 mm. long, with the edges greatly
thickened, so that the resulting mass is apparently solid. The
groove begins on the median side, fig. ii, and passes in a sinuous
course to the ventral side and along this diagonally to the very
tip. The apparent bifid nature of the stylet is due to an out-
growth from the median part, quite separate from the real end
of the organ, in which the groove is continued through its entire
length. We have then to describe a sinuous groove and its two
boundaries, which we will call the median mass and the external
mass ; and also the two tips. The external mass, seen from the ven-
tral side on the right of fig. i, shows a proximal part about 2 mm.
long and 1 mm. wide, bearing a marked ridge parallel to its sides
and continued up from the neck. And then it suddenly turns at
a large angle and becomes a rounded and gradually tapering ter-
minal part, something less than ^ nam. wide at first, and 6 mm.
long. This passes behind the slender protuberance of the median
mass to end as a flattened, horny tip together with the like ending
of the median mass. In other words both external and median
masses unite as the horny tip that we will call the canula. The
sudden change in direction of the mass is accompanied by a like
change in the groove whose edge it forms; this change of the groove
we will call the angle of the groove.

Seen from the outer face, fig. iv, the external mass is widely
swollen proximally, some 2^ mm. deep, and gradually narrows into
the distal part. The round canula is bent somewhat, ventrally.

ORGANS FOR SPERM-TRANSFER 249

On the dorsal face, fig. iii, the external mass is confluent with the
median mass, without boundary line. Thus the distinction be-
tween the two masses is useful chiefly on the anterior face where
they form the two sides of the groove.

In fig. Ill, the long triangular region running from the notch
that marks off the neck from the spiral region and ending distally
in the rounded and pointed canula, is to be regarded as made up
chiefly of the median mass, but the depressed part along the left
edge is part of the external mass.

On the median face, fig. ii, (which is unfortunately turned so
that part of the posterior face shows) the external mass shows
only its prc-ximal end along the side of the diagonal groove, and
into this groove the external mass here sends a narrow horny shelf,
dimly seen as light in fig. ii. The external mass has an angular
projection, or lip, at the very beginning of the groove which will
be described in connection with the orifice of the groove. At the
tip, part of the external mass is seen making the lower part of the
canula, to the right, that is, the curved strip of external mass seen
is flat and on a lower level than the median mass.

On the ventral face, fig. i, the median mass looks like a long
rounded white bone that begins suddenly without apparent con-
nection with the neck and, after running nearly straight for some
6 mm., turns externally across the external mass as a flat, curved
process that we will call the spatula. Beyond the spatula, w^hich
stands out freely as the second tip of the appendage, the median
mass continues as the narrow median edge of the canula. From
the external view, fig. iv, the visible part of the median mass, the
spatula is back of the external mass.

In the dorsal view, fig. iii, the main part of the spiral region is
median mass, forming a long triangle, beginning at a deep notch
near the neck and extending in the foreground as the vis ble part
of the canula and back of that as the spatula. At the notch may
be seen part of the lip on the external mass.

The median face best shows the median mass, but, fig. ii, being
not an exactly median view, does not do justice to it. In reality
this face is markedly flat where it comes against the like face of
of the other stylet of the pair. This flat face is a long ellipse, 2

250 E. A. ANDREWS

mm. wide and 5 long, and is smooth except for a roughened area
near its proximal end where there is a long tuft of finely plumose
setae which bend abruptly downward, that is, posteriorly, as if
an adjustment to the fact that they are pressed in between the
two stylets. These setae are so long as to be visible from all
points of view, cf. figs, i, ii, iii, iv.

The groove itself is seen only from the median and ventral
views. It is some 7 mm. long and begins as the orifice on the med-
ian face where it meets the ventral, fig. ii. The orifice is a con-
ical opening bounded by that depression of the neck that makes the
notch so conspicuous in fig. iii, by the rounded origin of the me-
dian mass, fig. ii, and by the overhanging lip of the external mass.
It is of such shape that the tip of the spout, fig. 1, can fit into it.
The groove leads from the orifice obliquely outward and distally
between the external and median masses some 3 mm. and then
turns to make a rounded angle, fig. i, toward the median line
some 3 mm. more. In this part of its course it is soon concealed
behind the median mass that is rising to form the base of the spat-
ula, but it still exists there and emerging again runs the entire
length of the spatula as a very narrow slit with horny edges.
The groove is thus a long double curve, bending abruptly outward,
then forward and slightly inward and finally outward again, as
seen from the ventral side. But it also bends in the vertical plane,
passing downward, then forward and upward and finally a little
downward at the tip. While the walls of the groove seem to be
merely hard rounded bone there projects into the groove from the
side a narrow shelf of horn that springs from the external mass
only. This will be seen in sections.

The spatula is a flat flagellum-like process some 2 mm. long, |
wide and perhaps | to iV thick. It is curved and pointed as seen
in the figures. It springs from the median mass where this sud-
denly narrows to help form the canula, fig. ii. In life the spatula
is milky white and pliable, not bony, more like leather. At its
base it passes suddenly into the bony walls of the median mass
and there can be bent as if in a socket. After drying it looks
more like a thin chitinous membrane over a dried contents. It
is somewhat concave at the base on the dorsal face. With methyl

ORGANS FOR SPERM-TRANSFER 251

green the horny tip of canula and the shelf in the groove stand out
clearly as distinct from the substance of the spatula.

The canula is some 3 mm. long and at base f mm. wide and thick.
It is a long cone, flattened somewhat from before back, bent up-
ward dorsall3\ and ending in a rather sudden point that bends
outward from the median side. The canula is made up of both
external and internal masses. Most of the length of the canula
is clear, yellow, horny matter, but at the base this is continued
as the white calcified material of the rest of the stylet. The bone
of the external mass stops rather suddenly, while that of the med-
ian mass is continued in the midst of the horny cap as a central
area, as seen from the median view. An enlarged view of the tip
of the canula, fig. 5, shows that both external and internal masses
make about the same amount of the canula, since the groove
continues sinuously almost to the exact tip of the organ, but yet
there is a greater prolongation of the external mass to form a short
ungrooved apex. This sketch is from a canula of the opposite
side of the body from that in fig. i. The two canular tips flare
away from one another.

The groove may be said to begin and to end on the median face
and to be shoved away from it through most of its course by the
ridge that we have called the median mass (fig. i.).

INTERNAL ANATOMY

When the stylet is macerated some days the entire contents
may be drawn out of the hard shell; such a cast of the shell has its
general long conical form with a short conical tip that came out
of the canula and a short flat plate that came out of the spatula.
It is made of connective tissue and blood covered with epidermis
with some red pigment cells and shows at the base some muscles
and at the middle some glands.

The muscles, as made out by dissection of fresh and preserved
crayfishes, are weak and run from the base of the stjdet into the
adjacent ridge of the sternum upon which the stylets articulate.
There is a wide thick fan of muscle that passes from the bony
articular plate of the anterior face of the stylet, fig. iii. When

252 E. A. ANDREWS

this is pulled the stylet is raised dorsally into its position of rest.
Since it lowers the organ into the groove on the thorax it may
be called the depressor, though it really swings the appendage for-
ward.

This depressor muscle is lodged in the protruding ridge of the
sternum from which the stylets spring, and its fibres are made
fast to the posterior wall of this ridge. There is also a smaller
muscle attached to the base of the stylet at its external edge which
would seem to antagonize the other and to tend to swing the sty-
let backward, that is, to raise it up from its horizontal position
of rest into the erect position of use; it may be called the erector
muscle.

The internal anatomy of the stylet as well as the character and
mode of use of the groove, were made clear from sections.

The diagram fig. 6 shows the ventral view of a left stylet as if
transparent, the extent of the glandular area being shaded; the
glands occur in both external and internal masses, but not in the
base of the stylet, and the}' extend from the neck to near the ori-
gin of the spatula, filling most of the cavity of the region in
which they occur.

The sections, (figs. 7 to 13, inclusive), were taken across the
stylet along the planes indicated by the like numerals in fig. 6.

The transverse section, (fig. 7) shows in black the exceedingly
thick shell with the depression on one side that forms part of the
orifice of the groove, overhung by the solid lip. Through the
thickness of the shell that forms this part of the orifice are seen
many fine tubes, passing from the internal glands to discharge on
the surface. The interior of the stylet is a delicate mass of con-
nective tissue, chiefly blood sinuses, crossed by few strands of
tissue, and bounded by the thin epidermis against the shell..
Scattered all through this are the tubular glands that bend and
are cut at various angles. These glands ultimately discharge by
the numerous fine ducts that penetrate the shell. In this section
the sharp angle above is the ridge {R) seen in figs. 6 and I passing
along the external mass. The angle to the right is the line be-
tween the ventral and external faces of the external mass.

ORGANS FOR SPERM-TRANSFEK

253

Sections 8 and 9 show the orifice passing into the groove; they are
cut obUquely transverse and, in addition to the section of the first
stylet, show also the section of the second stjdet as it lies locked in
the first. Disregarding for the present all but the lower part of
the sections we see that the stylet has widened out from the con-
stricted neck into a wide flattened mass sub-divisible with refer-
ence to the groove into the external and median masses. In

Fig. 7 Section across the stylet, in the region of the neck, just below the orifice,
on the level 7 of fig. 6. 72— the sharp ridge on the external mass, fig. 6 and i.
2. 90 mm. A.

Fig. 8 Cross section of the same at the level 8, showing the groove above the
orifice filled by the head of the accessory stylet, which is the separate. mass lying
to the left and above. 2. 90 mm. .4.

fig. 8 the orifice is so overhung by the lip as to be in section a C-
shaped bay, embracing the head and neck of part of the second
stylet. Here again the shell is remarkably thick, but is penetra-
ted by the ducts of the glands discharging on the surface that lines
the orifice. In fig. 8 the lower straight side to the left is the flat
face that is normally applied against its fellow on the outer side
of the body. Above is the angle {R) that represents the ridge of

254

E. A. ANDREWS

the external mass, just as in fig. 7. In the interior some of the
glands are very large. The section distal to this, (fig, 9) shows
the bottom of the groove receded from the surface and constricted
from the rest bj^ the continuation of the lip so that it forms a
rather elliptical hole with only a very narrow slit opening into the
deep groove that is seen from the surface. This surface groove
is bounded on the left by the greatly thickened shell substance of
the median mass and on the right by the thick shell of the external

M.m,

Fig. 9 Cross section of the same at the level 9, showing also the grasping second
stylet, above, and its wedge, to the left, where it is entering into the groove of
the spiral. 2. 90 mm. A.

Fig. 10 Cross section of the same at the level 10, showing the bottom of the
groove cut off by the shelf from the external mass. 2. 90 mm. A.

Fig. 11 Cross section of the same, at the level 11, showing, above, the base of
the spatula. 2. 90 mm. A.

ORGANS FOR SPERM-TRANSFER 255

mass. The cavity within the shell of the external mass is reduced
to a narrow space and the glands have become few.

Further along the stylet, (fig. 10) the groove has passed from
opening to the left (fig. 8), through the position shown in fig. 9,
to open more toward the right. The groove is a deep and narrow
one. Into it still open some few gland ducts from the remaining
glands of the median mass. As before the side walls of the groove
are made of very thick shell. The most unexpected fact is that
the bottom of the groove is shut off as a very minute hole overhung
by the continuation of the lip, which is now a horny shelf passing
all along the groove, near its bottom, and so nearly meeting the
opposite side as to practically shut off the bottom of the groove
as a special tube. This figure shows the form of the stylet at the
level, 10, of fig. 6. The flat side to the left is the flat face of the
median mass, while the rounded edges of the groove are the two
narrow parts of the external and median masses seen from the
ventral side in figs, i and 6, just proximal to the base of the spatula.

A section through the base of the spatula, (fig. 11) shows the
groove above overhung by the rising spatula that conceals it
from surface view, (figs, i and 6) but still allows access to the
groove from the right, in under the spatula base. The external
mass {Ex. m.) is now the greater, but it contains no glands, while
the median mass is reduced to a nearly solid shell prolonged as
the slightly hollow spatula. The tube at the bottom of the groove
is still there, overhung by the little chitinous shelf.

Near the apex of the organ, (fig. 12) the groove is again open
above, as we have passed beyond the base of the spatula, only the
tip of which is cut, lying well over to the right. This figure being
magnified twice the diameter of the preceding figures, shows
plainly the shelf that cuts off the bottom of the groove. The
median mass is a narrow and nearly solid shell that forms the left
wall of the straight, deep groove. The external mass is the main
part of the section and contains much very watery connective
tissue, covered with epidermis. In this section, the calcified
part of the shell is represented in black, as in the other sections,
while the chitinous or horny parts are dotted. From the surface
this region of the canula looks to be only chitin. Farther on the

JOURNAL OF MOBPHOLOGY, VOL. 22. NO. 2.

256 E. A. ANDREWS

calcified part of the shell fades away and only pure chitinous
matter is left, so that a section at the very tip of the canula, (fig.
13) is only chitin. This view is enlarged four times as much as
the preceding one and shows the disappearance of the superficial
part of the groove though the bottom, which is now close to the
surface, is still overhung by the shelf from the external mass.
That is the tube at the bottom of the groove can now discharge
by a slit to the surface at the tip of the canula; see fig. 5, where the
surface slit of the groove is represented by the black line and the
bottom of the groove, or the tube, is represented by the dotted
line, which comes finally to the surface at the tip as seen in the
section across the level 13. As fig. 5 is of *a right stylet and the
section 13 from a left stylet it shows the parts reversed; the main
bulk of the section is really of the external mass, as in fig. 12.

The specialization of the bottom of the groove had not been
expected till sections revealed it and suggested some special use.
Sections of stylets taken when being used in conjugation soon
showed that the tube at the bottom of the groove is the channel
for the transfer of sperm. Along this minute tube all the sperm
passes from the papilla to the sperm pocket of the female. A
section across the stylet where the median surface bears a tuft
of setae, between the levels of 8 and 9 of fig. 6, when sufficiently
enlarged, shows that the sperm is contained inside the tube of the
groove, as in fig. 14. This shows only the part of the shell about
the tube, with the sharp edge of the shelf above, jutting out to
almost meet the wall of the median mass (see fig. 9). The cavity
of the tube is full of a secretion containing at its centre a pear-
shaped mass of the peculiar sperm of the crayfishes. As was
shown (6) these sperm do not assume the star shape they have in
books as long as they are in the male and not even when in the
sperm receptacle of the female when normally protected from the
water, and in this section, where they are seen in transit, they
are still spherical, clear bodies with the peculiar bowl-shaped
central part that, as represented in the sketch, might be thought
a central nucleus. All along the groove above the orifice there is
thus a strand of sperm surrounded by a paste-like white mass
that fits tightly into the tube.

Fig. 12 Cross section of the same near tip, at level 12, showing the spatula cut
off to the right. 2. A.

Fig. 13 Cross section of the very tip of the stylet, at the level 13, of figs. 5 and
6, showing the groove coming to the surface of the horny canula. 2. D.

Fig. 14 Enlarged view of section across the tubule, between the levels 8 and 9
of fig. 6 showing the sperm cells enveloped in a secretion and shut in by the shelf
above. 2. D.

Fig. 15 Enlarged view of section of tubule and bottom of groove, about the
level 10 of fig. 6 showing the fewer sperm and little secretion in the tubule, sur-
rounded by a thick horny layer. The calcified skeleton is represented as black.
g. D.

257

258 E. A. ANDREWS

That this mass is run in under pressure seems indicated by the
way it tends to flow out at the narrow sUt leading up from the
tube into the groove and by the form of the sperm mass that
tends hkewise to copy the shape of the cavity that is filled, being
pointed toward the slit (fig. 14) . In successive sections this sperm
mass is found all along the length of the groove, always in the
bottom of the tube only, while the enveloping secretion for the
most part disappears. Thus in the fig. 15 from the level 10,
where there are still some secretion tubes coming through the
heavy shell of the median mass, (fig. 10) there are a dozen or so
sperm enclosed in the minute tube together with very little secre-
tion and the sperm seem to come into contact with the shell.
At this level, however, the thick and well-calcified shell (fig. 10)
is covered by a thick layer of horny substance that ntiakes the
shelf and continues on up the face of the external mass bounding
the groove, (fig. 15). The discharge of the milk-white sperm from
the tip of the canula, (figs. 5, 6 and 13) was seen in some males
separated from females in conjugation.

The anatomy of the stylet thus shows it to be a more refined and
specialized tool for sperm transfer than had been expected. It is
essentially a very fine tube receiving sperm at its larger base and
discharging it at its attenuated tip; but it has walls that give it
great strength and rigidity while allowing the tip some elasticity.
Moreover the receiving part of the tube is provided with glands
of problematic value.

In looking for further light upon the nature of this sperm trans-
fer organ we turn to its development in the individual.

ONTOGENY OF STYLET

We find that in Cambarus affinis the first and second larval
stages are externally alike, in both sexes, while the third shows
the male openings on the fifth legs, or the female on the third
legs. In the first stage, there are no abdominal appendages on the
first somite and but a crowding of epidermal nuclei under the
shell where the appendage will be. In the second stage, these
appendages are slight papillae. These indifferent stages are fol-

ORGANS FOR SPERM-TRANSFER 259

lowed by the third, in which the external openings are differen-
tiated but the appendages of the first somite are still simple papil-
lae, alike in both sexes, unless they be longer in the male.

In the fourth stage, which is about 11 mm. long, the pleopods
of the female still are simple papillae but little longer than in the
third stage, while in the male they are long, simple spines, point-
ing toward one another and but slightly forward, as indicated
in fig. 10 p. 127, Andrews, Ontogeny of Annulus, Biol. Bull.,
1906.

The ventral face of the left spine or slightly specialized first
pleopod, of a male 11 mm. long, is seen in fig. 16, magnified 430
in the camera sketch. This is from a larva killed July 1st, from
late spring hatching. The organ is like a club ; it is very simple,
nearly cylindrical and very blunt. It is not jointed, although
there is a faint groove marking off the base from what will be
the neck and spiral.

On the base there is a slight ridge with depressions on the median
side of it. Internally there are two muscles from the base into the
sternum of the abdomen. The distal part of the appendage is
slightly grooved along its ventral face, thus marking off an exter-
nal from a median mass. In cross section, fig. 17, the shell is
not very thick and beneath it is a well formed epidermis with
large nuclei, from which connective tissue strands traverse the
large blood space in which blood corpuscles float. This section
shows the groove on the lower side. The appendage is articulated
to a slightly elevated stump on the sternum that holds one of the
articular muscles and part of the other and ends in an elliptical
orifice into which the base of the stylet fits. This articulation
is so oblique that the stylet lies down and cross-wise towards its
fellow and is but little elevated or directed forward.

In the male of this stage, the openings on the fifth legs are short
slits, not a third of the width of the above simple stylet, and to
each slit there leads a strand of nuclei that represents the efferent
duct.

In males of 15 to 18 mm., in the fifth stage, the stylet (fig. 18)
is about 1 mm. long and is somewhat more specialized. The
base is set off from the terminal part by a more pronounced fur-

260

E. A. ANDREWS

Fig. 16 Posterior face of left first stylet of male 11 mm. long. Enlarged 21S
diameters.

Fig. 17 Cross section of the same stylet. 2. D.

Fig. 18 Posterior face of left stylet of male 18 mm. long. Length of stylet
1 mm. f. A.

Fig. 19 Section of stylet of a male 12 or 15 mm. long. 2. D.

ORGANS FOR SPERM-TRANSFER 261

row, but there is no movable joint. The organ is more pointed
and the groove is very deep from the rising up of its sides. Thus
in section fig. 19, the narrow median mass, {M.m) to the left,
rises high up beyond the groove and the groove itself is a narrow
space between the wide external and the narrow median masses.
In the surface view, (fig. 18) the bottom of the groove is indicated
by the broken line; it is already twisted so that the groove looks
towards the median side along its proximal part and then for a
short distance toward the observer, that is toward the ventral
side, and finally at the tip toward the median side again. Where
the groove is open ventrally the median mass is rising up as a
protuberance that will form the spatula. As yet the canula is
only the spoon-shaped end of the organ.

In a male 22 to 21' mm. long and probably in the sixth stage,
(fig. 20) killed October 4th, we find the same stage as in other
males of this size killed in July, this being an exceptional male that
failed to grow as the average do to be nearly two inches long in
October. Here the spatula is quite evident as a blunt rounded
finger-like elevation that crosses over the groove. As shown by
the dotted line the bottom of the groove is to the right of its mouth
along the proximal part of its course and to the left along under
the base of the spatula; that is, the sinuousness of the groove is
exaggerated by the fact that the sides not only rise up but grow
over the groove, the external mass overhanging toward the medi-
an line proximally and the median mass growing over away from
the median line, distally. The base of the stylet now bears a few
short acicular setae and is provided with three muscles at its at-
tachment to the sternal elevation upon which it stands. By this
time the stylets point forward under the thorax. The canula is
now a short rounded blunt termination of the stylet in which the
groove is no longer widely open but reduced to a slit by the up-
growth of its walls.

In an autumnal male 38 mm. long, (fig. 21) the stylet has be-
come much longer and more modeled but still shows the stiff
joint between the base and the partly-formed neck. The few
setae extend along the ridge of the base on to the proximal part
of the external mass. The median mass sticks out abruptly at

262

E. A. ANDREWS

Fig. 20 Posterior face of left stylet of male 22 mm. long in October. S. A.

Fig. 21 Posterior face of left stylet of male 38 mm. long in October. Enlarged
25 diameters. 2. 90 mm. A.

Fig. 22 Anterior face of left accessory stylet, somewhat turned to show part
of the external face: a view between vii and viii. 2. ao- On the left an en-
larged sectional view of the cup at the end of the radius and the wedge cut off.

ORGANS FOR SPERM TRANSFER 263

the notch, or orifice, and bears a tuft of short setae. The spatula
is long, flat and pointed. The canula is bluntly pointed and
turned outward.

Later when the animal is 64 mm. long, the false joint of the
stylet has disappeared and the tips become more sharp and long.
Even before this size the males are known to conjugate, when
about two inches long.

We thus find that the complex stylet of the adult starts from
a slender papilla that becomes slightly flattened and grooved so
as to form a very clumsy spoon with its depression rather more
median than ventral. Then the sides of this groove grow up and
make the groove into a cleft, which opens as before toward the
median face proximally and distally; but along the middle of its
course is forced to open ventrally and even externally by the over-
hanging growth of the median mass. The organ might be imi-
tated by taking a long strip of clay with a slight length-wise groove
on it and rolling the sides up over the groove, the median side
tending to roll over outside the other. How the shelf from the
external mass first grows out over the groove to cut off its inner
part as a tube was not made out, but it is evidently a secondary
specialization of the shell made by some special activity of the epi-
dermis in a line near the bottom of the groove after the groove has
becoriie deep.

THE SECOND OR ACCESSORY STYLET

The accessory stylets (figs, v-viii) are evidently specializations
of the common type of abdominal appendages, (fig. 26). They
are elevated only when in use in conjugation; and at rest are car-
ried forward under the thorax, horizontally, where they rest upon
the first stylets and are closely packed in with them inside the
special sternal groove of the male thorax.

Figs. V, VI, VII, VIII, represent the left second stylet as seen from
the ventral or posterior, the median, the anterior or dorsal, and
the exterior faces, respectively. Like the unmodified pleopods
this has a basal protopodite, an exopodite, an endopodite. The
exopodite is a slender offset with setae, while the endopodite is

264 E. A. ANDREWS

the complex large part of the appendage that bears a terminal
fiabelum and the remarkable side protuberance, found on no other
limbs, which may be called the triangle.

Describing the entire stylet from the base outward, we see
that the protopodite is chiefly a very strong flattened bony mass
extending diagonally inward so that while the endopodite and ex-
opodite are about parallel to the median line of the animal the
protopodite forms an angle of 45° with it. This makes it possible
for the endopodites of the two stylets to come together at the me-
dian line and for the endopodite of each side to lie upon the groove
of the base of the first stylet, like a lance in its rest, although the
bases of the two second stylets are fastened to the sternum of the
second abdominal somite some distance from the median line.
The protopodite is not entirely one-jointed but at its base is a soft
membrane where it is joined to the sternum and in this are two
large calcified plates, (figs, v and vi) besides two minute ones,
(fig. vii) all of which together make a narrow basal section of the
protopodite. Dissection shows there are muscles passing from
this base of the protopodite into the sternum that may depress
and elevate the appendage.

The protopodite is some 6 mm. long, 2 wide and 1^ thick. The
exopodite is a slender filament some 9 mm. long and ^ mm. thick;
a slightly flattened tapering cylinder set with long setae on Exter-
nal and median face. The setae are really plumose and together
form a sparse brush. The exopodite is obscurely divided into
some twenty segments. The basal 2 mm. is partly calcified, the
rest membranous. It articulates freely with the outer distal
corner of the protopodite so that it may be moved from the posi-
tion of rest parallel to the endopodite, outward through 90° and
swung back and forth some 45°. The tip of the exopodite often
lies dorsally within the cavity or hollow of the triangle, and may
have some use as a cleaning brush.

The endopodite is the stout calcified mass, roughly cylindrical
but flattened from before back, some 9 mm. long on the median
(fig. vi) and 7 mm. on the external face (fig. viii), and bearing at
its distal end a flagellum on the external side and the flat triangle
on the internal side. This bony mass is set on the protopodite

ORGANS FOR SPERM-TRANSFER 265

by a very stiff oblique joint at about 45° and allows of very little
lateral and rotative motion. It may be forced outward and in-
ward through but few degrees, its tip traveHng only 4 mm. It
may be twisted so that the triangle, from being almost concealed
dorsal to the end of the bony mass (fig. v), may be turned outward
a few degrees toward a horizontal position and present more of
its median face, somewhat as in fig. vi. The movement is com-
parable to that of a stiff arm that should allow only a little side-
wise movement and a verj- little twisting at the elbow with the
end result that the triangle, or hand, at the end, accomplishes a
little adjustment to the orifice of the first stylet. This is done as
if by supination, though done by the above twist at the elbow.

The flagellum is the real termination of the endopodite; it is
some 3 mm. long, 1 mm. wide and rapidly tapering, also flattened,
being a long triangular terminal tip to the essentially flat endopo-
dite. By the presence of white lateral areas in the otherwise mem-
branous flagellum, it is obscurely divided into 9 or 12 joints. At
the tip and along the sides it bears long plumose setae that are
often sparse or worn off along the outer side. The flagellum
springs from a socket in the bony shell of the wide end of the endo-
podite. The external angle of the edge of this socket, figs, v
and VIII, forms a hard protuberance at the end of a bony ridge
(the Guide). The setae along the flagellum as well as those
along the exopodite do not stand out horizontally, right and left,
but slant ventrally, or posteriorly, (fig. viii).

The most novel and characteristic part of the second appendage
of male crayfishes is the lateral outgrowth which we will call the
triangle. It is a form of the Decapod appendix masculina of
Boas. The triangle stands up dorsally so that at rest, it, with its
fellow of the other side of the body, fits into the squarish cavity
left between the two necks of the first stylets. It is not well
seen normally from the ventral view, (fig. v) but it may be pulled
outwards through 90° and then looks as in the median view (fig.
vi). It is a flat triangular outgrowth, partly calcified and partly
membranous. The edges are calcified and the centre membran-
ous, so that the whole suggests a bent arm or wing with skin
stretched across it. Each long side of the triangle is about 3 and

266 E. A. ANDREWS

the shorter base about 2 mm. The bony rhns of the triangle as
seen in fig. vi may be called the humerus and the radio-ulna.

The distal free part of the apparatus, ,(figs. vi, vii) is a tri-
hedral mass set with long plumose setae and might be likened to a
sort of hand at the end of the fore-arm. We will call it the wedge
from its appearance and use as seen in sections (fig. 9).

The humerus articulates at each end; proximally loosely with
the side of the exodopodite mass, (fig. vi) ; distally at the elbow,
firmly with the other firm edge of the triangle, the radio-ulna.
On the external or concave face of the triangle, (fig. viii) the hu-
merus is not as well separable from the membranous part of the
triangle, and between its proximal end and the bone of the main
mass of the endopodite there is more or less expanse of membrane.
On this outer face, (fig. viii) we find that all the concave aspect
of the triangle is membranous.

The humerus is wide and smooth and flat on the inner face,
fig. VI, but on the outer face forms only a narrow edge to the mem-
brane, fig. VIII,

The soft hollow face of the triangle in life is swollen with contained
liquid. The soft area is not only the outer face of the triangular
protuberance but also half of the dorsal face of the distal part of
the main trunk of the endopodite.

The whole darkened area of fig. viii might be compared to the
soft inside of the palm of a hand and it is this which comes against
the neck of the first stylet, in conjugation.

While the humerus is wider toward the base and slender at
the elbow end, the radio-ulna is the reverse; that is, it begins nar-
row at the elbow and widens to the hand or terminal part. . The
radio-ulna is a thick plate-like mass that is not in the same plane
as the humerus, but about 45° with it, so that it has the appear-
ance of a scroll rolling in over the depressed membranous outer
face of the triangle, (figs, vii, viii). The radius part is the free
rounded edge, (fig. viii) and this ends abruptly opposite the base
of the hand, which is back of it in the figure, while the ulna plate
runs on continuously in the background of this figure and passes
imperceptibly into the hand, or wedge, (fig. vii).

The radius stands free, away from the membrane, as a rounded

ORGANS FOR SPERM-TRANSFER 267

bony ridge much thicker than the ulna plate from which it is
faintly marked off by a suture. Thus in sections (fig. 8), the
radius looks like a head on a slender neck. The abrupt termina-
tion of the radius is very like the elbow end of the human radius,
a shallow cup. The actual cup is made by clear horny matter of
considerable thickness and is prolonged as a horny sharp ridge all
along the radial edge of the pyramidal wedge. The head of the
radius stands out as wider than the neck (fig. vii).

The ulna is but a vaguely defined thick area of the general
shell and it continues as the hand or wedge, which is, next to the
head of the radius, the most peculiar part of the triangle. This
wedge is a hard horny pyramid of three faces. One is rounded
and setose, two flat, meeting at a sharp edge, (see small sketch,
fig. 22). Its exposed rounded face (figs, vi, vii) is set with a dense
brush of plumose setae. The external or ulnar face (fig. vii) is
smooth bone, bearing setae along its right edge and ending, to the
left, in the sharp horny ridge that runs up from the head of the
radius and is shown as a dark shade in fig. v. The concealed
innermost face is bony and contains orange pigment; along its
left edge it bears setae (fig. viii), and its right edge is the sharp
horny membrane that runs up from the head of the radius. In
the union of this face with the soft membrane of the concavity of
the triangle there is a bony articular plate.

The photographs do not represent one feature of the triangle and
that is the small tuft of some five or six, or so, very wiry bent
plumose setae that spring from the elbow of the triangle and, for
the most part, curve so as to lie down close to the soft membrane.
These setae are roughly shown in fig. 22 at the elbow. This also
gives in the side sketch, an end view of the head of the radius as
seen when the base of the wedge was cut off and the stump of the
ulna and free end of the radius viewed from the face where the
wedge had been. This is intended to show the head of the radius
as a rounded saucer with flat bottom, not deep, but with flaring
and rounded sides that form a rim thicker than the neck of the
radius below. The cut off setae in this figure are the bases of those
on the union of ulna and wedge, just above the level of the line
23 in the main fig. 22.

268 E. A. ANDREWS

INTERNAL ANATOMY OF THE SECOND STYLET

Dissections and sections showed the presence of the same gen-
eral structures as in the case of the first stylet, with the important
difference that the special glands of the tube of the first stylet are
absent and on the other hand the intrinsic muscles that are ab-
sent in the first are well developed in the second stylet. The
muscles are arranged as in the younger stages (figs. 27, 28). Be-
sides the three muscles at the base that pass into the sternum of
the second abdominal somite a very short distance there are long
muscular strands within the stylet itself.

The protopodite springs from a considerable elevation of the
sternum and in the adult two muscles were found within this
elevated articular region. Pulling one tended to depress the sty-
let into its position of rest while the smaller muscle was thought
to be probably concerned with the erection of the stylet. Pull-
ing all the basal muscles made the stylet not only lie down but also
move toward the median line, which would enable it to fit in nicely
with the first stylet. Some of these extrinsic muscles extend a
distance into the protopodite itself, to be attached to the shell.
There are also long strands arising from the shell of the protopo-
dite and running to the exopodite and the endopodite. Those of
the exopodite seem associated with the basal muscles, so that pull-
ing the muscle in the sternum made slight twitching movements
of the exopodite, simulating those seen during conjugation, which
may thus be caused by contractions of the muscles that hold the
entire appendage in position. Pulling the muscles that are in the
distal part of the protopodite made both exopodite and endopo-
dite move dorsally and also away from the median plane.

The muscles that move the exopodite are better developed than
those of the endopodite. Within the exopodite there is a long
intrinsic muscle that would seem fit to bend the slender filament
slightly. Inside the endopodite, beside the slight muscles of the
base concerned with the movement upon the protopodite, there are
in the adult two slight threads that represent the muscle seen in
early stages (figs. 27, 28) passing from the terminal flagellum down

ORGANS FOR SPERM-TRANSFER

269

into the region whence springs the triangle. These muscles are
seen in the sections of the triangle (figs. 23, 24) as two black dots.
These sections, with those in figs. 8, 9, show the anatomy of the
triangle. Fig. 23 is a section along the Une 23 of fig. 22. The
great thickness of the calcified shell is shown by the black mass.
The membranous parts are shown by the thin black, as to the left
in fig. 24. The cavity within is blood space traversed by connec-
tive tissue strands and faced by epidermis against the shell. In

Fig. 23 Cross section of the triangle on the level 23 of fig. 22.
Fig. 24 Cross section of the same about the level 24 of fig. 22.

90 mm. A.
90 mm. A.

fig. 23 the triangle and the distal part of the protopodite are cut
across with the hollow face to the left. The dense shell mass to
the left above is the guide ridge, (fig. viii) which somewhat over-
hangs the cavity of the triangle and bears on its median face some
setae, (fig. vi), which are connected at the root with the epidermis
by the long canals of which one is seen to the left (fig. 23) pene-
trating the shell. Opposite this on the median face of the endo-
podite there are also a few setae which do not appear in the

270 E. A. ANDHEWS

photo^ruph (lig. v) but present one of their canals in the shell of lig.
23, to the right. In contrast to the excessive thickness of the shell
of this main stem of the endopodite, the triangle, as repre-
sented by the lower part of this section, is relatively thin shelled.
The radius is th(^ thick knob in the lower left corner. The shell
to the I'ight is the ulna, the thick mass against the concavity of the
triangle is in reality more membranous than calcified, but as yet
thick. But further toward the elbow (fig. 24) along the hne 24,
(fig. 22), the corresponding region is a thin membrane reaching
from tlu^ n(^ck of the radius across to the thick guide ridge. In
reality the elbow stands out more as in fig. viii so that the width
of the section 24 is nmch greater than fig. 28. Fig. 24 shows
clearly, on the riglit, the hinge-like line of demai-cation between
the outstanding triangle and the main stem of the endopodite,
being in fact cut at the edge of the proximal articulation of the hum-
erus (fig. vi), wliere there is a sudden change in level in passing from
the humerus to the main stem. In sections 23 and 24, the small
black dots above within the connective tissue, are the muscles
that iim up int-o the flagellum, much as hi fig. 28.

Section 8 shows the radius standing out from the flat triangle
with the thick mass of the humerus above in the figure, wliile
fig. 9 shows the thick end of the endopodite above and in the
groove of the first stylet the cut off wedge, as will be described
below in considering the :idjustm(Mits of the first and second stylets
during conjugation.

ON'iXKVKNY OF TITi; A(XM<]SS()HV, Oil Sl'X'ONI), STVLl'^T

Between the individual development of the first and the second
stylets there is this imi)ortant diffeivnce that while the first never
at any time looks like one of the ordinary pleoiK)ds but is of late
appearance and is also a dwarfed, specialized, or reduced append-
age from the first, the second appendage is present as soon as the
others are and is at first like the ordinary appendage and becomes
specialized by the addition of an outgrowth and not by the loss
of parts.

ORGANS FOR SPERM-TRAN8FER 271

The plcopods of the second, third, fourth and fiftli somites
of both males and females are represented at the time of hatch-
ing and all alike have the appearance seen in i\g. 2() whi(;h is mag-
nified 75 diameters and represents the anterior face of the third
left pleopod of a male 18 mm., in July, when in the fifth larval
stage. The pleopod is flat and translucent; the endopodite (En.)
is longer than the exopodite {Ex.) and both are fringed by long
setae that are really plumes, though not so figured. Both endo-
podite and exopodite are obscurely joint(Hl and the proto]x)dite
has a short annular segment as well as a long main segment.
Through the thin shell may be seen the muscles, represented by the
dotted lines. At the base are three large and one minute muscles;
two of the main three are posterior and one anterior, and appar-
ently the movement of the entire appendage would be a more
powerful backward swing and weak forward recovery, as in swim-
ming. Within the main segment of the protopodite are three
long muscles that would seem to aid in bending the appcnulage at
its base, while distally there are two muscles which both go to the
exopodite to move it. The endopodite is left with only intrinsic
muscles to move it at its base and with a long branched muscle
that can act only to bend the endopodite itself. Th(^ exopodite
has also intrinsic muscles at its base as well as i\w musck^s of the
protopodite to move it. There is likewise a long branched muscle
to bend the exopodite.

In the early stages the second appendage of the male is quite like
this third pleopod, but in a male of 21 mm. (probably in the same
larval stage as the male having the third appendage shown in fig.
26) we find th(^ pleopod of the second somite modified as in fig.
27, that is, there has been added to it the excrescence seen on the
median side of the endopodite. This is to become the triangle or
appendix mascuUna of the adult.

The first discovered trace of this outgrowth was seen in a larva
of the fourth stage, 11 mm. long, in July. This first beginning
of the triangle is the slight elevation {x) seen in fig. 25, on the side
of the endopodite. This figure represents only that part of the
endopodite which is not well jointed and forms a sort of base
beyond which is the more flabelliform distal part, (fig. 26). It

jotniNAi, OK Moupiioi.ofiy, vol,. 22, NO. 2

272

E. A. ANDREWS

will be noted that the row of plumes on the right, or median side
of the endopodite (fig. 25), is interrupted distally so that there is
a blank space where one would expect one or two setae, and in
this space there protrudes to the right a rounded elevation. The
position of this slight elevation with reference to the muscles leaves

Fig. 25 Posterior view of basal part of the endopodite of the accessory stylet
of a male 11 mm. long. Enlarged 215 diameters.

no doubt that it is the same thing as the larger elevation of the
next larval stage (fig. 27) . In the preparation the epidermis, not
here shown, grew out to form this elevation as a hollow outgrowth,
leaving no question as to the possible artificial nature of the
bulging of the cuticle shown in fig. 25.

OKGANS FOK SPERM-TRANSFER

273

In the fifth larval stage (fig. 27) the protopodite has become
wider and stouter and the basal part of the endopodite is much
expanded distally where the protuberance arises from it. The

26 27

Fig. 26 Anterior face of third left pleopod of a male 18 mm. long. 2. D.
Fig. 27 Anterior faoe of left accessory stylet of male 21 mm. long. Enlarged
75 diameters.

result is that the exopodite begins to take on that relative insig-
nificance in size, characteristic in the adult accessory stylet.

274 E. A. ANDREWS

The new growth on the median side of the basal part of the endo-
podite (fig. 27), is a sort of knob set on a neck and indined at
about 45° to the axis of the endopodite. Its form is not spherical
but rather more that of a short cylinder on a slightly shorter neck.
The long axis of the cylinder and of the neck is at an angle of 45
degrees to the side of the endopodite. Not only this protruding
knob must be reckoned as part of the future triangle but also the
neighboring widened area of the endopodite which is depressed
as indicated in the shadow in fig. 27 and which will be the de-
pressed anterior face of the future triangle. In fact this depres-
sion is accentuated by the position of the knob, which not only
stands out as represented in the figure but also rises up toward the
observer; that is, anteriorly away from the general plane of the
endopodite. The base of the flabeUiform distal part of the endo-
podite is continued on to the external distal corner of the basal
region of the endopodite as a ridge standing up above the de-
pressed area, and forpiing what will be the guide ridge of the per-
fected organ.

In a small male, 38 mm. long, in October, the second pleopod
had advanced to the state of perfection shown in fig. 28, which is
an external view of a left accessory stylet, which was about three
times as long as the one shown in fig. 27. The muscles in the
protopodite remain as before, though not so well seen from this
point of view, and the same is true of the endopodite and the exo-
podite. The protopodite and the exopodite have grown so large
and massive that the slender exopodite is much subordinated.
The great increase in the basal part of the endopodite, along with
the enlargement and specialization of the triangle, leaves the plu-
mose terminal part of the endopodite as a slender palp-like rem-
nant of the original end of the endopodite. The triangle is now
so much longer at its free edge than at its attached part that it
has the adult triangular form when seen from the median face ; or
more explicitly, the obliquely set cylindrical knob of fig. 27 has
grown so much longer at its free edge than at its attachment that
the length between its ends about equals the distance of the prox-
imal end or elbow from the main mass of the endopodite, which

ORGANS FOR SPURM-TRANSFER

275

29 28

Fis. 28 External face of left accessory stylet of male 38 mm. long in October,
enlarged 25 diameters.

Fig. 29 External view of the united first and second stylets of the left side of
an adult male, 110 mm. long. 2. 90 mm. Oq.

Fig. 30 View of the median face of the same. 2. 90 mm. Cq.

Infigs. 29, 30, 31:
I| = 1st stylet
II = 2nd stylet
5 = crossed fifth leg seen in section

5' = other fifth leg not crossed.
C =canula
Sp = spatula.

276 E. A. ANDREWS

gives the wide scalene triangle as seen from the median side (fig.
vi). The proximal elongation of the cylinder makes the elbow of
the triangle, while the distal elongation has made the pyramid
or wedge that runs up toward the flagellum of the endopodite.
As yet no setae were seen on the wedge. The triangle, however,
is not merely a flat plate that grows out diagonally, but from the
first it is thick through in the anterior-posterior direction, thus
producing the cylindrical edge seen in fig. 27, where the thick edge
is restricted and marked off by a less thick neck; moreover the
thickening of the cylinder is toward the anterior face. By the
stage shown in fig. 28 there is great thickening toward the external
face. Moreover the external free edge of this thickened cylinder
is now itself thickened as a ridge hanging out from the ventral
rim over the depressed area as indicated by the broken line in fig.
28. This rounded thick edge is the future radius. (Compare
figs. 28 and viii.) From this state it is an easy transition to the
more sculptured form of the appendage seen in adults.

The second pleopods of the male thus owe their special struc-
ture to a gradual emphasis of the endopodite and protopodite with
the addition of an outgrowth peculiar to these appendages, the
triangle. The triangle at first is a mere blister on the median
side of the endopodite but soon becomes an oblique plate that is
surmounted by a thickening. The plate grows anteriorly and the
thickening of its free edge becomes longer than the base of the
plate, with a resulting triangular form as seen from the median
face. The thick ridge grows out externally and this extension
itself acquires a thickened rim, posteriorly, which is the radius.

The triangle is thus a triangle only as seen from the median
face of the pleopod, in its entirety the triangle is a curved object
like a half open hand, and as such is capable of being applied to
the rounded surface of the first stylet. It is made of a cylinder
obliquely set along the edge of a plate and curving over it, like
fingers over the palm. A slip of paper if cut of angular form and
bent twice at right angles may be made to represent the stylet.

ORGANS FOR SPERM-TRANSFEK 277

USE OF STYLETS IN CONJUGATION

The way in which the various parts of the stylets are used in
the process of conjugation and sperm transfer has been found out
partly by direct observation, partly by experiment, and partly
by more indirect inferences that still leave some questions un-
answered.

The phenomena of conjugation in general have been described
elsewhere (5) and we will here consider chiefly the use of the sty-
lets. There is a stage in the early part of conjugation, where the
male has seized the female and clasped all her claws, when he
rises up away from her sufficiently to allow the pleopods to swing
back and forth. In this swinging the long stiff stylets and acces-
sory stylets take part and then are soon locked together, after
which the stylets are held by the crossed fifth leg so that hence-
forth they make a rigid mass which cannot be folded down against
the thorax again by any pressure until that fifth leg is removed.
The process of locking together of the stylets is as follows :

The swinging of the pleopods is caused by their basal muscles ;
and likewise the muscles in the bases of the stylets move them
slightly backward, or erect them, and forward, or depress them.
While both first and second generally move together and right
and left alike, they have been seen to move independently. By
a special movement of the second stylets they are clasped against
the first in such a way that the triangle is applied to the neck of
the first stylet. By arching the abdomen, cat-like, the second
stylet is drawn up dorsally along the first, and then, by partial
relaxation of the arch of abdomen, the second is shoved distally,
along the first, while held tight against it; the result is that the
wedge glides along in the groove of the stylet and the radius enters
into the inner tubule through the flaring orifice and is shoved in
so far that it remains fast. In sections (fig. 8) it is seen that radius
fits into the groove as in a socket and, all the walls being thick and
solid, the radius cannot be forced out again without running it
back along the orifice. The fact is that the locking is very firm
and when one tries to pull the second stylet backward the first
is dragged with it and only by pulling the second dorsally toward

278 E. A. ANDREWS

the base of the first can one separate the two, as by that means
the radius is brought to the orifice out of which it readily passes.

When the two stylets have been erected by their own erector
muscles and locked together by their muscular movements which
lead to this mechanical fastening of the edge of the triangle within
the groove, they form one organ, physiologically, which is to
transfer the sperm without any further muscular activity within
it.

The appearance of the two locked organs is indicated in the
somewhat diagrammatic sketches 29, 30. In 29 the external view
of the left stylets and part of the fifth thoracic legs is shown. The
second stylet, to the right of the figure shows the solid tip region
of the endopodite applied closely against the most protuberant
part of the posterior face of the spiral of the first stylet, while
the terminal flabellum runs along parallel to the canula and spatu-
ula. In fact .the tip of the bony endopodite seems to overlap
the contours of the spiral and this is due to the soft nature of the
depressed region of the median face of the end of the endopodite
as is seen in fig. viii. The guide ridge is the part seen external to
the spiral in fig. 29, while the soft surface is squeezed against the
rounded face of the spiral and the triangle is applied close against
the median face of the spiral so that it can be seen only from the
median view.

Turning to the median view we see, (fig. 30) the triangle lying
over the neck and extending out along the groove. The elbow
of the triangle lies over the orifice. The radial edge of the tri-
angle conforms with the obliquity of the groove since both the
wedge and the radius are firmly inserted in the groove.

Figures 29 and 30, show the supporting fifth leg in section, as
a rounded cross-hatched area. It will prevent the locked stylets
from being shoved forward, or closed up against the sternum an-
teriorly. It is also obvious that the movement backward toward
a vertical position will be hindered, not only by the inclination
and rigidity of the basal joint of the first stylet, but by a like join-
ing of the base of the second stylet, since one cannot move back
without the other, for the radius and wedge will go no further

ORGANS FOR SPERM-TRANSFER 279

toward tip of groove. The second forms a mechanical brace
tending to hold the first from going backward.

In order to separate the two the second must move toward the
animal and glide along the first till free from it. And this motion
is actually seen. The locking is not always done without trial
and may be broken and renewed during conjugation, so that
we often see two positions of the stylets, that of perfect locking,

Fig. 31 Same view when the accessory is drawn back into position of recession
showing the papilla at the mouth of the groove.

as in figs. 29 and 30 when the triangle is most advanced toward
the tip of the spiral, and a preliminary and alternate position of
recession when the triangle is applied against the base of the first
style proximal to the orifice. This position of recession is shown
in fig. 31. The triangle goes as far toward the basal end of the
first stylet as possible, till stopped by the knob on the base (fig.
ii). In this recession the orifice with the papilla meeting it, is
exposed and the ventral lip is seen.

It should be borne in mind that the back and forth play of
the triangle on the first stylet is limited not only by the knob
basally and the narrowness of the groove that prevents the radius
from going into it dorsally beyond the position of figure 30, but
it is limited laterally by the fact that the triangles of the two sides

280 E. A. ANDREWS

of the body are in contact and are held together by being placed
in the squarish hole between the necks of the first two stylets.

The two triangles play back and forth like two hands with bent
fingers, back to back, in a narrow space between the first stylets
and, like hands, each runs its palm or soft flat surface along the
median constricted part of the first stylet and the firm guiding
ridge — its thumb, as it were — along the external face of the stylet
(fig. 29). In one case from 3 to 4 seconds were taken to glide the
triangles back from the normal position to the recession (fig. 31) ;
there they remained four or five seconds and advanced strongly
in two seconds. Another recession took 12 seconds, but the ad-
vance occupied 2 seconds.

If we imagine figure v apphed to i, vi to ii and vii to iii,
VIII to IV, we will appreciate how nicely all the surfaces adjust
themselves. The oblique ridge of the external mass of fig. i is
overlaid by the soft depressed area, (figs, viii, 22) so that the
thumb-like guide shows external to the ridge as in fig. 29.

In life the two sets of appendages, right and left, are so closely
applied together that the median face of neither can be seen,
directly, without mutilation experiments on one side, but the pres-
ence of the guide ridge along the external face of the spiral (fig.
29) enables one to judge where the triangle must be at any stage
of advance or recession, a matter of importance in deciding as to
its use in sperm transfer.

That an application of the second, or accessory stylet, to the
first is necessary for the completion of normal conjugation and the
filling of the sperm pocket by transferred sperm, was determined
not only by the above facts of structure and use but by the follow-
ing experiments. The instincts of the male are so strong that,
when in the process of conjugation the second stylet on one side
was cut off, there was no immediate visible effect, except the
escape of some blood from the stump of the appendage. And
when on the next day all the stylets, both first and second, were
cut off, the male seized and turned a female and carried the
conjugation as far as possible in the absence of the organs of
transfer. The instincts thus go on without the means of carrying
them to completion.

ORGANS FOR SPERM-TRANSFER 281

It was then easy to get males to begin conjugation when the
accessory stylets had been removed from both sides. Three
such males made conjugation experiments with several females,
successively, but in no case was there an evidence that the annu-
lus had been filled by these mutilated males, through in one case
the union lasted for eight and one-half hours. In these attempted
conjugations it was not evident how the absence of the second
stylet prevented perfect sperm transfer. In one case the male
let fall three or four sperm masses, or pseudo-spermatophores,
about 1 mm. long on the telson of the female but it was not deter-
mined how this happened. Apparently this was from failure to
have a close union at the orifice, which would lead one to think
the failure due to absence of the triangle that normally holds the
papilla tight to the orifice. But the failure may have been due to
the absence of piston like movements of the radius. More expe-
riments should show what the uses of the different parts of the
triangle really are.

HOW THE SPERM IS FORCED ALONG THE TUBULE OF THE STYLET

The adjustment of the papillae, whose anatomy has been des-
cribed, to the stylets must now be considered in order to appreciate
the final use of the stylet.

As seen in fig. 1, the papilla juts out toward the median plane
so far that it can be placed across the narrowest part of the first
stylet where the notch is (fig. in); that is across the dorsal face
of the first stylet. But its tip turns abruptly inward far enough
to reach along the median face (fig. ii) as far as the orifice, into
which its tip fits. In figs. 30, 31, this position of the papilla is
crudely represented; in reality the tensely swollen translucent
spout is very nicely applied to the rounded faces of the entrance
.to the groove. The papilla is seen in this position when the
triangle is receded (fig. 31) and in the advance of the triangle its
tip becomes concealed, but it doubtless remains as before.

Returning to the actions of the combined stylets which embrace
the papillae we note certain 'tamping' movements. Besides the
advance and recession of the second stylet along the first, the first
and second together when locked, are seen to execute quick jerks

282 E. A. ANDREWS

that carry the tips of the first back and forth a part of a milli-
meter only. \Yhen the tips of the stylets have gained entrance
into the annulus, these thrusts may serve to introduce the tip far-
ther into its cavity. As in the movements of recession the force
here must be exerted by the muscles of the abdomen, as the sty-
lets themselves have no telescopic power; and actual twitching
of the anterior part of the abdomen were seen.

SPERM EMISSION AND CONDUCTION

In normal conjugation nothing is seen of the sperm so that its
transfer from the deferent duct to the cavity of the annulus is a
matter of inference. The papilla is applied to the orifice of the
tube of the first stylet so that it may discharge into it and sections
show the tube full of sperm, (figs. 14, 15) ; moreover in some abnor-
mal cases the sperm is seen to issue from the tip of the canula into
the water, and, as the tip of the canula is normally inside the
sperm pocket, it is evident that the sperm must pass along the
stylet from the papilla. The force that propels the sperm is no
doubt muscular contraction, but it is not clear at first what mus-
cles are concerned; there are none within the first stylet which
acts merely as a passive tube.

From such figures as 2, it is evident that the deferent duct has
powerful transverse muscles that could squeeze out the sperm with
force and this seems the main if not only motive force to carry
thie sperm through the papilla and all along the tube of the stylet
into the annulus.

The force necessary to propel the liquid sperm through a tube
that is only some 20 to 40 ^ in diameter (figs. 13, 15) is great and
attempts to force india ink through the tubule of the stylet with a
small hypodermic syringe failed. When the specially ground can-
ula was inserted into the orifice, while the radius was engaged in
the tubule, no ink could be forced out of the tip of the stylet.
It was inferred that the radius blocked the way, as it fits in so as
to nearly occlude the lumen (fig. 8) , but the same failure was met
with when the triangle was removed from the stylet, but then the
ink jetted out along the proximal part of the groove where the

ORGANS FOR SPERM-TRANSFER 283

triangle had been. Apparently the wedge of the triangle is well
fitted to hold the liquid in the tubule since it fills up the groove
external to it (fig. 9), where the sides of the groove are not as close
together as they are distally (fig. 10), which is beyond the wedge.
When the ink had been introduced into the tubule and not forced
out of the tip of the stylet the triangle was applied to the stjdet
and the radius worked back and forth like a piston in the tubule
with the result that some of the ink issued from the tip' of the can-
ula of the stylet.

This suggested that the radius might act like a piston in normal
sperm transfer and thus propel the sperm from the papilla along
the tubule to the annulus. We also saw that when a pair was
separated in conjugation the sperm that issued from the tip of the
canula of the stylet was mixed with bubbles of air when held out
of the water, which suggested some action at the base of the tubule
(at the orifice) to draw the air into the tubule. However, this
might be movements of the triangle or simply failure of the tri-
angle to hold a tight joint around the tip of the papilla and ori-
fice, for thus air could be drawn in by the stream of sperm ad-
vancing, driven by pressure of the muscles of the deferent duct.
When the radius was inserted into the orifice and shoved along
in the tubule, sperm was forced out of the tip of the canula, which
seemed to demonstrate the ability of the radius to act as a pro-
pelling piston.

We failed to detect any such piston motions during conjuga-
tion, but they would be of very slight extent and not readilj'^
observed. The movements of advance and recession described
above are of a much grosser magnitude than the piston movements
that might be supposed to take place. The movements 31, 30
are only for getting right adjustment of the enveloping triangle
over the papilla tip and the entrance of the radius into the tubule
so that the hand-like triangle may make such tight binding of the
papilla to the orifice that no sperm escapes or comes into contact
with the water. Yet the piston may then presumably be in posi-
tion to advance or recede a little. When we thrust the triangle
strongly so far along the stylet that the elbow was at the orifice,
(fig. 30), the triangle tended to spring slowly back out of the groove

284 E. A. ANDREWS

till only half of the length of the radius remained in the
groove, owing apparently to the elastic side walls of the groove
shoving against the wedge (figs. 9, 10) as these walls are the closer
together toward the tip of the stylet.

By this mechanical means the piston might tend to recede,
while the movements of the muscles of the abdomen might make
the entire second stylet advance enough to shove the piston along
the groove again. We can easily pump the radius back and forth
in the groove by moving the whole second stylet. The muscles
of the abdomen make the slight twitching back and forth jerks
of both first and second stylets above mentioned as tamping
movements. Now after the first stylet, with the second locked
to it, is introduced into the cavity of theannulus as far as possible,
these movements of tamping, if they be continued, could not ad-
vance the first stylet but may push the second further along the
first and so cause the piston to act on the sperm. The dish-like
head of the end of the radius (fig. 22) receives explanation upon
the assumption that it is useful in shoving the sperm along in the
tubule, in fact, the solid bone-like piston with horny cupped tip
provided with elastic flaring edge seems a remarkably well made
apparatus for pushing liquid along in a tube that it fits so well.

Some such piston movements might be expected from the state-
ments of Coste (C. R. 46, 1858), that Gerbe in his laboratory
saw the male Astacus apply the foliacious part of the second sty-
let to the first stylet and by reiterated back and forth motions
during the passage of sperm, keep as he thought, the trough of the
first stylet free from sperm that might harden there else. Schil-
linger, states that the second stylet is used to push the sperma-
tophores out of the first stylet.'

The groove and its concealed inner part that forms the tubule
are of course open to the water and if the sperm is to pass free
from contact with the water to the cavity of the annulus the as-
sumed piston movements of the radius may serve to clean out the
tubule and fill it with harmless secretions. The source of such
secretions may be surmised to be the glands in the tip of the pap-

' As reported by Ortmann in Bronn's Klassen und Ordnungen.

ORGANS FOR SPERM-TRANSFER 285

ilia (fig. 2) or those along the tubule itself (fig. 8). Possibly this
preparatory action of the radius is all that it has to fulfill and that
the pressure of the muscle of the efferent duct is all sufficient
to cause the sperm to run through the length of the stylet. In
connection with this question we have to bear in mind that the
sperm is in some way freed from its envelope of secretion made in
the efferent duct before it is laid away inside the sperm pocket
where it exists pure (1).

This separation of sperm from enveloping secretion takes place
in the tubule of the stylet. In the proximal part of the tubule the
secretion of the deferent duct (fig. 2), is still all around the strand
of sperm (fig. 14), but distally the sperm is almost pure inside the
tubule (fig. 15).

We found also that in one case a male, fallen on the side while
still holding a female, had the stylets only partly erected so that
they were free in the water and from the tip of each canula a very
fine stream of sperm, finer than the tip of the spatula, issued slowly
and coiled up in a small mass. From one canula the sperm then
slowly sank in ten minutes down in the still water as a fine thread
with a coil at the tip. Another male showed faint sperm jelly
on the tip of the flagellum of the endopodite of the second stylet
and this was pure sperm becoming modified by the water; there
was no secretion.

There are however besides these escapes of pure sperm, escapes
of sperm inside of secretions that resemble spermatophores. In
a male, in which the triangle was in the position of recession, (fig
31), there were such white sperm threads, | to 1 mm. long, about
the orifice of the groove. The pseudo-spermatophores that in
abnormal or interrupted conjugations were sometimes seen were
soft, paste-like tubes containing a central mass of sperm. The
short pieces of tube stick by their ends to the inside of a pipette
used to pick them up and to the shell of the crayfish on which they
fall.

The wall of these tubes is a very thin layer of secretion which
is vesiculate and stringy like dough and can be drawn out into
clear threads with minute droplets along them. These would
seem to be not normal spermatophores, which in Astacus have

286 E. A. ANDREWS

thick walls, but only rods of sperm enveloped in some slight se-
cretion from the deferent duct, or possibly that of the papilla
or of the glands of the spiral. The thin walls of these tubes break
open, hernia like, and sperms ooze out.

The separation of the sperm from the secretion of the deferent
duct may be due merely to the diminution in diameter of the tub-
ule; the pressure of the duct would drive the central part of the
current faster than the envelope and thus the central sperm might
flow out of the very narrow tip of the canula and leave the envel-
ope of secretion behind in the wider parts of the tubule. Finally,
when enough sperm had passed along to fill the annulus, the envel-
oping secretion might be forced out and this would make that
wax-like mass that fills the external parts of the annulus and pro-
jects in excess from its mouth as the so-called sperm plug. Pos-
sibly again the piston movements of the radius might come into
play to clean out the secretion from the stylet tubule and ram it
into the annulus. In fact in the last stages of conjugation of one
pair slow and repeated movements of advance and recession of the
triangle were seen which may be interpreted as concerned with
plug making.

The use of the glands of the spiral is not known. Possibly their
secretion cleanses the surfaces to be used in sperm transfer and
aids in keeping water from the sperm. Possibly the secretion
may help the enveloping secretion of the deferent duct to adhere
to the walls of the tubule of the spiral and thus hold it back till
the sperm has passed on into the annulus.

THE RIGHT AND LEFT DUPLICATION OF STYLETS

The striking fact of the exact duplication of both first and sec-
ond stylets right and left suggests questions as to the use of right
and left in conjugation. Are both sides used at each conjugation?

Again the remarkable dimorphism of the females of C. affinis
and of probably all other species of that genus, which expresses
itself in the occurrence of females with the vestibule of the sperm
pocket opening a little to the right of the middle line and of
females with the pocket opening to the left, so that the symmetry

ORGANS FOR SPERM-TRANSFER 287

of the two i.s reversed, raises the question as to whether the
males are adjusted, in habit, to these two kinds of females, so
as to use the right set of sperm transfer organs for a left-handed
female and vice versa, or not.

The crucial experiments to determine whether males actually
use the one stylet for right and the other for left-handed females
have not yet been made. However, some facts and considerations
make it improbable that a male is obliged to do so and indicate
that a male may adjust his stylets so as to use either right or left
on any form of female annulus, leaving the question still open as
to what is the normal habit of the males with reference to the two
forms of females.

In the first place we found that though the two first stylets seem
to be in the annulus they are never both firmly inserted. One is
fixed firmly by its tip while the other may be drawn away by a
pair of forceps. Moreover the one that is inserted has its tip
some f to If mm. in advance of the other and its base is locked
against the base of the other, diagonally, the abdomen being ad-
vanced more on one side than the other.

Observations showed that not only were there cases of the right
stylet in the left annulus but of right stylet in the right annulus
and of left stylet in left annulus and of left stylet in right annulus.
Whether in these cases the sperm was actually transferred was,
unfortunately, not made out. It is possible that a male may in-
sert one stylet and afterwards the other till finally the actual
sperm transfer takes place with some more definite reference to
the symmetry of the annulus than the above observations would
indicate.

That there is any alteration in the advance of the stylets was
not made out, but there is often an alteration in the use of the fifth
leg, right and left. At any one time many males will be found
with the left and others with the right leg crossed, but continuous
observations show that the male will change from right to left
in difficult cases especially, till a better adjustment is obtained.

It was at first thought that there was a relation between the
fifth leg and the advance and use of the first stylet so that these
were on the same side, that is, the stylet being advanced by the

JOURNAL OF MORPHOLOGY, VOL. 22, SO. 2

288 E. A. ANDREWS

use of the leg of that side of the body, but cases were recorded in
which the advanced stylet was on the opposite side from the
crossed leg. Males crossed the right leg with either right or left
stylet advanced and males crossed the left leg with either right .
or left stylet advanced. Here again there is the possible objec-
tion that the condition observed was not permanent or the one
employed in actual sperm transfer. More minute observation
of several normal cases are necessary.

One good case seems, however, rather conclusive. In this a
male, in November, crossed the left fifth and advanced the left
stylet, but after an hour of attempts to enter the annulus, crossed
the right fifth and five hours later the right stylet was one mm. in
advance of the other and the female had a sperm plug in a right
annulus. Here the leg and stylet used did coincide, but the
annulus was not the one to be expected.

Again in some conjugates killed by boiling while united it was
found that in one a right stylet was advanced to a right annulus
and in others a left stylet to a left and to a right annulus.

As far as the evidence goes it gives the impression that the male
is free to use either right or left stylet with either right or left
fifth legs till successful in getting some one tip of the stylets into
the vestibule of the annulus, which may be a right or a left one,
indifferently. Yet future observations may show that the lines
of least resistance are for the male to use the left stylet for the
right-handed female, and the reverse, and that this actually takes
place, in nature as the normal, though we doubt if it be at all neces-
sary. Observations show that both papillae are ready to dis-
charge sperm at the same time and it should be determined by ex-
periment whether the male uses both right and left sets of sperm
transfer organs, alternately, at each conjugation or not.

When the first and the second stylets were cut off from one side
of some sixteen males and, either at once or some weeks after,
these males were given females, the unexpected result followed
that in spite of many repeated attempts, one lasting nine hours,
the numerous conjugations of these unilaterally mutilated males
did not result in any clear cases of successful sperm transfer. In

ORGANS FOR SPERM-TRANSFER 289

many cases the annuli of the females were artificially cleared
out so that any new plugs would have been seen.

Among these cases there were males that alternately used the
fifth left and right legs in crossing, though some had only the left
series of stylets and others the right; the leg being crossed on the
side where there was no stylet and on the side where there was a
stylet. And these same cases were attempting conjugation with
females that were of both kinds, right and' left forms, so that there
was no agreement between the kind of annulus and the fifth leg
used.

In only one case was there any sperm seen and this was seen
twice in successive conjugations of the same male that seems to
have been peculiar. This sperm lay in pseudo spermatophores,
8 mm. long, upon the telson of the female under the left stylet,
and probably escaped from some imperfection of the closure of
the triangle.

While it was not found out why there was this apparent inability
to complete sperm transfer while the stylets of one side were miss-
ing it is thought that this is not due to the need of using sperm
from both sides of the body at each conjugation but rather to the
mechanical factor that the two sets of stylets are always applied
to one another so firmly as to hold the tips of the stylets at the
annulus, so that when one is absent the tip of the remaining one
lacking the usual support cannot be readily brought to the middle
line of the body. Moreover it is possible that the triangle will
not be well applied to the orifice unless the fellow triangle be there
to shove against it, as both are packed in side by side between the
necks of the first stylets.

SUMMARY

Though the sperm of the crayfish, Cambarus affinis, is injured
by exposure to water, it is transferred from the male to the female
under water and stored up in an external pouch.

The part played by the female in this insurance against injury
in transit has been elsewhere described.

290 E. A. ANDREWS

The present paper describes only those organs of the male that
are combined to form a safe conduit for the sperm from the male
to the receptacle on the female.

The actual sperm transit apparatus of the male consists of three
organs on each side of the body. The anatomy and use of these
three organs are here described in detail.

The 'papilla' or end of the deferent duct is provided with glands
and a valve. It is distended by blood and applied to fit accu-
rately to the beginning of a tube.

This tube is the innermost part of the groove of the first stylet,
or limb of the abdomen, and hitherto its existence and use has not
been described.

The first abdominal limb is, in action, a duct leading the sperm
uninterruptedly from the deferent duct into the receptacle of the
female. It contains large glands of problematical use, and relies
for mechanical support upon the habit of the male in using the
second abdominal limb as well as one of the fifth thoracic limbs
to insure the entrance of the first stylet into the receptacle of the
female.

The second stylet is accessory to the first in applying its hand-
like outgrowth over the papilla and insuring a tight joint. It
also gives mechanical support to the first stylet. How much it
may also serve as a piston for cleaning the tube or even for aiding
in sperm transfer is left undecided.

The ontogeny of the first stylet shows that it begins after the
other abdominal limbs and is from the first a simple unbranched
outgrowth which becomes a tube by the depression of its central
and elevation of its lateral parts to form a deep groove, the bottom
of which is ultimately isolated by a shelf.

The morphology of the organ, based upon its use, anatomy and
development, gives the basis for its utilization in defining species
and subgenera. The tip or canula that is inserted into the recep-
tacle to discharge sperm is the real tip of the organ and all other
tips are to be referred to lateral outgrowths from one or the other
side of the original groove.

The ontogeny of the second stylet shows that in the first larva
it is just like the following abdominal limbs; but its subsequent

ORGANS FOR SPERM-TRANSFER 291

fate is to add on a lateral outgrowth {appendix masculina) wliich
becomes the useful part of this organ when acting as a necessary
part of the sperm transit apparatus.

The duplication of all three organs, right and left, seems neces-
sary in, as far as removal of one set leads to the lack of necessary
mechanical support for the perfect functioning of the opposite
set.

The evidence is against the conclusion that the right and left
openings of different receptacles upon different females are neces-
sarily met by the males employing the stylets of one side rather
than an other. In each case the male may by trial obtain the
entrance of some one of the two stylets into the receptacle of the
female.

The extreme solidity of the shell of the stylets is to be correlated
with the amount of force exerted by the male in making a water
tight passage for the sperm from the deferent duct into the recep-
tacle of the female.

While all six organs are necessary for sperm transfer, most of
them may be removed without preventing the males from carry-
ing out many of the stages of conjugation that would normally
lead up to sperm transfer.

Many of the peculiarities of the form and structure of the trans-
fer organs are demonstrated to be of use, or even necessary.

The accurate interadjustment of the six organs is necessary for
the perpetuation of the species.

It is difficult to believe that in the evolution of Cambarus the
increasing perfection of these organs could have been decisive
in eliminating the less perfect organs. Astacus survives with
more simple organs and the majority of genera of crayfish. have no
stylets at all. The perfection of the organs, characteristic of
Cambarus may have been brought about from laws of change
that it will require much experimentation to discover.

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

292 E. A. ANDREWS

, IJTKRATURE CJTP:D

1 Andrews, K. A. U)()B The luumluw vcntralis. Proc. Boston 8oc. Nat. Hist.

vol. 32.

2 1908 The aiinulus of a Mexican crayfish. Biol. Bull. vol. 14.

3 1908 The sperm receptacle of the crayfishes Cambaius cubensis and C. para

doxus. Proc. Wash. Acad. Sci. vol. 10.

i 1904 Breeding habits of crayfish. Am. Nat. vol. 38.

o 1910 Conjugation in the crayfish C'ambarus^affinis. Jour. Exp. Zoo!, vol. 9.

6 1904 Crayfish spermatozoa, .\natoin. .\nz. vol. 2h.

PLATES 1, 2. 3, 4

EXPLANATION OK FIGXTRES

I. Photograph taken with a magnification of about ten diameters, of the
posterior face of the first stylet of the left side.

II. Photograph of the same, taken from the median side, but diagonally, so
that the posterior side is also shown in part.

III. Photograph of the same from the anterior face.
I\'. Photograph of the same from the external face.

V. Photograph taken enlarged about ten diameters, of the second, or accessory
stylet, of the left side of adult male. Posterior face.

VI. The same from the median face.
V'll. The same from the anterior face.
VIII. The same from tiie externa! face.

ORGANS FOR SPERM-TRANSFER

E. A. ANDREWS

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

294

OlUi.VxNS FUR, SPEUM-TJIAN8FKH,

MIEWS

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

295

ORGANS FOR SPERM-TRANSFER

E. A. ANDREWS

JINAI. Ol MOIiPHOI.OCiY, -VOL. T2 , NO

290

• lUJANS FOR SPERM-TRANSFEH

NAL OF MORPHOLOGY, VOL. 22, NO. 2

297

OVIPOSITION INDUCED BY THE MALE IN PIGEONS

WALLACE CRAIG

Department of Philosophy, University of Maiiie

The influence of the male upon the time of oviposition is a mat-
ter in regard to which pigeons differ from some other birds, notably
the domestic fowl. With regard to the fowl I have consulted a
number of poultry keepers and experts, chiefly Dr. Raymond
Pearl and Dr. Frank M. Surface, of the Maine Agricultural Ex-
periment Station, where the most extensive studies of the egg-
laying of fowls have been, and are being carried on. Dr. Pearl
and Dr. Surface tell me that the domestic hen, and also the hen
of the wild Gallus bankiva so far as can be ascertained, commence
their spring laying at an approximately fixed date which can nei-
ther be deferred by withholding the cock nor advanced by giving
the cock before the usual time.

Pigeons differ widely from poultry in this respect. If, from the
winter season onward, an old female piegon be kept unmated and
isolated, she refrains from egg-laying, in evident distress for want
of a mate, until the breeding season is far advanced; at length
she does begin to lay, but her laying without a mate manifestly
partakes of the abnormal. And a virgin pigeon, if kept isolated
from other pigeons, may postpone her laying for a still longer pe-
riod. On the other hand, a female pigeon, young or old, will lay
very early in the season if she be early mated. Moreover, there
is a pretty definite interval between the first copulation and the
laying of the first egg, namely six or seven days; if the egg be de-
layed much beyond this time, the fact indicates some indisposi-
tion on the part of the female. And as the pair rear brood after
brood throughout the season, this time-relation between copula-
tion and egg-laying is regularly repeated.

299

300 WALLACE CRAIG

The utility of this time adjustment in pigeons seems obvious.
The male pigeon takes his turn daily in the duty of incubation:
hence the female must not lay the eggs before he is ready to sit.
This aspect of the matter, which has to do with pigeon sociology,
has already been treated elsewhere (Craig '08) and will be dis-
cussed more fully in a book dealing with pigeon behavior. The
present paper is to show, not why the male should determine the
time of oviposition, but how he does determine it.

The thesis of the present paper is, that the influence of the male
in inducing oviposition is a psychological influence; that the stim-
ulus to oviposition is not the introduction of sperm, for the male
can cause the female to laj^ even though he does not copulate with
her. This is easily proven by an experiment, which requires only
pigeons, patience, and time, and I shall now recount seven repe-
titions of such experiment, the first two being accidental cases,
the other five being trials designed and carried out on purpose to
test the thesis.

Case 1 (1903). In the spring of 1903 I brought together a vir-
gin female dove (individual female no. 7, the species in all these
trials being the blonde ring-dove, Turtur risorius) and a young
inexperienced male, intending simply that they should mate in
the normal manner. The young male played up to the female, but
due to his inexperience and to other causes which need not be dis-
cussed here, his mating behavior was imperfect and he did not
copulate with her. Nevertheless, in due time (six days) she laid
an egg, and a second egg, as usual, forty hours later. This was the
first intimation to me that a male bird can stimulate the female to
lay, without copulating with her. Such an explanation seemed
so absurd at that time that I dismissed it with the assumption that
the birds must have copulated unobserved, and I did not even
test the eggs to see if they were fertile. Looking back on that
case now, however, and considering the observed behavior of that
male, I feel reasonably certain that he did not fertilize the eggs
but simply stimulated oviposition through the psychic (neural)
channels.

Case 2 (1904). A female dove (no. 5) had been kept alone ever
since her mate had died in November, 1903, and as time wore on

OVIPOSITION IN PIGEONS 301

she showed mtense anxiety to mate. She being a very tame bird,
I had often caught and held her gently, but she did not like to be
held, so one day in early March I tried tickling her head and pull-
ing the feathers about her neck somewhat as a courting male would
do it, and, finding that the poor lonely bird received these atten-
tions with intense pleasure and became still more tame, I contin-
ued to preen her neck daily. She now acted toward the hand as
if it were a mate, went through a nesting performance in her seed
dish, there being no nest in her cage, and to my astonishment laid
her eggs in due season. The first egg was laid March 11 and the
second March 13. There is no doubt in my mind that the caress-
ing of this bird's head and neck brought on oviposition. I once
tried to repeat the experiment with another female dove, but she
would not accept the touch of the hand as the former dove had
done. Yet there is other evidence indicating that, with a spe-
cially tamed bird, this experiment, inducing oviposition by the
hand, could be successfully repeated.

This case called to mind that of 1903, and suggested an ex-
periment to determine definitely whether the male dove can stim-
ulate the female to lay, without actual copulation. Opportunity
to try this experiment was not found till 1907 and following years,
when it was planned as follows.

Method of the regular trials

The experiment requires an unmated female dove that is not
laying eggs, preferably a young dove that has never laid. It is
best tried early in the season (e.g., in February), especially if an
old dove be used, for, as said above, if the female is kept too long
without a mate she may lay without one. Side by side with this
female, in a separate cage, is placed an unmated male, and the two
are given several days to become acquainted. When they act to-
ward one another like mate and mate, the doors separating them
are opened and they are allowed to come together for a time, under
constant supervision. When they attempt to copulate, a slender
rod which can be thrust between the bars of the cage is used to
keep them apart. Such attempts are made many times in a day,

302 WALLACE CRAIG

mostly in the afternoon, and are continued for several days in
succession; hence it is best that the experimenter should be able
to devote some hours a day for several days in succession to a
single pair or at most two pairs of birds. Whenever the birds are
not under surveillance they are shut apart, each in his or her own
cage. But they should be allowed to come together daily until
the egg is laid.

A factor which caused difficulty in one of my trials was the nest.
In cases 1, 2, 3 and 6, the bird laid without any nest at all (except
that in case 6 a nest was given just a few hours before the egg was
deposited). But in case 4 {q.v.) the female refused to lay without
a nest : it was then necessary to remove the male and make the trial
again, first giving the female a nest, and waiting long enough to
prove that the nest alone would not cause her to lay.

Results of the regular trials

Case 3 (1907). Female dove, no. 20. This bird had been
bought recently from a dealer, and it was not known whether she
had laid earlier in the season. But she was kept isolated for
some time, during which she showed no inclination to lay. She
was then given a male in the manner indicated. No nest given.

June 9. Male allowed in cage of female, and plays up to her.

June 15. First egg.

June 17. Second egg. (The second egg was of no special in-
terest. After the first egg was laid, I generally left the doors
open, allowing the pair to come together without surveillance.)

Case 4- (1908). Female, the same. She had not laid since the
close of last season. No nest given.

February 4. Male allowed to enter.

The female was unresponsive and showed by her behavior that
this time she was holding back for want of a nest. This deficiency
was supplied in the following manner (vide ut supra.)

February 8. Male taken away to another building.

March 10. Nest put in cage. Female paid practically no at-
tention to it. Many days were allowed to pass, in order to make
sure that the nest alone would not stimulate the female to lay.

OVIPOSITION IN PIGEONS 303

March 21. Male (after short period in sight of female, that
they might become re-acquainted) allowed to enter.

March 27. Egg laid.

Case 5 (1910). Female, the same as in cases 3 and 4. She has
laid no eggs since last season (1909.)

January 20. I begin to allow male in cage, at same time putting
nest in.

January 29. Egg laid.

Case 6 (1908). Female, no. 19. Virgin, has never laid. No
nest given. In this case, the date on which the female was first
given the requisite stimulus cannot be stated so definitely as in
the other cases.

July 12. Male,- in his cage, placed close to cage of female.
Cooing commences. Female so excited that she several times
assumes, and maintains in extreme degree, the copulation posture.

July 14. Male allowed into cage of female, but he fights her,
so that it is necessary to remove him (otherwise the female might
be painfully injured), and to allow the pair a few days more of
preliminary acquaintanceship.

July 18. Male allowed to begin his series of daily visits.

July 22. Egg laid.

Case 7 (1910). Female no. 19, the same as in case 6. She has
laid no eggs since last summer (1909.)

For several days before contact with the male, a nest was kept
in her cage; but she paid no attention to it, showing that the nest
alone would not stimulate her to lay.

January 20. Male allowed to enter.

January 26. Egg laid.

SUMMARY

1. In six cases, stimulation of a female dove by a male, without
copulation, was followed by oviposition; and in one other instance
(case 2), stimulation by the hand of man in imitation of a male
dove was followed by oviposition

2. In six of the seven cases (being all except case 3, in which the
previous history was unknown), it was known that the female

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

304 WALLACE CRAIG

had laid no eggs previously during the current year. In two of
these six cases the dove was a virgin aad had never laid.

3. It is true that the female may, if left without a mate, begin
to lay late in the season. Hence it might be suspected that the
sequence of stimulation and egg-laying in the seven cases was mere
coincidence. But this is precluded, first of course by the fact that
coincidences are not known to happen seven times in succession,
and further by the following considerations.

4. In some of the trials it was proven that the female when stim-
ulated by the male laid much earlier in the season than she did
when not so stimulated. This is shown in the following table.

Female, no. 20.

1908. (Case 4), stimulated by male, laid March 27.

1909. (Control), without male, began to lay May 13.

1910. (Case 5), stimulated by male, laid January 29.
Female, no. 19.

1909. (Control), without male, began to lay April 26.

1910. (Case 7), stimulated by male, laid January 26.

5. The interval between the first stimulation by the male, and
the laying of the first egg, was as follows:

Case 1. 6 days.

Case 2. (Male not used.)

Case 3. 7 days.

Case 4. 6 days.

Case 5. 9 days.

Case 6. 4 to 10 days, depending on what is regarded as the
first stimulation in this case.

Case 7. 6 days.

The average and the variation of these intervals tally closely
with the average and the variation of the interval in normal
breeding, between the first copulation and the laying of the first
egg.

6. There were no exceptions. Ovoposition never failed to fol-
low within nine days after the first contact with the male. (The
onlj^ partial failure was that of the first trial in case 4, which was
due to faulty experimental conditions.)

OVIPOSITION IN PIGEONS 305

CONCLUSION

These facts make it certain that the male dove can stimulate the
female to lay, without copulating with her.

Harper ('04) mentioned the fact that ovulation in the pigeon
does not take place until after the bird is mated, but he was in
doubt as to how far the influence of mating was a 'mental' one and
how far it was a matter of the introduction of sperm. The present
paper goes to show that the stimulus to the whole process of egg
development and laying is a psychic (neural) stimulus, not de-
pendent upon the introduction of sperm.

BIBLIOGRAPHY

Craig, Wallace 1908 The voices of pigeons regarded as a means of social con-
trol. Am. Jour. Sociol., vol. 14, pp. 86-100.

Harper, Eugene Howard 1904 The fertilization and early development of the
pigeon's egg. Am. Jour. Anat., vol. 3, pp. 349-386.

THE ANT-COLONY AS AN ORGANISM'

WILLIAM MORTON WHP]ELER

As a zoologist, reared among what are now rapidly coming to
be regarded as antiquated ideals, I confess to a feeling of great
diffidence in addressing an audience so thoroughly versed in the
very latest as well as the very oldest biological facts, methods and
hypotheses. I feel, indeed, like some village potter who is bring-
ing to the market of the metropolis a pitiable sample of his craft,
a pot of some old-fashioned design, possibly with a concealed
crack which may prevent it from ringing true. Although in what
I have to say, I shall strenuously endeavor to be modern, I can
only beg you, if I fail to come within hailing distance of the advance
guard of present day zoologists, to remember that the range of
adaptability in all organisms, even in zoologists, is very limited.

Under the circumstances, my only hope lies in appealing to our
permanent common biological interests and these, I take it,
must always center in the organism. But the point of view from
which we study this most extraordinary of nature's manifestations,
is continually shifting. Twenty years ago we were captivated by
the morphology of the organism, now its behavior occupies the
foreground of our attention. Once we thought we were seriously
studying biology when we were scrutinizing paraffine sections of
animals and plants or dried specimens mounted on pins or pressed
between layers of blotting paper; now we are sure that we were
studying merely the exuviae of organisms, the effete residua of
the life-process. If the neovitalistic school has done nothing else,
it has jolted us out of this delusion which was gradually taking
possession of our faculties. It is certain that whatever changes
may overtake biology in the future, we must henceforth grapple

1 A lecture prepared for delivery at the Marine Biological Laboratory, Woods
Hole, Mass., August 2, 1910.

307

308 WILLIAM MORTON WHEELER

with the organism as a dynamic agencj' acting in a very complex
and unstable environment. In using the term organism, there-
fore, I shall drop the adjective ' living,' since I do not regard pickled
animals or dried plants as organisms.

As I wish to describe a peculiar type of organism, I may be
asked, before proceeding, to state more concisely what I mean by
an organism. It is obvious that no adequate definition can be
given, because the organism is neither a thing nor a concept, but
a continual flux or process, and hence forever changing and never
completed. As good a formal definition as I can frame is the follow-
ing: An organism is a complex, definitely coordinated and there-
fore individualized system of activities, which are primarily
directed to obtaining and assimilating substances from an envir-
onment, to producing other similar systems, known as offspring,
and to protecting the system itself and usually also its offspring
from disturbances emanating from the environment. The three
fundamental activities enumerated in this definition, namely
nutrition, reproduction and protection seem to have their incep-
tion in what we know, from exclusively subjective experience,
as feelings of hunger, affection and fear respectively.

Biologists long ago constructed an elaborate hierarchy of organ-
isms. Those of a speculative turn of mind, like Spencer and Weis-
mann, postulated the existence of very simple organisms, the
physiological units, or biophores, which, though invisible, were
nevertheless conceived as combining the fundamental activites
above enumerated. These biophores were supposed to form by
aggregation the cells, which may exist as independent organisms
in the Protozoa and Protophyta or unite with other cells to form
more complex aggregates, for which Haeckel's term 'persons'
may be adopted. The person may be merely a cell-aggregate or
consist of complexes of such aggregates as the metameres of the
higher animals, for the separate metameres, according to a very
generally accepted theory, are supposed to be more or less modi-
fied or highly specialized persons. Somewhat similar conditions
are supposed to obtain in the composition of the vascular plants.
The integration both of the metameric and non-metameric Meta-
zoa may proceed still further, the simple persons combining to

THE ANT-COLONY AS AN ORGANISM 309

form colonies in which the persons are primarily nutritive and
acquire fixed and definite spatial relations to one another, whereas
the more specialized animals, like the social insects, may constitute
families of mobile persons with reproduction as the 'Leitmotiv'
of their consociation. In man we have families associating to form
still more complex aggregates, the true societies. Other compre-
hensive organisms are the coenobioses, or more or less definite
consociations of animals and plants of different species, which the
ecologists are endeavoring to analyze. Finally we have philos-
ophers, like Fechner, stepping in with the assertion, that the earth
as a whole is merely a great organism, that the planetary systems
in turn are colonies of earths and suns and that the universe it-
self is to be regarded as one stupendous organism. Thus starting
with the biophore as the smallest and ending with the universe as
the most comprehensive we have a sufficiently magnificent hier-
archy of organisms to satisfy even the most zealous panpsychist.
As biologists we may, for present purposes, lop off and discard the
ends of this series of organisms, the biophores as being purely
hypothetical and the cosmos as involving too many ultrabiological
assumptions. We then have left the following series: first, the
Protozoon or Protophyte, second the simple or non-metameric
person, third the metameric person, fourth the colony of the
nutritive type, fifth the family, or colony of the reproductive type,
sixth the coenobiose, and seventh the true, or human societ}'.
Closer inspection shows that these are sufficiently heterogeneous
when compared with one another and with the personal organism,
which is the prototype of the series, but I believe, nevertheless that
all of them are real organisms and not merely conceptual construc-
tions or analogies. One of them, the insect colony, has interested
me exceedingly, and as I have repeatedly found its treatment as
an organism to yield fruitful results in my studies, I have acquired
the conviction that our biological theories must remain inade-
quate so long as we confine ourselves to the study of the cells and
persons and leave the psychologists, sociologists and metaphy-
sicians to deal with the more complex organisms. Indeed our
failure to cooperate with these investigators in the study of ani-
mal and plant societies has blinded us to many aspects of the

310 WILLIAM MORTON WHEELER

cellular and personal activities with which we are constantly
dealing. This failure, moreover, is largely responsible for our
fear of the psychological and the metaphysical, a fear which be-
comes the more ludicrous from the fact that even our so-called
'exact' sciences smell to heaven with the rankest kind of material-
istic metaphysics.

Leaving these generalities for the present, permit me to present
the evidence for the contention that the animal colony is a true
-organism and not merely the analogue of the person. To make this
evidence as concrete as possible I shall take the ant-colonj^ as a
paradigm and ask you to accept my statement that the colonies
of the termites, social bees and wasps, which the limited time at
my disposal does not permit to consider, will be found to offer the
same and in some cases even more satisfactory data. I select the
ant-colony not only because I am more familiar with its activities,
but because it is much more interesting than that of the polyps,
more typical and less specialized than that of the honey bee, less
generalized than that of the wasps and bumble-bees, and has been
much more thoroughly investigated than the colonies of the sting-
less bees and the termites.

The most general organismal character of the ant-colony is its
individuality. Like the cell or the person, it behaves as a unitary
whole, maintaining its identity in space, resisting dissolution and,
as a general rule, any fusion with other colonies of the same or
alien species. This resistance is very strongly manifested in the
fierce defensive and offensive cooperation of the colonial personnel.
Moreover, every ant-colony has its own peculiar idiosyncrasies
of composition and behavior. This is most clearly seen in the
character of the nest, which bears about the same relation to the
colony that the shell bears to the individual Foraminifer or mol-
lusc. The nest is a unitary structure, built on a definite but
plastic design and through the cooperation of a number of persons.
It not only reflects the idiosyncrasies of these persons individually
and as a whole, but it often has a most interesting adaptive growth
and orientation which may be regarded as a kind of tropism. In
many species the nest mounds, which are used as incubators of the
brood and as sun-parlors for the adult ants, are constructed in

THE ANT-COLONY AS AN ORGANISM 311

such a manner as to utilize the solar radiation to the utmost.
In the Alps and Rocky Mountains we find the nests oriented in such
a manner that the portions in which the brood is reared face south
or east, and as time goes on the nests often grow slowly in these
directions, like plants turning to the light, so that they become
greatly elongated. This orientation is, in fact, so constant in some
species that the Swiss mountaineers, when lost in a fog, can use it
as a compass.

Every complete ant-colony, moreover, has a definite stature
which depends, of course, on the number of its component persons.
And this stature, like that of personal organisms, varies greatlj^
with the species and is not determined exclusively by the amount
of food, but also by the queen mother's fertility, which is constitu-
tional. Certain ants live in affluence but are nevertheless unable
to form colonies of more than fifty or a hundred individuals, while
others, under the same conditions, have a personnel of thousands
or tens of thousands.

One of the most general structural pecuharities of the person
is the duality of its composition as expressed in the germ-plasm on
the one hand and the soma on the other, and the same is true of
the ant-colony, in which the mother queen and the virgin males
and females represent the germ-plasm, or, more accurately speak-
ing, the ' Keimbahn,' while the normally sterile females, or workers
and soldiers, in all their developmental stages, represent the soma.
In discussing the question of the inheritance or non-inheritance of
acquired characters the Neodarwinians trace all the congenital
modifications of the worker and soldier phases to the queen,
just as in the personal organism all the congenital somatic char-
acters are traced to the germ-plasm of the egg. Since the homo-
logue of the reproductive organ of the ant-colony consists of the
virgin males and females, and since the males mature earlier than
the females, the colony may be regarded as a protandric hermaph-
rodite. Some colonies, however — and this is probably charac-
teristic of certain species — produce only males or females and are
therefore in a sense gonochoristic, or dioecious. And this protan-
dric hermaphroditism and gonochorism, like the corresponding
conditions in persons, may be interpreted as a device for, or, at

312 WILLIAM MORTON WHEELER

any rate, as an aid, in insuring cross-fertilization. The fecun-
dated queen of the ant-colony represents the first link in the
'Keimbahn' and therefore corresponds to the fertilized egg of the
personal organism. She produces both the worker personnel and
the virgin males and females, just as the fertilized egg produces
both the soma and the germ-cells. The colonial soma, moreover,
may be differentiated as the result of a physiological division of
labor into two distinct castes, comprising the workers in which the
nutritive and nidificational activities predominate, and the sol-
diers, which are primarily protective. Here, too, the resemblance
to the differentiation of the personal soma into entodermal and
ectodermal tissues can hardly be overlooked.

The structure of the ant-colony thus appears to be very simple
as compared with that of its component persons. The question
naturally arises as to the particular type of unicellular or per-
sonal organism which it most resembles. Undoubtedly, if we
could see it acting in its entirety, the ant-colony would resemble
a gigantic foraminiferous Rhizopod, in which the nest would rep-
resent the shell, the queen the nucleus, the mass of ants the
Plasmodium and the files of workers, which are continually going
in and out of the nest, the pseudopodia.

The ant-colony, of course, like the person, has both an onto-
genetic and a phylogenetic development; the former open to
observation, the latter inferred from the ontogeny, a comparison
of the various species of ants with one another and with allied
Hymenopterous insects, and from the paleontological record.
The fecundated queen, as I have stated, represents the fertilized
egg which produces the colonial organism, but she is a winged and
possibly conscious egg, capable not only of actively disseminating
the species, like the minute eggs of many marine animals, but of
selecting the site for the future colony. After finding this site
she discards her wings and henceforth becomes sedentary like the
wingless workers which she will produce. The whole colony rests
satisfied with the nesting site selected by its queen if the environ-
mental conditions remain relatively constant. If these become
unfavorable, however, the colony will move as a whole to a new
site. In most species such movements are rather limited, but the

THE ANT-COLONY AS AN ORGANISM 313

nomadic driver and legionary ants are almost continually moving
from place to place and must cover a considerable territory during
the year. After the queen has selected the nesting site, she im-
mures herself in some earthen or vegetable cavity, laj^s a number
of eggs, supplying them with yolk derived by metabolism from
her fat-body and now useless wing-muscles, and feeds the hatch-
ing larvae on her salivary secretion, which, though highly nutri-
tious, is, nevertheless, very limited in quantitj^ so that the off-
spring when mature are dwarfed and very few in number. They
are in fact, workers of the smallest and feeblest caste; but they set
to work enlarging the nest, break through the soil or plant tissues,
construct an entrance on the surface and seek food for themselves
and their famished mother. This food enables her to replenish her
fat-body and to produce more eggs. Her expansive instincts
and activities now contract, so to speak, and becoine reduced
henceforth to a perpetual routine of assimilation, metabolism
and oviposition. She produces brood after brood during her long
life which may extend over a period of ten to thirteen years. Her
workers assume the duties of foraging, of feeding the larvae and
one another, and of completing the nest. Their size and poly-
morphism increase with successive broods, till the soldier forms,
if these are characteristic of the species, make their appearance.
Then the individuals which correspond to the reproductive cells
of the personal organism, namely, the virgin males and females
develop, and the colonial organism may be said to have reached
maturity. Like the personal organism, it may persist for thirty
or forty years or,, perhaps, even longer without much growth of
its soma, since the workers and soldiers of which this consists are
exposed to many vicissitudes and live only from three to four
years and probably, as a rule, for a much shorter period. If
the queen grow too old or die the colonj^, as a rule, dwindles and
eventually perishes unless her place is taken by one or more of
her fertile daughters.

This is the ontogenetic history of most ant-colonies. It is so
similar to the phylogenetic history derived from the sources men-
tioned above that we have no hesitation in affirming that it con-
forms in the most striking manner to the biogenetic law. The

314 WILLIAM MORTON WHEELER

very ancient behavior of the soUtary female Hymenopteron is
still reproduced during the incipient stage of colony formation,
just as the unicellular phase of the Metazoon is represented bj^
the egg. A further correspondence of the ontogeny and phylog-
eny is indicated by the fact that the most archaic and primitive
of living ants form small colonies of monomorphic workers
closely resembling the queen, whereas the more recent and most
highly specialized ants produce large colonies of workers not only
verj^ unlike the queen but unhke one another.

In order to complete the foregoing account it will be necessary
to consider some interesting modifications of the usual method
of colony formation and growth, especially as these modifications
furnish additional and striking evidence in favor of the contention
that the ant-colony is a true organism. In many species, after the
colony has reached maturity and especially if the food-supply
continue to be abundant, several of the virgin females may be
fecundated in the nest, lose their wings and remain as members of
the colony. This may, indeed, contain half a dozen and in extreme
cases as many as forty or fifty or even more fertile queens. But
often the growth of the colonial organism becomes excessive through
an increase in the worker personnel and passes over into a form of
colonial reproduction, when the young fertilized queens, each
accompanied by a band of workers, start new nests in the vicinity
of the parental formicary. In this manner a very large and com-
plex colony may arise and extend over many adjacent nests. For
some time the new settlements may remain in communication with
the home-nest through files of workers, but eventually the daugh-
ter settlements may become detached and form independent
colonies. The resemblance of this method of reproduction, which
is essentially the same as the^swarming in the honey-bee, to the
asexual reproduction of many unicellular and multicellular organ-
isms by a process of budding, is too obvious to need further com-
ment.

The important role of nutrition in the development of the
colony will be clear from the foregoing remarks. It becomes even
more striking in the methods adopted by the queens of cer-
tain parasitic species in starting their colonies. Some European

THE ANT-COLONY AS AN ORGANISM 315

observers and myself have found a number of queen-ants that are
unable to found colonies without the aid of workers of allied spe-
cies. These queens may be separated into four groups, as follows:

1. The queen which enters a colony of an alien species and
decapitates its queen or is the occasion of her being killed off by
her own workers. The intrusive queen is then adopted by the
workers and a compound colonial organism arises, consisting of
the germ-plasm of one species and the soma of another. The queen
proceeds to lay eggs, which are reared by the alien workers, thus
relieving her of all the labor and exhaustion endured by the inde-
pendent typical ant-queen during the early stages of colony for-
mation. Pari passu with the development of the worker off-
spring of the intrusive queen, the worker nurses grow old and die,
so that the colony eventually comes to consist of only one species,
the soma of the host being replaced bj^ that of the parasite. This
method of colony formation, first observed among our American
ants and later among certain European and North African species,
I have called temporary social parasitism. Now many of the
species, which behave in this manner, have extremely small queens,
or queens provided with a peculiar pilosity or sculpture that tend
to endear them to the workers of the alien colonies which they
invade. If we regard the large fertilized queens of ordinary ants,
which are supplied with a voluminous fat-body and wing-muscu-
lature, as representing eggs provided with a great amount of yolk,
and the diminutive queens of the temporary social parasites as
the equivalents of alecithal eggs, we have another striking resem-
blance between the personal and colonial organisms, for the large
queens, like the yolk-laden eggs of many vertebrates, are produced
in small numbers but are able to generate the colonial soma inde-
pendently, whereas the small queens, which are produced in
great numbers, in order that some of them may survive the vicissi-
tudes of a parasitic life, correspond to the small yolk-less eggs of
many parasites, which have to be deposited in plant or animal
tissues in order that the imperfect young on hatching may be
surrounded by an abundance of food.

2. The queen of the blood-red slave-maker (Formica san-
guinea) adopts a different method. She enters the colony of an

316 WILLIAM MORTON WHEELER

allied species, snatches up the worker brood and kills any of the
workers or queens that endeavor to dispute her possessions. The
ants hatch with a sense of affiliation with their foster mother and
proceed to rear her eggs and larvae as soon as they appear. Here,
too, the colony is formed by a mixture of two species, but the work-
ers produced by the intrusive queen inherit her predatory instincts
and therefore become slave-makers. They keep on kidnapping
worker larvae and pupse from the nests of the alien species, carry
them home, and eat some of them but permit many to mature, so
that the mixed character of the colony is maintained. This, how-
ever, is not invariably the case, for old and vigorous sanguinea
colonies may cease to make slave-raids and the slaves may die off
and leave a pure colony of the predatory species. The advantages
of this method of colony formation arfe obvious, for the colonial
soma, being composed of two species, grows more rapidly and is
much more efficient as a nutritive and protective support to the
colonial germ-plasm, which is restricted to the predatory species.

3. The colony-founding queen of the amazon ants of the genus
Polyergus resorts to a modification of the method adopted by
sanguinea, as has been shown by Emery's recent observations.
She enters the colony of an alien species, perforates its queen's
head with her sickle-shaped mandibles and permits herself to be
adopted by the workers. She pays no attention to the brood but
begins to lay eggs, the larvae from which are carefully reared by
the workers. The Polyergus offspring inherit the pugnacity of
their mother, but, like the sanguinea workers, have the ability
to kidnap the brood of other ants. They are, in fact, slave-makers
of a very deft and ferocious type. Like their mother, however,
the}' are unable to excavate the nest, to care for their own young
or to take food except from the mouths of the workers that hatch
from the kidnapped larvae and pupae. The mixture of the two
species is therefore obligatory, and the slave personnel, which
represents the nutritive and nest-building portions of the colonial
soma, has to be maintained throughout the life of the colony.

4. Certain feeble queen ants belonging to a few aberrant
genera (Anergates, Wheeleriella) invade populous nests of an alien
species and are adopted in the place of their queens, which are

THE ANT-COLONY AS AN ORGANISM 317

destroyed by their own workers. The parasites then proceed to
lay eggs but these give rise only to males and females as the
worker caste is entirely suppressed. The colony retains a mixed
character, the parasitic species usurping the functions of the germ-
plasm, while the host is purely somatic. As there are no means
of prolonging the lives of the host-workers and as they do not re-
produce, the whole colony is short-lived and the maturation of the
parasitic sexual individuals has to be accelerated so that it will
fall within the brief life-time of the worker hosts. This condition
I have called permanent social parasitism.

These four peculiar types of colony-formation all lead to
the formation of compoand colonial organisms, comparable to
certain compound personal organisms which, with few exceptions,
can be produced only by artificial means. In temporary social
parasitism the colonial egg can develop its soma only when grafted
on to the soma of another species. This soma eventually perishes
and the colony then assumes a normal complexion. This condi-
tion reminds us of certain tropical plants, like the species of Clusia
and Ficus, which develop as epiphytes on other trees but after
killing their hosts take root in the soil and thenceforth grow as in-
dependent organisms. The slave-makers of the sanguinea or
facultative type are also unable to develop the soma except when
grafted on to the soma of another species, but in this case the co-
operation of both somas in nourishing and protecting the germ-
plasm is maintained for a much longer period. This kind of colony
may be compared with a graft made by uniting the longitudinal
half of one plant with that of another so that both take nourish-
ment through their roots. To make the resemblance more com-
plete one of the grafted halves would have to be pruned in such a
manner as to prevent flowering. In the amazons or obligatory
slave-makers and the permanent social parasites the alien soma
alone has a nutritive function, so that the conditions are like those
in ordinary vegetable grafts, in which the stock retains the roots
and the scion produces the flowers and fruit.

I have dwelt on the various methods of colony formation not
only because they give us an insight into colonial reproduction,
but because they throw light on the colonial organism from the

318 WILLIAM MORTON WHEELER

standpoint of parasitology. That the four types of queens and
their offspring are directly comparable with entoparasitic persons
is not so remarkable as the fact that in ants the host and para-
site form a mixed organism which could only be obtained
with persons by jumbling together the component cells of host
and parasite like two kinds of peas shaken in a bottle. Notwith-
standing this mixture the parasitic colony not only retains its
identity and the anticipatory character of its behavior but cas-
trates the host colony and constrains its soma either to cooperate
in many of its activities or to specialize as a purely nutritive or
nest-building auxiliary. The host is thus reduced to the status
of a nourishing or protective organ of the parasite. This behavior
has many striking analogies among persons. Giard long ago
called attention to the fact that when the cirriped Sacculina
settles under the abdomen of a male crab and sends its rootlike
haustoria into the tissues of its host, the latter undergoes cas-
tration, and its narrow abdomen expands to form a protection
for the soft-bodied parasite. In other words, the parasite acts
as if it were a mass of crabs' eggs and the male crab behaves as
if it had changed its sex and develops an abdomen of the female
type.

Not only are there ants, like those already considered, that may
be regarded as colonial entoparasites, but there are also a number
of species that may be called colonial ectoparasites. These form
the so-called 'compound nests,' in which two or more species
live amicably side by side, or may even mingle freely with one
another, but rear their broods in separate nests, thus indicating
in the clearest manner the integrity of the colonial organism. This
is also shown by the vast number of myrmecophilous insects,
which are, of course, ento- or ectoparasitic persons, and behave
towards the ant colony as if it were a rather incoherent and there-
fore more vulnerable, or exploitable personal organism.

Finally we come to what the neovitalists regard as the most
striking autonomic manifestations of the organism, namely the
regulations and restitutions, and face the question as to whether
these, too, have their counterpart in the colonial organism. I
believe that the following facts compel us to answer this ques-

THE ANT-COLONY AS AN ORGANISM 319

tion in the affinnative. If the worker personnel be removed from
a young ant-colony, leaving only the fertile queen, we find that
this insect, if provided with a sufficiently voluminous fat-body,
will set to work and rear another brood, or, in other words, re-
generate the missing soma. And, of course, any portion of the
worker or sexual personnel, that is removed from a vigorous colony
will be readily replaced by development of a corresponding portion
of the brood. On the other hand, if the queen alone be removed,
one of the workers will often develop its ovaries and take on the
egg-laying function of the queen. In ants such substitution
queens, or gynaecoid workers are not fertilized and are therefore
unable to assume their mother's worker- and queen-producing
functions. The termites, however, show a remarkable provision
for restituting both of the fertile parents of the colony from the
so-called complemental males and females. In ants we have a
production of fertile from normally infertile individuals, but the
incompleteness of the result does not disprove the existence of a
pronounced restitutional tendency.

Very striking examples of this tendency are exhibited when
colonies are injured by parasitic myrmecophiles. I shall consider
only the case of the peculiar beetle Lomechusa strumosa, which
breeds in colonies of the blood-red slave-maker (Formica san-
guinea) . Though the beetle and its larvse are treated with great
affection, the latter devour the ant larvse in great numbers,
so that little of the brood survives during the esirlj sumAier months
when the colony is producing its greatest annual increment to the
worker personnel. The ants seem to perceive this defect and en-
deavor to remedy it by converting all the surviving queen larvse
into workers. But as these larvse have passed the stage in their
development when such an operation can be successful, the result
is the production of a lot of pseudogynes, or abortive creatures
structurally intermediate between the workers and queens and
therefore useless in either capacity. It is instructive to com-
pare this case with the regeneration of the lens from the iris in the
Amphibian eye. In his recent analysis of the stimuli of restitu-
tion in personal organisms Driesch reaches the conclusion that
"the specificity of what is taken away certainly forms part of the

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

320 WILLIAM MORTON WHEELEK

stimulus we are searching for, and it does so by being communi-
cated in some way by something that has relations to many, if
not all, parts of the organism and not only to the neighboring
ones." He also says that "each part of the organism assigns its
specific share to an unknown something and that this something
is altered as soon as a part is removed or absolutely stopped in
its functional life, and that the specific alteration of the something
is our stimulus of restitutions." These quotations and Driesch's
further discussion of the problem are even clearer in their applica-
tion to the colonial than to the personal organism, for in the
former it is much easier to see how each individual insect "can
do more than one thing in the service of restitution" than it is
to understand how each cell of the person can do more than one
thing in restoring a lost organ.

I fear that I may have wearied you with this long attempt to
prove that the ant-colony is a true organism, especially as this
statement must seem to some of you to be too trite for discussion,
but when an author like Driesch writes a large work in two volumes
on the "Philosophy of the Organism" and ignores the colonial
organisms altogether, an old-fashioned zoologist may perhaps be
pardoned for calling attention to a well-founded, though some-
what thread-bare, biological conception.

If it be granted that the ant-colony and those of the other social
insects are organisms, we are still confronted with the formidable
question as to what regulates the anticipatory cooperation, or
synergy of the colonial personnel and determines its unitary and
individualized course. The resemblance of the ant- or bee-colony
to the human state long ago suggested a naive reply to this ques-
tion. Aristotle naturally supposed the colonial activities to be
directed and regulated by a ^aaCkebs or riyeixuv, because these
personages managed affairs in the Greek states. After the sex
of the fertile individual had been discovered by Swammerdam,
the word 'queen' was naturally substituted for (SaatXevs or 'king/
and as queens in human states do not necessarily govern and are
often rather anabolic, sedentary and prolific persons and the
objects of much flattering attention, the term is not altogether
inapt when applied to the fertile females of insect colonies. It

THE ANT-COLONY AS AN ORGANISM 321

has been retained although everybody knows that these colonies
represent a form of society very different from our own, a kind of
communistic anarchy, in which there is '' neither guide, overseer
nor ruler," as Solomon correctly observed. In this respect too, the
colony is essentially the same as the personal organism, at least
in the opinion of those who do not feel compelled to assume the
existence of a 'soul' in the scholastic sense. For it is clear, that to
primitive thinkers the soul was supposed to bear the same rela-
tion to the person as the iSaatXevs to the insect colony and the
king to the human state. This supposition is still held though in
a more subtle form, by writers of the present day. Some of these,
like Maeterlinck, clothe the postulated controlling agency in a
mystical or poetic garb and call it the 'spirit of the hive.' The
following passage from the Belgian poet's charming account of
the honey-bee will serve to illustrate this method of meeting the
problem:

What is this 'spirit of the hive' — where does it reside? It is not like
the special instinct that teaches the bird to construct its well planned
nest, and then seek other skies when the day for migration returns.
Nor is it a kind of mechanical habit of the race, or blind craving for life,
that will fling the bees upon any wild hazard the moment an unfore-
seen event shall derange the accustomed order of phenomena. On the
contrary, be the event never so masterful, the 'spirit of the hive' still
will follow it, step by step, like an alert and quickwitted slave, who is
able to derive advantage even from his master's most dangerous orders.

It disposes pitilessly of the wealth and the happiness, the Hberty and
life, of all this winged people; and yet with discretion, as though governed
itself by some great duty. It regulates day by day the number of births,
and contrives that these shall strictly accord with the number of flowers
that brighten the country-side. It decrees the queen's deposition or
warns her that she must depart; it compels her to bring her own rivals into
the world, and rears them royally, protecting them from their mother's
political hatred. So, too, in accordance with the generosity of the flowers,
the age of the spring, and the probable dangers of the nuptial flight
will it permit or forbid the first-born of the virgin princesses to slay in
their cradles her younger sisters, who are singing the song of the queens.
At other times, when the season wanes, and flowery hours grow shorter,
it will command the workers themselves to slaughter the whole imperial

322 WILLIAM MORTON WHEELER

brood, that the era of revolutions may close, and work become the sole
object of all. The 'spirit of the hive' is prudent and thrifty, but by no
means parsimonious. And thus, aware, it would seem, that nature's
laws are somewhat wild and extravagant in all that pertains to love, it
tolerates, during summer days of abundance, the embarrassing presence
in the hive of three or four hundred males, from whose ranks the queen
about to be born shall select her lover; three or four hundred foolish,
clumsy, useless, noisy creatures, who are pretentious, gluttonous, dirty,
coarse, totally and scandalously idle, insatiable, and enormous.

But after the queen's impregnation, when flowers begin to close sooner
and open later, the spirit one morning will coldly decree the simultaneous
and general massacre of every male. It regulates the worker;^' labours
with due regard to their age; it allots their task to the nurses who tend
the nymphs and the larvae, the ladies of honour who wait on the queen
and never allow her out of their sight ; the house-bees who air, refresh, or
heat the hive by fanning their wings, and hasten the evaporation of the
honey that may be too highly charged with water ; the architects, masons,
wax- workers, and sculptors who form the chain and construct the combs ;
the foragers who sally forth to the flowers in search of the nectar that
turns into honey, of the pollen that feeds the nymphs and the larvae,
the propohs that welds and strengthens the buildings of the city, or the
water and salt required by the youth of the nation. Its orders have gone
to the chemists who ensure the preservation of the honey by letting a
drop of formic acid fall in from the end of their sting; to the capsule
makers who seal down the cells when the treasure is ripe, to the sweepers
who maintain public places and streets most irreproachably clean, to the
bearers whose duty it is to remove the corpses ; and to the amazons of the
guard who keep watch on the threshold by night and by day, question
comers and goers, recognize the novices who return from their very first
flight, scare away vagabonds, marauders and loiterers, expel all intruders,
attack redoubtable foes in a body, and, if need be, barricade the en-
trance.

Finally, it is the spirit of the hive that fixes the hour of the great annual
sacrifice to the genius of the race: the hour, that is, of the swarm; when
we find a whole people, who have attained the topmost pinnacle of pros-
perity and power, suddenly abandoning to the generation to come their
wealth and their palaces, their homes and the fruits of their labour;
themselves content to encounter the hardships and perils of a new and
distant country. This act, be it conscious or not, undoubtedly passes the
fimits of human morahty. Its result will sometimes be ruin, but poverty

THE ANT-COLONY AS AN ORGANISM 323

always; and the thrice-happy city is s'cattered abroad in obedience to a
law superior to its own happiness. Where has this law been decreed
which, as we soon shall find, is by no means as blind and inevitable as
one might believe? Where, in what assembly, what council, what in-
tellectual amd moral sphere, does this spirit reside to whom all must
submit, itself being vassal to an heroic duty, to an intelligence whose eyes
are persistently fixed on the future?

It comes to pass with the bees as with most of the things in this world;
we remark some few of their habits; we say they do this, they work in
such and such fashion, their queens are born thus, their workers are
virgin, the}^ swarm at a certain time. And then we imagine we know
them, and ask nothing more. We watch them hasten from flower to
flower, we see the constant agitation within the hive; their life seems very
simple to us, and bounded, like every life, by the instinctive cares of
reproduction and nourishment. But let the eye draw near, and endeav-
our to see ; and at once the least phenomenon of all becomes overpower-
ingly complex; we are confronted by the enigma of intellect, of destiny,
will, aim, means, causes; the incomprehensible organization of the most
insignificant act of life.

Other authors like Driesch, give the postulated controlling
agency the sharper outlines of a would-be scientific but in reality
metaphysical entity and call it the 'entelechy.' It is true that
the entelechy is deduced by Driesch from the autonomic peculiari-
ties of the personal organism, but as the colony has all the essen-
tial attributes of the organism, he would undoubtedly assign it an
entelechy, which according to the definition would have to be
nonspacial, but working into space, nonspsychic, but conceivable
only after analogy with the psychic, and non-energetic, but never-
theless capable of determining the specificity of the colonial
activities through releasing and distributing energy.

I confess that I find the entelechy quite as useless an aid in
unravelling the complex activities of the ant-colony as others have
found it in analyzing the personal organism. This angel-child,
entelechy, comes, to be sure, of most distinguished antecedents,
having been mothered by the Platonic idea, fathered by the Kant-
ian Ding-an-sich, suckled at the breast of the scholastic forma
substantialis and christened, from a strong family likeness, after
old Aristotle's darhng evTeXexeta, but nevertheless, I believe that

324 WILLIAM MORTON WHEELER

we ought not to let it play about in our laboratories, not because
it would occupy any space or interfere with our apparatus, but
because it might distract us from the serious work in hand. I
am quite willing to see it spanked and sent back to the metaphys-
ical house-hold.

But, speaking seriously, it seems to me that if the organism be
inexplicable on purely biological grounds, we should do better to
resort to psychological agencies like consciousness and the will.
These have at least the value which attaches to the most imme-
diate experience. And even the subconscious and the super-
conscious are more serviceable as explanations than such anaemic
metaphysical abstractions as the entelechy. Of course, psychic
vitalism is one of Driesch's pet aversions and he will have none of
it, because he is a solipsist, but the fact that he is compelled to
operate with a 'psychoid' and with an entelechy conceivable only
jper analogiam with the psychic, shows the inconsistency of his
position.

Before we can adopt any ultrabiological agencies, however,
except in a tentative and provisional manner, an old and very
knotty problem will have to be more thoroughly elucidated. I
refer to the problem of the correlation and cooperation of parts.
If the cell is a colony of lower physiological units, or biophores,
as some cytologists believe, we must face the fact that all organisms
are colonical or social and that one of the fundamental tendencies
of life is sociogenic. Every organism manifests a strong predelec-
tion for seeking out other organisms and either assimilating them
or cooperating with them to form a more comprehensive and effi-
cient individual. Whether, with the mechanists, we attribute this
tendency to chemotropism or cytotropism, or with the psychic
neovitalists, interpret it as conscious and voluntary, we certainly
cannot afford to ignore the facts. The study of the ontogeny of the
person, i.e., the person in the process of making, in the hands of
recent experimentalists, has thrown a flood of light on the pecu-
liarities of organization, but the animal and plant colony are in
certain respects more accessible to observation and experiment,
because the component individuals bear such loose spacial rela-
tions to one another. Then too, the much simpler and more primi-

THE ANT-COLONY AS AN ORGANISM 325

tive organismal type of the colony, as compared with that of the
person, should enable us to follow the process of consociation and
the resulting physiological division of labor more successfully. In
the problem, as thus conceived, we must include, not only the
true colony and society, and the innumerable cases of symbiosis,
parasitism and coenobiosis, but also the consociation and mutual
modification of hereditary tendencies in parthenogenetic and
biparental plants and animals, since in all of these phenomena our
attention is arrested not so much by the struggle for existence,
which used to be painted in such lurid colors, as by the ability
of the organism to temporize and compromise with other organ-
isms, to inhibit certain activities of the aequipotential unit in the
interests of the unit itself and of other organisms ; in a word, to
secure survival through a kind of egoistic altruism. ^

2 Since this paragraph was written I have found that several recent authors
have given more explicit expression to a very similar conception to the role of
cooperation and struggle in the development of organisms. Especially worthy
of mention in this connection are Kammerer (Allgemeine Symbiose und Kampf
urns Dasein als gleichberechtigte Triebkrafte der Evolution. Arch. f. Rass. u.
Ges.-Biol.6, 1909, pp. 585-608), Schiefferdecker (Symbiose. Sitzb. niederrhein.
Ges. f. Natur. u. Heilk. zu Bonn, 13, Juni, 1904, 11 pp.), Bolsche (Daseinskampf
und gegeuseitige Hilfe in der Entwicklung. Kosmos, 6, '1909); and Kropotkin
(Mutual aid, a factor of evolution, London, 1902).

SEXUAL ACTIVITIES OF THE SQUID, LOLIGO
PEALII (LES.)

I. COPULATION, EGG-LAYING AND FERTILIZATION

OILMAN A. DREW

From the University of Maine, Orono, Maine

THIRTEEN FIGURES
FOUR PLATES

This account, which deals with some of the sexual activities
of the squid, is based upon observation made on specimens kept in
glass sided aquaria at the Marine Biological Laboratory, Woods
Hole, Mass. Specimens caught in the fish traps of the immediate
vicinity may, by careful handling, be kept in aquaria in fairly
good condition for a number of days. Such specimens occasion-
ally copulate and eggs are sometimes laid.

There are two methods of copulation. By one method the sper-
matophores ejaculate their contents so the sperm reservoirs thrown
from them are attached in a special depression on the inner side
of the outer buccal membrane opposite the junction of the two
ventral arms (figs. 8 and 10). They then slowly emit sperm,
which are carried to and stored in, a special sperm receptacle
that opens near this depression and is imbedded in the tissue of
the outer buccal membrane (figs. 10 and 11). In this receptacle
the sperm are mixed with a secretion and are not active. How
long the sperm may be retained in the receptacle is not known,
but there is some reason to think that they may be retained for
at least some weeks. Females with eggs that can be fertilized
may be found during the four months, June to late September,
that I have worked at Woods Hole. Without exception every
adult female that had not spawned had the sperm receptacle filled
more or less completely with sperm, although in many cases the

327

328 OILMAN A. DREW

eggs were far from mature. This, together with the dormant
condition of the sperm in the receptacle, and the fact that they
seem to be poured out only during egg laying, point to a possible
long retention. It is certain that the same female may have
sperm reservoirs attached near this receptacle a number of times
after it has been filled, and it is possible that the same sperm do
not continue long in the receptacle. There seems, however, to
be no evidence that they are discharged except during the period
of egg laying.

The other method of copulation results in fastening the sperm
reservoirs of the ejaculated spermatophores near the end of the
oviduct (fig. 8, s) usually directly on its walls but sometimes on
the mantle, gill or visceral mass. There is no special receptacle
for the sperm from these sperm reservoirs. They escape into the
water, becoming active as they escape, and pass out with the water
through the funnel. The escape of the sperm is rather rapid but
there are vast numbers in each reservoir, from which they are
constantly poured like smoke from a chimney until the reservoir
is empty. It is not known how long it takes to empty a reser-
voir but by keeping reservoirs from spermatophores that ejacu-
lated in dishes of sea-water, and by examining reservoirs normally
attached to the oviducts and buccal membranes of females, it
seems probable that the sperm do not all escape for two or more
days.

In aquaria I have seen rather more cases of copulation where
the spermatophores are inserted into the mantle chamber than
where the sperm reservoirs are attached to the buccal membrane.
This may be because of the limited quarters in aquaria. In the
larger floating tanks, in which specimens are sometimes kept be-
fore they are brought into the laboratory, the buccal membrane
copulation seems proportionally more common than in aquaria,
but even here the mantle chamber copulation seems to be rather
more frequent.

The same individuals may copulate several times in the course
of a few hours. In general the male is aggressive. The female
may attempt to escape or she may be quite passive. Spermato-
phores seem to be inserted in the mantle chambers of only those

SEXUAL ACTIVITIES OF THE SQUID 329

females that are nearly ready to deposit their eggs. In the large
number of trials made it was found that the eggs of these individ-
uals were so nearly mature they could be artificially fertilized.
Females that are nearly ready to deposit eggs have the nidamental
glands considerably swollen and the accessory nidamental glands
are highly colored with bright red. Wherever the spermatophores
were inserted in the mantle chamber these glands were in this con-
dition.

Before copulation both female and male are usually especially
active and may be known as sexually excited animals by their
peculiar movements. The female in swimming seems to be ner-
vous or excited. She throws short but rapid puffs of water from
the funnel, moves the tail fin very rapidly and, leaving the arms
quite limp, spreads them apart and frequently throws them to one
side. This gives the arms a jerky or trembling motion not shown
in ordinary swimming. Except during the most rapid movements
of the female, the male solemnly swims by her side, an inch or two
away, but parallel, and with his head in the same direction. He
frequently manipulates his arms, spreading them apart, commonly
with the two dorsal arms elevated nearly or quite to a perpendic-
ular position, and the third arms spread far to the sides (fig. 3).
This position is not infrequently accompanied by localized activ-
ity of chromatophores. A spot may appear near the base of each
third arm and a smaller spot on each second arm a little further
from its base. These spots do not remain continuously while
the male is in this attitude but suddenly appear with each increase
of activity on the part of either the male or female. Occasion-
ally blushing is quite general over the head and anterior end of the
body and sometimes includes the whole body but the bodies of
both animals generally remain colorless except for the special
spots mentioned on the male. The attitude of the male, with ele-
vated and spread arms, is not continuous but is assumed every
few minutes, or in some cases seconds, and the arms may be brought
into the usual position of a swimming animal for periods of many
minutes.

Males do not all respond equally to the presence of sexually active
females. Not uncommonly one male in an aquarium containing

330 OILMAN A. DREW

several males will follow the females around by the hour while
the other males remain entirely inattentive. Usually when a
male begins to show sexual activity he will follow a single female
although other females that show similar activities are present
in the aquarium. Occasionally he may change to another indi-
vidual but he nearly always returns after a few minutes to the
one to which he has been paying chief attention.

A few males have been observed that were so sexually excited
they followed individuals around quite indiscriminately. Under
such conditions I have upon three occasions seen a male catch
another male and insert spermatophores into his mantle chamber.
Two of the three instances were between the same individuals, the
second performance being only a few minutes after the first. In
each of these cases the male seized made great efforts to get away
and finally to get hold of the male that was holding him but was
unsuccessful. Upon killing the male that received the spermato-
phores, sperm reservoirs were found attached to the base of the
left gill and to the adjacent visceral mass. Such exceptionally
active males may copulate repeatedly with a single female. In a
few cases this has been carried so far that the female has actually
been killed. Even after the female has beco-me entirely inactive
and apparently dead the male may copulate with her several times.
In one case, a male that had been several days without food, after
copulating with a weakened female, retained his hold and killed
her by eating a considerable hole through the mantle.

The male always uses the same arm for transferring the sperm-
atophores. This arm, the left ventral, is not greatly modified,
but a short distance from its tip some of the suckers, especially
those in the row farthest from the midline of the body, and a ridge
between the rows of suckers show modification (fig. 4, h). The
peduncles of a dozen or more of the suckers of the outer row are
considerably elongated and the sucking discs of a few, (six or eight)
are greatly reduced in size or entirely absent. In both directions
from these, the discs become increasingly normal until no modi-
fication is apparent. The suckers of the row toward the midline
of the body are somewhat modified, the peduncles being somewhat
shorter than those of the other suckers in the row, and the suck-

SEXUAL ACTIVITIES OF THE SQUID 331

ing discs somewhat smaller, but in none of the suckers of this row
are the sucking discs entirely absent. A glandular plaited ridge
extends lengthwise between the suckers of this region and gives
off branches that join each of the peduncles. This ridge is highest
and broadest opposite the suckers that are most modified and grad-
ually disappears as the suckers become normal. At its highest
point it hag about the same elevation as the shortest modified
suckers, which are adjacent. Sections of the modified portion of
the arm show that the ridge and suckers mentioned are covered by
a thick columnar epithelium that stains deeply. Many of these
epithelial cells are filled with large rounded granules that stain
with eosin. The cells that cover other portions of the arm are
flattened or cubical, do not stain very deeply, and do not contain
granules. It seems probable that the cells of the hectocotylized
region secrete a substance that aids the arm in holding the sperm-
atophores. The modified suckers probably make the bending
and grasping necessary for the transfer of the spermatophores
more easily accomplished.

The positions of the animals during copulation are rather hard
to determine as the whole process generally does not occupy more
than ten seconds and during this time the animals are usually
swimming and the arms are changing positions, but by carefully
focusing attention during different acts upon first one arm and
then another, the positions and movements have been determined
with some accuracy I think. Fig. 1 represents the positions of
the animals while the arm of the male that bears the spermato-
phores is inserted into the mantle chamber of the female. This
figure is the result of my conception of positions after having care-
fully observed copulation more than twenty times. Since draw-
ing the figure many other observations have been made and the
positions always seem to be essentially as given.

The male usually grasps the female while both are sw^imming.
Occasonally the female maybe resting on the bottom in the charac-
teristic attitude, with the tips of the arms and the posterior
end of the body touching and the head and funnel region somewhat
elevated. If not swimming, she usually, when grasped, starts
to swim, but in a few cases that I have observed she made no effort

332 OILMAN A. DREW

and left the bottom only as she was lifted or turned by the male.
In every case the male attached from the left side of the female.
He frequently swims close to her and brushes the tips of his
arms along her head and mantle. Just before attaching, if both
are swimming, he sinks slightly beneath her and grasps her
body with his arms so that his right arms are all on the right
side of her body and his left arms are all on her left side. The
body of the male is seldom exactly ventral to the female but usu-
ally slightly toward the left side. Attachment is evidently made
as nearly as possible in the required position but when the female
darts ahead, as she frequently does, the male is likely to attach too
far posteriorly . In such cases he does not let go his hold but crawls
rapidly forward, arm over arm, until the right position is attained.
Naturally the positions of the individual arms differ somewhat
but in general the arrangement is reasonably well shown in fig. 1.

For about a second after his position is attained the arms seem
busy in making firm attachments, then with a rapid sweep his
left ventral arm is passed by the end of his funnel and is im-
mediately inserted into the mantle chamber along the left side
of her neck, near the funnel. During the act both animals are
usually quite without color and the inserted arm of the male
may be seen fairly distinctly inside the mantle chamber.

The movement of the arm past the funnel is rapid and only once
have I actually seen the grasping of the spermatophores and their
transference to the mantle chamber. In this case while watch-
ing squid in an aquarium that was placed so the squid were between
me and a window, a male grasped a female that was resting on the
bottom. The female, contrary to the usual custom, did not move.
As the male had attached far back on the body, opportunity was
given me to get into position for observation before the male
could crawl forward. As the female made no attempt to get free,
the male seemed far more deliberate than usual. Just before the
arm was passed by the end of the funnel, the penis could be seen
protruding into it. A number of spermatophores appeared in the
opening of the funnel and were grasped by bending the tip of the
arm around them. With a rapid sweep of the arm they were
immediatelv inserted into the mantle chamber of the female

SEXUAL ACTIVITIES OF THE SQUID 333

where they were held about five or six seconds. The arm was
then withdrawn and in about five or six seconds more the empty
cases of the spermatophores passed out of the funnel of the female
with a respiratory jet of water. These spermatophore cases
were pretty closely attached to each other by having the tubes of
their ejaculatory apparatus twisted together. They were re-
covered and found to be 41 in number. To the cluster were at-
tached five sperm reservoirs. Examination of the female later
showed that most of the other reservoirs were attached near the
end of the oviduct. While the number of spermatophores used
in an act of copulation varies greatly, the observations that have
been made, indicate that this may be a little, but not much above
the average.

The animals nearly always separate almost immediately after
the arm is withdrawn. Beside the male which started to eat the
female, a very few individuals have remained attached for from
some seconds to nearly a minute after the arm has been with-
drawn.

After copulation the female frequently seems considerably fa-
tigued and may settle to the bottom and rest some minutes before
becoming active again. I am rather inclined to think that this
is due to her struggles, for when the female remained quiet, the
apparent fa,tigue did not seem so marked. The male does not
seem greatly affected, but is likely to continue to be very active
for some time.

The copulation that leads to the filling of the sperm receptacle
on the buccal membrane does not seem to be preceded by special
movements. Although I have observed it several times the ab-
sence of preparatory movement has left me rather unprepared
for the observations that must necessarily be made so quickly,
for in this, as in the other form of copulation, the animals are sel-
dom in contact more than ten seconds. In the cases I have ob-
served my attention has been attracted by the sudden dart of one
squid, the male, from one end of the aquarium directly at another,
the iemale. Before the dart the squid face each other, and are
separated by thirty centimeters or more. The movement was
always exceedingly rapid and was probably due in each case to the

334 GILMAN A. DREW

expulsion of a single jet of water. The male seemed to reach the
female before she had time to move much, although she has given
me the impression of attempting to dodge as if frightened. The
two animals become attached head to head with their arms inter-
mingled, each grasping the other (fig. 2). Then as in the other
method, the male sweeps his left ventral arm past the end of the
funnel and grasps the bundle of spermatophores. These are im-
mediately thrust between the ventral arms of the female and held
there for a few seconds. The animals then separate and exami-
nation has shown fresh sperm reservoirs attached to the receiv-
ing depression on the buccal membrane of the female. The empty
cases of the ejaculated spermatophores may be held between the
arms several minutes but they are finally dropped. Here, as in
the other method of copulation, only the sperm reservoirs are
retained for any length of time.

The spermatophores begin to ejaculate immediately after leav-
ing the penis and the whole process is completed in a very few
seconds. Pulling the filament attached to the ejaculatory end of
a spermatophore is all that is needed to start its ejaculation. As
the ejaculatory end of the spermatophore leaves the penis last and,
as the spermatophores in the penis and the spermatophoric sac
are imbedded in a viscid secretion, there is every reason to believe
that the pull given the spermatophores by the arm with which they
are grasped, when this arm starts to transfer them from the penis
to the mantle chamber or to the buccal membrane, is sufficient to
start ejaculation. The arm carries the spermatophores into the
position necessary for the attachment of the sperm reservoirs
while they are ejaculating and holds them there until the ejacu-
lation is complete and the reservoirs are attached.

The structure of the spermatophores and the mechanics of
ejaculation which lead to the attachment of the reservoirs will
be treated in another paper. It should, however, be understood
that the spermatophores are never attached as such, but they
ejaculate and the sperm reservoirs are attached. As the reser-
voirs are attached by cement carried inside the spermatophores
and liberated by the ejaculation, they may be stuck anywhere.

The sperm slowly escape from these reservoirs and may then

SEXUAL ACTIVITIES OF THE SQUID 335

become free in the water, as when they are attached in the mantle
chamber, or may be stored in a special receptacle, as when they
are attached in the special depression on the outer buccal mem-
brane. They are mixed with a viscid secretion in the reservoir
and probably also before entering the reservoir, although I am not
certain about the latter. The epithelium of the region is abund-
antly supplied with goblet cells which very possibly supp'y se-
cretion for this purpose.

The depression in which the sperm reservoirs are mostly at-
tached is supplied with a deeply staining columnar epithelium
which is covered by a mass of rather hard material, evidently se-
creted by these cells, that shows distinct markings parallel with
the surface of the epithelium (figs. 11 and 12). These markings
seem to indicate that the material is secreted intermittently and
thus is formed in layers. This material forms a suitable place for
attachment of sperm reservoirs and probably serves no other pur-
pose. Reservoirs are sometimes attached to other portions of the
buccal membranes or to the tentacles but they are far more abun-
dant in the depression than anywhere else. The sperm that es-
cape from the reservoirs that are not attached in this depression
probably do not find their way into the sperm receptacle.

The sperm receptacle has the shape of a compound alveolar
gland (fig. 11). It is imbedded in the outer buccal membrane and
opens on the inner surface of this membrane at a point opposite
the junction of the two ventral arms. Simple cubical epi-
thelium lines the deeper alveoli of the receptacle, and cubical
epithelium with many goblet cells the portion nearer the open-
ing. Some, but not many, cilia have been seen on these cells.
The killing fluids used may not have preserved them, for the tails
of sperm in the reservoirs are not often individually visible in the
sections. With the exception of the tails of the sperm and the
possible cilia on the cells the material gives evidence of good pre-
servation. A layer of muscle fibers surrounds the receptacle as
a whole and bundles of fibers run between and around the indi-
vidual alveoli.

It was not determined whether the sperm are active in the in-
terval between their discharge from the reservoirs and their en-

JOURXAL OP MORPHOLOGY, VOL. 22, NO. 2

336 OILMAN A. DREW

trance into the receptacle or not. That they are not active while
stored in the receptacle is shown by opening filled receptacles on
dry slides. The sperm are invariably quiet, but immediately
become active when sea-water is added. In specimens killed soon
after copulation, sections show the sperm entering the receptacle
in narrow streams and not spread out as one might expect them to
be if the sperm were active (fig. 11). It was not possible to remove
all the sea-water from living specimens in which the receptacles
were being filled without causing disturbances in the vicinity of
the reservoirs and that made it impossible to determine the normal
condition of the sperm in transit from one to the other. In the
sections that show sperm entering the reservoir the tails all point
in the same direction, as would be the case if they were not swim-
ming actively but were being moved by an outside force. The
heads go first and the tails all trail behind. Swimming sperm
usually move in all directions but there may be some directive
cause that would account for their positions even if they are stored
through their own activities.

As previously stated, a female that is nearly ready to deposit
her eggs can be told by her peculiar nervous movements and the
way she manipulates her arms. Frequently the borders of the
accessory nidamental glands, which are very red at this time, may
be seen through the semi-transparent mantle and thus form a
further indication that the eggs are nearly ready to be deposited.
The nidamental and oviducal glands of such an animal are always
somewhat, and frequently greatly, enlarged. Immediately after
the eggs have been deposited these glands, while still large, are
soft and flabby.

As is well known the squid deposits her eggs imbedded in strings
of a jelly-like substance which vary in size with the size of the ani-
mal depositing them but which probably average about 8 mm. in
diameter and 90 mm. long. The jelly consists of an inner mass
that surrounds the eggs and a thick, rather tough but still jelly-
like sheath that forms the outer covering. The inner jelly is se-
creted inside the oviduct by the oviducal glands. The outer jelly is
secreted by the nidamental glands and is apparently moulded into
shape as it passes through the funnel. The accessory nidamental

SEXUAL ACTIVITIES OF THE SQUID 337

glands, which he j List in front of the anterior ends of the nidamental
glands and open by wide openings near the narrow openings
of these glands, are very active during this period and secrete
a viscid material. What the special function of this secretion is
has not been determined. It would seem from position and activ-
ity that the secretions from both sets of glands must be mixed as
the}' are poured out.

Until recently eggs have not commonly been deposited in aqua-
ria at Woods Hole. This maybe due to the way the animals have
been handled. Squid will not stand rough handling, either in
capture or transportation, and live well in aquaria afterward.
When captured in fish traps, quickly and carefully transferred to
live cars where they are supplied with an abundance of water, and
transported to the aquaria with as little excitement and as good
water as possible, they may be kept several days in pretty good
condition, but they wear themselves out by constantly bumping
against the walls of the aquaria and are not vigorous many days.
During each of the months I have worked at Woods Hole, June
to late September, specimens have been obtained that have de-
posited eggs in aquaria. During the first three months speci-
mens ready to deposit eggs are rather easy to get. In September
only a small proportion of those captured still contained eggs.

Eggs are somewhat more frequently deposited in aquaria at
night than during the day, but this may be due to the frequent
if not nearly continuous disturbance to which they are subjected
during the day in a laboratory where many people are working.
The usual number of strings deposited by a female in what would
seem to be a continuous laying period ranged from one to six.
These strings were commonly delivered from fifteen to forty min-
utes apart, the time between any two strings being quite variable
in an individual. One specimen, however, deposited twenty-
three strings in an hour and thirty-five minutes. These were laid
during a comparatively dark day when the laboratory was quiet.
Possibly the small number deposited by other females was due
to disturbance.

The end of the egg string begins to protrude from the end of
the funnel while the female rests upon the bottom in the attitude

338 OILMAN A. DKEW

habitually assumed by resting squid (fig. 5). When from one to
two centimeters of the egg string protrudes from the funnel, the
female leaves the bottom and begins to swim slowly backward.
This swimming is apparently due both to movements of the tail
fin and to small jets of water forced from the funnel along the sides
of the egg string. The jets of water cause the egg string to be
protruded gradually. The protruding end is now caught by the
ends of the two dorsal arms, which are bent ventrally between the
other arms for this purpose (fig. 6), and as the string is ejected from
the funnel, it is drawn between the circlet of arms. It usually
takes from half a minute to a minute for the egg string to pass
through the funnel and to disappear between the arms. It is
then held between the arms about two minutes or sometimes a
little longer. While the string is held between the arms it is com-
pletely enclosed by them and their free ends keep twisting around
each other. In this position they form a cone with the apex at
the ends of the arms (fig. 7). At other times the arms are held
so they form a dorso-ventrally flattened expansion that serves
somewhat as a rudder or anterior fin. The arms while enclosing
the eggs are never entirely still but move slightly upon each other
and are probably busy in moving the string about. While the
string is thus held the animal slowly swims back and forth, never
rapidly but continuously.

Toward the end of the period during which the string of eggs
is held, the animal shows an increasing tendency to turn the body
into a nearly perpendicular position to bring and keep the tips of
the arms in contact with the bottom (left animal in fig. 9). With
the arms held quite rigid and the tail fin moving rapidly she goes
bounding along on the tips of her arms, dorsal side foremost, with
a movement somewhat similar to the bounces that may be ob-
tained by pushing a lead pencil, held by one extremity and slightly
inclined from the perpendicular, over a table. This action is
generally repeated several times. She occasionally catches hold
of objects with her suckers, finally catches some object firmly,
draws down into close contact with it for two or three seconds
(right animal in fig. 9) and, when she releases her hold, leaves the
string of eggs fastened to the object she had laid hold of. At

SEXUAL ACTIVITIES OF THE SQUID 339

this time the jelly of the string is soft and sticky. It hardens
quite rapidly and soon will not stick to objects, but at this time
it adheres readily. The position of the string when taken between
the arms indicates that the string is finally stuck by the end that
first leaves the funnel.

After sticking a string of eggs the female rests upon the bottom
some minutes before another string makes its appearance. She
usually selects some protruding object like a stone, shell, or water
pipe upon which to stick the egg strings. Having stuck one
string she usually, but not always, returns to the same place to
stick later strings. If strings are present when a female begins
to deposit she usually attaches to these strings, or to nearby ob-
jects. This no doubt accounts for the very large clusters, with
strings containing eggs in various stages of development, that
are sometimes found. Upon several occasions clusters in fish-
traps a ad live-cars have been found that would not go into an ordi-
nary ten-quart pail. Such clusters are of course formed by many
females.

It is evident that the eggs may be fertilized in the oviduct, in
the mantle chamber, or between the arms. Examination of the
contents of the oviduct have in no case given evidence of sperm.
Eggs taken from the oviduct may easily be fertilized by placing
them in sea-water containing sperm, but in no case did eggs taken
from the oviduct show evidence of fertilization although many
sperm reservoirs that were giving off active sperm were attached
to the walls of the oviduct and to surrounding organs.

There can be no doubt, however, that eggs may be and are fer-
tihzed in the mantle chamber and also between the arms. That
the eggs may be fertilized in the mantle chamber is indicated by
reason rather than by obervation. When sperm reservoirs are
attached in the mantle chamber the sperm are constantly liber-
ated in the water in this chamber as long as the supply lasts. The
eggs upon leaving the oviduct also pass into the mantle chamber
and, as before stated, when eggs and sperm are mixed in sea-
water, fertilization results.

That fertilization may be delayed until the egg string is formed
and held between the arms is indicated by observations made on

340 OILMAN A. DREW

the female already mentioned that deposited twenty-three strings.
She was in a rather large aquarium with a number of other squid.
Copulation had occurred several times but this particular squid, .
which had been under observation some hours, had not been seen
to copulate. Dissection later showed that there were no sperm
reservoirs attached in her mantle chamber. Because of disturb-
ance she upon six occasions failed to get the egg string between
her arms. When she reached for the string with her dorsal arms
she was each time disturbed so she dropped the string and ejected
it directly into the water. Four of these strings were recovered
as quickly as possible after they were dropped, and placed in dishes
of fresh sea-water where the proportion of fertilized eggs could
be determined. From 40 to 50 per cent of the eggs in the strings
developed. More than 99 per cent of the eggs m strings that had
been held between the arms and then placed in similar dishes
developed. As already mentioned there had been copulation
among other squid in the aquarium and as the reservoirs were
attached in the mantle chambers there must have been many
free sperm in the water of the aquarium. It seems probable that
enough of these sperm reached the strings that were dropped, be-
fore they could be removed from the aquarium, to fertilize a por-
tion of the eggs. Microscopic examination of these strings imme-
diately after they were dropped revealed very few sperm, but the
strings that were held between the arms were swarming with them.
Sperm were able to penetrate and move actively about in the soft
jelly of a recently formed string, but the jelly soon hardened so
fresh sperm brought in contact with it were not able to work their
way in.

A "curious bit of habit reflex was exhibited by this squid each
time she dropped a string of eggs. Immediately after the dis-
turbance she took the attitude she would normally have taken had
the egg string been successfully lodged between the arms. The
arms were held in the form of a cone, the tips were twisted together
and she passed on through each of the succeeding phases even to
drawing down tight against an object as if to attach the egg string
that had never been between the arms. After this she rested until
the next string was formed, but she never interrupted the orderly

SEXUAL ACTIVITIES OF THE SQUID 341

sequence of her activities because she had accidentally lost a string
of eggs.

The methods of copulation of cephalopods have attracted the
attention of observers from very early times but the act of copu-
lation has not been actually seen for many species and where ob-
servations have been made they have for the most part been in-
complete. Aristotle makes several statements regarding the
breeding habits of cephalopods. It is quite possible that he saw
something of the act of copulation for some species, but his state-
ments are hard to follow and are evidently inaccurate. The most
important statements are here quoted to show the curious medley
of facts and fiction. In chapter 5, book 5, he says:

1 . All the malacia, as the polypus, sepia and teuthis, approach each other
in the same manner, for they are united mouth to mouth: the tentacula of
one sex being adapted to those of the other; for when the polypus has fixed
the part called the head upon the ground, it extends its tentacula which
the other adapts to the expansion of its tentacula, and they make their
acetabula answer together. And some persons say that the male has an
organ like a penis in that one of its tentacula which contains the two
largest acetabula. This organ is sinewy, as far as the middle of the tenta-
culum, and they say it is all inserted into the nostril of the female.

2. The sepia and loligo swim about coiled together in this way, and
with their mouths and tentacula united, they swim in contrary directions to
each other. They adapt the organ called the nostril of the male to the
similar organ in the female; and the one swims forwards, and the other
backwards. The ova of the female are produced in the part called the
physeter, by means of which some persons say that they copulate.

Again in chapter 10, book 5, he says:

1. The malacia breed in the spring, and first of all the marine sepia,
though this one breeds at all seasons. It produces its ova in fifteen days.
When the ova are extruded, the male follows, and ejects his ink upon them
when they become hard. They go about in pairs. The male is more
variegated than the female, and blacker on the back. The sexes of the
polypus unite in the winter, the young are produced in the spring, when
these creatures conceal themselves for two months. It produces an
ovum like long hair, similar to the fruit of the white poplar. The fecund-

342 OILMAN A. DREW

ity of this animal is very great, for a great mumber of young are produced
from its ova. The male differs from the female in having a longer head,
and the part of the tentaculum which the fishermen call the penis is white.
It incubates upon the ova it produces, so that it becomes out of condition,
and is not sought after at this season.

Part of these statements, such as ''The sepia and loligo swim
about coiled together in this way, and with their mouths and ten-
tacula united, they swim in contrary directions to each other"
would seem to be based upon such observations as could be made
from above but the further statement that they adapt their nos-
trils (funnels) together, probably indicates the ease with which
observation and supposition can be mixed. It is not necessary
further to analyze Aristotle's statements. No doubt much was
based upon fishermen's stories but he evidently did study the an-
atomy and habits of these animals and recognized the probability
that one of the arms of the male is used in copulation.

While the modified arm of the male thus early received atten-
tion, the true hectocotylus that separates entirely from the male
and attaches itself in the mantle chamber of the female escaped
notice for many centuries. To quote from the Cambridge Natural
History :

The typical hectocotylus seems to have entirely escaped notice until
early in the present (last) century, when both Delle Chiaje and Cuvier
described it, as detected within the female, as a parasite, the latter under
the name of Hectocotylus octopodis. KoUiker, in 1845-49 regarded the
Hectocotylus of Tre?noctopus as the entire male animal, and went so far
as to discern in it an intestine, heart, and reproductive system. It was
not until 1851 that the investigation of V^rany and Filippi confirmed a
suggestion of Dujardin, while H. Miiller in 1853 completed the discovery
by describing the entire male as Argonauta.

While nearly all male cephalopods show some modification of
one or more arms, the only ones that have been reported with de-
tachable arms are Argonauta, Ocythoe, and Tremoctopus.

Extended studies have been made on the modification of the
arms of cephalopods, and there have been a few observations upon

SEXUAL ACTIVITIES OF THE SQUID 343

the functional activities of these arms, but most of the observa-
tions have consisted in finding sperm reservoirs recently attached
to various portions of females.

In 1869 Lafont described copulation in Sepia. A translation
of that portion that deals with the act itself is as follows:

In copulation the male and female precipitate themselves upon one an-
other, hold together by their arms which are twined together, and remain
thus, mouth to mouth, for a variable time, which may last for two or
three minutes. This act is followed in the female by a state of very
marked general prostration, while in the case of the male the general
excitation is greatly prolonged and for a considerable time it keeps the
splendid appearance these animals show as the result of the accomplish-
ment of the function of reproduction.

He supposed that while the animals were attached by their
arms, head to head, the male ejected a packet of spermatophores,
which ejaculated while in his mantle chamber and the sperm reser-
voirs were then thrown from the funnel of the male into the bran-
chial chamber of the female with the current of water entering her
branchial chamber.

Sepia, like Loligo, has a receptacle for the storage of spermato-
zoa in the buccal membrane, and the position observed by Lafont
of animals attached head to head was doubtless a true position
of copulation, but it seems probable that the spermatophores were
not disposed of in the way suggested, but were transferred to the
buccal membrane by one of the arms of the male. Lafont found
sperm reservoirs attached in the mantle chamber of the females
near the mouths of the oviducts, so it seems probable that in this
form, as in Loligo pealii, both methods of copulation occur.

Racovitza (1894, a) observed copulation inSepiola. The male
seized the female, turned it over and inserted his first pair of arms
into the mantle chamber. Copulation lasted eight minutes dur-
ing which the female struggled to free herself. He speaks of the
spermatophores being fixed on the folds of a large pocket situated
on the left side of the pallial cavity of the female. These ejacu-
late and the freed reservoirs deliver their sperm into the pocket

344 OILMAN A. DREW

which m turn ejects them (from his description I take it they are
not stored up in this pocket as in the receptacle on the buccal mem-
brane of a squid) into the pallial cavity where they are supposed
to meet the eggs as they are laid.

The most complete account of copulation that I have seen
for anycephalopod was given by Racovitza in 1894 (b) for Octopus
vulgaris. He observed copulation in an aquarium and gives a
figure showing the positions of the animals. The copulation differs
markedly from that of Loligo, as might be expected, for Octopus
has a hectocotylized arm that is much more differentiated than
that of Loligo. The animals were some distance apart in the aqua-
rium. The male reached over with the hectocotyhzed arm, which
for this species is the third on the right side and, after caressing
the female with its tip, introduced its end into her mantle chamber
by the side of the funnel. Here it remained for something more
than an hour. During this time the female remained quiet, ex-
cept for certain spasmodic movements, while the male showed
only slight movements of the hectocotylized arm which were sup-
posed to be associated with the movements of spermatophores
down the longitudinal groove of this arm. Although it was not
possible actually to see the spermatophores in transit, examina-
tion of the female after copulation showed numbers of the sperm
reservoirs, derived from the ejaculated spermatophores, within
the oviducts.

Evidently there are at least three methods of copulation prac-
ticed by cephalopods. A method of caducous hectocotylism in
which the charged hectocotyl is liberated in the mantle chamber
of the female; a method in which the arm does not liberate any
special portion but is so modified that it can transfer spermato-
phores by a mechanism within itself to the region of the oviduct
of the female; and finally a shght modification of the arm that
simply enables it to grasp the spermatophores which are then trans-
ferred directly to the female by moving the arm. Where the lat-
ter method is employed there may be two kinds of copulation, as
in Loligo pealii.

Racovitza, (1894, c) in commenting on the copulation of Rossia
believes that, although special receptacles are found outside the

SEXUAL ACTIVITIES OF THE SQUID 845

mantle chamber of this species, they cannot be considered as nor-
mally functional. He seems led to this conclusion by finding
sperm reservoirs attached to various portions of the bodies of the
animals as well as in the immediate neighborhood of the mouths
of the oviducts. It would seem more likely in the light of the ob-
servations here recorded for Loligo, that a copulation that leads to
the filling of these receptacles is normal and that the sperm so
stored may be used in fertilizing the eggs.

It is certainly hard to conceive by what steps a complicated
method of transferring sperm that has led to the formation of a
hectocotylized arm and complicated spermatophores might be
perfected. The modification of different arms for copulation by
different cephalopods further increases the difficulty in under-
standing the history of hectocotylism as a whole.

While evidence that bears directly upon the history of the hecto-
cotylism seems to be lacking, such complications are so frequently
considered to be impossible to explain by known evolutionary
factors that it may be well at least to consider the great difficul-
ties presented by such structures. It must not be supposed that
in so doing I put myself in the position of defending a thesis. This
would be too much like the methods employed by many of the
Greek philosophers who needed little or no basis of fact upon which
to build. My only reason for considering the matter here is to
show that, with all the difficulties, the condition of hectocotylism
among modern cephalopods cannot be considered beyond the pos-
sible range of evolutionary factors.

Among the Dibranchiata the arms that show hectocotylism are
the first, the third and the fourth on both sides of the body. Some-
times more than a single one is affected. In such cases the modi-
fied arms may be symmetrically placed on the two sides of the
body, or they may be adjacent arms on the same side of the body.
Steenstrup attempted to base the classification of cephalopods
upon their hectocotylized arms but Brock and Hoyle have shown
that forms whose general body structure would seem to indicate
relationship, do not always have homologous arms modified.

While the arm is usually constantly on one side for all members
of a genus, unless both sides are modified as not infrequently hap-

346 GILMAN A. DREW

pens, a genus whose general body structure indicates near rela-
tionship may have the similar arm of the other side modified.
The position of the arm on the right or left side of the body is not
generally considered very significant. The somewhat frequent
occurrence of genera showing hectocotylism of arms symmetri-
cally placed on the two sides of the body may indicate a primitive
paired condition that has been replaced among the majority of
existing cephalopod genera, by specializing on one side and drop-
ping out on the other. Whether this accounts for the condition
or not, the shifting of a modification from one side of the body
to the other, sometimes involving modifications of other body
structures and sometimes apparently not, is not uncommon among
animals, and even if not easily explained, evidently has no very
great phylogenetic significance. Shifting in series is not so com-
mon and when we find in the same family, genera with the fourth
and others with the first arm hectocotjdized it becomes difficult
to imagine ancestral conditions that made this posisble.

Wherever known, male cephalopods use one or more of the
arms to transfer sperm to the female. Copulation has not been
described for many of the species but the presence of more or less
modified arms in more than half the recogaized families may be
taken as an indication that either these animals or their ancestors
used their arms in copulation.

If the spadix of Nautilus is used in copulation we have a pos-
sible indication that a number of arms may have been employed
in the transfer of sperm by primitive cephalopods. It is of course
possible that all the arms were used for this purpose and that the
present diversity can be accounted for by the specialization of
one or the other of the arms involved in this primitive condition.
This, however, does not seem reasonable when the diversity within
the limits of a single family is considered.

The arm that is used, and the way in which it is used, is asso-
ciated with the character of the spermatophores and the position
of their final discharge. The Octopoda show the greatest struc-
tural modification in their hectocotylized arms. While two of the
families of this group give no evidence of hectocotylism, none of
the genera of the remaining families are known to be free from it.

SEXUAL ACTIVITIES OF THE SQUID 347

and wherever found it is always the third arm that is involved.
Sometimes this arm is on the right and sometimes it is on the left
side. In three genera it is known to be caducous and in a fourth
(Alloposus) it is supposed to be. In the remaining genera in which
the hectocotylized arm has been studied, the modifications, while
not resulting in the actual separation of the arms, are of an exten-
sive nature. In Octopus, for instance, they involve not only
changes. in size, form, and the condition of suckers, but a special
groove is present through which the spe matophores are supposed
to be carried from the base, presumably from the penis to the tip.
The tip in turn is modified so it is supposed to function in placing
the spermatophores in position for ejaculation.

The Decapoda do not show such extensively modified hectoco-
tylized arms. The changes are here chiefly confined to some of
the suckers and their immediate vicinity. In Loligo this modi-
fication apparently serves to aid the arm in grasping the spermato-
phores, which are then transferred by the movement of the arm.
While the actual grasping of the spermatophores has not been
previously observed, there can be little doubt that other forms of
the Decapoda use the arms in a similar manner. Where copula-
tion has been observed the movements of the arms indicate that
they are used in the transfer, and the positions of the sperm reser-
voirs that have been found attached to the females indicate that
some arm must have functioned in getting them into position. As
there is no special transferring mechanism, this must have been
accomplished by the free movements of the arms.

Where structural modification is shght and the placing o the
spermatophores is due to dexterity, there is less difficulty in under-
standing how the function may be shifted from one arm to another
in response to changes in the position of the attachment of the
reservoirs on the female, than would be the case were great
structural changes involved. It would be much more difficult to
understand how there could be a shifting in series of arms as highly
modified as those of the Octopoda, where only the modified arm
could possibly perform the function.

It mast not be understood that habit formation requiring such
dexterity is considered easier to originate than modification in

348 OILMAN A. DREW

structure that will perform similar acts. When, however, the
habit and dexterity have been acquired, it is not inconceivable
that they might be shifted to another closely similar appendage
if the position of this appendage becomes more suitable for the
purpose. The modification is so slight in the arms of most of the
Decapoda, and the modification varies so greatly in the different
genera, that it may have been functionally acquired in each case.
So far as can be seen it would be mechanically quite possible for
a squid to use an unmodified arm, instead of the one that shows
the modification, for the transfer of the spermatophores. The
spermatophores might not be so tightly or compactly held but
the normal suckers would hardly seem to interfere greatly in the
performance of the function.

There is still another question involved. Is there any genetic
relation between these two methods of transfer and if there be,
which, if either, most probably came first?

A special method of copulation that requires the use of arms and
complicated spermatophores is not found among animals often
enough to make it at all probable that it has arisen in this group
more than once, so we can hardly doubt that the two methods
are genetically related.

At first sight the squid's method of grasping the spermatophores
and transferring them directly might be considered the simpler
process, but there is some reason to doubt that this method was
at the beginning of the series. While it would be hazardous to say
that Octopoda were the ancestors of Decapoda, there is much
reason to believe that the ancestors of the latter lived upon the
bottom and were far less active than the modern animals. Such
animals would not seem to be so well adapted for the transfer of
spermatophores by dexterous movements as the more active, free-
swimming forms. It is at least certainly true among modern ceph-
alopods that those that show great modifications in the structure
of the hectocotylized arms are found entirely among the less active
bottom forms. If the method of transferring sperm by means of
the arms originated before the Decapoda became free-swimming
animals, and this seems the only explanation of its prevalence

SEXUAL ACTIVITIES OF THE SQUID 349

among both Decapoda and Octopoda in modern times, it would
seem that structural modification probably came early.

Possibly this modification was based upon the use of one or more
arms as guides for the transfer of the sperm. It is possible that
having first used the arms as guides, structural modifications and
dexterous movements were developed as divergent methods. If
the two methods form a linear series, there is some reason to think
structural modifications came first. It would seem much easier
to explain modifications that lead to the change in the structure
of appendages for the transfer of spermatozoa, as the grooved hec-
tocotyhzed arm of Octopus or the modified abdominal appendages
of certain Crustacea, than to explain a sudden change that would
result in a practically unmodified arm functioning by grasping
spermatophores of a very specialized kind, transferring them
quickly and accurately to the required position and holding them
there until they have had ample time to ejaculate and fix their
contents. It seems more reasonable to suppose that an arm modi-
fied as a machine to perform this process, with its tip serving to
place the spermatophores in position, might in time acquire the
necessary dexterity and then lose the modifications previously
acquired, than to look at this as the beginning of the series. Again
we find that in such cases as the squid, where the arm is little modi-
fied but very dexterous, there is a special receptacle at some dis-
tance from the opening of the oviduct that is norma ly filled with
sperm during the breeding season. This would certainly seem to
be a comparatively recently acquired receptacle, so the copulation
leading to its being filled would also be considered comparatively
recent. That this receptacle is concerned in the fertilization of
the eggs is shown by observations made while the eggs were being
laid.

With no personal knowledge of the breeding habits of other
cephalopods than the squid, it would seem more reasonable to
consider the method of using the detachable hectocotyl of such
forms as Tremoctopus as one extreme, the method used by Loligo
in grasping spermatophores and transferring them directly as an-
other extreme and the condition shown by Octopus as the modern
greatly specialized product of a modification such as early cephal-

350 GILMAN A. DREW

opods probably developed. This would mean that the detachable
hectocotyl is an extreme specialization in structure and that the
modification shown by the squid represents possibly a degenera-
tion in structure but a remarkable specialization in habit.

Why a form should have two methods of copulation is not at
all clear. Certainly the introduction of the spermatophores into
the mantle chamber to a position near the oviduct is to be con-
sidered more primitive than their being placed in a position to
fill a receptacle outside of the mantle chamber, but why mantle
chamber copulation should be retained after the receptacle has
been perfected is not clear. That mantle chamber copulation is
not absolutely necessary for the fertilization of the eggs I think is
proved by my observations; that it is common is certain. That
the sperm receptacle is an improvement over the free attachment
of the sperm reservoirs in the mantle chamber is evident from the
longer possible retention of the sperm in the receptacle. In a lim-
ited period after the sperm reservoirs are freed from the sperma-
tophores, as when deposited in the mantle chamber, the sperm all
escape and are wasted unless oviposition takes place in the mean-
time.

SUMMARY

Squid have two methods of copulation. By one method sperm
reservoirs are attached in the mantle chamber on or near the ovi-
duct and immediately begin to discharge their contents freely in
the water. By the other method sperm reservoirs are attached
to the outer buccal membrane and the sperm become stored in a
special receptacle in the membrane, which is placed opposite the
junction of the two ventral arms and opens on its inner surface.

The left ventral arm of the male is always used in transferring
the spermatophores, which are grasped by the arm and transferred
by its free movement. Ejaculation of the spermatophores is evi-
dently started by the pull given their filaments when the arm starts
to transfer them from the penis to the mantle chamber or buccal
membrane. The transfer requires rapidity and dexterity and the
spermatophores are held in position until ejaculation is complete
and their sperm reservoirs are fastened. As many as forty sperm-
atophores may be transferred at a time.

SEXUAL ACTIVITIES OF THE SQUID 351

The egg strings are composed of two kinds of jelly. One kind
is supplied by the oviducal gland and the other by the nidamental
and probably accessory nidamental glands. The string is appar-
ently molded into shape by passing through the funnel. The
jelly is at first soft and sticky but soon becomes tough and loses
most of its stickiness.

From the funnel the egg string is drawn between the circlet of
arms, where it is held two or more minutes. In sticking the
string the female grasps some object with her arms and draws
down tight so the string is evidently crowded against it. When
she releases her hold the string is left sticking to the object.

Fertilization evidently does not take place inside the oviduct.
It doubtless may take place in the mantle chamber when sperm
reservoirs are present there, and is known to take place while the
egg string is held between the arms. The sperm are hberated
from the receptacle while the eggs are between the arms.

Notwithstanding complications, the conditions of hectocotyl-
ism shown by cephalopods need not be considered beyond the
influence of factors of evolution.

LITERATURE CITED

The cephalopod literature is very extensive. Only those papers directly re-
ferred to are here given.

Aristotle History of animals. Trans, by Richard Cresswell. 1891.

Brock, J. 1882 Anat. u. Syst. d. Cephalopoden. Z. f. wiss. Zool. 36.
1884 Mannchen d. Sepioloidea lineata. Z. f. wiss. Zool. 40.

HoYLE, W. E. 1907 Presidential address of Zoological Section. Rept. Brit.
Ass. Adv. Sci.

Lafont, M. a. 1869 Observations sur la fecundation des Mollusques Cephalo-
pods der Golfe de Gascogne. Ann. des Sci. Nat. (5) 11.

Racovitza, Emile. G. 1894a Sur I'accouplement des quelques Cephalopods
Sepiola rondeletii (Leach), Rossia macrosoma (d. Ch.) et Octopus vul-
garis (Lam.). Comp. Rend. I'Acad. des Sci. 118.

1894b Notes de Biologic. I. Accouplement et Fecondation chez
rOctopus vulgaris Lam. Arch. d. Zool. Exper. et Gen. (3) 2.
1894c Notes de Biologie. III. Moeurs et Reproduction de la Rossia
macrosima (D. Ch.). Arch. d. Zool. Exper. et Gen. (3) 2.

Steenstrup, J. J. S. 1856-57 Hectocotyl. hos Octopodstsegterne. Vid. Selsk.
Skr. (5) 4, Translated Ann. N. H. (2) 20.
1881 Sepiadarium og Idiosepius. Vid. Selsk. Skr. (6) L.
1887 Nota; Teuthologica? 7. Overs. Vid. Selsk. Forh.

JOURXAL OF MORPHOLOGY, VOL. 22, XO. 2

352 OILMAN A. DREW

EXPLANATION OF FIGURES

All of the figures that represent the attitudes of squid were drawn from memory
after repeated observations. While each figure is thus really a composite, and must
represent impressions received rather than the actual positions of particular in-
dividuals, much care has been given to the preparation of the figures and it is be-
lieved that the general attitudes are reasonably well represented. Sexually mature
squid are usually as much as 15 cm. and may exceed 40 cm. in length.

ABBREVIATIONS

hni, inner buccal membrane n, nidamental gland

bmo, outer buccal membrane na, accessory nidamental gland

d, depression in which sperm reservoirs o, oviduct

are attached r, rectum

g, gill s, sperm reservoirs (ejaculated from sper-

h, modified (hectocotylized) portion of matophores)

arm sr, sperm receptacle

j, jaws sro, opening of sperm receptacle

PLATE 1

EXPLANATION OF FIGURES

1 Copulating squid showing the positions taken by the animals when the sperm-
atophores are inserted into the mantle chamber. The figure shows the animals
during the period the arm of the male is inserted in the mantle chamber of the fe-
male. Drawn from memory after many observations.

2 Copulating squid showing the positions of the animals when the spei'mato-
phores are placed so that their reservoirs become attached to the outer buccal
membrane. The figure .shows the male in the act of grasping the spermatophores
with the tip of his arm as they are ejected through the funnel. Drawn from mem-
ory after many observations.

3 A common attitude of a sexually excited male. The arms are not kept rigidly
in a set position, but are frequently spread as shown in the illustration and held
thus for from a few seconds to a minute or more at a time. The drawing is based
upon sketches made of active animals.

• 4 Photograph of the two ventral arms of a male squid, showing the slight modi-
fication (h) consisting of enlarged peduncles, reduced sucking discs and a ridge
between the suckers, toward the tip of the left arm. The wrinkles on the arms are
due to shrinkage. A bit of the outer buccal membrane shows between the arms.
The arms from which the photograph was made are 9^ cm. long.

^FiXUAL ACTIVITIES OF THE SQUID

JODRNAL OF MORPHOLOGY, VOL. 22, NO. 2

353

PLATE 2

EXPLANATION OF FIGURES

5 A female at rest with the egg string beginning to protrude. Drawn from
memorj' and hurried sketches after many observations.

6 A female after she starts to swim, reaching for the egg string with her dorsal
arms. With these arms she draws the string between the circle of arms as it is
ejected from the funnel. Drawn from memory after many observations.

7 A swimming female, showing the positions of the arms while they surround
the egg string. They are held in this position, with the tips somewhat twisted to-
gether, for two or three minutes. While the arms closely surround the egg string
they show slight individual movements that may be of service in moving the egg
string so sperm will be more evenly distributed over it. Drawn from memoiy
after many observations.

8 A female squid with the mantle cut and spread o])eii and the arms separated
to show the position of attached sperm reservoirs (s) on the oviduct (o) and the
sperm receptacle (sr) in the outer buccal membrane.

354

IXTAI- ACTUITIKS OF THE SQUri)

lill.MAX A. DHKW

mi'HOI.OCY, V<

PLATE 3

EXPLANATION OF FIGURE

9 The specimen on the left 8ide shows a female in the position she assumes as
she bounces over the bottom on the tips of her arms just previous to selecting a
place for sticking the egg string. The specimen on the right side shows the posi-
tion of a female during the act of sticking an egg string to a rock. Only a few sec-
onds are required to stick the string. The positions of the animals are drawn from
memory after many observations.

356

.sp:xual activities of the squid

OII.MAX A. DRRW

9
3o7

PLATE I

EXPLANATION OF FICUKKS

10 Jaws and buccal moinbrane of a female squid, with the out er luemliranc [b/no)
pulled ventrally to expose the sperm receptacle (the opening of which is shown at
.sTo) and the adjacent depression (d). Several sperm reservoirs (.s), ejaculated from
spermatophores, arc shown attached in the depression. Magnified about 7
diameters.

11 Section of the outer buccal membrane taken through the sperm receptacle
(.s/-). This was taken from an animal shortly after the sperm reservoirs (.s) had
been attached and shows sjierm in transit from reservoirs to receptacle. Magni-
fied about 22 diameters.

12 Section through the e])ithelium and secretion lining the depression in which
sperm reservoirs are attached. Magnified about 300 diameters.

13 Section through an alveolus of a sperm receptacle. The clear spaces in the
epithelivmi are goblet cells. Traces of the Hagella on the sperm and possibly cilia
on some of the epithelial cells were visible but they were not definite enough to be
put in the drawing. Magnified about 300 tliameters.

358

■iKXiAi. Acii\ ]rii:s OF jiiK s(Hii

/ 4^^::^^.

^;/

MORPHOl.OUV,

%• -*

Drew, del.

359

STUDIES OF FERTILIZATION IN NEREIS

I. THE CORTICAL CHANGES IN THE EGG! II. PARTIAL FERTILIZATION

FRANK R. LILLIE

From the Hull Zoological Laboratory, University of Chicago

TEN FIGURES
ONE PLATE

I. THE CORTICAL CHANGES IN THE EGG

In many animals one of the immediate effects of fertiliza-
tion is to cause the egg to throw off a membrane, which is there-
fore known as the fertilization membrane. This is the case for
instance in the eggs of echinids and nematodes. In other cases,
where a definite vitelline membrane exists prior to fertihzation,
cortical changes occur in the egg immediately after insemination,
leading to the formation of a space, the so-called perivitelline
space, between the protoplasm of the egg and the vitelline mem-
brane. This is the case for instance in the eggs of at least many
annehds, molluscs and vertebrates, and it is unquestionably
a more common phenomenon than the formation of a fertiliza-
tion membrane. There can be little doubt that these apparently
different phenomena are simply varying expressions of a change
in the cortex of the egg, which is of the same nature in all cases.
Loeb's studies ('09) have thrown much light on the nature of
these cortical changes. In the case of the egg of Nereis they are
relatively obvious in their character and readily followed.

The ovocyte of Nereis is somewhat flattened in a polar direc-
tion, measuring about 87.5 x 100^; it is girdled by a double
equatorial zone of large oil drops. The large germinal vesicle is
central and somewhat elongated in a polar direction.

In his description of the unfertilized egg, Wilson ('92) distin-
guished two membranes: a delicate outer vitelline membrane,

361

362 FRANK R. LILLIE

and a subjacent membrane or layer, about 6-7^ in thickness,
which he called the zona radiata. As will appear from the sequel
however, the latter is not a membrane in the usual meaning of
the word, but a cortical, coarselj'' alveolar layer of the egg. Jt is
transparent and somewhat granular, and the granules tend to be
arranged in radiating lines. There is no perivitelline space in
the unfertilized egg.

In sections of unfertilized ovocytes fixed in Flemming's fluid,
the zona radiata is seen to be a coarsely alveolar layer with
homogenous alveolar contents (fig. 1). The walls of the alveoli
are continuous internally with the potoplasm of the egg, and
unite externally to form a protoplasmic la^'er applied to the
vitelline membrane. The alveoli are closed externally (figs.
1 and 2). The zona radiata is in fact a coarse emulsion or foam-
structure.

Unfertilized eggs of Nereis are entirely devoid of jelly and they
lie in immediate contact in the sea-water. If India ink be ground
up in the water, the particles come in contact with the vitelline
membrane. ILsich fertilized egg, on the other hand, is surrounded
by a thick layer of colorless transparent jelly; If many eggs are
contained in the dish, fusion of the contiguous gelatinous mem-
branes binds the eggs into a mass; the cortical layer (zona radiata)
is absent in fertilized eggs, and there is a narrow perivitelline
space between the vitelline membrane and the surface of the egg
(fig. 3).

The jelly is formed by the extrusion, or diffusion, of the alveo-
lar contents of the cortical layer through the vitelline membrane ;
the egg of Nereis, in fact, secretes its own jelly, as may be readily
demonstrated in life by inseminating under the microscope with
excess of sperm. If the sperm be added to closely placed eggs
and a cover glass applied so as to force the eggs into a single
layer, aiid the preparation examined with no loss of time, the sperma-
tozoa will be seen in large numbers in contact with the vitelhne
membrane. In one or two minutes the spermatozoa are moved
away from the surface of the membrane by some invisible repel-
ling substance, and, if the eggs be numerous, the spermatozoa
unite in three to five minutes to form lines that bound hexa-

STUDIES OF FERTILIZATION

363

gonal areas with the eggs in the centers of the hexagons (fig. A).
The substance that sweeps the spermatozoa away from the sur-
face of the eggs is the jelly. Synchronously with its formation,
the alveoli of the cortical layer are emptied and the alveolar walls
now appear as delicate lines crossing a wide perivitelline space^
(fig. B).

However, not all of the spermatozoa are thus carried out by the
secreted jelly, but in the case of each egg a single spermatozoon
remains attached to the vitelline membrane. This is very pret-
tily demonstrated if the eggs are under some pressure, so that

Fig. A. Diagram of fertilization with excess of sperm. The outflow of jelly
from the eggs has carried the supernumerary spermatozoa away from the surface
of the eggs (see text). In the case of each egg the single effective spermatozoon
remains attached. From a sketch of the living object.

the spermatozoa are prevented from reaching the eggs above or
below. In this case one can discover the single spermatozoon
attached to the vitelline membrane in practically every egg (fig. A).
All stages of the disappearance of the cortical layer may be
readily and rapidly observed. The alveolar walls, however,

1 Wilson ('92) states that "from twenty to thirty minutes after fertilization the
striae of the zona suddenly become indistinct and in the course of two or three
minutes the zona itself entirely disappears, leaving only the outer membrane."
But inasmuch as he was under the impression that the unfertilized eggs possess a
transparent, thick, gelatinous envelope like the fertilized ones, he failed to observe
the interesting phenomenon of formation of the jelly described here.

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

364 FRANK R. LILLTE

remain as delicate strands of protoplasm uniting the vitelline
membrane to the surface of the egg. The jelly, therefore, repre-
sents the alveolar contents only of the cortical layer of the unfer-
tihzed egg, and the perivitelline space is nothing but the contracted
alveoli of the cortical layer filled with fluid. The perivitelline
space must, therefore, be regarded as intraovular with a delicate
external cytoplasmic wall lining the vitelline membrane; this we
may distinguish as the plasma membrane; it is comparable in
some respects to the ferflization membrane of sea-urchins.

Unfertilized eggs allowed to stand in the sea-water form no
jelly and retain the cortical layer; the germinal vesicle remains
intact; but, if they be strongly centrifuged, the jelly forms, the
perivitelline space arises, the germinal vesicle breaks down and
both polar bodies are formed; but parthenogenetic development,
usually at least, does not take place. Similarly, the addition of
a fairly strong solution of potassium chloride to the sea-water
causes formation of the jelly and the perivitelhne space while the
eggs are still in the solution; maturation takes place after trans-
fer to sea-water but cleavage does not occur (in my experiments;
cf . Fischer) , though some differentiation may take place without
cleavage. It would appear, then, that conditions that so alter
the permeability of the plasma and the vitelline membranes as
to permit the outflow of the alveolar contents of the cortical layer
initiate development, but that the normal continuation of devel-
opment is dependent on other factors.

In the normal fertilization of Nereis the stimulus of the sperm-
atozoon causes the formation of the jelly and the perivitelline
space iong before it has penetrated the membrane; in fact pene-
tration does not take place until 40 to 50 minutes after insemina-
tion. However, mere contact of the spermatozoon with the mem-
brane is apparently not sufficient; but actual attachment of at
least a single spermatozoon is required ; this is shown by the fact
that the effective spermatozoon is not carried away from the
membrane with the unsuccessful ones by the outflow of jelly.
Yet the effect of the locafized stimulus of the attached spermato-
zoon is practically instantaneously effective over the entire

STUDIES OF FERTILIZATION 365

extent of the membrane; it is more like an electrical discharge
or some other physical disturbance than a chemical effect.

J. Loeb ('09) has formed the hypothesis that the cortical layer
of the egg, especially of sea-urchins, is an emulsion which is
rendered stable by a third substance consisting of lipoids, espe-
cially cholesterin. The emulsion becomes unstable on solution of
the lipoids; this enables the albuminous drops, which he conceives
to form one phase of the emulsion, to take up water; hence the
layer liquefies and the perivitelline space arises; the fertilization
membrane is thus formed. Hence, according to Loeb, the action
of lipoid-dissolving substances is to cause the formation of the
fertilization membrane. Without committing ourselves to these
specific views of the nature of the cortical emulsion, which Loeb
himself does not hold very strongly, we may admit that Loeb's
hypothesis, that the formation of the fertilization membrane is
due to the breaking down of a cortical emulsion, fits the case of
Nereis very well. If we go further, however, we must note an
important lack of agreement with Loeb's hypothesis. As Loeb
himself points out, the theory implies that the membrane of the
egg is permeable for sea-water and crystalloid substances, and on
the other hand impermeable for colloids; in Nereis the contents
of the cortical alveoli are unquestionably colloid, as Loeb's
hypothesis requires, but it is perfectly certain that they diffuse
through the membrane to form the external jelly ; at the same time,
unquestionably, sea-water enters to take the space previously
occupied by the colloid. The membrane is therefore permeable for
both crystalloids and colloids at this time. I have not, however,
investigated farther the properties of the egg membranes and
must leave this problem to those who are better qualified as phys-
iologists to make such a study.

It would appear that the presence of this colloid substance
in the cortex is an inhibition to the maturation of the egg, because
as soon as it is removed, maturation processes are set in motion
and both polar bodies are formed. In what manner it inhibits
is of course problematical. In the egg of Ascaris megalocephala
there is a similar excretion of a cortical colloid which forms, in
this case, the thick resistant perivitelline membrane. The ap-

366 FRANK R. LILLIE

pearance of the fertilization membrane of echinids might be
similarly due to excretion of a cortical colloid which is removed
by diffusion and hence is not detected. It is a problem worthy
of careful investigation whether the loss of cortical colloids is not
the first step in fertilization geuerally.

II. PARTIAL FERTILIZATION

Two functions of the spermatozoon in fertilization may be
sharply distinguished. The first is the initiation of the develop-
ment and the second is the transfer of paternal qualities to the
fertilized ovum (heredity from male parent). The first function
alone is under consideration in these experiments.

We have seen that in Nereis the immediate effect of attach-
ment of the spermatozoon is essentially the same as a mechanical
shock (centrifuging), or a chemical stimulus (KCl); that is, it
causes the breaking down of the cortical emulsion and consequent
formation of the gelatinous envelope of the egg. But apparently
the resemblance extends no farther, for in the case of mechanical
or chemical stimulation the impulse to development is lost or
greatly weakened after maturation has occurred; and the eggs
do not segment. On the other hand the normally fertilized egg
does not stop after maturation, but proceeds with its develop-
ment in a normal fashion. Now the cause of this difference
might be either: (a) because the stimulus of the spermatozoon
is qualitatively different from, or stronger than, mechanical or
chemical stimulation, or (b) because the fertihzing action of the
spermatozoon is not completed with the cortical changes but
continues after its entrance into the egg. If the first alternative
were correct, then the elimination of the spermatozoon after
membrane formation should not prevent the normal cleavage and
development of ova which had once been stimulated by it; but
if the second alternative were correct and the sperm nucleus
were prevented from entering the egg after it had induced mem-
brane-formation, then such ova should proceed no further in
their development than those mechanically or chemically stimu-
lated.

STUDIES OF FERTILIZATION 367

I have been able to perform this experiment on the eggs of
Nereis and have found that eggs in which the spermatozoon is
removed after the cortical changes have occurred proceed but
little farther in their development than eggs mechanically or
chemically stimulated, and they do not undergo segmentation.
Fertilization is therefore still incomplete after the formation of
the fertilization membrane.

It will be seen that if the results above indicated be demon-
strated, the process of fertilization is obviously something more
than a beginning of cytolysis or a mere alteration of permeability
of the peripheral cell membrane. It would appear to be a pro-
gressive change, starting at the periphery and gradually involving
the more central portions of the cell. We would, at least, have
to distinguish two stages in the fertihzing action of the sperma-
tozoon, one before and the other after penetration.

I shall consider first the evidence for the statement that in
the egg of Nereis elimination of the spermatozoon after membrane-
formation leaves the process of fertilization incomplete. In
the second place I shall note the respects in which the com-
pletely fertilized egg differs from the partially fertilized egg, and
finally, shall consider the bearing of the facts on the theory of
fertilization. Inasmuch as it will be necessary to make frequent
comparisons with the normal fertihzation, a brief account of
the salient features of this process will be given first.

A. Salient features of the normal fertilization

The egg of Nereis is difficult to fix in a thoroughly satisfactory
fashion; owing, no doubt, to the presence of the large oil-drops
and yolk-granules, uneven fixation with shrinkage is hard to
avoid. The eggs appear likewise difficult of penetration, owing
probably to the rather viscid jelly from which they cannot be
separated; this also makes any considerable number of eggs very
bulky and the killing fluid is apt to be much diluted if used in
ordinary amounts. After considerable experimenting with
picric acid, corrosive subhmate and osmic acid fixing fluids, I
finally found one which gives practically perfect results in all

368 FRANK R. LILLIE

stages of maturation and fertilization. This is Meves' modifi-
cation of Flemming's fluid made as follows: chromic acid, 0.5
per cent, 15 cc; osmic acid, 2 per cent, 3.5 cc; glacial acetic
acid 3 drops. The eggs were left in the fluid from thirty to forty-
five minutes. Fixation in this fluid causes no shrinkage, the oil
is so changed that it is not dissolved out by subsequent imbedding
in parafline; the sections stain beautifully m iron haematoxylin,
and certain substances are clearly differentiated which can be
detected only with the greatest difficulty after fixation in any
other fluid tried.

a. The penetration of the spermatozoon. It was noted in the first
part of tjiis paper that a single spermatozoon becomes attached
to the egg-membrane immediately after insemination, and that
the breaking down of the cortical layer, secretion of the jelly
and formation of the perivitelline space follow immediately,
though the actual penetration of the spermatozoon is delayed
forty or fifty minutes.

The act of penetration involves no motile activity on the part
of the spermatozoon ; after the latter has become attached to the
vitelline membrane all movement of the spermatozoon ceases
and it remains absolutely immotile throughout the forty or fifty
minutes that elapse before it is taken into the egg. The events
of this period as seen in the living egg are as follows :

1. The jelly is formed by outflow of the alveolar contents of
the cortical layer-as already noted; although a large amount of
jelly is formed in two or three minutes, yet the process lasts ten
or fifteen minutes before the deeper alveoli are emptied. There
is then a very wide perivitelline space crossed by the alveolar
walls which are attached to the plasma membrane, presenting a
very striking appearance (fig. B).

2. The protoplasm of the egg immediately beneath the at-
tached spermatozoon then forms a rounded elevation, the entrance
cone, which gradually moves across the perivitelline space and
comes in contact, and fuses, with the membrane (fig. B, a). This
condition is usually fully attained about fifteen to seventeen
minutes after insemination.

STUDIES OF FERTILIZATION 369

3. The entrance cone then gradually retracts, drawing the
membrane down to form a depression in which the spermatozoon
is included. At this stage one may easily imagine that the sper-

Fig. B. History of the fertilization-cone as been in the living egg. Four cam-
era drawings of the same egg: —

a Seventeen minutes after insemination,

b Nineteen minutes after insemination,

c Twenty-two minutes after insemination,

d Twenty-four minutes after insemination.
The fertilization-cone is shown at the height of its development in a, its gradual
recession and the simultaneous formation of a depression in the membrane is
shown in b, c and d.

matozoon has been taken into the egg, as it is apt to be concealed
in the depression of the membrane; but this is not the case. The
stage of best development of the depression, corresponding to

370 FEANK R, LILLIE

the complete retraction of the entrance cone, is about twenty-two
to twenty -five minutes after fertihzation (fig. B, b, c, d).

4. The perivitelhne space then narrows around the entire
egg, and the depression of the membrane in which the spermato-
zoon is seated disappears; in consequence, the spermatozoon
again becomes prominent externally.

5. It remains prominent for ten or fifteen minutes (about
forty to fifty minutes after insemination), and then disappears
rather abruptly within the egg as though some resistance had
given away. Its penetration coincides with the late anaphase of
the first maturation division; in a few cases it may be a little
earlier or a little later.

The egg is changing form at this time and in consequnce the
perivitelline space is often widened locally, especially in the ani-
mal hemisphere; if this happen in the region of penetration, which
may be any part of the egg, strong cytoplasmic strands are drawn
out between the membrane of the egg and the point of penetra-
tio.a, showing that the egg membrane and the cytoplasm are
actually fused here.

To repeat and extend the observations on the living egg several
series of eggs were preserved every five minutes from the time of
insemination in Meves' fluid. The study of the sections confirmed
and extended the above observation on the living egg as follows :

Ten minutes after insemination the entrance cone is quite
well formed and the spermatozoon is clearly seen outside, separ-
ated from the entrance cone only by the thickness of the vitelline
membrane with which it is in contact.

Fifteen minutes after insemination essentially the same con-
dition persists. The entrance cone is homogeneous, shading off
into the surrounding yolk-filled protoplasm. It stains very dark
in iron haematoxylin. The head of the spermatozoon appears
exactly as before.

Twenty minutes after insemination the entrance cone has flat-
tened out, but the spermatozoon is stiH external to the membrane.
The substance of the entrance cone is, however, as readily recog-
nized as when it projected above the surface of the egg.

STUDIES OF FERTILIZATION 371

About thirty-seven minutes after insemination (metaphase
of first maturation division) the sperm is still readily found on
the exterior of the vitelline membrane external to the substance
of the entrance cone which is now lens shaped. The substance
of the entrance cone is homogeneous and it stains less than before ;
it is sharply marked off from the unaltered egg cytoplasm by a
layer of small basophile granules. In the center of its external
face is a sharply differentiated granule which stains intensely
black in iron haematoxylin, and which is connected to the sperm
head by a fine thread passing through the vitelline membrane;
penetration has already begun.

Forty-three minutes after insemination (late metaphase of
the first maturation division) the entrance cone sinks into the
egg-cytoplasm, and the head of the spermatozoon begins to be
drawn within the egg in the form of a thick thread, less than one-
third the diameter of the sperm head, however. The sperm
nucleus is being drawn through the small perforation in the vitel-
line membrane.

Forty-eight minutes after insemination (stages of anaphase
of the first maturation division), nearly all of the sperm head is
drawn into the egg in the form of a thick thread several times
longer than the original sperm head. Before the head is entirely
within the egg its inner end begins to swell and becomes vesicular.
The entire entrance cone penetrates the egg-protoplasm always
retaining its original connection with the apex of the spermato-
zoon, so that the original orientation of the sperm is preserved
and may be readily recognized after penetration.

Fifty-four minutes after insemination (telophase of the first
maturation division), the entire head of the spermatozoon is
within the egg. The tail and middle piece usually remain without.

As I intend to publish a separate account of the interesting
details of penetration of the spermatozoon, and as the later stages
do not concern the present problem, I shall simply say, there-
fore, that as the united sperm-head and entrance cone penetrate
farther into the egg cytoplasm, they rotate in such a way that the
entrance cone which was originally in advance of the sperm nu-
cleus comes to lie behind it. During the rotation the sperm

372 FRANK R. LILLIE

aster arises from the pole of the sperm nucleus opposite the en-
trance cone, thus in the position of the original middle piece.

Morgan has recently ('10), with entire justice as it appears
to me, taken a stand against the current view that penetration
of the sperm is due to mechanical boring into the egg. He believes
that the presence of the sperm calls forth a reaction on the part
of the egg that leads to the absorption of the former. There can
be no question that the latter conception fits the case of Nereis
much better than the former. In the first place the spermatozoon
is absolutely motionless after its attachment to the membrane ; in
the next place the formation of the entrance cone shows a verj^
decided reaction on the part of the egg to the presence of the
spermatozoon; in the third place the retraction of the spermato-
zoon into a depression of the membrane is due to the retraction
of the entrance cone; and finally, as I shall show in a subsequent
cytological study, the inclusion of the spermatozoon within the
egg appears to be due to activity of the substance of the entrance
cone, and not to active penetration by the spermatozoon. The
spermatozoon does not penetrate the egg, it is drawn in or en-
gulfed.

6. The later history of the sperm nucleus. The sperm amphi-
aster is visible in the preparations all through the period of the
second maturation division (fig 4). After the formation of the
second polar body the sperm-nucleus begins to enlarge and the
amphiaster gradually wanes, but it may be recognized up to
the time of contact of the germ nuclei. The centrosomes of the
first cleavage spindle then begin to appear. Whether or not they
are continuous with those of the sperm amphiaster is a question
which I shall take up in the next study of this series. The cleav-
age asters rapidly become very large and conspicuous (figs. 5 and
6). During the growth of the germ nuclei a considerable number
of large granules staining strongly in iron haematoxylin appear in
their immediate vicinity.

The main point of these observations on the normal fertili-
zation, both in the living eggs and also in section, is to demon-
strate for elucidation of the experiments following: (1) That
membrane formation precedes penetration of the spermatozoon
by a long time. (2) That the spermatozoon does not penetrate

STUDIES OF FERTILIZATION 373

the vitelline membrane and enter the egg until at least forty to
fifty minutes after insemination, although its attachment to the
membrane takes place immediately. (3) That the presence of
the sperm-nucleus is readily demonstrable in all stages after
penetration.

B. Removal of the spermatozoon after membrane formation

In the summer of 1909 I was studying the effects of centri-
fuging on the egg of Nereis with the aim of getting more data
on the problem of polarity and the theory of formative stuffs.
It soon became apparent that the effects of centrifuging varied
with the stage of development, and so several series of experi-
ments were made in which the eggs were centrifuged at regular
intervals from before fertilization up to the time of the first
cleavage.

The effects of centrifuging may be divided into three cate-
gories: (1) A certain proportion of centrifuged eggs develop
approximately normally, the percentage varying greatly with
the stage of centrifuging. (2) A certain proportion of eggs,
varying at different stages, segment more or less abnormally,
sometimes extremely so {e.g. meroblastic), and produce embryos
with more or less pronounced abnormalities. (8) At certain
stages of centrifuging a variable proportion of eggs fails to carry
out even the first cleavage. The investigation of the causes of
such failure to segment revealed the fact that it was owing to
the removal of the spermatozoon after membrane-formation.
It is the evidence for this statement that is now under considera-
tion.

The results with reference to failure to segment were, in gen-
eral, as follows :

1. If unfertilized eggs were centrifuged and then fertilized,
all segmented, and a large percentage tended to develop quite
normally.

2. A disturbing factor comes in shortly after insemination,
owing to the fact that when the jelly is first secreted by the eggs
it is so viscid that the eggs stick together in the bottom of the
centrifuge in a mass which cannot be separated into its constitu-

374

FRANK R. LILLIE

ent eggs. The extreme viscidity of the jelly gradually dis-
appears, and after about twenty minutes from insemination, the
eggs no longer mat together. It is therefore difficult to investi-
gate the effects of centrifuging on the developmental capacity
of the eggs during the first ten or fifteen minutes after insemin-
ation. However, when the viscid stage begins to pass away
and eggs can be separated from the mass for examination, the
majority are found to undergo segmentation, as many as 98 per
o

40
60
80

f' \

20 30 40 50 60 70

Fig. C. The effects of centrifuging on the power to segment in Nereis. The ab-
scissae represent minutes from the time of i semination, the ordinates the percen-
tage of eggs dividing. At position a penetration of the spermatozoon is just
completed in most of the eggs. At position b the first polar body is extruded.
Data from experiment 2, 1910.

cent in one case (experimeat 2, 1910) twenty-one minutes after
insemination.

3. For about the next thirty minutes (twenty-one to fifty-
three minutes after insemination) centrifuging tends to inhibit
the cleavage of a certain proportion of the eggs which gradually
increases up to about thirty-seven minutes after insemination
and then decreases again. For instance, in experiment 2 of
1910, of the eggs centrifuged twenty- one minutes after insemin-
ation 98 per cent segmented; twenty-six minutes after insemin-
ation 36 per cent segmented; thirty-two minutes, 33 per cent;
thirty-seven minutes, 21 per cent segmented; forty-three min-
utes, 26 per cent segmented; forty-eight minutes, 52 per cent
segmented; fifty-three minutes, 75 per cent segmented; fifty-

STUDIES OF FERTILIZATION

375

eight minutes, 90 per cent segmented; sixty-three minutes,
90+ per cent segmented; sixty-nine minutes, 95+ per cent seg-
mented; control eggs, 99 per cent segmented. (See Fig. C.)

4. From this time on nearly all of the eggs segment after
centrifuging until, during the anaphase and telophase of the first
cleavage spindle, centrifuging again tends to inhibit cleavage.

The following table (Experiment 29) gives a fairly typical
series of results. There were twelve such experimental series
in all, more or less complete, giving concordant results except
that in some at the period corresponding to 29D, 90 to 98 per
cent were so affected that they failed to segment. On either
side of this critical period there is a decreasing susceptibility to
such injury by centrifuging.

Experiment 29
September 8, 1909

DESIGNATION

29..

29 A.

29B.
29C.

Control (not centri-

fuged)
Before insemination

20 minutes
30 minutes

29D i 41 minutes

29E.
29F.
29G.
29H.
291.
29 J.
29K.

51 minutes
66 minutes
79 minutes
95 minutes
114 minutes
121 minutes
127 minutes

29L 149 minutes

PERCENTAGE SEGMENTED

100 per cent

100 per cent practi-
cally
Majority
Majority
20 to 30 per cent
70 to 80 per cent
90 to 95 per cent
100 per cent
100 per cent
100 per cent
Some unsegraented
Less than majority
and these irregu-
lar
Most segmented fur-
ther

Eggs matted loosely

Centrifuged during
process of Ist
cleavage

Centrifuged in 2-cell
stage

Two major processes are going on in the egg at this time,
viz. : maturation and fertilization. The injury is not primarily
to the process of maturation, for the eggs that do not segment
form the polar bodies; nor is it probable that there is a general

37(

FRANK R. LTLLIE

systemic injury to the egg protoplasm at this time not received
at other times, when the fact that maturation continues and polar-
ity is preserved in these eggs, is considered. It would, there-
fore, appear probable that the injury is to the process of fertiliz-
ation itself, and this conjecture is completely confirmed by cyto-
logical study. The most conclusive experiment in this respect
is no. 27, the details of which are as follows:

The eggs were fertilized at 9:28 a.m., September 4, 1909.
Some of these were kept for control and all segmented normally.
The remainder were centrifuged about 60 x 120 revolutions at a
radius of 6 cm. in one minute, at the following times : 27A at
9 :58 A.M. ; 27B at 10 :03 ; 27C at 10 :08 ; 27D at 10:12; 27E at 10:16.
About 20 per cent of 27A segmented, 5 to 10 per cent of 27B, 20
per cent of 27C, 50 to 60 per cent of 27D, and 75 to 90 per
cent of 27E. Samples of the control and of each of 27A, 27B,
27C, 27D and 27E were preserved at 10:31 and 10:43 a.m.-

EXPERIMENT 27

Eggs fertilized at 9:28 a.m., Sept. 4, 1909

TIME AFTER

DESIGNA-

CENTRIFUGED

INSEMI-

SAMPLES PRESERVED

LIVING EGGS

TION

NATION

27 Con-

Xot centrifuged

(1) 10:311 A.M.

All segmented

trol..

(2) 10:45 §A.M.

27A . . .

60 X 120 rev. in

(1) 10:30 A.M.

About 20 per cent

1 min. 9:58 a.m.

30 min.

(2) 10:431 A.M.

segmented

27B....

60 X 120 rev. in

(1) 10:30J A.M.

About 5-10 per

1 min. 10:03 a.m.

35 min.

(2) 10:431 A.M.

cent segmented

27C....

60 X 120 rev. in

(1) 10:30iA.M.

About 20 per cent

1 min. 10:08 A.M.

40 min.

(2) 10:44 A.M.

segmented

27D...._

60 X 120 rev. in

(1) 10:31 A.M.

About 50-60 per

1 min. 10:12 a.m.

44 min.

(2) 10:441 A.M.

cent segmented

27E....

60 X 120 rev. in

(1) 10:31 A.M.

About 75-90 per

1 min. 10:16 a.m.

48 min.

(2) 10:45 A.M.

cent segmented

^Since the above was written this experiment has been repeated (Exp. 2, '10),
with the added precaution of preserving a sample of the normal eggs corresponding
to each stage of centrifuging, in order to make certain of the stages of fertilization
in each case. The results completely confirm those of experiment 27, and the
sperm was found to be external in the most critical stage (37 minutes after insemi-
nation in this experiment; see page 371).

STUDIES OF FERTILIZATION 377

Cytological studj' of the twelve lots of preserved eggs showed
stages ranging from the metaphase of the second maturation
spindle to the prophase of the first cleavage, the earlier stages
of course being found in lot 1 in each case.

In the control lots it was easy to demonstrate the sperm
nucleus at all stages to the formation of the first cleavage spin-
dle. The sperm nucleus is rendered particularly conspicuous
during the second maturation division by the large amphiaster
that accompanies it (fig. 4), both lying in the yolk-free proto-
plasm. After the formation of the second polar body the sperm
amphiaster gradually fades, but the sperm nucleus can be
recognized by its position and by the remnants of radiations
up to the time of union of the two germ nuclei; and in the later
stages its presence may be inferred by the degree of development
of the cleavage amphiaster and the number of chromosomes.
There is, therefore, no time from the beginning of the second
maturation division up to the formation of the first cleavage spin-
dle when the presence of the sperm nucleus cannot be readily
demonstrated.

In the study of the serial sections of the control eggs I found
no egg in which, all sections being present, the sperm nucleus
could not be demonstrated. In the serial sections of 27A, the
sperm nucleus could be recognized in only about 37 per cent of
the eggs; in 27B in only 10 per cent to 20 per cent; in 27C in about
25 per cent; in 27D in about 53 per cent; in 27E in about 76 per
cent. The stages of maturation of lots A to E corresponded very
closely with the stages of maturation of the control eggs killed
at the same time.

It is a relatively simple matter to demonstrate the presence
of the sperm nucleus, for a single positive observation suffices; but,
to be sure of the absence of a sperm nucleus from any particular
egg, it is necessary to examine practically every section of the
egg, and the absence of two consecutive sections is sufficient
reason for excluding an egg from the count. This may be one
reason why the number of eggs in the different lots shown to
contain sperm nuclei tends to be somewhat larger than the esti-
mate of the number of eggs that segmented. Another reason

378

FRANK R. LILLIE

probably is that a sperm nucleus may persist to a certain
extent even if injured and unable to produce the full fertilizing
effect and cause cleavage.

A third reason for discrepancy in the results is that abnorm-
alities of maturation may be produced by centrifuging which
render the determination of the sperm nucleus more difficult
than usual. It frequently happens that, as the first matura-
tion spindle is driven from the surface by the centrifugal force,
it divides before it reaches the surface again, producing two
maturation nuclei within the egg. The two second maturation
spindles may then unite to form a tetraster, one pole of which
approaches the surface and a single polar body is formed, leaving
three nuclei within the eggs (fig. 7). Such eggs were readily
recognized by the absence of the first polar globule, and by the
presence of the extra nuclei. But it was sometimes difficult to
determine in certain stages whether there were only three nuclei,
the sperm nucleus being absent, or four, the sperm nucleus being
present.

A fourth reason for a certain discrepancy between the esti-
mate of the number of eggs that segmented and the number
determined to have sperm nuclei might be that at the time the
experiment was made the importance of exact determination of
the number of eggs that segmented was not realized, and the
determination was made only roughly. Putting the results
side by side we have :

PERCENTAGE OF EGGS
OBSERVED TO DIVIDE
IN LIVING CONDITION

PERCENTAGE DETERMINED
BY SERIAL SECTIONS TO
POSSESS SPERM NUCLEI

27 Control

All

About 20 per cent
5 to 10 per cent
About 20 per cent
About 50-60 per cent
About 75-90 per cent

All

27A

About 37 per cent
About 10-20 per cent
About 25 per cent
About 53 per cent
About 76 per cent

27B

27C

27D

27E

Considering the various sources of error, the agreement is
very close except in 27A. But in this case we do not have to

STUDIES OF FERTILIZATION 379

explain why eggs in which the sperm nucleus was absent segmented
but on the contrary, why certain eggs that possessed the sperm
nucleus failed to segment, which is a very different thing. There
is no evidence that any egg in which the sperm nucleus was absent
succeeded in dividing.

The general conclusion that removal of the spermatozoon at the
times noted in the experiments involves incomplete fertilization,
is sufficiently demonstrated by the results.

Let us call the stage at which the spermatozoon is eliminated
in the greatest proportion of eggs, the critical period. The
exact number of minutes from the time of mixing eggs and sperm
to this stage varies of course through the season, owing to the
variations of temperature. Moreover, it is not exactly deter-
mined in all experiments, for in some the stages of centrifuging
fall on either side of it. This being understood, we may note that
in eight experiments the critical period occurred at from 25 min-
utes to 40 minutes after fertilization. This is quite a wide
variation, but when the time is represented as a fraction of the
entire period between fertilization and the first cleavage, it
is found that in all cases the period up to the critical period is
between 27 and 33 per cent of the total time up to the first cleav-
age. It is obviously a corresponding stage in all cases, for the
observed differences fall within the chances of error, viz : that the
critical period is hit exactly in only very few experiments, and
that the time of beginning of the first cleavage must be stated
rather arbitrarily on account of the variation in rate of individ-
ual eggs.

The critical period occurs shortly before the penetration
of the spermatozoon into the egg. We noted in the first part
of this paper that the penetration of the spermatozoon is extreme-
ly gradual; my observations on this point, both from the study of
the living egg and also of sections, show that it requires forty to
fifty minutes for the head of the spermatozoon to disappear
through the membrane.

As the most critical period comes in the great majority of
experiments from thirty-five to forty minutes after insemina-
tion, it is obvious that the spermatozoon is in some way prevented

JOURNAL OF MORPHOLOGY, VOL.

380 FRANK R. LILLIE

from entering the egg. The explanation is comparatively simple;
the spermatozoon is imbedded in the jelly by which the egg is
surrounded. When the jelly is first formed it is very viscid, and
adheres to the eggs during centrif uging so that they mat together
in the centrifuge. However, this stage passes and the result of
centrif uging is then to separate the jelly from the eggs. In many
cases the jelly carries off the attached spermatozoon with it.
After penetration this can of course no longer happen. The curve
of variation of the per cent of eggs centrifuged before penetration
that fail to segment is due to the following factors: (a) At about
twenty-five minutes the sperm head is sunk in a deep depression
of the membrane and hence is less likely to be torn away by the
jelly; (fe) the change in consistency of the jelly presumably
extends from without inwards; hence at first the innermost layer
in which the spermatozoon is imbedded tends for a time (also
presumably) to remain with the egg; (c) the time of penetration
varies somewhat in any lot of eggs. These facts together would
explain why the percentage of eggs that fail to segment after cen-
trifuging rises to a maximum and sinks again to a minimum in
the successive stages of centrifuging.

The fact that centrifuging inhibits cleavage in a small per cent
of the eggs from fifty to fifty-five minutes after insemination,
leads me to suspect that in some cases the sperm may be destroyed
after its penetration into the egg. In experiment 2, 1910, for
instance, cleavage was inhibited in 25 per cent of the eggs centri-
fuged fifty-three minutes after f ertihzation ; in the control eggs
killed at the time of centrifuging, penetration of the sperm is
completed in the great majority of eggs, though it is found exter-
nal still in a very few ; it is difficult to estimate the per cent of the
latter, but the impression is that it is less than 25 per cent. How-
ever, it is impossible to confirm this, and I mention the matter
here to call attention to the error in my first paper read before the
Central Branch of the American Zoological Society (Abstract in
Science, vol. 18, p. 36, May, 1910), in which I stated that the de-
struction of the sperm nucleus followed penetration. A renewed
study of the penetration has proved that this is not the case,
usually at least.

STUDIES OF FERTILIZATION 381

We are not, of course, free to infer that fertilization is complete
as the stimulus to development immediately after penetration
of the spermatozoon. The experiments prove directly that in
the egg of Nereis the stimulus of the spermatozoon as the impulse
to development consists of two distinct parts: (1) an external
stimulus that causes membrane formation and which is sufficient of
itself to induce the maturation ; (2) an internal stimulus dependent
on penetration of the spermatozoon. How long after penetration
fertilization is still incomplete cannot be decided on the basis of
these experiments.

In concluding this section, we may note that the cause of
failure to segment following centrifuging during the anaphase of
the first cleavage is an entirely different one. The cause in
this case is the breaking up of the karyokinetic figure and func-
tion, and dispersing the chromosomes. This is readily demon-
strated in sections. In the case of eggs centrifuged at the ' critical
stage', the sections show that the maturation spindle receives
no injury from centrifuging, but appears coherent and normal in
the sections. The sections of eggs centrifuged during the ana-
phases of the first cleavage show the chromosomes dispersed
through the cytoplasm and the cleavage spindle no longer
coherent, but broken up. In the first case the cause of failure
to segment is elimination of the sperm-nucleus, as shown by
study of series 27. This cannot be the cause in the second case,
and a sufficient explanation of the failure to segment is found in
the destruction of the karyokinetic figure.

C. Comparison of co??ipletely and partially fertilized eggs in later

We have noted so far that definite proportions of eggs centri-
fuged at definite periods in the process of fertilization fail to
develop a sperm-nucleus, and that similar proportions of the
same lots of eggs when left to develop fail to undergo segmenta-
tion. The facts (1) that all the control eggs of the same lot seg-
ment, and (2) that the centrifuged eggs that fail to segment,
nevertheless had formed the fertilization membrane and under-

382 FRANK R, LILLIE

gone maturation, prove that the unsegmented eggs had received
at least the first stimulus of fertilization. It was also shown that
the critical period for suppressing segmentation by centrifuging
occurs at a time shortly before entrance of the spermatozoon, and
that it is due to prevention of penetration. The partially fer-
tilized eggs, therefore, resemble the normal ones in the fact that
membrane formation and the first stimulus to development are
called forth by action of the spermatozoon, and they differ from
the normally fertilized eggs in that the internal egg protoplasm has
not received the direct stimulus of the spermatozoon. A cy to-
logical examination of such eggs could not fail to be of interest
and might give some clue to the internal function of the spermato-
zoon in fertilization.

Both polar bodies form regularly in such eggs as already noted,
and the egg-nucleus (female pronucleus) arises and attains the
same size as in normally fertihzed eggs. The chromosomes of the
first cleavage spindle then form in the usual fashion and at
the usual time, accompanied by disappearance of the nuclear
membrane. But, whereas, in the presence of a sperm nucleus,
cytoplasmic asters accompany these processes and a spindle
rapidly arises during the prophases of the first cleavage, in the
absence of the sperm nucleus there is absolutely no sign of c\^tas-
ters or evidence of spindle formation. The chromosomes lie
naked in the cytoplasm surrounded by a clear area (fig. 7).

Each chromosome then splits longitudinally in the usual
fashion, but the halves do not separate. At the time of the
telophase of the normal first cleavage there is a tendency to
scattering and breaking up of the chromosomes. When the nor-
mal eggs have reached the two and four-celled stages, the scat-
tering and breaking up of the chromosomes have progressed
much farther in the unsegmented eggs, and in the course of two
or three hours there remains no differentiated nucleus or chro-
mosomes, but only numerous chromatic granules scattered
throughout the cytoplasm.

The behavior of the partially fertilized eggs may be compared
on the one hand with that of normally fertilized eggs and on the
other with that of eggs caused to mature by centrifuging. As com-

STUDIES OF FERTILIZATION 383

pared with the former, the chief difference observable by cytologi-
cal methods is the entire absence of the achromatic part of the
karyokinetic figure. The differences in later stages may be
conceived as secondary effects of this defect or of the conditions
determining such defect. When eggs are caused to mature by
centrifuging the process begins as in normally fertilized eggs by
the breaking down of the cortical layer and formation of the
jelly; the germinal vesicle ruptures and the two maturation
divisions follow. After the completion of the maturation
the chromosomal vesicles of the egg nucleus usually fail to unite
perfectly, and in a httle while they separate and scatter in the
cytoplasm without formation of chromosomes, so that each egg
appears to possess a considerable number of small nuclei. In a
few cases the first indications of chromosome formation may be
observed in the vesicles shortly after maturation but not later.
Subsequently the chromosomal vesicles appear to dissolve in the
cytoplasm liberating small chromatic nucleoli.

The partial stimulus of the spermatozoon is thus somewhat
more effective than the mechanical shock of centrifuging, though
both produced the same initial changes, apparently equally well.
This may possibly be due to entrance of a small amount of matter
from the spermatozoon; for at the critical period the perforator-
ium of the sperm has penetrated the membrane and is imbedded
in the entrance cone.

D. General discussion

The general conclusion that the function of the spermatozoon
in the stimulus to development involves at least two factors
has already been clearly stated by Boveri ('07) and Loeb ('096) :
According to Loeb, one factor is the cytolysis of the ''very thin
cortical layer of egg"; but while this stimulates development,
the latter is often abnormal and therefore usually comes to a
halt. A second process is necessary to ensure more normal and
lasting development (Loeb '096). Apparently Loeb is not very
clear concerning the nature of the second factor, but is inclined
to regard it as inhibiting the cytolysis which he conceives to be

384 FRANK R. LILLIE

begun as the first factor of the developmental stimulus. This
conclusion was formed as a result of two kinds of experiments:
In the first Loeb found that the best results in artificial parthen-
ogenesis are obtained, in the egg of a Californian sea-urchin, by
a double treatment : first using a cytolytic agent and then follow-
ing it by treatment with hypertonic sea-water, or by inhibiting
oxidation for a while. In the second class of experiments Loeb
and Elder found that mere membrane formation might be induced
in sea-urchin eggs by external contact of starfish spermatozoa,
but farther development did not take place except in the relatively
few cases in which the spermatozoon entered the egg (Loeb
'09b, p. 249), or unless the eggs were treated after membrane
formation with hypertonic sea-water. Although this experiment
is complicated by the hybridizing, yet it demonstrates the same
distinction between external and internal functions of the sperma-
tozoon in fertilization that I have shown for Nereis by a different
method.

Artificial parthenogenesis may be induced in the sea-urchin
egg without membrane formation and this fact appears to me to
indicate that the internal function of the spermatozoon is prob-
ably at least as fundamental as the external function (membrane
formation), though, as Loeb points out, development without
membrane formation takes place in a less normal fashion than
after membrane formation. But inasmuch as we may have
membrane-formation without development, and development
without membrane formation, it would seem premature at least
to consider membrane formation as the chief function of the sper-
matozoon in fertilization.

The experiments described in this paper show that m Nereis
fertilization by the spermatozoon is incomplete after the forma-
tion of the membrane. The question then arises, when is the
function of the spermatozoon in fertilization completed? Zieg-
ler's and Wilson's experiments show that it is incomplete even
some time after entrance of the spermatozoon. Ziegler's experi-
ments ('98) consisted "in so constricting the egg of the sea-urchin
after penetration of the spermatozoon that the one part contains
the sperm nucleus, the other part the female sex-nucleus. The

STUDIES OF FERTILIZATION 385

part that contains the sperm nucleus undergoes cleavage and
develops farther; in the other part the female sex-nucleus under-
goes remarkable transformations, dissolving and reappearing, a
process which is repeated several times." In spite, therefore,
of the presence of the sperm-nucleus in a constricted part of
the same egg, the part containing the egg nucleus was not
fully fertilized. It made abortive attempts at division, but the
karyokinetic figure was too feeble to carry the process through.

Wilson observed ('03) that if the fertihzed eggs of Cerebrat-
ulus be cut in two shortly after the penetration of the sperm-
atozoon ''only a single fragment develops even though the
fragments be refertilized immediately after the operation. In
such cases it is almost invariably the nucleated fragments that
develop, but in a very few cases I have observed that the enu-
cleated fragment develops, ' while the nucleated one forms the
polar bodies, but proceeds no further." By the nucleated frag-
ment in this case Wilson means the fragment containing the
maturation spindle. Farther on he adds ''The few cases in
which the enucleated fragment of the bisected fertihzed egg
develops are doubtless those in which the plane of section sep-
arates the sperm-nucleus from the egg-nucleus." This is indeed
the only possible explanation. In such cases the fragment con-
taining the egg-nucleus is only partially fertilized.

Boveri has also observed that if freshly fertilized sea-urchin
eggs be broken into fragments by shaking, certain of the frag-
ments contain the egg nucleus alone. If such fragments are not
subsequently entered by a spermatozoon, the nucleus enlarges,
dissolves and reappears again; but they do not segment ('96).
Later he observed that such pieces may divide at least twice ('02).

These experiments demonstrate that fertilization is still
partial even some time after entrance of the spermatozoon into
the egg; but they do not show at what stage it is complete.
Boveri's very interesting observations on 'partial fertilization'
in the sea-urchin egg ('88 and '90) carry the solution of the prob-
lem a step farther (cf. also Teichmann '02). In the experiments
which furnished the material for his observations both eggs and
sperm were weakened, the former by standing for fourteen hours

386 FRANK R. LILLIE

in sea-water and the latter by treatment with KOH prolonged
to a stage in which only a small percentage of the spermatozoa
continued to move. Under these conditions in a large number
of eggs the sperm aster separated from the sperm nucleus, which
was usually left on one side, and proceeded alone to conjugate
with the egg-nucleus. Thereupon the cleavage spindle formed
with the egg-nucleus alone, and segmentation of the egg ensued.
In the four-cell stage usually, but sometimes in the two or eight-
cell stage, the sperm nucleus united with one of the segmentation
nuclei. Boveri concludes from this and other results that the
fertihzing action of the spermatozoon consists in the introduction
of a centrosome into the egg.^ When this has united with the egg
nucleus, with or without participation of the sperm nucleus, fer-
tilization would be complete; or with the sperm nucleus alone in
merogony it likewise completes fertilization.

Boveri has used the term 'partial fertilization' for the phenom-
enon just described, although he admits that it is a misnomer.
It is unfortunate that such a significance should have come to be
attached to the expression, because, as has been shown, my own
results and those of Ziegler, Wilson and Boveri himself prove that
partial fertilization in the literal sense really occurs. The vari-
ous stages of partial fertilization as shown by the results in the
literature on the subject are:

1. External contact alone by the spermatozoon producing,
a. Formation of the fertilization membrane, (Loeb and Elder

for sea-urchins, Lillie for Nereis).

h. Maturation and formation of the chromosomes from the
egg nucleus without spindle, (Lillie for Nereis).

c. Maturation and cleavage to stereoblastula, (Bataillon:
hybrid union of eggs of Pelodytes punctatus and Bufo calamita
with sperm of Triton alpestris).

2. If the spermatozoon be removed shortlj'' after entrance,
a. Maturation alone may result, (Wilson on Cerebratulus) .

^Herbst ('07 p. 202 and '09 p. 277) interprets Boveri's 'partial fertilization' as
a combination of parthenogenesis and fertilization. Such an interpretation does
not, however, explain Boveri's account of the behavior of the sperm-centrosomes.

STUDIES OF FERTILIZATION 387

b. An abortive karyokinetic figure may form with the egg
nucleus alone, (Ziegler and Boveri on sea-urchins).

c. In some cases at least two cleavages may result, (Boveri
on sea-urchins). Boveri's so-called 'partial fertilization, is really
full fertilization, so it does not come in this series.

The difference in reaction of the egg-cytoplasm to its own
nucleus and to the introduced part of the spermatozoon can be
explained on only one of two grounds; either in the general sense
of Boveri's theory on purely morphological grounds, or on the
ground of a chemical difference, presumably of sexual origin,
between the egg on the one hand and the sperm on the other.
The latter form of interpretation seems to me preferable because
it is a physiological interpretation which takes cognizance of the
sexual factor in fertilization,

Boveri's theory of fertilization rests on the identification of the
centrosomes of the sperm aster with a definite formed element
(centrosome) introduced into the egg by the spermatozoon;
but it has never been demonstrated in any case in all the liter-
ature on the subject of fertilization that the centrosomes of the
cleavage spindle, or indeed of the sperm aster, are derived from
any definite formed element of the spermatozoon. Boveri
himself admits this in his sixth cell study ('07, page 266) ; and so
long as definite proof of the continuity of the so-called sperm-
centrosomes from the spermatid up to the formation of the first
cleavage spindle is lacking, all of Boveri's observations are open
to another interpretation than the one he has given; to the inter-
pretation, namely, that the sperm asters represent a reaction of
the egg cytoplasm to a male element, or at least a foreign element,
represented for the most part by the sperm nucleus.

The biological analysis of fertilization seems to me to rest now
upon the problem of the origin of the sperm aster in the egg.
More crucial evidence is needed on this point, and I do not believe
that any refinement of cytological technique will give the result.
Experimental evidence is needed; either, as Boveri ('88) sug-
gested, the introduction of a non-nucleated spermatozoon in the
egg to prove whether or not the sperm asters would arise without
the nucleus and fertilize the egg, or the introduction of only the

388 FRANK R. LILLIE

anterior part of the sperm into the egg to prove whether or not
the sperm nucleus without the centrosome, which is contained
in the middle piece of the spermatozoon, would cause the produc-
tion of asters in the egg-cytoplasm.

As Boveri, among others, has pointed out, there is not only-
one, but several stages of inhibition in the history of the egg.
This may be illustrated by noting the stages at which in the eggs
of various animals the need for fertilization arises. In some eggs
it is before the rupture of the germinal vesicle {e.g., Nereis),
in others at the time of the mesophase of the first maturation
spindle {e.g., Chaetopterus and Cerebratulus), in others again
after the formation of the first polar body {e.g., Amphioxus,
amphibians), in others again, not until after the formation of
both polar bodies {e.g., sea-urchin). There is no doubt that the
last stage of inhibition is the most difficult one to overcome, both
because many eggs pass by the earlier stages without apparent
specific stimuli and also because it is possible to cause eggs that
normally stop at the first or second stage of inhibition to pass on to
the last stage by stimuli that are ineffective when this stage is
reached, {e.g., Nereis as noted in this paper and Chaetopterus as
noted in various earlier papers).

The nature of the inhibition that causes the need for fertiliza-
tion is a most fundamental problem. Is it the same in all these
cases, i.e., a gradually increasing inhibition that may be effect-
ive before maturation, but in some cases not until maturation
has progressed a certain distance, or even until it is complete?'*
If it is the same cause at all these stages then it is certain that the
need for fertilization is not due to any defect of the egg-centro-
somes, for the pause takes place in Chaetopterus (e.g.) while the
egg-centrosomes are at the very height of their activity. If,

^Bataillon ('10) holds the view that the nature of the inhibition is the
same whether the arrest is at the stage of the resting nucleus or in the height of
karyokinesis. He holds the view that the inhibition is due to accumulation of ex-
cretory products and that the stimulus to development is essentially a process of
elimination. Bataillon's paper was received after my own was completely written.
His interesting results will be considered more fully in my next paper. In Part
I of the present paper (last paragraph) I have presented a view similar in some
respects to Bataillon's.

STUDIES OF FERTILIZATION 389

therefore, we are to hold to the theory of Boveri in its Hteral
sense, we must beheve that there are different kinds of inhibition.
However, it is, I beheve, simpler and more logical to hold that
the inhibition differs only in intensity at these various stages;
and this point of view seems to be supported by the fact that the
same stimulus which at a lower intensity will cause only matur-
ation to take place in Chaetopterus, at a higher intensity will
cause differentiation also to proceed, though in this case without
cleavage. Boveri ('07), however, holds that there are different
kinds of inhibition, that the postulated degeneracy of the egg-
centrosomes after maturation is in a sense the more primitive,
and that other kinds have been secondarily acquired, a point of
view that gives a more or less definitely teleological aspect to
the question.

From a physiological point of view we might inquire, what are
the conditions that cause the postulated sudden degeneration
of the egg-centrosomes? Such a condition if found, would be
nearer the fundamental cause of inhibition of the egg and it might
turn out to be the same cause that conditions in so many cases
an earlier arrest of activities in the egg.

The experiments on artificial parthenogenesis are sometimes
regarded as involving the entire problem of fertilization. But
if it be true, as many believe, that biological fertilization, (if I
may be pardoned such an expression) is fundamentally a sexual re-
action, then the physico-chemical analysis of fertilization must
compass the entire problem of sex, which is much wider than the
problem of parthenogenesis. The physico-chemical analysis
of fertilization has dealt, up to the present exclusively, with the
latter problem, and for this reason the earlier title of such studies
'artificial parthenogenesis', seems to me much more fitting than
'chemical fertilization' which is sometimes loosely used. From
the zoological point of view, at least, parthenogenesis and fertil-
ization are not interchangeable functions. There is a factor
present in fertilization which is absent in parthenogenesis, and
the latter is never the exclusive mode of reproduction among
animals. The biological analysis of fertilization therefore
involves problems that do not occur in the physico-chemical
analysis of parthenogene>-is.

390 FRANK R. LILLIE

LITERATURE CITED

Bataillon, E. 1910 Le probleme de la fecondation circonscrit par I'impreg-
nation sans amphixie et la parthenogenese traumatique. Arch,
de Zool. exp. et gen. 5 ser. Tome 6.

BovERi, Th. 1888 Ueber partielle Befruchtung. Sitz'b. d. Ges. fiir Morph.
u. Phys. in Munchen. Bd. 4, H. 2.

1890 Zellenstudien. Heft. 3. Ueber das Verhalten der chromatischen
Kernsubstanz bei der Bildung der Richtungskorper und bei der Befruch-
tung. Jena. pp. 32 ff.

1896 Zur Physiologic der Kern und Zelltheilung. Sitz'ber. d. Phjs.-
Med. Ges. zu Wurzburg, (cited from Teichmann — unfortunately the
original paper was inaccessible to me).

1902 Das Problem der Befruchtung. Jena, G. Fischer.

1907 Zellen-Studien. Heft 6. Die Entwicklung dispermer Seeigeleier.
Ein Beitrag zur Befruchtungslehre und zur Theorie des Kerns. Jena.
Gustav Fischer.

Fischer, Martin H. 1903 Artificial parthenogenesis in Nereis. Am. Jour.
Physiol, vol. 9, pp. 100-109.

Herbst, Curt 1907 Vererbungsstudien V. Auf der Suche nach der Ursache der
grosseren oder geringeren Ahnlichkeit der N^chkommen mit einem der
beiden Eltern. Arch. f. Entw'mech. Bd. 24, p. 185.

1909 Vererbungsstudien VI. Die cytologische Grundlagen der Ver-
schiebung der Vererbungsrichtung nach der miitterlichen Seite. I
Mittheilung. Arch f. Entw'mech. Bd. 27, p. 266.

LoEB, Jacques 1909a Ueber das Wesen der formativen Reizung. Berlin, Julius
Springer.

1909b Die chemische Entwicklungserregung des tierischen Eies.
Berlin, Julius Springer.

Morgan, T. H. 1910 Cross and self-fertilization in Ciona intestinalis. Archiv
f. Entw'mech. der Organismen. Bd. 30, ii, Theil.

Teichmann, Ernst 1902 Ueber Furchung befruchteter Seeigeleier ohneBeteil-
igung des Spermakerns — Jen. Zeitschr. f. Naturw. N. F. Bd. 30, p!105.

Wilson, E. B. 1892 The cell-lineage of Nereis — A contribution to the cytogeny
of the annelid body. Jour. Morph. vol. 6.

1903 Experiments on cleavage and localization in the Nemertine egg.
Arch. f. Entw'mech. vol. 16, pp. 417-418.

ZiEGLER, H. E. 1898 Experimentelle Studien tiber die Zelltheilung, II. Arch.
f. Entw. mech. vi.

PLATE 1

EXPLANATION OF FIOURKS

1 Axial section of an unfertilizod normal ovocyte of Nereis, fixed in Flem-
ming's fluid, weaker solution. In this fixing fluid the yolk granules swpll and tend
to run together. The oil drops are dissolved out in the preparation and are
represented by emptj^ spaces, v. in. vitelline membrane, c.l. cortical layer, from
which the jelly is formed.

2 Section of an egg of Nereis, fixed in Aleves' fluid five minutes after insemina-
tion. The Cortical layer is already somewhat reduced in thickness. The yolk
granules are not swollen. The oil drops are not dissolved out. The section is
approximately horizontal, c.l., remains of cortical layer; v.m., vitelline membrane.

3 Section of an egg of Nereis, fixed in Meves' fluid fifteen minutes after insem-
ination. The cortical layer has entirely disappeared, and the perivitelline space
is formed. The germinal vesicle is breaking down and the first maturation spindle
is forming, p.v. perivitelline space, v.m., vitelline membrane.

4 Section of an egg of Nereis, fixed in Meves' fluid fifty-seven minutes after
insemination. Only one centrosome of the sperm amphiaster is shown.

5 Section through the first cleavage-spindle of Nereis, normal, one hour and
twenty-seven minutes after insemination.

6 Tetrapolar second maturation spindle of Nereis. See text for description
(p. 378). Three egg nuclei are formed in such a case.

7 Egg nucleus of egg of Nereis in which the spermatozoon was removed by
centrifuging. The chromosomes of the first cleavage are formed, but there are
no asters. Cf. fig. 5.

All figures drawn with the camera lucida with Zeiss comp. oc. 6 and 2 mm.
horn, oil im. obj.

391

STUDIES OF FERTILIZATION

FRANK R. LILI.IE

JOURNAL OF MORPHOLOGY, VOL. 22, NO.

392

•i 4

1^ *^^i^

393

THE GROWTH AND DIFFERENTIATION OF THE

CHAIN OF CYCLOSALPA AFFINIS

CHAMISSO

WM. E. RITTER and MYRTLE E. JOHNSON

From the Laboratory of the Marine Biological Association of San Diego

TWENTY-FIVE FIGURES
FOUR PLATES

CONTENTS

Purpose of the research 396

1. Special 396

2. General 396

Brief description of the species 398

Measurements of the zooids of the wheels and of a portion of the chain not yet

transformed into wheels 399

Treatment of the quantitative data 406

Attempt to connect the formation of wheels with morphological, physiological,

and mechanical phenomena presented by the animals 414

1. Segmentation of the stolon, and the deploying point 414

2. Position and relation of the zooids in the chain from the deploying point '

to the twist 414

a Shifting of the zooids 414

b Peduncles and foot-pieces 417

c Emergence of the chain to the outside world 417

d Twist in the chain 418

e Reduction of the foot-pieces, first break in the chain, and formation

of the first wheel 418

3. Comparison of the rate of growth of the chain as a whole with the rate of

other animals 419

Discussion of the observations from the causal standpoint 420

1. Cause of the twist * 420

2. Unequal growth of zooids and foot-pieces as a factor in the breaking up of

the chain 422

3. Impossibility that the character of the blood supply to the zooids can be

the cause of the size scheme within the wheels 427

4. Unlikelihood that the wheel arrangement of the zooids in Cyclosalpa has,

as believed by Brooks, anything to do with the position of the first four
blastozooids of Pyrosoma 429

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

395

396 W. E. RITTER AND M. E. JOHNSON

The larger significance of such studies 431

1. Supplementing biological with quantitative observations 431

2. Natural periodicity in organisms and exacter methods in biological re-

search 432

3. The inadequacy of treating periodicity, generally, as an aspect of fluc-

tuating variation 440

Bibliography 444

PURPOSES OF THE RESEARCH

1. Special

One of us (Johnson, '10) has shown that the individuals of the
blocks into which the chains of blastozooids of Salpa fusiformis-
runcinata, S. cylindrica, and S. zonaria-cordiformis become dif-
ferentiated, fall into size schemes, or systems.

The question naturally arises, how general is this phenomenon
among salpae? The possibility that the wheel grouping in Cycio-
salpa corresponds to the block grouping in Salpa proper, occurs to
one rather readily in spite of the conspicuous differences between
the two. If this conjecture be right, we should expect to find a
size scheme of zooids in the wheels of Cyclosalpa similar to that
in the blocks of Salpa. That such a scheme exists in the wheels
even more pronouncedly than in the blocks, the sequel will show.

2. General

This much more evidence is consequently adduced favorable
to the idea of correspondence between the wheels and the blocks.
But what do we mean by correspondence? In a general sense the
blocks and wheels undoubtedly correspond: both are groups of
shnilar organisms similarly located with reference to the parent
zooid. This much of correspondence is recognizable to cursory
inspection. Does the discovery of a similar size scheme among
the zooids in the groups in the two species advance our interpreta-
tion of these organisms much if at all ? Does it amount to any-
thing more than a recognition of one more resemblance? Accord-
ing to the meaning that interpretation' and 'resemblance' have
in most later biological writing, we must probably say no. We

CHAIN OF CYCLOSALPA AFFINIS 397

must hold that unless the new correspondence includes some-
where what we hold as one or more 'causal factors' not much has
been accomplished. If, for example, we extend the inquiry to the
question of the dependence of the size scheme of the zooids upon
growth and other internal factors on the one hand, and upon en-
vironmental factors on the other; and if here, too, we find further
correspondence, our belief in the essential identity as we might
say, unless standing for extreme exactness of expression, would
be reached.

We shall see that the size scheme of the zooids in the wheels is
almost certainly foreshadowed before the wheels themselves are.
If this be so, then the resemblance between the Cyclosalpa chain
and the Salpa chain is considerably closer before than after the
wheels appear in the former. But it is difficult, if not impossible,
to attribute the block production in the Salpa chain entirely to
other than inherent factors, of which growth seems to be the most
immediate. So far, therefore, as we can rely upon our evidence
for the marking off of the Cyclosalpa chain into groups before the
wheels are formed we seem to have placed the notion of corre-
spondence between the wheels and the blocks on firm ground.

Evidence still more convincing perhaps, that the wheels and
blocks correspond in a strict biological sense, in the sense that both
are expressions of periodicity in growth, is found in the fact that
growth and development are observed to be periodic in so wide
a range of living beings. The growth of plants for example, ap-
pears to be nearly if not quite always of that nature. Finally,
belief in the correspondence would, so far as we can see, reach
high water mark, should it be finally made very probable that not
only growth and development but all strictly biological processes
whatever, are periodic. We are undoubtedly a long way from this
last conception. Certain it is, though, that we now have sufficient
facts to make the hypothesis of periodicity as warrantable as its
opposite, namely, that certain phenomena are continuous in the
sense of not being automatically interruptive and group-wise.
We may now proceed to the handling of our data.

398 W. E. RITTER AND M. E. JOHNSON

BRIEF DESCRIPTION OF THE SPECIES

The species Cyclosalpa affinis (Chamisso) was taken in
abundance at La Jolla from May to November, 1909, during
which time most of the observational portion of this research was
made. The longest chains and the largest wheels were brought
in during the earlier part of this period, while the later catches
yielded many specimens of the solitary form with short chains,
and many medium sized single wheels.

Although the salpae do not survive for more than a day or two
in the ordinary aquarium, the material has been sufficiently
abundant to admit of considerable work, with the living specimens.

The two generations of this species differ markedly, as do those
of all members of the genus. The intestine of the solitary form
(fig. 11) is straight, extending nearly the full length of the animal,
the anal opening being just back of the ganglion. The intestine
of the aggregate generation, on the contrary, projects from the
ventral side of the creature as a large, almost circular loop, (fig.
12) the anus being only a little to the left of the oesophageal
mouth.

The solitarj^ form has eight body muscle bands, according to
our system of enumeration, while the aggregate generation has
five on the dorsal and six on the ventral side. The hypophysis
(%p.), endostyle (end.), and gill (gi.) present much the same appear-
ance in both forms. The orifices are also similar except that the ,
solitary form possesses short, tail-like appendages one on each
side of the atrial orifice. Our records show the maximum length
of specimens of the solitary generation to be 15 cm. and of the ag-
gregate 8 cm. In both generations the test is thin, soft, and highly
transparent, without special thickenings. In the aggregate gen-
eration, projecting from the ventral side, is the broad, thin pedun-
cle (ped.) by which the zooids are united to form the wheel, and
within the pharyngeal cavity on the right side, two thirds of the
way back, is the embryo. In the young sohtary individual, the
eleoblast, near the heart, and the remnant of the placenta, about
one-third of the way back from the oral orifice, are both opaque,
nearly spherical bodies and are very prominent.

CflAIN OF CYCLOSALPA AFFINIS 399

The stolon originates just above the heart and extends straight
forward along the median line. The zooids, as in other salpa
chains, are first in single file, but at a certain point, the deploying
point, shift to double file. The deploying point in this species
occurs close to the anterior end of the heart, 3 or 4 mm. from the
root of the stolon. At a point about two-thirds of the way be-
tween the heart and the branchial orifice, the chain bends down-
ward and passes to the outside world through an opening in the
test. Immediately outside the opening, the chain doubles back
under the parent and turns over so that it appears to be greatly
twisted at this point. Before the twist, the zooids are arranged in
two nearly parallel rows along the common stolonic blood vessel.
After the twist, they are arranged in wheels which are connected
tangentially and contain six to sixteen zooids each. These wheels
show a gradual increase in size toward the distal end (fig. 11).

MEASUREMENTS OF THE ZOOIDS, OF THE WHEELS AND OF A

PORTION OF THE CHAIN NOT YET TRANSFORMED INTO

WHEELS

Serial measurements were made of the zooids of a number of
chains with and without wheels as well as of the zooids of separate
wheels. Lengths only were taken.

The measurements of the wheels were made with dividers
and the results are given in millimeters. Those of the unbroken
chains were made with the micrometer eyepiece used in the Zeiss
binocular microscope. A unit in the tables represents 0.1 nam. of
actual length.

The zooids of Chain I were separated from the chain for measure-
ment, but the others were measured while still on the chain. It
appeared that the latter is the more accurate method, since separa-
ting the zooids, besides being a tedious process, is apt to distort and
mutilate them. In these measurements, the posterior extremity
was taken at the atrial orifice. The intestine was not included,
as a slight difference in its inclination would make a difference in
the apparent length.

In table 1 are given the lengths of the zooids of the unbroken
portion of Chains I-VII. Table 2 gives the same data for the

400

W. E. RITTER AND M. E. JOHNSON

>OOC^-^OOGOCnO<MCOt— 050C

>oooo^ — »

I I

(M«>-«<«o«o«)i>-ooo-.cc^tOQO-icqeo>otDC»a.o«'J'too-. occiraooo=cowooOMt~«>occ

oodo«offl»t~t-t-t-t-t~oooo'ooooooooooa.«n<»OiO>oooo'o-. — — c^e^^^SS

M 1

tD=o'a5cotoa=r-t-t-t-.->~-t-t^t~oooOQOQOoooi0050>ooooo^.--.--.--«e222222;

^OlO-H-tl^OC

O 00

>CDh-OOOiO'-iC4CC'*»OeCb-OOOSO^HCQCC'*iC(Ot^OOOiOT-<C<)eO'**'»C«Ob-.OOOiO

CHAIN OF CYCLOSALPA AFFINIS 401

OO'^t^COQOcOCOCOOiCOCOQOOi-'J'OOOOOOQOiOiOC

)e0C0O-^OC^C0C0C0»OOOOOC^Tt«00C0iO00»O(MG0»CO'-He0C0«0»0 00O0000OC0»O00OC

222Si^j::222g§g;3;3?^?SSSS^?5?5Sgg?

?^S?£;SS5??5S5:^§gSSSSSSS£§^;2

OOTj-b.-HO — COt-C^.O--0>'»<OOOtONOt-u:>OOCOOOC

0-Ht-O-J.U5O«)OMinO00'OO(M>0M00Ot^t~OMro

10»-HC^C^CO'^»0(OCOOOOOCSO'— 'C)'^"^iCC0»Or-.00OCiC'3C0UDf~^r~.tO00'-'O^

o.«t-

-*t^-H0UC<HO(Mt^'-l

-HM

Mcocq

^T)< oq

>oo>ooo

ccototoo

oou-

ooooco

U5OO00O

coo

OiO

•* oooooo

-^^

^-H

c5cq"

sssgs

g^???

S35?ss;

?:;S^g5

ssss

o^c^

eoc<ou5-*Tt<>o>o;ocD

t^CO

tOQOOOOOOlM

ust^ooro

<MCM«,-*t-

CO OT O O "5 QO O -H

^^^^J^

coco

TfUi

lOU^UO

o to «5 m lo ^ >o .c «:>

■raira

■oio>o

lOCOCO

CO CO to CO

t-oooc

00«)<3=<»

0505000

o —

rt ^

C-J c^] ca CO

MeOCO-*U5>0 00 050

O-H

<N(NM

-#tOf~

OS-HM^

«!C~00OO

ast-t~o>-«'tt>oo

ococoooo

o>— .

ThcOOJO

U5IOX5

>010»OIOlO>0>0>OtO

coco

CO CO CO

cocococot-r~t-t*t-t-QOoo

oooooo

00 05 05 OS

2222"

««

«CJ

S2S3S2

0««.

t~ooooOrt,-i-*iocoa,o

<M««)

-.«"5

OOOOQO

-.cot- 00 -H

co<occ

woot-o

^ooooco

C-Jf.

oootoooto

oioco

o.o......

.=0

waoooooicn

OiOOO

CO CO CO

coco-*^

iracocor-t-

QOO

gS?3?J

COCOOO^O.OC<OC.COO«COC^.OC.O>^C

«oc..o^^o«^-c.coco=-.oooco.oocooc.c..o.-o<o.„«oo«

2;3;2;sS2S2^^^-;;2S222g;^;

-HC0C0'^iOC0r^000005C?>OC^C<IC0'^«0COt— cr5O»-(C^,^i0i>.|>.00t_3OSC3

jc^cqcac^cacq(Mcqe->rqcococococococococo^^^^^T».lriS;t^™S

Ti<t-a>oot~coeo(M-Hiooo<M>oc<it~coooooc

oOrtcoQOOOtooococo— iOoocoococjcoocoioooOcoeo>rai-oooo

-S^S222^222222gg^g?3S?

o^^cdt^ododoio'-Nco'O-^'cdcdcocdcdt^os-'coir. cc'ooododo»C3c»

«««oooo>^.oo«o.ooo^o.ooo

OlOCOtOOOC

0C0i0OC000>0O'OiOO>0O>0C0OO<_00u

100.001

■5 00W5OC

2S3^l2222^^22gN?3S3c

3?5g5S^S

5gSS^gg???3S?SS??^SS^^^5

ggSS£

sssss

402

W. E. RITTER AND M. E. JOHNSON

TABLE

Length measurements of the zooids of wheels of chains VII and VIII
Horizontal lines indicate end of wheels.
Double lines indicate one wheel lost.
Unit — 1 nam.

NO.

CHAIN VII

CHAIN VIII

CHAIN VIII— continued

CHAIN vin— continued

7.3

6.4

6.1

13.7

12.6

20.5

20.9

7.6

7.7
7.7

7.6
7.8
8.2
8.1

6.7
6.9
7.1

6.8
7.0
6.7

13.1
12.9

12.0
12.2

19.2

18.6

18.6
19.8
20.7

22.7

18.6

13.0
13.9
14.0

12.9
13.0
13.4

17.2

7.8
8.2

7.4
7.4

6.8
6.8

19.8

8.3

19.8

8.6

8.4

7.5

7.4

14.2

13.5

21.9

20.2

8.4

8.7

7.6

7.5

14.2

14.1

22.5

20.4

8.7

8.8

6.7

7.1

14.3

14.0

21.9 20.5

8.9

8.7

8.8
8.7

6.8

7.2

i 51

14.3

12.8

19.8
21.1
22.9

17.5

7.2

8.7

7.8
7.6

14.9
15.2

12.4
14.0

20.0

8.2

8.8

21.0

9.5

9.7

8.2

8.3

15.3

14.7

24.2

23.2

9.5

10.0

8.5

8.6

15.7

15.3

22.8

23.2

9.7

10.0

8.8

8.2

! 55

16.0

14.3

24.1

21.6

9.8

10.2

9.4

8.4

16.4

15.3

23.0

19.9

9.6

10.3
10.0
9.3

9.2
9.3
10.6

8.7

t 59

15.0

14.8
14.0

21.3

20.6

9.4

9.8

8.4
9.1

15.4
15.5

21.2
23.7

20.9

13.9

21.3

10.6

11.0

9.8

8.9

16.6

13.2

100

24.5

22.6

10.9

11.1

9.7

9.1

16.2

15.4

101

24.6

23.4

10.9

11.4

10.5

9.0

15.3

14.7

102

23.6

10.8
10.1

11.4
10.9

9.7

8.8

i 63

15.2
14.0

15.4
14.3
13.6

103

104

105

106

24.0
23.9

22.5

9.8
9.6
10.2

8.9
8.9

8.5

22.1

9.7
11.0

10.6
10.7

16.0
16.3

23.0
25.1

22.1

18.3

22.2

11.1

10.5

9.8

8.1

18.8

17.9

107

26.3

23.1

11.4

9.8

8.5

19 6

17.7

108

26.1

23.7

11.8
11.5
11.1

10.6
11.2
11.2

10.1

9.2
9.1

18.4
18.0
15.7

17.7
17.1
14.7
14.7

109

110

Ill

26.4
25.5
22.4

23.4

9.6
10.9
11.8
12.1
12.3
12.3
11.6

22.9

10.2
9.9
10.5
10.8
10.7
11.5

21.3

11.7
11.1
12.8
13.0
12.9
12.2
12.5

11.3
11.4
12.4
11.8
11.8
11.7

16.2
18.8
19.4
19.7
19.5

17.3
19.4
19.4
18.3
18.4
16.1
19.1
20.0

12.0
12.7
13.4
13.6

11.5
11.4
11.4
11.6

18.6
19.3
20.5
22.2

CHAIN OF CYCLOSALPA AFFINIS

403

wheel portions of Chains VII and VIII. In all, the unbroken
portions of seven chains were measured. The number of zooids
used depended upon the miminum size measurable. In four cases,
90 zooids were taken but in the other three chains only 80, 70 and
52 zooids respectively were large enough to be measured accurately.
[.The measurements of all series are given but only two are plot-
ted, the right-hand series of Chain II in fig. 1, and the right-hand

S 3 13 17 Bl ZS t9 J3 J? 41 'ti 4a S3 37 tl 6} 69 73 77 61 63 89

Chain II Right side

Upper line Length of zooids
Dotted line Length if peduncles

Lower lint Length of foot pieces

1 1 1 1 1

•\

!
t

\ I

r— '

1\v

-^ J

Fig. 1 Plot of length measurements of the zooids, peduncles, and foot-pieces
of chain II, right series. Vertical distances represent length. Horizontal dis-
tances represent position in the chain.

series of Chain VII in fig. 2. In the latter figure, 1-90 are the
zooids of the unbroken part of the chain while 91-108 are wheel
zooids, the divisions between wheels being indicated by vertical
dotted fines. Table 3 gives the measurements of zooids of several
short chains of wheels which furnish figures for comparing graphs
of wheels of various sizes. In each series of results here given
except length of Chain I, two measurements were taken and these

404

W. E. RITTER AND M. E. JOHNSON

Length measurements of zooids of various small groups of toheels.
Unit — 1mm.

Group A

Group B

Group C

Group D

POUB WHEELS

FOUR WHEELS

POUR WHEELS

TWO WHEELS

10.9
11.7
12.1
12.1
12.6
13.2
14.7
15.3
15.6
16.3
14.8

10.7
10.8
11.1
11.4
13 2

19.0

18.4
19.1
19.0
20.5
17.9

18.5
18.6
19.1
19.. 6
20.7
19.9

16.7
18.0
19.1

19.7
20.1
21.6
21.1

20.8 27.9

21.5 29.4
21.2 1 29.2

21.9 i 29.8

22.6 ' 30.1
21.9 29.3

27.2

26.4

30.0

30.7

30.8

13.5
13.6
14.5
15.7
15.8
15.1

29.5

19.6
21.1
21.6
22.0
22.5
22.2

19.8
19.9
20.2
22.1
22.4
23.0
21.8
21.8
23.5
24.2
24.3
23.1

22.0
23.4
24.2
24.2
23.7
24.2
23.0

28.9
30.9
30.8
40
30.4
30.3

30.7

22.3
22.6
22.6
22.5
21.0

31.8

32.9

31.9

33.2

16.6
17.0
16.9
17.1
18.5
18.1
12.7
18.5
19.4
19.4
18.9
15.1

15.1
17.1
16.7
17.9
18.5
18.9
15.9

31.3

13 '. ...

22.3
23.3
22.3
24.6
24.1
23.0

19.9
22.3
24.0
25.5
23 5

24.0
24.1
25.0
25.2
25.7
25. S

25.0
26.5
27.6

28.4
27.8
27.8
23.8

15.7

18.8
18.9
18.6
19.8

22.8
24.6
24.9
25.8
25.7
24.9

22.3
24.0
26.1
25.4
25.7
25.7
24.5

24.0
26.6
27.3
28.5
28.4
28.0
25.9

26 ...

CHAIN OF CYCLOSALPA AFFINIS

405

TABLE 3— CONTINUED

Gboup E Group F

Group G

Group H

THREE

VTHEBLS

THREE WHEBI-S
K. L.

TWO WHEELS

one WHEEL

23.4
24.6
25.2
24.7
25.1
23.6

23.0 22.3 22.9
24.8 23.8 23.3
25.7 24.4 23.9

26.1 , 24.8 1 23.9
26.5 25.0 24.7

21.3

21.9
21.6
22.3
22.4

21.0
22.2
22.6
24.2
22.6

25.4
25.1
25.0
24.9
23.9

23 1

24.5

25.1

24 8

25.3 26.3

25.1
25.0

21.3
22.0
22.5
24.3
25.3
25.7

23.2
25.2
25.9
26.8
26.8
26.1

24.4

24.2

24.9
26.8
27.6
27.2

27.7
27.2

24.1
26.9

248

24 2

27.2 1 27.2 25.9
27.5 ' 28.7 25.1
27.7 28.6 ! 26.0
27.2 28.4 26.8

27.4
28.8
28.7
29.7
29.2
27.6

27.5
28.3

28.4
27.2

27.9
27.2
27.2
29.2
29.8
29.9
30.3

30.0
29.8
29.2
30.3

28.2

26.6
29.9
29.8
29.3
30.4

406

W. E. RITTER AND M. E. JOHNSON

were averaged for the final result. Where the first and second
measurement differed by more than 10 per cent, a third measure-
ment was taken and the three figures averaged.

if,

104

132

'■

170

■■

■

117

;

r^> ■

7^1 ]

104

\' \

inn

\l\\

: 1

f t^

IIR

;

^/\

•i

7«

;

; 1

LtngThofZco/di

Chain VII Right jide

30-iee- IVneo/ fcrhanjf cha,'.

l,hr^ufon,r^i^Mr^,,„„

: 1

;

~!"

;

-]-

■

;

! ■

■

N |i

Fig. 2 Plot of the length measurements of the zooids of chain VII, right side,
including wheels.

TREATMENT OF THE QUANTITATIVE DATA

At first glance, one sees a resemblance between the curves for
the wheels of Cyclosalpa affinis and the blocks of Salpa fusiformis-
runcinata. In both cases, the end zooids are smaller than those
nearest them, the maximum values lying somewhere between,
usually nearer the distal end.

CHAIN OF CYCLOSALPA AFFINIS

407

/^' ^v,

\l 1

f
/

^^Tn

^ — ^

^^'

-r

Fig. 3 Mean curves for wheels of various sizes. Vertical distances represent
length of zooids. Horizontal distances represent position in the wheel.

408 W. E. RITTER AND M. E. JOHNSON

Some variation in the graphs of wheels of different sizes was
noted, and to make sure of its general trend, the data for all the
wheels were considered. The wheels were first grouped accord-
ing to size, Group A included wheels whose zooids averaged 5-10
mm. in length; Group B, 10-15 mm.; and so on. In Group A
were ten wheels. Not only does the number of zooids in a wheel
vary, but the number in one-half of a wheel is not always the same
as in the other half. For this reason the ten wheels were re-
garded as twenty half wheels.

Among these twenty half wheels of Group A were three contain-
ing four zooids; one with five zooids; nine with six zooids; and seven
with seven zooids. The corresponding values of the three four-
zooid half wheels were averaged, the three first zooids together,
the three second zooids, the three third, and the three last zooids.
The result was a typical curve for a four- zooid half-wheel whose
zooids have an average length of 6-10 mm. The five, six, and
seven-zooid half-wheels were averaged in the same way. Similar
computations were made for the other four groups and the results
plotted. The graphs were smoothed and those for each size were
averaged in order to get the typical curve for that size. These
curves (fig. 3) show that the size differences between the zooids
of a half-wheel greatly increase as the zooids grow and that the typical
form already noted becomes increasingly evident.

Passing now to the unbroken portion of the chain, we find that
the zooids increase in length very slowly at first and more rapidly
later; also that though the curve is fairly smooth at first, it be-
comes quite irregular toward the end. Upon closer examination
of fig. 2 and the graphs of other chains, we surmise that these
irregularities are the forerunners of the groups making up the
wheels; in other words that the periodicity shown so plainly in
the wheel part of the chain extends back into the unbroken part.
Were this found to be true, the fact could hardly be ignored in
considering the problem of the break-up of the chain and the pro-
duction of wheels.

In order to test the conjecture more critically we submitted
the measurements to Mr. George F. McEwen, the mathematical
expert of the Marine Biological Station of San Diego for examina-

CHAIN OF CYCLOSALPA AFFINIS 409

tion. Out of this examination has come the graphs shown in figs.
4, 5, 6, 7, and 8.

A curve was computed to fit the graph (fig. 2), as nearly as pos-
sible. From the equation of this smooth curve we get a 'calcu-
lated value' for each zooid; that is the length of each zooid, if the
series were as smooth as our calculated curve. We next subtract
the observed length of each zooid from the calculated length, and
get a series of values, some plus and some minus according as the
irregular graph went below or above the smooth curve. When we
plot these plus and minus values above and below a horizontal line
we have the graph fig. 7. It shows that the values follow the
curve fairly well at first and then vary more and more; in other
words, that we have a periodic curve of increasing amplitude. ^

'Mr. McEwen gives the following summary of the method used: The sizes
for each of the points corresponding to the numbers 45, 50, etc., to 90 were taken
as the ordinates of a curve whose abscissae were 1, 2, etc., to 10. It was assumed
that the above curve corresponded to an equation of the form

y = a+bxi+cxi

and the most probable values of the coefficients a, b, and c were computed accord-
ing to the method of least squares. By substituting {2x — 8) for Xi in the above
equation, the equation

y = a + b (2x — 8) + c (2x — 8)'

was obtained in. which, if y^ of the number of the point is substituted, will equal
the computed value of the corresponding size. (This equation was used to calcu-
late the corresponding values of y, which were used in connection with the observed
values for computing the algebraic sum of the residuals and the probable error,
for the purpose of determining if the equation was a proper expression for
measured values of y.)

It was assumed that this equation, determined from the 10 points was very
nearly the same as if it had been computed from the 45 actual points, and there-
fore represented the relation between the number and the average size of all the
points. This assumption was verified in one case by including all the points and
comparing with the result when only 10 points were used.

The observed values of y were subtracted from the corresponding computed
values and these differences were plotted as ordinates against the numbers as abs-
cissae, thus giving a representation of the deviation of the observed values from
those given by the equation. These deviations are due to errors in the measure-
ments, and to the fact that the assumed equation was not a true expression for
the relation. As the error in measurement was =*= 0.1, it is evident that the devia-
tions are due mainly to the latter fact.

The periodic character of these curves shows that the true law is a periodic
fluctuation of increasing amplitude about a mean value increasing in a regular
manner with the number of the point.

410

44 48

W. E. EITTER AND M. E. JOHNSON
32 56 60 64 68 72 76 80 84

"^^^

'^-

ii^

»<s

:z^

rrx

^'T

—

1 1 1 1 1 1

Plot of Differences

Chain IV. flight side

areraged for eferif 3rd pom

"9'"

or.

'"■f

Fig. 4 Plot of differences for chain IV, right side.

v::

\ /

-.4

/».

'v.

-I.Z

^..,

-16

- P/ot cf Differences

Chain VI.. IfighT side.

-20

ifi

-i-

— \

-.4

S-

-0

J-

-1.2

— P/ot of Differences
_ Chain Vt.. Left jic/e.
■IS -90

-1.6

-to

-24

1 1 1 1 1 1 1

Fig. 5 Plot of differences for chain VI, right and left sides.

CHAIN OF CYCLOSALPA AFFINIS

411

Chain IV, whose plot of differences is shown in fig. 4, is one of
the smaller chains and in it one would expect to find the grouping
less evident than in the larger chains. However it can be plainly
seen even here. The right and left sides of Chain VI are shown in
fig. 5. With Chains IV and VI, the differences were figured only
for the zooids 45-90. In figs 6 and 7, the two sides of Chain VII

—

P/ot of Pifferences
Choin V/l Leff jide'
. Upper 3tne,. 45-30 _
Loivtr series- I-1S

. II

-8

.?A

-N

-■—

-y

"^y

2 6 10 (4 18 tZ J* 30 i1 M <a 46

Fig. 6 Plot of differences for chain VII, left side

are given entire, the curves and the differences being figured
separately for the two parts of the chain, since it can be fitted
better when but half is considered at one time. The complete
series being given, one can more readily see how the amplitude of
the waves increases toward the end.

It will be remembered that in computing the differences, the
observed values were subtracted from the calculated values.
Hence upward curves in fig. 2 appear as downward curves in fig. 7,
To make the comparison with the wheel graphs easier the signs

JOURNAL OF MORPHOLOGY, \0L. 22, NO. 2

412

W. E. RITTER AND M. E. JOHNSON

were reversed for fig. 8 so that values greater than the correspond-
ing ones in the fitted curve he above the x axis while smaller ones
lie below. Vertical dotted lines have also been drawn to indicate
a possible grouping of the zooids. Irregularities appear, it is true,
but since irregularities often appear in the wheels also it is to be
expected here. Moreover, with such small values the chances
for error are so great that one would expect considerable variation.

70 74 78

—

Plot of Oifferences
Chain V/I...1f,ght Jide.
. Upper merits.... 4S-90.
LoMer series t—^S

1.6

-A

A'l

-1.2

-fA

V "^

■^

—

{

-8

6 10 14 18 22 Z6 X }A X 4

Fig. 7 Plot of differences for chain VII, right side

What we get then from these plots of differences is the probable
fact that the unbroken paH of the chain really shows a periodicity
or incipient grouping closely resembling that of the wheeled portion
of the chain, the groups including four to eight zooids each, which
is the number found in the completed wheels.

The plots of differences brought to light an aspect of the matter
which had not been anticipated, namely, the existence of another
toave with a longer period, shown in all the curves. The plots in

CHAIN OF CYCLOSALPA AFFINIS

413

figs. 4 and 5 have been smoothed by averagmg for every ninth
point in order to make this curve more plain. The curves are
not just the same for the two sides of Chain VI, the difference
probably being due to a difference in the way in which the com-
puted curve fits in the two cases.

With what, if any, other biological phenomena in the species
this newly discovered periodicity is connected we do not know.

7zoo/ets. : 7zooids \6 zooids. \S zooidj. \S zooids. • S zooids\ AS zooids. \ 4zooids

■<6-S2. \53-J9. \60-eS. \66-70. '7/-76. .77-81. \]«-fftf. '^7-90

Fig. 8 Plot of differences for chain VII, right side
Inverted and possible groups indicated

Of its existence however, there seems to be no doubt; and it is
certainly interesting to recognize that we have here an instance,
by no means uncommon in organic phenomena, of waves, so to
speak, of one size riding upon those of another size. It is highly
desirable to take these cases in hand with a view to finding their
connection with other phenomena.

414 W. E. RITTER AND M. E. JOHNSON

ATTEMPT TO CONNECT THE FORMATION OF WHEELS WITH MOR-
PHOLOGICAL, PHYSIOLOGICAL, AND MECHANICAL PHENOMENA
PRESENTED BY THE ANIMALS

1 . Segmentation of stolon and the deploying point

To find other factors entering into the wheel production, a study
of the structure of the chain was made. The portion of the chain
in which the zooids are in single file is of the same general form as
that of other species. The incipient zooids, marked off by the
infolding ectoderm, have their aboral ends uppermost, and the
dorsal side of each against the ventral side of its neighbor, the
dorsal sides being towards the base of the stolon. The blood sup-
ply passes out through one-half of the large axial blood vessel and
back through the other half. The segmentation of the stolon in
some cases extends to the root of the stolon, in others not quite so
far. A possible significance of this variation will be pointed out
in another connection. Where segmentation extends to the root
of the stolon, the more proximal segmentation lines are very
irregular and this fact may be of considerable interest in a way we
shall not stop to consider here. Judging by some chains, one
would say that the segmentation begins at the sides of the stolon,
but others lead us to suppose the beginning is above and below
(along the genital rod and the neural tube) while in still others
it seems to be equally advanced in all parts of the circumference.
The latter condition is probably the usual one.

2. Position and relation of the zooids in the chain from the
deploying point to the twist

a. Shifting of the zooids. What we call the deploying point, is
the point, or region where the zooids, by moving alternately to
the right and left shift from single to double file. While the zooid
is moving out, it also moves upward and begins to turn, so that
its dorsal side faces out instead of toward the base of the stolon.
These changes .take place gradually. The oral ends shove out
and begin to turn and before the turn is complete, the aboral ends

CHAIN OF CYCLOSALPA AFFINIS 415

shove out and turn. The sketch of the deploy mg point will make
this clearer. Figs. 14, 15 and 16 are dorsal, lateral and ventral,
views of the deploying point of one chain. These drawings were
outlined with the aid of the camera and much care was taken to
make them accurate. -

Calling the zooid whose oral end has just begun to shift, no. 1,
the aboral ends of nos. 8 and 9 (numbering on one side only) are
beginning to do the same. No. 25 (not shown in the figure)
seems to have reached the final position with the rearrangement
of the internal organs complete. We find now that the right
sides of the right-hand zooids (considering those to be right-hand
zooids that correspond to the right-hand side of the parent) and
the left sides of the left-hand zooids are toward the base of the
stolon. This statement applies to the chain before it emerges
from the parent. The orientation of the older, extruded part of
the chain is given later.

All of the observations on the early growth and differentiation
of the chain agree with those made by Brooks for C. pinnata with
the exception of the orientation. Brooks ('93 p. 79) says of the
single file zooids:

At this stage each Salpa is bilaterally symmetrical, and its plane of
symmetry is the same as that of the stolon, while its long axis is at right
angles to that of the stolon, which becomes converted into a single row
of Salpae, so placed that the dorsal surfaces of all of them are toward
the base of the stolon, their ventral surfaces towards its tip, their right
and left sides on its right and left respectively, their oral ends at its top
or neural side, and their aboral ends at its bottom or genital side.

Again in his description of the double row he says :

The single row of Salpae becomes converted into a double row, which
consists of a series of right-handed Salpae and a series of left-handed ones,
placed with .... the left sides of those on the right and the right
sides of those on left towards the base of the stolon.

-The loop-like structures seen at the oral extremity of the zooids in figs.
14 and 15 might easily be mistaken for the intestine. They are not this struc-
ture, but indicate very nearly where the oral orifice will appear.

416 W. E. HITTER AND M. E. JOHNSON

This, as will be seen by comparing it with our description, is
the opposite of the condition found inC. aifinis, since Brooks places
the oral ends of the zooids uppermost while we find the aboral
ends up. This mistake was probably due to lack of sufficient
material for the study. He says (p. 87):

In all my preserved specimens the tip of the stolon had been so much
flattened by contact with the side of the bottle, in transportation, that
I have not been able to study in detail the way in which this wheel-like
arrangement is acquired, and the subject should receive the attention
of those who are able to stud}^ living specimens.

It is a point upon which one could easily go astray if hampered
by a lack of material.

As the changes in internal organization seem to correspond with
those of C. pinnata, and as Brooks'description is so clear and com-
plete, we need not go into the subject, but refer to his account
(Brooks, '93 pp. 80-106.)

When the zooid has moved into its secondary position it lies
upon the stolonic blood vessel rather than to the side or around
it. With this change, two small vessels develop for each zooid,
one leading to it from each half of the stolonic vessel (fig. 24, ihv.).
The blood flows along one-half of the main vessel (say the upper
half) out through the upper small vessels to each zooid and returns
by the way of the lower set of small vessels to the lower half of
the main vessel where it joins the inflowing current. These cur-
rents are reversed with the reversal of the blood current in the pa-
rent. The zooids now increase in size very rapidly, lengthening
out more above the upper level of the vessel than below it, so that
at the twist the oral ends extend but a little way below the vessel,
while the aboral ends extend far above it. As a result,the aboral
ends of the zooids of opposite rows come in closer contact than do
the oral ends. Since the zooids of the two rows are arranged al-
ternately, each zooid will lie against two of the opposite row. As
growth continues and the zooids, through their increased size,
move outward as well as upward, they are forced farther apart,
but the connection is retained through peduncles which now
develop.

CHAIN OF CYCLOSALPA AFFINIS 417

b. The peduncles and foot-pieces. These structures play so
important a role in the production of the wheels that they must
be described in some detail. Almost all the figures show the pe-
duncle in one stage or another. Fig. 12, best gives its relation to
the full grown zooid, showing that it is a thin flap or sheet extend-
ing out from the ventral median line. The diagram (fig. 17)
shows that the peduncles of the series are parallel throughout the
first part of the chain, and that each by means of its 'foot-piece'
{fp.) is in contact with four others, its two neighbors in each row.
These foot-pieces are also well shown in the right-hand portion of
figs. 19 and 22. As the zooids grow and extend out farther from
the blood vessel, the peduncles lengthen, and the foot-pieces
grow longer as the zooids grow wider, at least until the region of
the twist is reached.

Along with the great increase in the size of the zooids and the
development of the peduncles comes a change in the circulatory
system. The two individual blood vessels coalesce to form one
vessel with two channels (fig. 24, ihv.). The blood current has the
same course as before except that the incoming and outgoing
currents of each zooid pass through one vessel. Observation of
the blood currents in the living animals made the task of working
out the circulation much easier and more certain than it would
have been if confined to preserved specimens. The cross section
(fig. 25) shows well the relation of the zooids to the blood vessel,
the foot-pieces joining the zooids above the vessel, (fp.) and the
individual vessels leading from the zooids to the two parts of the
large vessel (ibv.)

c. The emergence of the chain to the outside world. Through the
first part of its course, the chain is enclosed within a definite
tube in the test just below the endostyle, this tube ending at a
point just posterior to the placental vessel and anterior to the first
body muscle band. This first muscle bends posteriorly here so
that its insertion is along the lower part of the tube opening.
The placenta usually disappears before the chain reaches this point.
There seems to be more or less of a cavity left in the test where
the placenta was, and the chain, as it reaches this point, no longer
being held in its horizontal position by the tube, following the

418 W. E. RITTER AND M. E. JOHNSON

line of least resistance, turns down into the cavity, and by the
rapid growth of the zooids, soon breaks through the thin wall to
the outside, the tip bending downward.

d. The twist in the chain. The general character of this part of
the chain naay be seen from figs. 11 and 13, while the peduncles
and blood vessels of the region are shown in the diagrams figs. 17
and 1,8. Before the twist, we have within the parent, a straight
double row of zooids with oral ends down. After it, the chain is
turned back under the parent, and the zooids are again found with
oral ends down. Until the zooids break through the test to the out-
side the chain has not begun to twist, the zooids still lying symme-
trically along both sides of the median line. In fig. 13 thirty-six
zooids are outside and the twist is just complete. The presence
of two rows of zooids in the chain makes the turn appear more
complicated than it really is. The chain simply doubles back under
and then turns over, this turn being almost invariably to the left.
This leaves the zooids with aboral ends again uppermost, but the
row that was before on the left is now on the right side of the
parent.

e. Reduction of foot-pieces, first break in the chai7i,and formation
of the first wheel. The first visible intimation of the break-up of
the chain comes in the peduncles and foot-pieces. The foot-
pieces (fig. 17) gradually grow longer toward the distal end of the
chain, coming to their maximum length a little before the end of
the unbroken part is reached. After the maximum they shrink
(fig. 22). The decrease is much more rapid than the increase,
there being only about sixteen to twenty-four zooids in the di-
minishing series. Fig. 19 shows that the first group consists of
nine zooids whose peduncles have broken loose from the rest.
The foot-pieces have shrunk still more and the distal ends of
the peduncles have been drawn closer together. But while the
foot-pieces, by which the zooids are held in the axial line of the
stolon, become successively and rapidly smaller just before the
beginning of the break in the chain, the zooids themselves are be-
coming constantly larger. A consequent crowding of the zooids
results. This brings about a pushing of the bodies of the zooids
forward in the chain beyond the foot-pieces. The strain to which

CHAIN OF CYCLOSALPA AFFINIS 419

the series of adhering foot-pieces is thus subjected results inevit-
ably in a pulling apart of the foot-pieces somewhere. As a matter
of fact the break produces groups and not single pairs. These
groups then promptly shape themselves into the wheels.

3. Comparison of the rate of growth of the chain as a whole with
the rate of growth of other animals. As a matter somewhat to
one side of the main problem, we have thought it worth
while to compare the rate of growth of the chain as a whole with
what is known of the growth of other organisms. This was done
by the method employed by Minot ('91) in his study- of the rate
of growth of guinea pigs; namely, by finding the per cent of increase
throughout the chain. The values of the corresponding zooids
of Chains I, II, III, VI, and VII were averaged. (Chains IV
and V being so much shorter were omitted.) The per cent of gain
of the secoiwl over the first, third over the second, etc., was then
computed and the values plotted. The result is a very ragged
line showing a gradual increase through two-thirds of its length
and a more sudden drop at the end. To get a graph whose course
was more evident, the increment was computed again, this time
taking the series in groups of five. The first value here is the
per cent of increment of the second five over the first five, etc.
(table 4, fig. 9). The gradual increase, maximum toward the
end, and rapid decline is here plainly shown in spite of the limited
data.

This result seems strikingly different from that for the guinea
pig and other animals of higher order, where the per cent of in-
crement is a diminishing one from birth on. The difference may,
however, be more apparent than real since, to make the compari-
son more correct, it would seem that stages in the mammalian
development preceding birth would have to be used.

The drop in rate of increase, when the wheel part of the chain is
reached, may be significant for the comparison, but we do not
consider our observations carried far enough into the life of the
chain as a whole to warrant any speculation based upon them.
A study of the growth of still younger and still older, larger zooids
will have to be made to meet the requirements here.

420

W. E. RITTER AND M. E. JOHNSON

rn%

10 IS ZO iS JO J5 40 45 50 5} 60 iS 70 75 SO 85 SO SS ^

30 40 50 60 70 aO SO 100 1

—

10-90 Unbroken part of efioin

Fig. 9 Percentage of increment throughout chain

DISCUSSION OF THE OBSERVATIONS FROM THE CAUSAL STAND-
POINT

1. Cause of the twist

Although salpae do not move through the water very rapidly,
still there is enough motion to make the end of the chain double
back imder the parent, as soon as it projects into the water. The
reason for its turning over is less evident. The zooids begin to
pulsate some time before the twist is reached and, having been with
aboral extremities uppermost in the original or normal position,
we may suppose they tend to assume the same position again when
the normal state of things is interfered with by the bending
back of the chain. Observation of the living animals shows that
the chains of wheels and the separate wheels (at least the smaller
ones) usually move along with aboral extremities uppermost.

It is easier to say that zooids ' tend to assume the normal posi-
tion' than to show the cause of this tendency. It may be that the
specific gravity of the oral ends is greater, or that the pulsation
may have something to do with it, or there may be some tropism

CHAIN OF CYCLOSALPA AFFINIS

421

TABLE 4
Per cent of increment throughout the chains

SERIAL NOS. OF THE ZOOIDS

AVEBAGE SIZE

FEB CENT -h 2

1- 6

6-10
11-15
16-20
21-25
26-30
31-35
36-40
41^5
46-50
51-55
56-60
61-65
66-70.
71-75
76-80
81-85.

9.2
9.9
10.3
9.3
11.7
13.3
13.4
12.6
15.1
14.9
17.9
16.7
17.4
14.2
15.1
16.7
14.5

Wheel portion

422 W. E. RITTER AND M. E. JOHNSON

involved; but we have no observations under this head. As the
chain Hes, with aboral ends of the zooids uppermost, the propul-
sion of the zooids drives them away from the ventral side of the
parent, while if they were with oral ends up the pulsations would
drive the oral ends up against the ventral side of the parent. In
one specimen in which the, chain had not yet emerged, the end of
the chain was turning to get around the placental blood vessel.
Had our observations been limited to this one instance, we might
conclude that the twist is initiated in this way. But after look-
ing over a large amount of material and finding that the placenta
usually disappears before the chain reaches that point, it is evi-
dent that this in no way accounts for the twist.

2. Unequal growth of zooids arid foot-pieces as a factor in the
breaking up of the chain

We seem to have found a cause sufficient for the present re-
search, for the break, in some way, of the chain of zooids. This is,
as already pointed out, the unequal growth of the bodies and foot-
pieces of the zooids. The first question that arises when we
attempt to push the analysis farther is, why is the break into groups
rather than into single pairs of zooids? Nothing in the differen-
tial growth recognized appears to bear upon this question. So
far as that is concerned we should suppose the zooids would be
picked off one by one, or at most in single pairs.

Just how constant these groups are, may be seen from the fre-
quency polygon (fig. 10). We see that of ninety-two half wheels
seventy-three contained six or seven zooids each, while only two
contained eight, and five contained four zooids. This constancy
is to be expected when we regard the breaking apart as a growth
phenomenon depending upon constant causes rather than upon
chance.

In some waj^ the wheel phenomenon is clearly dependent to a
large extent on the strength of the adherence among the foot-
pieces, which are but parts of the central ends of the peduncles.
As may be seen by fig. 18, the radial blood vessels, the other
main connection of the zooids, break apart early in the life of the

CHAIN OF CYCLOSALPA AFFINIS

423

No. of zoo/e/a... Ma/flV/iee/. Whole hf heel.

Fig. 10 Frequency polygon showing the number of zooids in the wheels

wheel, so that in the fully grown wheels the zooids are held together
almost entirely by the peduncles. The observed facts certainly
suggest group adherence among the foot-pieces themselves. Can
direct evidence of any such thing be obtained?

Having proved the existence of a pronounced size grouping of
the zooids in the wheels it naturally occurred to us that there may
be something of the same sort in the peduncles and foot-pieces.
We consequently made a considerable number of measurements
on these structures. Some of the numbers are given in table 5, and
in fig. 1; the graphs of peduncle lengths (dotted hne) and foot-
piece lengths (lower continuous line) are presented. It is doubt-
ful if these show anything. We have not assumed that they do.
The difficulties in the way of making the measurements are too
great for the methods employed. It should, however, be borne
distinctly in mind that these negative results prove no more than
the insufficiency of the measurements. The fact that the foot-
pieces do cling to one another in groups, and that the zooids to
to which they belong are demonstrably different in size, appears
to make it probable, a priori, that the adhesive power of the foot-
pieces is of the gradational, or periodic sort, in spite of our failure
to find it. The suggestion is that the graded size of the zooids
is reflected in the adhesive power of the foot-pieces. Could this

424

W. E. RITTER AND M. E. JOHNSON

OiCOiOOiCOiOOOOiOCCiOiCOCOOOiCOiOifDOOOOO

si U

0»000000>OOeOOOOOOiCOOOOOO»C"20fOCOOO

oooooiciooiocxicooo>oo»co>oiooococciioioioofcoo

OOiCiOOOOO»OiC»^inOiOOLOOO>OOOOOOiOiOO

I0»00i00000i0000000i00 0000»0000

lOOiOOOOiOOiOiOiOOOOOOiOiOiOiOOiOiOiOOiOOO

i <N CC .-I 1 '

'000»C»OiOOO»OiOi.OOiOOiOiOOOlMOiOOOOiOOO

CHAIN OF CYCLOSALPA AFFINIS 425

00 "3 00
.-i^oJ

cooo
00 t~-t>-

00 o

OOiOiOOiOiOOOOiO^OOLOiOiOiOiOiOOOOOOOOOiOOOOOOOiO

0»C»Ci»00»OiOOOiO'OOOOOOiOiOOiOOOO»OOOtOOOOOOiOOiOO

<x)oot^!£!t^Mc<ic<^(^^l-Hco■*■^co■<4^co(^^coeoc<^Meoc<^<^^'-^c^^.—^,— (.-HC],— looocio

^•-ooLOiooio»ooicu^>o-*T}<Tt<-*-*TtiTti'^TtiTficocoirc>(roforoeocococc(

426 W. E. RITTER AND M. E. JOHNSON

conjecture be proved true, an exceedingly important biological
point would have been made.

And now as to the evidence that a periodicity corresponding
to the future wheels does exist in the chain before its break-up.
In discussing the results of our treatment of the data pertaining
to the unbroken part of the chain, we said the curves, as shown in
fig. 2, for example, 'probably' show a periodicity. We permitted
ourselves to doubt to this extent, in the interest of conservatism.
We wish now to sum up the evidence for periodicity. Its strength
lies in the fact that it is cumulative rather than in the sufficiency
of any one piece.

In the first place, does not the undoubted fact of periodicity in
the wheels themselves, and the groups that immediately precede
them, make the presence of periodicity in the rest of the unbroken
part of the chain probable a priori? It would seem so. In the
second place mathematical treatment of the quantitative data
makes it almost certain that a periodicity corresponding to
theory actually does exist. Third and finally the probable exten-
sion of the periods far back into the young part of the chain, leads
us to suspect that this fact is connected with another observation
of quite a different order, an observation, that is, which strongly
indicates that the periodicity is really established at least as early
as the segmentation of the stolon itself.

One of us has shown that in Salpa fusiformis-runcinata the very
early segmented part of the stolon may be interrupted by an unseg-
mented part (Johnson, '10, p. 154 and fig. 8). While such inter-
ruptions have not been observed in Cyclosalpa affinis attention
was called, when speaking of the first stages in the segmentation
of the stolon, to the fact that in some cases the segmentation
reaches to the very root of the stolon, while in others a stretch of
unsegmented stolon exists. May not this difference indicate a
periodicity in the segmentation corresponding to the periodicity
in growth that we have found?

The reader may think that the grouping, as shown in the plots
of differences, is too variable and indefinite to warrant the con-
clusions we have drawn. True, the groups here are not as regular
as the wheel graphs shown at the end of the curve (fig. 2) , but though

CHAIN OF CYCLOSALPA AFFINIS 427

the small groups appear to be more irregular on account of their
riding on the secondary waves, they are of the same sort. It
must be remembered, too, that the values are very small and
the chances of error are large. In fact, such a uniformity of re-
sult throughout all the graphs examined, in spite of small values
and difficulty of measurement, is very convincing.

The transformation of the groups of zooids into wheels is easily
understood : The moment the break occurs so that the pressure of
the zooids upon one another in the group can exert its effect back-
ward as well as forward, the hindmost pair swings in toward the
axial line, each of the other pairs up to the transverse middle line
of the group following in its proportional amount. Since by this
time the foot-pieces have wholly or almost wholly disappeared and
the central ends of the peduncles have become closely appressed,
the swing of the zooids disposes the peduncles in the form of the
spokes of a wheel, the hub being represented by a small elliptical
space. The course of things here described is illustrated in fig. 17.
That the pressure tending to force the mid-zooids of the groups
outward is considerable is obvious from the zig-zag form into which
the axial vessel is thrown, due to the pull on the radial vessels,
as seen in the second group of fig. 18. The disappearance of the
axial vessel in the older wheels may be supposed to be partly due
to the same cause, although probably the vessel is actually
in course of degeneration.

3. Impossibility that the character of the blood supply to the zooids
can be the cause of the size schemes within the wheels

No study involving the growth of the zooids could be complete
without attention having been given to so fundamental a matter
as that of the blood supply. For example, the question naturally
arises, does not the break-up of the chain into groups so affect
the common blood vessel of the stolon that the zooids do not
share alike in nutriment received, and is not this inequality respon-
sible for the disparity in size among the zooids?

The changes in the circulatory system are best shown by the
diagram fig. 18. At the end of the continuous part of the chain,

JOURNAL OF MORPHOLOGY, VOL. 22, NO. 2

428 W. E. RITTER AND M. E. JOHNSON

the individual blood vessels are arranged at regular intervals along
the large vessel. The arrangement is the same for the first wheel,
but with the second or third wheel the axial vessel begins to shrink.
As the vessel remains in connection with the individual lateral
vessels, while growing smaller, it comes to have a zig-zag course,
due to the opposite but alternate pulls upon it by the growing zo-
oids. The shrinkage of the vessels goes on so rapidlj^ and to such
an extent that in the next wheel the vascular connection between
the central zooids is lost. The portion of the main vessel which
joins two wheels together persists for some time. In fact this and
the transparent cellulose envelope which forms around the wheels,
filling in the spaces between the zooids, are all that hold the chains
together and very long chains of wheels are sometimes found. The
small remnants of the individual vessels gradually disappear.
Though these vessels end blindly, the blood may still be seen in
them for some time, flowing out one side and back the other. After
the disappearance of the main vessel at the center of the wheels,
short circuits are maintained between the zooids connected at
any point. Thus in fig. 18, zooids 1, 2, 12, 6^, and 7^ have a circuit
of their own. Thus it would seem that if any of the zooids of the
wheels have an advantage over others the end ones would be fa-
vored as against the middle ones, but the middle ones are on the
whole larger. Hence inequality in blood, supply seems to be
excluded from being a determining factor in the size relations
observed.

If there be any communication between the zooids of the un-
broken chain or of the wheels, other than by the circulatory sys-
tem just described, it must be through the peduncles. The ves-
sels in the peduncles are irregularly arranged but they are dis-
tinctly larger toward the edges and reach part way into some of
of the papillae. They are easily followed in the living specimens.
To test the question of blood communication between zooids,
injections were made. Methylene blue in sea water was used,
which could plainly be seen in the transparent peduncles and in
the bodies of the salpae. The first attempt was on two wheels
whose stage of development was the same as the third and fourth
in fig. 18. The needle was inserted in the stolonic vessel half way

CHAIN OF CYCLOSALPA AFFINIS 429

between the two wheels. The color shot out through the small
vessels to the peduncles of the zooids still remaining in contact.
It went throughout the vessels of the peduncles of the zooids but
stopped cleanly at the edge of the peduncle. No zooids whose
connection with the main vessel had been lost showed any touch
of the color. However, the wheel was again examined about fif-
teen minutes later. The stained zooids had died and dropped
away from the wheel, the peduncle dropping away with the zooid.
A slight stain was found around the papillae of the peduncles of
one or two of the other zooids where they had come in contact
with the stained ones. We conclude that there is no direct vascu-
lar connection here but that there is possibly some interchange by
absorption through the thin ectoderm. Another injection was
made in the peduncle of one of the zooids in a wheel. The color
flowed throughout the peduncle and into the zooid but did not
enter other zooids of the wheel. We therefore seem driven to con-
clude that the hlood supply is not a factor in the size differentiation
of the zooids of a wheel.

4. Unlikelihood that the wheel arrangement of the zooids in Cyclo-
salpa has, as believed by Brooks, anything to do with the
position of the four first blastozooids of Pyrosoma

Brooks was firmly convinced that the radial, or wheel arrange-
ment of the asexually produced zooids in Cyclosalpa is homolo-
gous with the radial disposition of the first four blastozooids
of Pyrosoma. This he regarded as one of the strongest evidences
of the close relationship between the two genera. Thus he says
(Brooks '93, p. 133) :

The opinion that Salpa and Pyrosoma are closely related does not
however, rest upon superficial resemblances, but upon their fundamental
identity of structure, although one of the details, the resemblance in their
asexual multiplication, is so complete as to be almost enough in itself
to establish their affinity.

The same view he expresses with only a little less assurance in
several other connections. We had no thought, in entering upon

430 W. E. RITTER AND M. E. JOHNSON

the present study, of considering this point, nor do we propose
now to go into it extensively. However, our results on the growth
and mechanical factors involved in producing the wheels of Cy-
closalpa seem to have so much bearing on the question, that we
can hardly pass it by without notice. The resemblance between
such a figure of the Cyclosalpa wheel as, for example, that given
by Brooks ('93, pi. 1, fig. 2), and reproduced by Delage and H^-
rouard (p. 203, fig. 151) and a figure of an early Pyrosoma colony
like 15, (pi. 31), by Huxley ('59) is considerable and not unnaturally
suggests true heredity kinship. The moment, however, one comes
to look into the details of how each group comes about ontoge-
netically rather than phylogenetically, he finds them so different
that his imagination is balked at an attempt to interpret them
as both referable to a common hereditary operation. In the first
place Brooks seems never to have observed the fact that the
Cyclosalpa wheel is at the outset bilateral. None of his published
figures give any intimation of this, nor does he refer to it in his text.
For instance, the two figures, 8 and 9, pi. 2, of his latest publica-
tion (Brooks, '08) represent wheels of C. fioridana, and C. pinnata
as though they were perfect — as though the zooids were disposed
in exactly the same way throughout the circuit. We would not, of
course, assert that he did not draw just what he saw in these
two instances, especially since we have had no chance to examine
the wheels of C. fioridana, and have seen but a single one of C.
pinnata. In the one specimen of C. pinnata which we have,
attentive study finds that two zooids on opposite sides of the cir-
cuit have slightly different positions from the others. These
probably indicate where the axis of the chain lay; but the de-
parture from perfect regularity is so slight and of such a char-
acter that it might be easily overlooked had one not discovered,
by studying the formation of the wheels, what their real nature
is. In C. afiinis the bilaterality of the wheels is probably never
wholly obliterated.

The first four ascidiozooids in Pyrosoma, on the contrary,
stand in single file as do the Salpa zooids before the deploying
point is reached and the radial order is taken on by the swinging
around of the file so that number four comes to be adjacent to

CHAIN OF CYCLOSALPA AFFINIS 431

number one. Further there is no opportunity in the Pyrosoma
g^roup for the differential mechanical action caused in Cyclosalpa
by the growth and crowding of the zooids while the foot-pieces
diminish in size. Neither is it possible seemingly, for the periodic
phenomenon to play any such part in the arrangement of the
Pyrosoma zooids as it appears to in Cyclosalpa.

THE LARGER SIGNIFICANCE OF SUCH STUDIES

1. Supple7nenting biological with quantitative observations

We venture to call attention to the way in which morphological
and physiological observations and considerations join hands
with quantitative observations in this research. Numerous
structural details in the adult individuals of both sexual and
asexual generations, in the chain of zooids as a whole, in the in-
dividual wheels and the individual zooids composing the wheels,
and in the unbroken part of the chain both as a whole and as to
its individual elements had to be attended to. On the functional
side not only growth in several of its aspects, but the mode of
swimming, certain facts pertaining to the circulation of the blood,
and some points about nutrition have come in for consideration.

All this sort of thing is so familiar to modern biologists as to
need no special mention. Not so with what we have done in a
quantitative way. It seems to us that in this we have entered
a region of research that biologists will be compelled to regard
vastly more seriously in the future than they have in the past or
do now. The case in hand furnishes a rather striking illustra-
tion of what the quantitative method can do. It can enable us
to see facts we cannot see otherwise. It amounts to a great
increase in the power of our eyes just as does the microscope.
This statement is to be taken literally, not figuratively. One
may easily imagine a magnifying instrument that would so en-
large the wheels as to make visible the size differences between the
zooids. It would seem that this is what the application of math-
ematics in physical science very frequently does. We should
never have suspected from ordinary examination size differences

432 W. E. RITTER AND M. E. JOHNSON

of a systematic character among the zooids of the chains. It
was only from certain biological considerations combined with
aid from this instrument, that the existence of the system was
made certain. And it should be specially noted how our results
would have been affected by failure to recognize this fact. The
breaking up of the chain in some way, and the production of wheels
from the breaking, could have been inferred from the unequal
growth of the bodies and foot-pieces of the zooids; but why the
breaking should be into groups rather than into single pairs would
have remained with no definite answer but for the discovery of
the periodicity in growth in the unbroken as well as in the broken
part of the chain.

2. Natural periodicity in organisms and exacter methods of
research

But promptly comes the question from some of the foremost
biologists, What of it? What particular good is there in knowing
that growth is periodic so long as we have no explanation of why
it is so? Our real interest, they say, is in the causes not the mere
facts of organic phenomena. This objection displays, in our opin-
ion, one of the most pervasive and fundamental weaknesses in the
biological philosophy of the day. Looked at critically, it is found
to mean that facts of nature, in order to be interesting and
deemed really worth while, must be prejudged; that an explanation
of them must be ready at hand before they are observed in order
that they may be attractive. The issue must be looked squarely in
the face. It is in fact the old, old issue between the inductive
and the deductive methods of interpreting nature; between ob-
servation and reason going hand in hand, and the power of reason
alone; between the a posteriori and a priori modes of reasoning.
The objection carries the implication that great numbers of facts
of nature can be explained without having been themselves ex-
amined; that the unobserved causes of many observable effects
may be sufficiently inferred from observations on other effects
than the particular ones under consideration. In a word the
meaning is implied if not expressed, that some time nature may

CHAIN OF CYCLOSALPA AFFINIS 433

be fully known without having been fully studied. This concep-
tion of nature and the knowledge of nature is always and every-
where the begetter of dogmatic assertion on the part of leaders,
of subserviency to authority on the part of followers, and of idol-
atry to certain facts and neglect of others by everybody. This
is not the place to go into the logic, or rather, the epistemology,
of biology. The case under treatment does, however, justify us in
a few observations and reflections on procedure in research.

Why is it that the biological sciences are designated as obser-
vational and descriptive, to distinguish them from the physical
sciences which are called quantitative and exact? Surely no
present-day student of nature would contend that living objects
are qualitative alone and so must be dealt with in terms of quality,
while non-living objects are quantitative and are to be dealt
with in terms of quantity ! There is surely no structural part or
activity of any organism that does not exist in some quantity
or other, and hence is not susceptible of being measured in some
way. Contrarywise, there is surely no inorganic body or sub-
stance that has not qualities of some sort by which it is described
and defined. Yet why is it that in spite of the brave effort made
by a few distinguished men of science during the last half century
to introduce conceptions of quantity and the methods of mathe-
matics into biology, these efforts have met with only limited
success at best, and are ignored in practice and frowned upon in
theory by many of the foremost bilogists? Only a few months
ago a distinguished investigator declared in the presence of the
senior author of this paper that the quantitative method in biol-
ogy is dead, and this student suiting practice to theory, though
working in fields where quantitative conceptions and exact de-
terminations are particularly important, rarely attempts to meas-
ure in any rigorous way the biological phenomena with which
he deals. Attention cannot be called too strongly to the extent
to which much of what is esteemed the very highest type of recent
biological work has laid stress on accurate quantitative determi-
nation of certain environmental factors of organisms, but has
ignored almost wholly quantitative determinations of the vital
phenomena themselves. There can be no question about the

434 W. E. R/TTER AND M. E. JOHNSON

importance of exactness in the determination of external factors.
So far these methods are admirable; but, it appears to us, it
must be recognized that when exactness has gone thus far it has
gone at best not more than half the way. Nothing less than equal
exactness all along the line will do to fulfil the highest demands
of physical science.

Let one recall the degree of refinement with which physicists
and chemists are measuring the phenomena with which they deal:
the wave lengths and angles of refraction of light; the quantity
of heat generated in chemical reactions; diffusion rates of gases
and liquids; atomic weights and combining ratios, and innumer-
able other things. Then let him compare these with the ridicu-
lously crude quantitative determinations made in nearly all
departments of biology. A few aspects of physiology, as for
instance, the temperature of the human body; and a number of
phases of the psychology of higher animals — reaction times, for
example — have been brought under mensurational treatment
comparable with the standards of exactness long demanded in
physics. But the vast fields of morphology, of general physiol-
ogy, of individual and race growth and decline, of propagation,
of variation, of automatic and responsive action, etc., have hardly
been touched quantitatively as physics and chemistry would
understand this term. As yet we in biology have hardly heard
of anything corresponding to physical constants, units of measure-
ment, coefficients of change, etc. Yet will any one, fully alive
to the spirit of modern physical science, venture to maintain
that inorganic phenomena are so utterly different from organic,
that conceptions and practices so enormously fruitful in the one
realm are wholly inapplicable in the other?

It is a significant fact that many biologists, the most ardent
in defence of the so-called mechanistic or materialistic view of
living things, are farthest away from, even most hostile to, the
very methods for biology proper that have so largely made the
physical sciences what they are. One looks in vain through num-
bers of technical writings by biologists of this school for anything
like exact, comprehensive accounts, either qualitative or quan-
titative of organsims or parts of organisms, or even functions

CHAIN OP^ CYCLOSALPA AFFINIS 435

of organisms, dealt with. Yet how these writings bristle with
such expressions as 'differs considerably/ 'constant results.'
'as a rule/ 'very similar/ 'normal segmentation/ 'normal nuclear
spindle/ 'normal blastulae/ 'normal animal/ 'practically iden-
tical/ 'essential features,' 'increases in exact proportion,' and so
on!

Two rejoinders are frequently made to this demand for carry-
ing more exact methods into biology. One is on the purely theo-
retical ground that it is not necessary; that 'mere quantity' is of
no great moment in life phenomena; that slight differences are
of the purely 'fluctuating' or individual sort, so have no large
significance. To answer this objection in full would take up much
farther into philosophical discussion than we can go here, but it
may be the more warrantably passed by because the attitude of
mind that makes it is seen to be obviously hostile to the whole
trend and spirit of physical science. If the history of progress
in science can be relied upon to furnish any clue as to how progress
is to be continued in the future, the man of science, who holds a
general view of nature that makes many facts insignificant and
negligible, is bound to come to grief sooner or later.

The other objection is more practically justifiable. It is that
the phenomena of living beings are so complex and subtle,
and that animals, especially, are so sensitive to changes in exter-
nal conditions as to make it impossible to apply to them in more
than a very limited way, the exacter methods of the physical
laboratory. Our answer to this is two-fold. In the first place,
we are persuaded that exact methods could be applied far more
widely than they are, and they undoubtedly would be, did our
general conceptions call for such applications. The other an-
swer is that if it be true, as it well may be, that many life processes
are too subtle and involved to submit to measurement on an
exact and large scale, then the only course open for the inter-
pretation of such processes is to iritroduce no considerations that
involve the conception of accurately measured quantity. The ex-
tent to which this principle, seemingly so obvious and unes-
capable, has been violated in much biological theory during
the last quarter century or more, is seen to be remarkable once

436 W. E. RITTER AND M. E. JOHNSON

one comes to think about the matter. For example, reflect on
the extent to which theories of development and heredity have
made use of the notion of equation and reduction nuclear divi-
sions of the germ cells ; yet who has determined in any rigid quan-
titative way the elements that enter into the hypothetical equali-
ties and inequalities? How familiar is the textbook statement
that the chromatin of the male fertilization nucleus is 'exactly
equal' to that of the female nucleus with which it fuses! But on
what sort of determinations does this assertion rest? On scarcely
another thread of evidence than that they ' look equal !' And here
we come upon the almost incredible naivete with which biologists
in most things eminently sound, have gone down before this fal-
lacy! Only a short time ago while discussing this point with a
number of biologists, one of them, a man of excellent standing and
great carefulness in nearly all scientific matters, replied to my
strictures, "if chromosomes look equal why are they not equal?"
The words were hardly off the man's tongue when he saw what
a remarkable statement he had made. The incident illustrates
the straits to which one may blindly go in following a theory.

We conclude this topic with a quotation from John Tyndall.
In his well-known address on the "Scientific Use of the Imagi-
nation," he says:

Let me say here that many of our physiological observers appear to
form a very inadequate estimate of the distance which separates the
microscopic from the molecular limit, and that, as a consequence, they
sometimes employ a phraseology calculated to mislead. When, for
example, the contents of a cell are described as perfectly homogeneous
or as absolutely structureless, because the microscope fails to discover
any structure ; or when two structures are pronounced to be without dif-
ference, because the microscope can discover none, then, I think the mi-
croscope begins to plaj' a mischievous part.

In view of the vast amount of evidence now before us from so
many aspects of biolog;^^, that vital processes are periodic in their
most fundamental manifestations, it appears unwarrantable to
assume without proof that any whatever are not so.- But see
what periodicity means ; it means that the phenomena are increas-

CHAIN OF CYCLOSALPA AFFINIS 437

ing and decreasing ; that they have phases ; that the time element
being considered, they change in value from moment to moment.
How then can we treat any particular phase, or stage of such phe-
nomena so as to meet the demands of rigorous science without
considering each phase in relation to the other phases? So far
as they are treated without such reference the procedure would
seem to be of the nature of 'random observations' — of the 'grab-
sample' kind — that always, whether in common life, business, or
science finally proves to be inadequate if not disastrous. Astron-
omy, physics, chemistry, and in general geology, have passed quite
out of this portion of their careers.

Taking it as established that biology is allied in essential nature
with these older, less complex sciences, does it not seem inevit-
able that it too must move on and leave its cruder, haphazard
methods behind? Does it not look as though this very fact of
periodicity, this gradual come-and-go of things in the operations
of organisms is to be one of the chief if not the chief way out?
To press the inquiry a little closer, does it not look as though the
wide prevalence of repetitive parts in reproduction and growth,
which though like one another still differ from one another by
some regular quantity, is to be one of the most important, though
only one, of these exits?

It appears to us that cell division, for example, including the
division of all cell parts subject to this process will have to be
looked at sooner or later from this standpoint. Take the Fora-
minifera, for instance, unicellular organisms (according to the
current interpretation) the bodies of great numbers of which be-
come divided into many sections called nodes and chambers.
In the great majority of species, as a glance at figures enables one
to see, these divisions fall into quantitatively differentiated series.
To make the point more cogent we introduce figures of two species
Reophax membranaceus Brady (fig. 21) and Peneroplis arietinus,
Batsch. sp. (fig. 20). Now let one compare these organisms with
the salpa chain, the one, for example, represented in fig. 18, and
catechise himself something like this : surely there is some resem-
blance between these objects. Both are composed of a considerable
number of sections rather regular in form and much like one an-

438 W, E, RITTER AND M. E. JOHNSON

other, though obviously differing from one another in size. Both
objects are hving, and both have come to be what we see them by
a process of organic growth. Can we properly ignore these sim-
ilarities in our efforts to interpret the organism, because on the
whole the differences between them are more numerous and con-
spicuous than are the resemblances? Is it not at least possible
that by turning to these few correspondences seriously they may
serve as the starting point for the discovery of still others, and
finally result in the detection of laws of organic growth and func-
tioning that would greatly broaden our conceptions of, and hold
upon, life phenomena?

One reason for selecting the Foraminif era as a group with v.'hich
to make the comparison is the fact that the comparison of these
organsims with higher ones in somewhat the same way has been
made by several other zoologists. For instance, Schaudin ('95)
speaks of the production and breaking off of parts in Calcituba
polymorpha Roboz. as having "eine gewisse Ahnlichkeit mit der
Strobilation."

But the most interesting comparison from our standpoint, of
Foraminifera with other organisms was made by L. F. de Pourtales
in 1850. At the meeting that year of the American Association
for the Advancement of Science Professor L. Agassiz presented a
short communication from this young zoologist in which Agassiz
said:

Mr. Pourtales has, for the first time, pointed out a direct, well sustained
analogy, which is to be found in the order of succession of the cells in for-
aminiferae of the genera Textularia, Candima, Biloculina, Triloculina,
and Quinqueloculina. This succession agrees fully with the succession
of leaves in plants — so fully that it can be expressed by the same frac-
tions with which botanists are