PRELIMINARY REPORT ON THE EARLY HISTORY 
OF THE EGG AND EMBRYO OF CERTAIN 
HYDROIDS. 
CORA JIPSON BECKWITH. 
Within the last few years some of the hydroids (Pennaria, 
Clava leptostyla, Endendrium, Tubularia crocca) have been de- 
scribed as differing widely in their processes of maturation, 
fertilization, and early cleavage from what we have come to 
think of as typical. Other members of the same group (Tubu- 
laria mesembryantheum, Clava squamata, H\dra, Gonothyrea, 
JEquorea, Tiara, Gonioneinits} have been described as perfectly 
typical. That certain of the hydroids should conform to the 
type and others not, seemed improbable. The fact of the low 
organization of the group which is used by Hargitt to account 
for such variation seemed hardly sufficient. On this account at 
the suggestion of Professors Morgan and Wilson, I undertook 
the reexamination of two of the forms, Pennaria and Clava 
leptostyla, which varied most from the type, to discover if pos- 
sible the relation between the aberrations described in these forms 
and the usual type. The results on Pennaria were worked out in 
the winter of 1907 at the Zoological Laboratory of Columbia 
University, on material collected at Woods Hole during the 
previous summer. Those on Clava were obtained in the summer 
and autumn of 1908. 
A brief review of Hargitt's work on Pennaria and Clava lepto- 
styla will be necessary to recall the points which I wished to clear 
up. In both these forms, according to his results, the processes 
of maturation and fertilization of the egg are very obscure and 
incapable of demonstration. The germinal vesicle just, before the 
time that maturation should take place moves to the periphery 
of the egg where it loses its staining capacity, the nuclear mem- 
brane breaks down, and the nuclear substance becomes diffused 
throughout the egg, where it is no longer recognizable, due 
probably to some chemical or physical change in the egg. The 
183 
184 CORA JIPSON BECKWITH. 
maturation and fertilization processes were supposed to take 
place while the chromatin is in this unstaining condition, the 
sperm having also lost its staining capacity after entering the 
egg. The first evidence of nuclei in the egg after the disappear- 
ance of the germinal vesicle was found in the appearance of 
' nuclear nests " or groups of small vesicles scattered throughout 
the egg, several such nests usually occurring in an unsegmented 
egg. In Pennaria not less than four such nests appear simultane- 
ously indicating four centers of nuclear reconstruction. In 
Clava, as I understand the description, this condition of several 
nuclear groups occurs occasionally in an unsegmented egg. More 
often, however, eggs were found already segmented with a single 
resting nucleus in each cell which I take it were believed to arise 
as described for Pennaria. This reappearance Hargitt thinks is 
explicable on the ground that after fertilization the chemical or 
physical conditions again change so that the chromatin once more 
responds to stains. The chromatic material that was scattered 
at the time the nucleus disappeared collects again, forms vesicles 
which especially in Pennaria occur in groups or nests, each nest 
finally fusing into a single nucleus. Thus a syncytium arises 
without mitosis and with no apparent evidence of maturation or 
fertilization having taken place in the egg. In Pennaria after 
these nuclear groups are formed, nuclear proliferation is by 
mitosis. In Clava leptostyla, however, up to the sixteen-cell 
stage Hargitt describes nuclear proliferation by amitosis, later 
cleavages being mitotic. 
The points in question accordingly are: (i) The nature of 
maturation and fertilization processes, (2) the formation of 
nuclei de novo, and (3) the role of amitosis and mitosis in early 
cleavages. 
Pennaria and Clara Icptostyla were preserved at Woods Hole 
where Hargitt obtained his material. Also the same killing fluids 
and stains were used so that the question of method is eliminated. 
Material was killed every hour of the day and night and every 
half hour in the early morning hours which proved to be the 
most important period. 
I found in both Pennaria and Clava leptostyla that in material 
killed between the hours of 4 and 6 A. M., it was possible to 
EARLY HISTORY OF EGG OF HYDROIDS. 185 
demonstrate the maturation processes and that they take place 
with the utmost exactness and in the typical manner, as reference 
to the figures will show. Fig. i shows the germinal vesicle of an 
egg of Pennaria in which the inner wall is breaking down, the 
nucleolus passing into the cytoplasm where it is lost, while the 
chromatin is grouping itself around the periphery of the nucleus 
to form the chromosomes. Figs. 2 and 1 1 are nearly correspond- 
ing stages of the first polar spindle in Pennaria and Clava lepto- 
styla respectively. In both forms the chromosomes are evidently 
bipartite and the number is determined to be one half the somatic 
number. In the figure of Pennaria (which is in the late pro- 
phase) the spindle has not yet swung around into position. A 
comparison of Figs. 3 and 12 shows again practically identical 
conditions of the second polar spindle in the two forms. The 
first polar body lies outside the egg, the second polar spindle is 
in the late anaphase. I have found numerous intermediate 
stages. Figs. 4 and 13 show corresponding stages of the recon- 
structed egg nucleus with the polar bodies lying outside the egg. 
Figs. 5 and 6 give two stages in the fertilization of Pennaria. 
The two-germ nuclei lying side by side at the periphery of the 
egg later move toward the center of the egg where they form the 
fusion nucleus at the ends of which astral radiations appear. 
The origin of the first cleavage spindle is not determined. For 
lack of proper stages the fusion of the two-germ nuclei has not 
been demonstrated with certainty in Clava leptostyla. 
From this point forward the two forms differ slightly. In 
Pennaria, after the first cleavage the two nuclei are reconstructed 
by the formation of chromosomal vesicles as shown in Figs. 7 
and 8. For some reason, possibly the rapidity of nuclear divi- 
sions, the chromosomal vesicles often fail to fuse into a single 
nucleus but give rise to a " nuclear nest " which subsequently 
gives rise directly to the chromosomes of the following cleavage 
figure. Fig. 9 shows two such vesicles passing on to a spindle, 
one vesicle already broken up into the individual chromosomes, 
the other still in the vesicular stage. Fig. 10 shows an equatorial 
view of such a group of vesicles, some of the chromosomes 
already forming an equatorial plate. 
A shortening of the resting stage between the nuclear divisions, 
1 86 CORA JIPSON BECKWITH. 
it seems to me, might account for Hargitt's interpretation of the 
" nuclear nests." The rapidity of nuclear division is accom- 
panied by slow cytoplasmic division so that the former constantly 
outruns the latter, the result being that an unsegmented egg 
often contains several such groups of vesicles or " nuclear nests." 
In Clava the cytoplasmic cleavage does not lag so far behind 
the nuclear division, but in fact keeps pace with it. For this 
reason it was possible as shown by Figs. 14, 15, 16, 17, 18 and 19 
to demonstrate the first cleavage spindle, and the successive pas- 
sage of the two-cell stage into the four-cell, eight-cell and sixteen- 
cell stages. The nuclear reconstruction takes place here again 
by the formation of chromosomal vesicles ; but Clava differs from 
Pennaria in that, as a rule, the vesicles all fuse into a single 
nucleus between successive cleavages So far as I am able to 
tell as yet, it seems probable that the cleavage in Clava is fairly 
regular. 
SUMMARY. 
In the two hydroids (Pennaria and Clava leptostyla) under 
question, the maturation and fertilization processes take place 
in a perfectly typical fashion and form no exception to the gen- 
eral rule in this regard. 
The conclusion that the " nuclear nests " indicate the forma- 
tion of nuclei de novo is shown to be untenable. The occurrence 
of these nests is explained by the conditions of nuclear recon- 
struction after cleavage, the chromosomal vesicles failing to fuse 
between successive divisions in Pennaria and the cytoplasmic 
division lagging behind nuclear division gives a syncytium with 
several nuclear groups. 
Maturation and the early cleavages take place by means of 
mitosis and not amitosis. No evidence whatever of amitotic 
division was found. 
My results regarding the maturation and fertilization phe- 
nomena make it very probable that Hargitt's failure to observe 
these stages was due simply to the fact that the eggs were not 
obtained at the right time of day. In eggs collected at the 
proper time (4-6 A. M.) there is no difficulty in proving the 
typical stages of maturation and fertilization. 
VASSAR COLLEGE. 
POUGHKEEPSIE. N. Y. 
EARLY HISTORY OF EGG OF HYDROIDS. 187 
BIBLIOGRAPHY. 
Allen, Carrie M. 
'oo The Development of Parypha crocea. Biol. Bull., Vol. I. 
Bigelow, H. B. 
'07 Studies on the Nuclear Cycle of Gonionemus Murbachii. Bull, of 
Museum of Comp. Zool. at Harvard College, Vol. XLVIII., No. 4. 
Boveri, T. 
'90 Zellen Studien (Tiara), Ueber das Verhalted der Chromatischen Kern- 
substanz bei der Bildung der Richtungskorper bei der Befruchtung. 
Jena Zeitschr., Bd. 24, Heft 3. 
Brauer, A. 
'91 Uber die Entwicklung von Hydra. Zeitschr. f. Wiss. Zool., LII. 
'91 Ueber die Entstehung der Geschlechtsproducte und die Entwickelung 
von Tubularia messembryantheum Allm. Zeitschr. f. Wiss. Zool., Bd. 
LII. 
Haecker, 
'92 Die Furchung des Eis von ^Equorea. Arch. f. Mikr. Anat., Bd. 40. 
Hargitt, C. W. 
'04 The Early Development of Pennaria. Arch. f. Entwickls. Mechan. der 
Organismen, Bd. XVIII. 
'04 The Early Development of Eudendrium. Zool. Jahrb., Bd. XX. 
'06 Development of Clava leptostyla. Biol. Bull., Vol. X. 
Harm, Karl. 
'02 Entwicklungsgeschichte von Clava squamata. Zeitschr. f. Wiss. Zool., 
Bd. LXXIII., Heft i. 
Morningstein, P. 
'01 Untersuchungen iiber die Entwickelung von Cordylophora lacustris 
Allm. Zeitschr. f. Wiss. Zool., Bd. LXX. 
Wulfert, J. 
'02 Die Embryonal Entwicklung von Gonothyrea loveni Allm. Zeitschr. f. 
Wiss. Zool., Bd. LXXL, Heft, z 
I 88 CORA JIPSON BECKWITH. 
EXPLANATION OF PLATE I. 
Pennaria. All drawings X 1,300. 
FIG. i. Early prophase of germinal vesicle of egg of Pennaria; the inner 
nuclear wall is breaking down ; the nucleolus passing into the cytoplasm ; the 
chromosomes forming. 
FIG. 2. Later prophase of first polar spindle of egg of Pennaria; some 
of the chromosomes bipartite ; some quadripartite. Reconstructed from two 
consecutive sections. 
FIG. 3. Anaphase of second polar spindle of egg of Pennaria. 
FIG. 4. Female germ nucleus and polar bodies in egg of Pennaria; chro- 
matin is in fine reticulum. 
FIG. 5. Fertilization of Pennaria egg ; male and female germ nuclei at 
the periphery of the egg and of equal size. Reconstructed from three con- 
secutive sections. 
FIG. 6. Egg of Pennaria showing the fusion nucleus with astral radia- 
tions at either end of the nucleus. 
FIG. 7. Telophase of ist cleavage of egg of Pennaria; nuclear reconstruc- 
tion by the formation of chromosomal vesicles. 
FIG. 8. " Nuclear nests " formed by the partial fusion of the chromosomal 
vesicles lying at the ends of the spindle (probably second or third cleavage). 
Pennaria. 
BIOLOGICAL BULLETIN, VOL. XVI. 
PLATE I. 
"*".;. xi^vv/^^Wi.'".- '.'.* '' "", 
... 
&*z 
V., 
>. - -"e.^ 1 J > * 
' -' .iV'V'v * 
y 
-:.*:;. 
4 
"V..VV i '!, .*'.' I ''.' ' " 
V- " : ' \ 
OCR text unavailable for this page.OCR text unavailable for this page.I9O CORA JIPSON BECKWITH. 
EXPLANATION OF PLATE II. 
Pennaria. Figs. 9 and 10, X 1,300. 
Clava leptostyla. Figs, n to 13, X 1,300; Fig. 14, X 350. 
FIG. 9. Third or fourth cleavage spindle with " nuclear nest " passing 
on to it to form the equatorial plate, one vesicle broken up into chromosomes. 
Pennaria. 
FIG. 10. Polar view of a cleavage spindle showing a "nuclear nest" 
as it is breaking up to form the equatorial plate, slightly later than the 
above, some of the chromosomes already free in the cytoplasm. Pennaria. 
FIG. ii. Metaphase of first polar spindle of egg of Clava leptostyla. 
FIG. 12. Anaphase of the second polar spindle of egg of Clava leptostyla. 
FIG. 13. Female germ nucleus and polar bodies of egg of Clava leptostyla. 
FIG. 14. Egg of Clava leptostyla showing the first cleavage spindle. 
BIOLOGICAL BULLETIN, VOL. V 
PLATE 
10 
. .,. 
A. 
_ 
' . 
12 
,-' 
->-! ^\ 
rV>- 
OCR text unavailable for this page.OCR text unavailable for this page.I Q2 CORA JIPSON BECKWITH. 
EXPLANATION OF PLATE III. 
Clava leptostyla. Figs. 15 to 19, X 350. 
FIG. 15. Same, showing reconstruction of daughter nuclei by chromosomal 
vesicles. 
FIG. 1 6. Two-cell stage of Clava leptostyla passing into the four-cell 
stage, second cleavage spindles showing. 
FIG. 17. Four-cell stage of Clava leptostyla with resting nuclei. 
FIG. 1 8. Four-cell stage of Clava passing into the eight-cell stage, spindles 
showing in two cells. 
FIG. 19. Lateral view of eight-cell stage of Clava (five cells showing in 
section), some of the cells with spindles, some with resting nuclei. 
BIOLOGICAL BULLETIN, VOL. XVI. 
OCR text unavailable for this page.ON THE SEX OF HYBRID BIRDS. 
MICHAEL F. GUYER, 
UNIVERSITY OF CINCINNATI. 
In a former paper 1 I have noted the difficulty of obtaining 
female hybrids from pigeons or doves of widely different parent- 
age. Of the seven hybrid offspring of very distinct species then 
in hand, six were male. Since that time, through the courtesy of 
the Museum d'Histoire Naturelle in Paris and the Museum of 
Natural History in London, I have had the opportunity of exam- 
ining a number of different hybrids in the family Phasianidse and 
among them also I have found a remarkable predominance of 
males. In the following tabulations the sex of each hybrid, when 
known, and the parentage, is given, together with the date the 
individual was placed in the museum. It has been impossible to 
give the specific name always because a number of the specimens 
bore only the popular names. The letters placed after the year 
of accession indicate the respective locations of the specimen in 
question; thus, B =: British Museum (Museum of Natural His- 
tory) ; P = Museum d'Histoire Naturelle, Paris ; C = Museum, 
University of Cincinnati. 
GUINEA-FOWL X CHICKEN. 
Sex. Date and Location. 
Guinea-fowl X common fowl = ? 1902 B. 
Guinea-fowl X common fowl = $ 1899 B. 
Pintade X Poule = ? 1854 P. 
Black Langshang Cock X Guinea-hen = $ 1903 C. 
Black Langshang Cock X Guinea-hen = <$ 1903 C. 
Black Langshang Cock X Guinea-hen = <$ 1903 C. 
Black Langshang Cock X Guinea-hen = $ 1908 C. 
Black Langshang Cock X Guinea-hen 1909 C. 
Thus, of eight guinea-chicken hybrids, the sex is known in six 
cases and it is invariably male. 
1 Guyer, M. F., " Spermatogenesis of Normal and of Hybrid Pigeons," 
Dissertation, University of Chicago, 1900. Also published as Bui. 22, Uni- 
versity of Cincinnati, 1903. 
193 
194 MICHAEL F. GUYER. 
PHEASANT X CHICKEN. 
Chrysolophus pictus X Bantam fowl = <$ 1890 B. 
Phasianus colchicus X Game bantam = $ 1902 B. 
Phasianus colchicus X Spanish fowl = ? 1845 B. 
Phasianus colchicus X Common fowl =: <$ 1884 B. 
Phasianus colchicus X (Japanese 
long-tailed cock X common hen) = $ 1905 B. 
Faison X Poule = ^ 1851 P. 
Faison X Poule = 1845 P. 
Faison X Poule = J 1836 P. 
Faison X Poule = 1855 P. 
Faison X Poule = ^ 1813 P. 
Faison X Poule = c? 1851 P. 
Faison X Poule = <$ 1851 P. 
Faison X Poule =: <$ 1846 P. 
It will be seen that of thirteen pheasant-chicken hybrids, the 
twelve of which the sex is recorded are all male. 
PEAFOWL X CHICKEN. 
Paon X Poule Cochinchinoise = <$ 1907 P. 
Paon X Poule Cochinchinoise = 1907 P 
From the foregoing it will be observed that of the total of 
twenty-three hybrids from markedly different parentage (guinea 
X chicken, pheasant X chicken, and peafowl X chicken), each 
one of the twenty of which the sex is known is male. 
PEAFOWL X PEAFOWL. 
Pavo cristatus X Pavo muticus = B. 
PHEASANT X PHEASANT. 
Chrysolophus pictus X Phasianus reevesi = <$ 18876. 
Hybrid P. colchicus-reevesi X Gennceus nycthemerus = ? B. 
Chrysolophus pictus X Phasianus colchicus = c? 18556. 
Hybrid P. colchicus-reevesi X Gennceus nycthemerus = <$ i94 B. 
Gennceus horsfieldi X Phasianus versicolor <$ 1866 B. 
Hybrid P. reevesi-colchicus X Gennceus nycthemerus = <$ 1902 B. 
Phasianus colchicus X Gennaus nycthemerus = <$ 19026. 
Hybrid C. pictus-amherstia: X Phasianus colchicus <$ 18976. 
Chrysolophus pictus X Gennans nycthemerus = <$ 1906 B. 
Phasianus colchicus X Chrysolophus pictus = <^ 19046. 
Phasianus colchicus X Gennceus melanotus = c? 18656. 
Phasianus colchicits X Chrysolophus amherstia = 18986. 
Lophophorus impeyanus X Euplocamus 1 melanotus = ? 1893 P- 
1 Euplocamus is a synonym of Gennceus. 
- Presumably C. pictus. 
THE SEX OF HYBRID BIRDS. 195 
Faisan dore 2 X Faisan commun 3 <$ 1842 P. 
Faisan a collier 4 X Faisan argente 5 1886 P. 
Faisan commun 3 X Faisan amherst 6 1837 P. 
Faisan dore 2 X Faisan ordinaire 3 = $ 1853 P. 
Faisan commun 3 X Faisan argente 5 = <$ 1837 P. 
Faisan commun 3 X Faisan argente 5 = J 1 1843 P. 
Phasianus mongolicits X Phasianus colchicns = <$ 1906 B. 
Phasianus colchicus X Phasianus reevesi - $ 19043. 
Phasianus colchicus X Phasianus torqnatus = <3 18946. 
Phasianus colchicus X Phasianus reevesi = <$ 18946. 
Chrysolophus amherst ice X Chrysolophus pictits $ B. 
Phasianus colchicus X Phasianus reevesi <$ 18976. 
34 Chrysolophus amherstia: X /4 Chrysolophus pictus = <$ 18876. 
Euplocamus 1 su'inhoii X Euplocamus nycthemcrus = <$ 1882 P. 
Euplocamus sivinhoii X Euplocamus nycthcincnis = $ 1875 P. 
Euplocamus lineatus X Euplocamus nycthemcrus = ^ 1887 P. 
Euplocamus lineatus X Euplocamus nycthemerus = $ 1878 P. 
Euplocamus horsfieldi X Euplocamus lineatus = c? 1819 P. 
Euplocamus horsfieldi X Euplocamus lineatus c? 1869 P. 
Faisan amherst 6 X Faisan dore 2 1882 P. 
Faisan amherst X Faisan dore 1902 P. 
Faisan commun 3 X Faisan a collier 4 1860 P. 
Faisan commun X Faisan a collier = g 1858 P. 
Faisan commun X Faisan a collier = < 1843 P. 
Of a total of thirty-seven hybrid pheasants, nineteen have been 
from parents sufficiently widely separated to be ranked by syste- 
matists as separate genera or subgenera, and of these nineteen, 
fifteen were of known sex, namely, fourteen males and one 
female. Of the remaining eighteen there were twelve males, three 
females and three of which the sex was undetermined. 
Thus of a grand total of sixty-one hybrids, the sex is known 
in fifty-one cases and among these there are only four females in 
all. Furthermore, three of these females were hybrids between 
species of the same genus, the other one, between species from 
genera not widely divergent. In hybrids between individuals of 
distantly related genera or between individuals from different 
subfamilies (e. g., guinea )( chicken) where the sex has been 
recorded it has been invariably male. 
There are three possible sources of error in these data. In the 
3 Presumably P. colchicus. 
4 Presumably P. torquatus. 
5 Presumably G. nycthemerus. 
6 Presumably C. amherstice. 
196 MICHAEL F. GUYER. 
first place it is known that sterile females sometimes, although 
rarely, take on the male plumage, and it may be urged that there 
is no means of knowing certainly that the sex was determined 
beyond all doubt by opening the abdominal cavity and finding the 
testes. However, since the specimens had to be partially dis- 
sected before the skins could be mounted, it is reasonable to sup- 
pose that in the vast majority of cases the sex was thus accurately 
determined. The five guinea-chicken hybrids as well as the six 
dove and pigeon hybrids mentioned in my former paper were all 
dissected by me personally and consequently I am sure of their 
sex. 
In the second place the objection may be raised that possibly 
the museums have preserved only the males, inasmuch as they 
make handsomer specimens and are not as similar in appearance 
as female pheasants. There is, of course, a possibility of this, 
especially in the case of hybrid pheasants from closely related 
species. Hybrids from widely different parents are so rare, how- 
ever, that there is every probability that if there had been females 
they as well as the males would have been preserved. As a 
matter of fact, the few female pheasant hybrids that I have been 
able to find in museums are not similar in appearance nor do they 
resemble the males. As hybrids they are as interesting in every 
way as the. males and it seems probable, therefore, that had there 
been more of them they would have been preserved. When due 
allowance is made for all errors the facts still indicate that there 
is a marked tendency for hybrids, especially those from widely 
separated parents, to be male. 
Lastly, there is the remote possibility that there has been 
a greater mortality among the females in early life. In the few 
cases (guinea-chicken hybrids and various pigeon hybrids) of 
which I have data regarding the number of eggs laid and the 
history of the young, there is no evidence of such mortality. 
It may be noted in passing that in the collections of the British 
Museum there is to be seen a hybrid between individuals of two 
different families, namely, a penelope (Family Cracidae) and the 
common fowl (Family P hasianidse). This hybrid resembles 
more the fowl than the penelope. Unfortunately the sex is not 
recorded. 
THE SEX OF HYBRID BIRDS. 197 
In looking over the literature of the subject to see if anything 
had been recorded concerning the sex of hybrids outside the 
group Phasianidse, I found that in general little attention had 
been paid to it. Some mention is made of the sex of hybrids in 
Suchetet's 1 voluminous work on hybrid birds. In speaking 
(p. cxvii) of hybrids and mongrels, he asserts that among the 
former he believes there are more males than females,, and he 
cites various authorities in substantiation of his belief. Thus, 
according to data collected by Buffon, there are more male than 
female mules and Buffon asserts, furthermore, that among hybrid 
birds the number of males exceeds very much that of females. 
Suchetet cites the following figures from Buffon : the proportion 
of males to females in hybrids between the he-goat and the ewe 
are 7 to 2 ; between the dog and the wolf, 3 to i ; between the 
goldfinch and the canary, 16 to 3. Suchetet cites still further 
examples from other authorities, but he seems not to have gone 
over his own extensive notes on hybrids with this question of sex 
in mind. For example, on pages cxxi cxxxiv he gives a state- 
ment in tabular form of data collected from some eighty-five pub- 
lic and private museums concerning in all 234 2 specimens of 
hybrids between wild birds (i. e., not domesticated) or of forms 
reputed to be such hybrids. Since in many cases the sex of these 
hybrids has been given, I have gone through the tables and 
arranged the birds according to sex as far as it is indicated, with 
the following results: 
Of hybrids between species bearing the same generic name 
there are in all 124, of which 72 were male, 18 female and 34 
of undetermined sex. The remaining no hybrids were between 
individuals bearing different generic names and of these 74 were 
male, 13 were female and 23 were of undetermined sex. Thus 
it will be seen that the males far outnumber the females in each 
case. Furthermore, this would remain true in the proportion of 
about 3 to 2, even should it be counted that all those of undeter- 
mined sex were female ! 
In his later amplifications of this list he discusses (p. 507) 48 
1 Suchetet, Andre, " Des Hybrides a L'Etat Sauvage ; Oiseaux," Vol. I., 
1896, Lille. A large volume of over 1,000 pages. 
- Suchetet states the total as 236 but he has made an error of 2 in his 
addition on page cxxxii. 
198 MICHAEL F. GUYER. 
additional hybrids between Tetrao tetrix and Tetrao urogallus, 
of which 40 are male and 8 female. Again, page 573, he lists 
20 hybrids of Lagopus albus- and Tetrao tetrix, of which 13 are 
male and seven are female. 
As to the general bearing of these facts upon any one of the 
numerous theories of sex-determination, the writer does not feel 
disposed to dogmatize, although certain suggestions present them- 
selves. For a general and unbiased statement of our present 
knowledge regarding the question of sex-determination, the 
reader may consult the recent publications of Thomson 1 or of 
Morgan. 2 
Both of these writers agree that when all the evidence is con- 
sidered it does not seem improbable that the conditions which 
regulate the development of sex may be different in different kinds 
of animals. Regarding the sex-determining influence of nutri- 
tion and temperature, either directly on the developing organism 
or through its parents, Thomson points out that while the evi- 
dence in any given case is inconclusive, still when all the cases 
are taken together, " they have a certain cumulative suggestive- 
ness which would warrant further experiment particularly as 
regards the lower animals and the indirect influence on offspring 
through the parents" (1908, p. 490). 
In general, where the experiments tend to show that nutrition 
is a factor after the period of fertilization, it has been the produc- 
tion of females that was supposedly favored by such increased 
nutrition ; the question being apparently one of increased con- 
structive metabolism. It would follow that anything tending to 
retard or hold at a low ebb the constructive phases of metabolism, 
especially during early embryogeny, would be inimical to the pro- 
duction of females. Now in the case of hybrids, and particularly 
those from widely separated parents, there would in all proba- 
bility be more or less default in the metabolic processes because 
of the incompatibilities which must necessarily exist between two 
germ-plasms so dissimilar. It seems not improbable, therefore, 
that this might be the determining factor in the production of an 
excess of males in the case of such hybrids. 
1 Thomson, J. Arthur, "Heredity," London, 1908. 
2 Morgan, T. H., "Experimental Zoology," New York and London, 1907. 
THE FEMALE CHROMOSOME GROUPS IN 
SYROMASTES AND PYRROCHORIS. 
EDMUND B. WILSON. 
The conditions seen in Syromastes marginatus L. are of inter- 
est on account of the light that they throw on those observed by 
Morgan in Phylloxera, as reported by him at the December meet- 
ing of the American Society of Zoologists, and recently published 
in the issue of Science for Feb. 5, 1909. In both forms the 
" accessory " chromosome is not a single but a double body, and 
the female chromosome-groups contain two more chromosomes 
than the male. 
A reexamination of the spermatogenesis of Syromastes 1 led to 
a confirmation of Gross's result that the spermatogonial number 
is even (22), and that the " accessory " chromosome is formed by 
the union of two chromosomes that are separate in the sperma- 
togonia. This double element divides equationally in the first 
spermatocyte-division but passes undivided to one pole in the 
second, so that half the spermatozoa receive two more chromo- 
somes (12) than the other half (10). This led me to the infer- 
ence that the female somatic groups should have two more 
chromosomes than the male i. e., 24 instead of 22, as had been 
described by Gross. I had at that time no female material, but 
through the kindness of Professor Boveri have since obtained an 
abundant supply of the ovaries. Examination of this material 
demonstrated the correctness of my earlier inference. A consid- 
erable number of ovaries have been sectioned, many of which 
contain numerous and very fine division-figures, showing the 
chromosomes with great clearness. Whenever a good view of 
the equatorial plate can be obtained, 24 chromosomes are unmis- 
takably seen to be present, as is clearly shown in photographs. 
Two figures (Fig. i, c, d} are appended, both of which were 
drawn upon enlarged photographs by the method described in my 
fourth " Study." Syromastes is an exceptionally favorable form 
1 " Studies on Chromosomes, IV., " Journ. E.vp. Zool., VI., i, 1909. 
199 
2OO EDMUND B. WILSON. 
for study, and the diagrammatic clearness with which the chromo- 
somes appear in many of the sections precludes, I' think, the 
possibility of error in respect to the number. 
In my description of the male groups I emphasized the fact 
that the two components of the " accessory " are of slightly un- 
88 
*> AA** 
FIG. i. a, b, Syromastes marginatus L., spermatogonial chromosome- 
groups ; c, d, ovarian groups of the same ; e, f, Pyrrochoris apterus L., ovarian 
chromosome-groups. 
equal size (as is shown in the photographs accompanying the 
paper), and that they are recognizable in the spermatogonial 
groups as two separate chromosomes which are the second and 
third smallest of all the chromosomes. In the female groups 
each of these chromosomes is represented by a corresponding pair 
(black in the figures). In most of the ovarian groups the smaller 
two are readily recognizable, and in some cases, though not al- 
ways, this is also true of the larger pair. The numerical and 
size-relations are such as to show that after maturation the egg 
must contain one member of each of these pairs. Though noth- 
ing is directly known of the maturation-process in the female, it 
may be inferred with probability that in synapsis the two larger 
and the two smaller of these pairs unite to form two correspond- 
ing bivalents, which may be designated as aa and bb (II and ii. 
CHROMOSOME GROUPS IN SYROMASTES. 
2O I 
in the terminology of my former paper). By the subsequent 
disjunction of each of these pairs the mature egg receives a and 
b in addition to 10 other chromosomes. Fertilization of the egg 
by a spermatozoon containing the "accessory' 1 (a -\- b} will 
therefore give the characteristic female group (a, b, a, b -j- 20), 
while fertilization by one that lacks a -{- b will give the male group 
(a, b-\-2o}. For the sake of comparison two of the sperma- 
togonial groups are reproduced in Fig. I, a, b. In the figures of 
both sexes the chromosomes identified as a and b are made black. 
The essential relations in both sexes are shown in the diagram, 
<f 
a 
010 
6 6 
c 
a 
88 
00 
d 
FIG. 2. Diagrams of maturation in Syromastes. a, spermatogonial chromo- 
somes (actual number 22) ; b, c, spermatocyte-division ; d, ovarian chromo- 
somes (actual number 24) ; e, f, maturation division (inferred). 
Fig. 2, in which a is cross-barred, b black, and the other chromo- 
somes (of which only four are represented) are in outline. 
It is evident that, except for the different total number of 
chromosomes in the two species, these phenomena are essentially 
similar to those seen in Phylloxera falla.v, which likewise has two 
" accessories " and two more chromosomes in the female somatic 
groups (12) than in the male (10). In Phylloxera cary&caulis 
the two " accessories " are of unequal size, as in Syromastes, but 
the phenomena are complicated by the fact that these two chromo- 
somes are often united in the somatic groups of the male, and are 
apparently always thus united in those of the female. In the 
male they are always united at the time of the spermatocyte- 
2O2 EDMUND B. WILSON. 
divisions to form the " accessory," which is in reality double, like 
that of Syromastes, and sometimes separates into its two com- 
ponents as it moves to the' pole. Morgan therefore concludes 
that the true numbers of the chromosomes in the two sexes of 
this species are eight and six respectively, though the female seems 
to show but six and the male either five or six (according as the 
two " accessories " are united or separate). 
The point to which I would call attention is the similarity 
between Phylloxera carycscaulis and Syromastes in respect to the 
mode of synapsis. In the spermatogenesis of both forms the two 
unequal " accessories " unite in synapsis (if the process can prop- 
erly be so called) to form the bivalent ab, consisting of two un- 
equal components, which passes into the female-producing sper- 
matozoa. In the maturation of the male-producing egg of 
Phylloxera, however (and apparently the same must be true of 
the sexual eggs), a different process takes place, the two larger 
and the two smaller components uniting to form the bivalents aa 
and bb, again exactly as there is reason to conclude in the case of 
the egg of Syromastes. In Phylloxera, as Morgan points out, this 
involves a redistribution of the four chromosomes, since in the 
somatic groups they are united to form ab and ab, but recombine 
at the maturation period to form aa and bb. This remarkable 
redistribution, I think, loses much of its anomalous character on 
comparison with the facts in Syromastes, where a and b are al- 
ways separate in the somatic groups. 
These facts, together with those determined by Payne in Fitchia 
and other forms (now in press in this journal) and my own 
earlier ones on Thy ant a, 1 which shows essentially the same con- 
ditions as in Fitchia, lead me to a somewhat different interpreta- 
tion of the " accessory " chromosome in Syromastes from that 
given in my fourth " Study." In that paper I adopted the con- 
clusion that the two components of the bivalent " accessory " were 
identical respectively with the large and small ; ' idiochromo- 
somes " of such forms as Metapodius or Lygceus. I did not then 
see that all the facts are equally consistent with the view that 
these two components, taken together, represent the single odd 
1 Reported at the meeting of the American Society of Zoologists in Decem- 
ber, 1906, but still unpublished. 
CHROMOSOME GROUPS IN SYROMASTES. 2O3 
chromosome of Anasa, Protenor and other similar forms, and 
that the small idiochromosome has disappeared. Payne discov- 
ered that in the reduvioids, where a single small idiochromo- 
some or " Y-element " 2 is always present, the large idiochromo- 
some is in some species a single chromosome (Diplocodus}, in 
others is represented by two (Fitchia, Conorhinus} or three chro- 
mosomes (Prio nidus') , and in the galgulid genus Gelastocoris 
(Galgulus} by four chromosomes, which behave in maturation as 
a single unit (X-element) that is obviously comparable to a single 
large idiochromosome in its relation to sex-production. Payne 
concludes, with great probability, that the double or multiple 
X-element in these forms has arisen by the separation of an orig- 
inally single large idiochromosome (such as still exists in related 
species) into two or more components. In Conorhinus, where 
the X-element is double, the two components are unequal in size, 
and by the disappearance of the Y-element a condition would 
arise closely similar to that seen in Syromastes. 
Whether such has been the actual mode or origin in Syro- 
mastes or not, it seems probable that here too the double " acces- 
sory " was originally a single chromosome that has separated into 
two parts, which still act as a unit in the maturation divisions and 
retain the same relation to sex-production as the original one. 
Phylloxera carycccaulis may plausibly be regarded as in process 
of transition from the condition in which a single " accessory " 
chromosome is present (as appears to be the case in the aphids) 
to one in which it has separated into two parts, as in Syromastes. 
1 will add a brief account of the female groups in Pyrrochoris 
aptcrus L., material for which has also been obtained through 
Professor Boveri. A reexamination of the male groups (Wilson, 
Study IV.) showed the spermatogonial number to be 23, including 
a single unpaired idiochromosome ("accessory" chromosome) 
which is at once recognizable from the fact that it is nearly twice 
the size of any of the other chromosomes. This passes into half 
the spermatozoa, which receive 12 chromosomes, while the others 
receive but n, as was originally described by Henking. The 
2 In a recent general discussion I have used the terms " X-element " and 
"Y-element" to designate respectively the large and small idiochromosomes, 
or their homologues, whether they consist of a single, chromosome or of more 
than one. See Science, XXIX., 732, January 8, 1909. 
2O4 EDMUND B. WILSON. 
female groups clearly show 24 chromosomes, of which two are 
of the same relative size as the unpaired one of the male. Two 
of the ovarian groups, typical of many others observed, are shown 
in Fig. i, e, f. These facts show that Pyrrochoris conforms to 
the ordinary type in which the male has an odd chromosome, as 
in Anasa, Protenor, Alydus, Largus and. many others. 
COLUMBIA UNIVERSITY, 
January 23, 1909. 
OBSERVATIONS ON THE GERM CELLS OF HYDRA. 
GEO. W. TANNREUTHER. 
The germ cells of hydra, which pass through stages comparable 
to those of higher animals, possess some peculiarities that have 
not been previously mentioned. Some investigators advocate the 
specificity of the germ cells, while others claim that the sex cells 
are the immediate derivatives of the ordinary interstitial cells. 
Kleinenberg (4), Hertwig (3), Brauer (i) and others consider 
the sex cells as interstitial in origin and state that the egg first 
becomes recognizable after the ovary has begun to develop. 
Downing (2) claims that there is continuity of germ plasm in the 
sense of a specific line of germ cells, that the egg (Hydra fusca} 
is always present even before the interstitial cells begin to form 
the ovary and that the egg may grow rapidly and take the initia- 
tive in its formation. He furthermore believes that the egg is 
recognizable as such in the adult hydra and in general that in 
some stage in the embryonic development certain cells are stamped 
with sex characters, so that they and their progeny form the sex 
cells distinct throughout the life of the individual. 
A careful examination of sections from Hydra sp.? (Brauer) 
gives pretty conclusive evidence that not only the egg but the 
sperm as well is interstitial in origin. There can be no question 
in case of the sperm, as the different stages in development can 
readily be traced from the interstitial cells to the mature sperm. 
Furthermore, the progenitors of the spermatozoa have no special 
characters by which they can be recognized as germ cells. The 
cells that give rise to the eggs are interstitial in position and can 
be distinguished in the adult hydra from the ordinary interstitial 
cells by their large nucleus, nucleolus and abundance of chroma- 
tin, even before the growth of the ovary begins, as Downing 
states. This is especially true during the breeding season. If 
these sex cells could be distinguished during the budding season 
as well, it would at least suggest specificity of the germ cells. 
In the dioecious form Hydra sp. ? (Brauer) the egg cells can- 
205 
2O6 GEO. W. TANNREUTHER. 
not be distinguished from the ordinary interstitial cells except 
during the period of sexual reproduction. Fig. i, a and b, rep- 
resents two adjacent eggs at the stage of development when they 
first become recognizable. They have begun to enlarge and can 
readily be distinguished from the adjoining cells by their size and 
vacuoles next the nucleus. The above figure was taken from an 
adult hydra in which six eggs were present in different stages of 
development, ranging from the interstitial cells to the undisputed 
egg. These eggs are isolated or found in groups. When two or 
more are found adjacent the cell walls become dissolved and one 
persists as the developing egg (Tannreuther). In a few in- 
stances observed two of these adjacent cells persisted and gave 
rise to two mature eggs. These cases, however, are extremely 
rare in proportion to the number of eggs produced. 
The above results do not warrant the view that there is con- 
tinuity of germ plasm in Hydra. Until sex cells distinct from 
the somatic can be traced through successive generations, we have 
no positive evidence of such a continuity. 
The different generations of the sex cells of Hydra are distinct 
and can readily be recognized. In the formation of the egg there 
is a distinct growth period. Reduction occurs at the end of the 
growth period just before the first polar body is formed. The 
polar bodies remain attached after cleavage has begun, by means 
of a single cytoplasmic thread. The first polar body is larger 
than the second. 
The spermaries are composed of an indefinite number of cysts. 
The individual cysts originate from a single or several interstitial 
cells. Fig. 2 represents a longitudinal section of a single cyst 
containing spermatogonia, which have originated from a single 
interstitial cell. After the spermatogonia have divided a few 
times (the number of divisions varies in different individuals) 
those found at the distal end of the cyst become transformed into 
the spermatocytes of the first generation without growth. In 
Fig. 3 the spermatocytes of the first generation are recognized by 
their dense chromatin mass. Different stages of development are 
found in a single cyst, ranging from the spermatocytes of the 
first generation to mature sperm, as shown in Fig. 4. The suc- 
cessive zones of the developing sperm are distinct. The nuclei 
OBSERVATIONS ON GERM CELLS OF HYDRA. 
2O7 
at the extreme proximal end of the cyst cease dividing and become 
the nuclei of new interstitial cells after the spermaries have dis- 
appeared. 
My observations on the nuclear changes in the different stages 
of development confirm those of previous investigators. Reduc- 
tion occurs after the last spermatogonial division just before the 
spermatocytes of the first generation divide. In the division of 
the nuclei of the spermatocytes of the first and second genera- 
tions, the cell wall remains intact and produces a multinucleate 
cell (Figs. 5-7), and give rise to four sperm within a common 
FIG. i. Longitudinal section showing two developing eggs, a and b, eggs. 
FIG. 2. Longitudinal section of a single cyst. 
FIG. 3. Section of cyst little later than preceding, a, spermatogonia ; 
b, spermatocytes of the first generation. 
FIG. 4. Longitudinal section of cyst showing maximum development. 
c, spermatocytes of the first and second generations ; d, mature sperm. 
FIGS. 57. Formation of spermatids within common vesicle. 
FIGS. 8io. Developing sperm. Figs. 510, from living cells. 
FIGS, ii AND 12. Cysts removed from living spermary. 
2O8 GEO. W. TANNREUTHER. 
vesicle. The vesicular wall is very thin and the four sperm pass 
through it when mature and become free within the distal end of 
the cyst (Fig. 4). The mature sperm are very active within the 
cyst and finally escape to the exterior through a small temporary 
opening of the spermary. After the mature sperm have all 
escaped, the opening closes until more sperm mature. When the 
spermaries are extremely large, this process continues from forty 
to fifty hours. 
In order to show the four sperm within a common vesicle, it 
is necessary to dissect the living spermary apart, as the sperm 
always escape from the common vesicle before passing to the 
exterior. The individual cysts when separated become more 
spherical (Figs, n and 12), and the different stage of develop- 
ment in the living sperm can easily be distinguished. 
Kleinenberg in his description of the sperm in Hydra viridis 
definitely states that the sperm are formed from interstitial cells 
that have divided a number of times; ultimately the nucleus of 
the cell (the spermatocyte of the first order) disintegrates while 
the cell substance becomes granular, and in place of the nucleus 
there appears from one to four refractive bolies, which give rise 
to the sperm. The refractive bodies referred to beyond doubt 
result from the two divisions of the nucleus, which he thought 
disintegrated. The four nuclei (spermatids) within the common 
vesicle do have the appearance of refractive bodies, especially in 
'the living material. 
Korotneff (5) gave similar results. He states that the sperm 
form directly from the nuclei of a multinucleate mother cell. 
Downing (2) does not mention this interesting phenomenon, 
which is found in Hydra viridis and Hydra sp. ? (Brauer). 
The mature sperm possess extreme vitality and may remain 
active from one to three days after escaping from the spermary. 
The mature sperm of any individual spermary possess about the 
same degree of fertility. No degenerating spermatogonia were 
found. 
In the monoecious form Hydra viridis the spermaries become 
mature before the ovaries. Occasionally sperm and ova on one 
individual would ripen at the same time, making self-fertilization 
possible. In order to prove that self-fertilization did occur, indi- 
OBSERVATIONS ON GERM CELLS OF HYDRA. 2CX) 
viduals with both spermaries and ovaries at the same stage of 
development were isolated and placed in distilled water for one 
hour, in order to kill any mature sperm that might adhere from 
other individuals. The Hydra were then placed separately in 
water free from sperm. The spermaries and ovaries matured 
and self-fertilization took place. The cleavage was normal. 
SUMMARY. 
1. Hydra sp. ? (Brauer) does not show continuity of the germ 
plasm. 
2. The eggs can be distinguished from the interstitial cells in 
the adult Hydra before the ovary is formed. This is especially 
true during the breeding season. 
3. The different generations in the formation of the germ cells 
are comparable to those of higher animals. 
4. In the division of the spermatocytes of the first and second 
generations the cell wall remains intact and the four spermatids 
are formed within a common vesicle, each producing a mature 
sperm. 
5. Self-fertilization occurs in Hydra viridis. 
ZOOLOGICAL LABORATORY, 
UNIVERSITY OF MISSOURI. 
BIBLIOGRAPHY. 
1. Brauer, A. 
'91 Ueber die Entwicklung von Hydra In. Z. f. Wiss., Zool., 52, 2. 
2. Downing, E. R. 
'04 The Spermatogenesis of Hydra. Zool. Jahr., Band 21. 
'08 The Ovogenesis of Hydra fusca A preliminary paper. Biol. Bulletin, 
Vol. XV., No. 2. 
3. Hertwig, R. 
'06 Ueber Knospung und Geschlechtentwicklung von Hydra fusca. Biol. 
Centralbl., Vol. 26. 
4. Kleinenberg, N. 
'72 Eine anatomisch-entwicklungsgeschichtliche, Untersuchung. Leipzig. 
5. Korotneff, A. 
'83 Zur Kentniss der Embryologie der Hydra. Z. Wis. Zool., Bd. 38. 
6. Tannreuther, Geo. W. 
'08 The Development of Hydra. Biol. Bulletin, Vol. XIV., No. 5. 
BUDDING IN HYDRA. 
GEO. W. TANNREUTHER. 
Budding in hydra has received the attention of only a few 
investigators. Our principal information as to the origin of the 
bud is derived from Lang's 1 account. He describes the bud as 
beginning by an increase in volume and division of the interstitial 
cells. After the ectoderm becomes thickened, the mesoglea dis- 
appears and the cells pass from the ectoderm into the endoderm. 
This process continues until the ectoderm becomes reduced to its 
normal thickness. The mesoglea then reforms and a cavity ap- 
pears in the thickened endoderm, which becomes the enteron of 
the new individual. 
The main object of the present paper will be to give a brief 
account of the origin and development of the buds in hydra, more 
especially their manner of growth and what cells contribute most 
to their rapid formation. 
The species studied, Hydra viridis and Hydra sp.? (Brauer) 
differ somewhat from Lang's account. The mesoglea does not 
disappear and the ectodermal cells do not pass into the endoderm. 
The bud, however, begins by an increase in volume and division 
of the interstitial cells. After they have increased once or twice 
in volume, as shown in Fig. I, there is a slight outbulging of 
the ectoderm, which is scarcely perceptible. Fig. 2 represents 
the condition of the interstitial, ectodermal and endodermal cells 
in the origin of the bud as they appear more highly magnified. 
A few mitotic figures are visible. No amitotic divisions were 
observed. The endodermal cells contain numerous food particles, 
which may pass intact through the mesoglea into the ectoderm. 
The cells directly concerned in the formation of the bud are well 
supplied with food, while the remaining cells of the parent hydra 
show a scarcity. Many of the endodermal cells in the distal half 
of the hydra have a glandular appearance and are most active in 
1 Lang, Albert, " Uber die Knospung bei Hydra und einigen Hydropolypen," 
Z. wiss. Zool., Vol. 54, 1892. 
2IO 
BUDDING IN HYDRA. 
211 
the secretion of the digestive fluid, which aids in the breaking up 
of the food in the enteron, while those endodermal cells in the 
region of growth, " the formation of buds and sexual organs," 
are the most active in ingesting the partly digested food from the 
enteron and preparing it for diffusion into the ectodermal cells. 
The ectoderm and endoderm in the early formation of the bud, 
which are quite uniform, soon become differentiated into two dis- 
FIG. i. Longitudinal section of hydra showing the origin of bud at a. 
FIG. 2. Part of preceding figure at a. X 80. 
FIG. 3. Longitudinal section through wall of hydra showing curvature of 
mesoglea and beginning of enteron in forming bud. m, mesoglea ; a, enteron. 
FIG. 4. Stage of development a little later than preceding, a, the point 
of junction between parent and bud. 
FIG. 5. A portion of preceding figure at a, which includes the entire area 
in the formation of bud. end., endoderm. X 90. 
FIG. 6. Diagram of a longitudinal section of parent hydra with bud. 
a, active growing region in formation of bud. 
212 GEO. W. TANNREUTHER. 
tinct regions. The cells more centrally placed (Fig. 3, a), which 
correspond to the distal end of bud, become inactive, while those 
on either side of the central -region continue active, divide rapidly 
and contribute almost entirely to the rapid growth of the bud. 
There is a slight curvature in the mesoglea (Fig. 3, m), and the 
enteron of the new individual becomes apparent. In a stage little 
later than the preceding (Fig. 4), the relation of bud to parent 
and the condition of cells is more plainly represented. Fig. 5, 
a portion of Fig. 4 at a, represents the entire area that contributes 
directly to the growth of the bud. The cells at the apex of the 
forming bud are small. Those on either side are larger and more 
active in the process of division and growth. Their contents is 
very similar, showing an abundance of food material and cyto- 
plasmic granules. 
The ectoderm and endoderm at the junction of the parent and 
bud (Fig. 7) divide very rapidly and become the most active 
growing region in the production of the new individual. The 
remaining cells of the bud seldom divide. The formation of the 
tentacles is similar to that of the buds. The cells corresponding 
to their basal ends take the most active part in their growth. The 
mouth is formed at the distal end of the bud by a breaking 
through of the ectoderm. 
The rate of growth of the bud is determined by the amount of 
food present. In an active feeding hydra, the buds are often 
completely formed in thirty to forty hours. While in those 
hydras with a moderate supply of food the buds grow very slowly 
and may require four or five weeks or even more time for their 
complete development. In the latter instance after the buds are 
nearly formed, they will be absorbed in the absence of food. In 
the process of absorption the cell walls of the bud become imper- 
ceptible and the cell contents presents the appearance of a com- 
plete syncytium. When the buds are nearly absorbed, if the 
hydra is again supplied with food, the buds very seldom reform. 
In a few instances the buds were neither absorbed nor reformed, 
but remained attached to the parent as permanent individuals. 
Tentacles were formed later. 
When the buds have reached their complete development the 
ectodermal cells at the proximal end undergo a rapid change 
BUDDING IN HYDRA. 
213 
(Fig. 8, cct). They become more narrow, elongated and present 
the appearance of glandular cells. They have the power to 
FIG. 7. Longitudinal section at junction of parent hydra and bud, taken 
at a in Fig. 6. end., endoderm. X 100. 
FIG. 8. Longitudinal section through base of mature bud and wall of 
parent hydra. ;,, mesoglea of bud; -m, mesoglea of parent; ect., glandular 
ectoderm ; ent.. enteron of bud. 
FIG. 9. Longitudinal section showing change of mesoglea in separation of 
parent and bud at a, and formation of new mesoglea at b and c. The endo- 
dermal cells at proximal end of bud have united, a, mesoglea between bud 
and parent becoming thinner ; b. new mesoglea of bud ; c, new mesoglea 
of parent. 
FIG. 10. Longitudinal section showing completion of mesoglea at b and c 
and persistence of old mesoglea at a. 
214 GEO - W - TANNREUTHER. 
secrete a sticky substance before the bud becomes separated from 
the parent. 
The first step in the process of separation of bud from parent 
occurs in the mesoglea. Fig. 8 represents the condition of meso- 
glea and cells before separation begins. The mesoglea, which 
connects the bud with parent, is uniform throughout. But almost 
immediately it becomes thinner and thinner until it is indistin- 
guishable from the ordinary cell wall (Fig. 9, a). The enteron 
leading from the parent to the bud becomes discontinuous by the 
union of the endodermal cells. New mesoglea is now formed at 
the extreme basal end of the bud (Fig. 9, b} and at the point of 
former union with the parent at c. The mesoglea, which is more 
of a gelatinous nature, increases in thickness "by means of secre- 
tion from the endodermal cells and soon reaches its normal con- 
dition (Fig. 10, b and c). After the formation of the mesoglea 
is complete, the bud remains attached to the parent by a few ecto- 
dermal and endodermal cells, as shown in Fig. 10. The former 
connecting mesoglea is represented by the dotted lines at a. The 
cells between the dotted lines, which are endodermal, become 
external to the newly formed mesoglea and take the position of 
ectoderm. Whether these cells persist or not and function as 
ectoderm is difficult to say, as there is no possible means of fol- 
lowing them in the process of separation of bud from parent. 
SUMMARY. 
1. The initial step in the formation of bud in hydra is found in 
the interstitial cells. 
2. The most active region of growth in the formation of the 
bud is found at the junction of the forming bud and parent, where 
the cells divide very rapidly and contribute almost entirely to its 
growth. 
3. When the bud is nearly formed the ectodermal cells in the 
basal region become transformed into granular glandular cells, 
which later secrete a glutinous substance for attachment of hydra. 
4. The rate of growth in buds is controlled by the amount of 
food present. Starvation after buds are nearly complete often 
causes their complete absorption by parent hydra. 
ZOOLOGICAL LABORATORY, 
UNIVERSITY OF MISSOURI. 
Vol. XVI. April, /pop. No. 5. 
BIOLOGICAL BULLETIN 
THE NUCLEOLI IN THE SPERMATOCYTES AND 
GERMINAL VESICLES OF EUSCHISTUS 
VARIOLARIUS. 
TvATHARINE FOOT AND E. C. STROBELL. 
I. In the resting first spermatocyte of many insects there 
appears a spherical, chromatic body which invariably stains as 
intensely as the chromosomes. 
McClung ('02) was the first investigator to draw special 
attention to this structure by his suggestion that it may function 
as a sex determinant. 
A majority of the cytologists who have studied this structure 
interpret it as one of the chromosomes which persists through 
the rest stage of the first spermatocyte, and they claim that its 
presence or absence in a spermatozoon is the determining factor 
of sex, McClung assuming that its presence in a spermatozoon 
causes the fertilized egg to produce a male, whereas Stevens, 
Wilson and others affirm that its presence in a spermatozoon 
causes the fertilized egg to produce a female, and in those cases 
where two (a large and a small) chromatin nucleoli are present, 
the larger is the female-determining, and the smaller, the male- 
determining factor. 
Identifying the chromatin nucleolus as a chromosome enlarges 
at once the sphere of the problem, involving this structure in the 
maze of hypotheses and theories associated with the chromosomes. 
This identification of the chromatin nucleolus as a chromosome 
is of great importance, because it presumably demonstrates that 
a definite chromosome retains its individuality throughout the 
rest stage, and it presents very strong evidence for the theory 
of the individuality and continuity of the chromosomes-- the 
theory which is the corner-stone of all hypotheses involving the 
215 
2l6 KATHARINE FOOT AND E. C. STROBELL. 
claim that the chromosomes are the cause rather than the ex- 
pression of cell activities. 
We believe that the evidence we find in Enschistus, as well as 
in Anasa ('07), is distinctly opposed to the interpretation that the 
chromatin nucleolus of the resting first spermatocyte is a phase 
of one or more of the chromosomes, but we shall reserve the pub- 
lication of this evidence in huscliistns until we can control it by a 
more complete comparison with other forms. In the present 
paper we shall limit ourselves to the relatively simple question, 
is the chromatin nucleolus a structure associated with the male 
cell only, as claimed by McClung, or is it, as claimed by Wilson, a 
chromosome received from the egg, which during the rest stage 
appears in the form of a chromatin nucleolus ? 
Two chromatin nucleoli unequal in size have been demonstrated 
by Montgomery ('98, 'or, '06) in the first spermatocyt.es of 
several varieties of Euscldstus, including Euschistus variolaritts, 
and Wilson ('05) has demonstrated these structures in other Hem- 
iptera, naming them chromosome nucleoli, because he believes 
them to be chromosomes. In describing them in Cccnus and 
Lygcens he says : " Throughout the whole of the growth period 
in Ccemis, and from stage e onward in Lygceus, at least one of the 
idiochromosomes can always be distinguished as a compact, 
spheroidal, intensely staining chromosome nucleolus and fre- 
quently both idiochromosomes are distinguishable in this form 
in all of these stages. ... It is clear beyond all question that 
at least the large idiochromosome may retain its identity through- 
out the whole growth period, with the small idiochromosome the 
case is not so strong" (p. 389). Of Brochymena he says : " There 
can be no doubt that when only one chromosome nucleolus is 
present it is to be considered as a bivalent body arising by the 
fusion or synapsis of the two idiochromosomes " (p. 392). 
In the resting first spermatocyte of Enschistiis variolarius the 
chromatin nucleolus is the most deeply stained and conspicuous 
structure in the cell (Photos I ro). As a rule, only one is present, 
but quite frequently there are two, sometimes equal, but in most 
cases, very unequal in size ; exceptionally we have found three, 
or even four, nucleoli in the same nucleus. 
We roughly estimated the number of the nuclei with one or 
NUCLEOLI IN EUSCHISTUS* VARIOLARIUS. 2 I/ 
more chromatin nucleoli, by counting 625 cells in one testis, and 
of these cells 591 showed one nucleolus, 27 two nucleoli, and 7 
three nucleoli. These 625 cells represented only a small area of 
the preparation and we did not repeat the experiment on another 
testis the estimate, therefore, offers merely an indication that in 
a large majority of cases only one chromatin nucleolus is present. 
By a comparative study of the spermatocytes and oocytes of 
the same form we ought to find an answer to the question, 
whether the chromatin nucleolus is associated with the male cell 
only, or whether it has its origin in the oocyte. 
As far as we are aware, the germinal vesicles shown in Plates 
I., II. and III. are the first that have been demonstrated in any of 
the Hemiptera heteroptera. A great deal has been written about 
the spermotocytes and important generalizations have been drawn 
from the behavior of the chromatin nucleolus in these male cells, 
but not one word has been said about the corresponding, very im- 
portant stages in the female, although we have been confidently 
assured that the chromatin nucleolus in the male has been derived 
from the egg. If this is true, it would seem at least logical to 
expect to find some evidence of its presence in the egg at the 
stage of development corresponding to the stage in which it is so 
conspicuous in the male cell. 
Even if we assume for the sake of argument that the chromatin 
nucleolus is a chromosome, we still do not avoid the logical ex- 
pectation of finding it in the oocytes, for if one of the chromo- 
somes of the spermatocyte possesses the distinguishing trait of 
persisting through the rest stage in the form of a chromosome 
nucleolus, and if that chromosome is a chromosome contributed 
to the male cells from the mother, we should expect to see its 
distinguishing characteristic manifested in even a more marked 
degree in the oocyte. And further, as the chromosome nucleolus 
of the spermatocyte is claimed to be only one of a pair of chromo- 
somes --its mate being contributed by the egg at the time of 
fertilization -- we should expect to find in the resting oocyte a 
pair of chromosomes represented by two univalent, or one biva- 
lent chromosome nucleolus. 
In the case of Enschistus we are told that the larger of the two 
chromatin nucleoli of the spermatocyte is the homologue of the 
2l8 KATHARINE FOOT AND E. C. STROBELL. 
accessory chromosome of other forms, and if this interpretation is 
correct we may expect to find a large bivalent or two univalent 
chromatin nucleoli in the growing oocytes. 
II. We find the chromatin nucleolus of the spermatocyte per- 
sisting through the entire growth period, Photos I to 10. All 
the resting spermatocytes shown in these photos, are from testes 
which contain many first spermatocyte metaphases, in which 
the idiochromosomes show the typical inequality in size. We 
have selected preparations of the resting spermatocytes showing 
variations in the relative size of the two chromatin nucleoli 
(when present) in order to demonstrate their frequent lack of con- 
formity to the size relations of the idiochromosomes. Photo 10 
shows a nucleolus persisting until the chromosomes are formed 
and we have several photographs demonstrating its presence to a 
still later stage, where the seven bivalents can be counted, but we 
shalLreserve the discussion of these later stages for a subsequent 
paper. 1 
As the chromatin nucleoli of the first spermatocyte of EitscJiis- 
tiis persist through the entire growth period, we certainly have a 
right to expect to find in the egg the equivalent of the large 
chromatin nucleolus during these same stages, if as asserted it 
owes its origin to the egg. Again, if the so-called male sex- 
determinant (the small chromatin nucleolus) persists through the 
growth stages in the male cell, there seems to be no clear reason 
why the so-called female sex-determinant (the larger chromatin 
nucleolus) should not persist through the growth period of the 
oocyte. 
A study of the oocytes shown in Photos 1 1 to 30 demonstrates 
that in none of these cells can be found a structure resembling in 
any way the chromatin nucleolus shown in the spermatocytes of 
Photos I to 10. This is clearly demonstrated in the sperma- 
tocytes and oocytes of Plate I. There is no structure in the 
oocytes of Photos n to 19 which resembles in the least the 
chromatin nucleolus of the spermatocytes of Photos i to 10. The 
only nucleolus in the oocytes is a relatively .large achromatic 
structure, first clearly demonstrated in the older oocytes. It is 
1 We have a number of photographs showing the transition stages between Photos 
9 and lo, but lack of space prevents their reproduction on these plates. 
NUCLEOLI IN EUSCH1STUS VARIOLARIUS. 219 
clearly seen in the germinal vesicles of Photos 18 and 19, Plate 
I., and in the germinal vesicles of Photos 20 to 25, Plate II., 
and in Photos 26 to 30, Plate III. All these preparations show, 
that as a rule its position is peripheral and this is also true 
of the principal nucleolus of the germinal vesicles of Allolobo- 
pJiora (Photo 31). The absence of a chromatin nucleolus is 
conspicuous not only in the germinal vesicles of Euschistus but in 
the young oocytes as well. Compare the young oocytes of Photos 
1 1 and 12 with the young spermatocytes of Photos I to 5. 
In none of the oocytes, from the earliest to the latest stages 
do we find a chromatin nucleolus, and its absence in the ger- 
minal vesicles of Euschistus is made more conspicuous by the 
fact that a chromatin nucleolus is invariably present in the ger- 
minal vesicles of AllolobopJiora. If we could find such a struc- 
ture in the germinal vesicles of Euscliistus, in addition to the 
achromatic nucleolus, the advocates of the sex determination 
hypothesis could justly claim it as convincing evidence in support 
of their theory of its female origin, even if a demonstration of its 
relation to one of the chromosomes was lacking. 
The absence of a chromatic nucleolus in these germinal vesi- 
cles cannot be due to the technique since in this case the more 
delicate achromatic nucleolus would show some disturbance. 
This point is exemplified in the germinal vesicle of Allolobophora 
(Photo 31), for in this preparation the pricking has disturbed the 
large, less chromatic nucleolus, while the small, denser, chro- 
matic nucleolus remains perfectly intact, and this condition holds 
true for the hundreds of germinal vesicles of Allolobophora we 
have preserved. 
It therefore seems fair to assume that the technique is not 
responsible for the absence of a chromatin nucleolus in the ger- 
minal vesicles of EnscJiistus. 
In an earlier paper : we ventured to call attention to the like- 
ness of the chromatin nucleolus in the egg of AllolobopJiora to 
the chromatin nucleolus of the first spermatocytes of insects. 
Compare the chromatin nucleolus in EnscJiistns, Photos i to 10, 
with the chromatin nucleolus in the egg of AllolobopJiora, Photos 
3 i and 32. In both these forms the chromatin nucleolus stains as 
'Foot and Strobell ('05). 
22O KATHARINE FOOT AND E. C. STROBELL. 
intensely as the chromosomes, and in both forms we sometimes 
find one and sometimes two chromatin nucleoli. 1 The comparison 
may be carried even a step further. We may compare the 
chromatin nucleolus of Alloloboplwra to the chromatin nucleolus 
of some forms where it is figured at one pole of the first spindle 
as an undivided chromosome. The chromatin nucleolus of 
Allolobophora often persists through the first division, and in such 
cases it is either free in the cytoplasm, near the spindle, or at one 
pole of the spindle in which position it strongly resembles the 
undivided chromosome figured in so many forms. Photo 32 
shows a section through the first maturation spindle of an egg of 
Allolobophora, and a comparison of the chromatin nucleolus at 
the upper pole of this spindle, with the chromatin nucleolus in 
the germinal vesicle of Photo 31 leaves no doubt of the identity 
of the two structures. The identification is further confirmed by 
the clear demonstration in this and the adjoining sections of all 
eleven chromosomes in the equatorial plate of the spindle. If the 
chromosomes of Allolobophora were spherical as they appear in 
sections of so many forms, we could not so easily identify the 
chromatin nucleolus, for it stains exactly like the chromosomes. 
Such a case is described by Arnold ('08) in the spermatogenesis 
of the Coleopteran Hydrophilus piceus. During the rest stage 
the chromatin nucleolus of this form is easily recognized, but he 
finds it impossible to differentiate it from the chromosomes during 
the first prophase, because the chromosomes at this stage are 
also spherical and both structures stain alike. He is able to 
identify it again at the first metaphase, where it lies in the cyto- 
plasm near the spindle, or within the spindle at one pole. It 
persists through the first division, but disappears before the 
second mitosis. 2 It is clear that in its origin, behavior and time 
of disappearance the chromatin nucleolus of the spermatocytes of 
HydropJiilits closely resembles the chromatin nucleolus of Allo- 
lobophora, with the exception that the latter does not invariably 
persist through the first division. Arnold ('oS) supports Moore 
and Robinson's ('05) conclusion that the chromatin nucleolus of 
1 Foot and Strobell ('05), Photo 115. 
2 Wheeler ('97) figured the nucleolus of the germinal vesicle of Mysostoma 
glabrum persisting through the second cleavage and he says that in some cases it 
persists through the next stage, and perhaps even later, page 35. 
NUCLEOLI IN EUSCHISTUS VARIOLARIUS. 221 
the first spermatocyte is not a chromosome and does not take 
part in either division. His evidence for this is very convincing, 
he says, if it is to be interpreted as an accessory chromosome 
which remains undivided in the first spindle it should divide in 
the second spindle, and half the second spindles should show one 
more chromosome than the other half. Arnold finds, on the 
contrary that the number of chromosomes in the second spindle 
is as constant as the number in the first. He gives in the follow- 
ing table the results he obtained : 
Out of 100 counts of first meiotic : 
85 per cent, gave 15 gemini and the nucleolus. 
ij it over T C " " " " 
8 " " " under 15 
Out of another 100 counts of first meiotic : 
90 per cent, gave 15 gemini and the nucleolus. 
3 " " " over 15 " " " " 
7 " " " under 15 " " " 
Out of 100 counts of second meiotic : 
91 per cent, gave 15 chromosomes. 
6 " " " over 15 " 
3 " " " under 1 5 " 
The resemblance between the chromatin nucleolus of the first 
spermatocyte of insects and the chromatin nucleolus in the ger- 
minal vesicles of some forms has been noted by Hacker ('07) who 
says, that in those cases in which the heterochromosome is fig- 
ured as a cap-shaped mass on a large pale plasmosome, the two 
structures are extraordinarily like the two kinds of nucleoli in 
the germinal vesicles of the lamellibranch type. He says : " So 
kann man sich dem Eindruck nicht entziehen, das z. B. das, 
' akzessorische Chromosom ' in den ruhenden Spermatocyten 
von Scolopendra (Blackman, 1905, tab. 2, fig. 10-15) fruherein- 
fach als ' Nucleolus ' beschrieben worden ware, und man ist gen- 
eigt, sich zu fragen, ob denn in alien vorliegenden Angaben 
bereits eine reinliche Scheidung zwischen den Heterochromoso- 
men und den echten Nucleolen (Plasmosomen) durchgefiihrt ist. 
Es sei mirauch gestattet, darauf hinzuweisen, das die von Wilson 
u. a. gegebenen Bilder, in welchen ein Heterochromosom als 
222 KATHARINE FOOT AND E. C. STROBELL. 
kappenformige Masse einem grossen Plasmosom aufgelagert er- 
scheint, eine auserordentliche Ahnlichkeit mit den Befunden in 
den Keimblaschen des Lam'ellibranchiaten-Typuszeigen, d. h. in 
denjenigen Keimblassen, in welchen sich zweierlei Nucleolen von 
verschiedener Lichtbrechung, Quellbarkeit und Tingierbarkeit 
befinden." 
It is a suggestive fact that in Allolobophora we have an her- 
maphrodite form, and if one is determined to regard the chromatin 
nucleolus of the spermatocytes as a sex-determinant, some inter- 
esting conclusions might be drawn from a comparison of the two 
types of nucleoli observed in this hermaphrodite form, with the 
two types characteristic of the male and female cells in Euschistus. 
Compare for example the large nucleolus in the germinal vesicle 
of Allolobophora (Photo 31) with the large achromatic nucleolus 
of the germinal vesicles of Euschistus (Photos 1 8 to 30), and com- 
pare further the smaller chromatin nucleolus of Allolobophora 
(Photos 31 and 32) with the chromatin nucleolus of the sperma- 
tocytes of Euschistus (Photos i to 10). 
Even if we could be sure that in Euschistus the achromatic 
nucleolus of the egg cells is not represented in the male cells, and 
that a striking likeness exists between the male and female nu- 
cleoli of Euschistus and the two types of nucleoli in the germinal 
vesicles of Allolobophora, we would still be unable to claim from 
the comparison any results of general significance, for the reason 
that the observations of many investigators of the spennato- 
genesis of insects offer contradictions to an assumption that the 
achromatic nucleolus is associated solely with the female cell 
these observers having figured a pale, achromatic nucleolus 
in the spermatocytes, in addition to the chromatic nucleus char- 
acteristic of these cells. As opposed to these facts, however, 
other investigators have found no structure in the spermatocytes 
that can be interpreted as a nucleolus, other than the character- 
istic chromatin nucleolus. 1 These conflicting observations, if 
equally reliable, compel the conclusion that individual forms may 
differ as to the absence or presence of an achromatic nucleolus. 
1 In his " Studies on Chromosomes IV.," Jour. Exp. Zool., Vol. VI., No. i, 
1909, Wilson in describing Syromastes, does not mention a large pale plasrnosome, 
such as he has described in other forms. In Pyrrochoris he states that there are from 
one to three nucleolar-like bodies, which on account of the staining reactions, he 
believes to be plasmosomes, page 82. 
NUCLEOLI IN EUSCHISTUS VARIOLARIUS. 223 
In Euschistus we have not been able to demonstrate its pres- 
ence at any stage of the growth-period of the spermatocytes. In 
sections we often find faintly staining areas that might be inter- 
preted as an achromatic nucleolus, but in view of the possibility 
of artefacts in such preparations, we hesitate to interpret them as 
true nucleoli, unless we can support the interpretation in our 
smear preparations. Until this point can be settled we are not 
justified in drawing any conclusions from the obvious differ- 
ence in type between the nucleoli in the male and female cells of 
Euschistus, though we have here quite as marked a sexual dif- 
ference of the nucleoli, as any that has been shown for the 
chromosomes, a difference that appears in no way associated with 
the maturation divisions of either germ cell. 
In comparing the achromatic nucleolus of the oocytes of 
Ensc/iishis with the principal nucleolus of AllolobopJwra we find 
some individual differences between the two. The principal 
nucleolus of Allolobophora is clearly differentiated in the young 
oocytes as a small, dense chromatic body, and it can be traced 
uninterruptedly through the entire growth period of the oocytes. 
During this period it increases in size, proportionately to the 
growth of the nucleus, gradually becomes less dense and less 
chromatic, and finally, when the germinal vesicle has reached its 
maximum size the principal nucleolus often appears as in Photo 
31. In some cases, however, its dense and chromatic character, 
distinctive of the earlier stages, persists until the chromosomes are 
fully formed or again it may so completely disintegrate that only 
a clear space remains as evidence of its existence. 1 
The achromatic nucleolus of Euschistus differs from the prin- 
cipal nucleolus of Alloloboplwra in not being present in the young 
oocytes as a small dense chromatic nucleolus. We have been 
unable to demonstrate any such chromatic body in the young 
oocytes of Euschistus. In the youngest stages in which a nucle- 
olus is found, it appears as clearly achromatic as when it has 
reached its maximum growth in the germinal vesicle compare 
Photos 13 and 16 with Photos 18 to 30. There are also indica- 
tions that it is often not formed until after the oocytes have attained 
a definite growth, for in such clear preparations as are shown in 
1 Foot and Strobell ('05), Photo 122. 
224 KATHARINE FOOT AND E. C. STROBELL. 
Photos ii, 12, 14, 15 and 17, there is no evidence that such a 
structure is present. 1 
If the nucleolus in the germinal vesicle of Euschistus corre- 
sponds to the principal nucleolus of Allolobophora it must be con- 
ceded that there are marked individual differences between the 
two, differences quite as marked as those existing between the 
chromosomes of these two forms. 
INDIVIDUALITY AND CONTINUITY OF THE CHROMOSOMES. 
The investigators who question the individuality and con- 
tinuity of the chromosomes urge the necessity for further study 
of the critical stages bearing on this problem the stages occur- 
ring between the end of one mitosis and the beginning of the 
next. Meves ('08) holds that during these stages it is impossible 
to recognize the chromosomes. He says : " Man sucht hier, 
wie O. Hertwig 1890 geschrieben hat, in den Kern etwas hinein- 
zudemonstrieren, was kein unbefangener Beobachter in seiner 
Struktur erkennen wird. 
Pick ('07) claims, that a study of the rest stages proves that 
the theory of the individuality and continuity of the chromosomes 
is untenable. 
Tellyesniczky ('07) also makes a strong plea for more thor- 
ough and careful work on these stages. We are in sympathy 
with him in his skepticism concerning the individuality and con- 
tinuity of the chromosomes, but our results differ from his in 
some, perhaps unimportant, details --differences which may 
normally exist in different forms. For example Tellyesniczky 
('05) believes that a homogeneous distribution of the nuclear sub- 
stance throughout the nuclear vesicle precedes every mitosis, and 
the young spermatocytes of our Photos I, 4 and 5 lend support 
to this claim. The young oocyte of Photo 13 also indicates a dif- 
fused condition of the chromatin, but as a rule the chromatin of 
the young germinal vesicles appears as fine granules often very 
evenly distributed throughout the nucleus as in Photos n, 14 
and 1 7. Possibly this granular condition may be an artefact 
one phase of a precipitate left after drying, for in some cases the 
1 Among recent observations on this point, Deton ('08) finds in Ihysanozoon 
Brochii that the nucleolus is absent in the young oocyte, first appearing at the begin- 
ning of the growth period. 
NUCLEOLI IN EUSCHISTUS VARIOLARIUS. 225 
chromatin of the young oocytes is as homogeneous as that of 
the spermatocytes of Photos I, 4 and 5. 
We find nothing in these germinal vesicles answering to Telly- 
esniczky's karysomes, unless he would homologize the aggre- 
gations of chromatin shown in Photos 12, 15 and 16 with these 
structures. These aggregations of chromatin are too numerous 
to be interpreted as the seven bivalent chromosomes of the later 
stages, or as their fourteen component univalents. In the germi- 
nal vesicle of Photo 15, for example, there are about fifty clumps 
of chromatic substance. Perhaps these clumps of chromatin cor- 
respond more closely to the segregated granules Wassilieff ('07) 
describes in the nucleus of Blatta. These clumps, he says, are 
too numerous to be interpreted as representing individual chro- 
mosomes. Later he finds that they disintegrate and are scattered 
throughout the nucleus like a fine dust ; the delicate threads of 
o 
the later stages being formed at the expense of this dust-like 
substance. 
In the germinal vesicles of Euschistns variolarius the chromo- 
somes first appear as extremely fine threads, which seem to arise 
as a concentration of the distributed granular mass, these granules 
gradually disappearing as the threads become more defined. 
Photo 17 shows a germinal vesicle at the stage just before the 
chromosomes appear the chromatin is still distributed as gran- 
ules throughout the nucleus, although the size of this nucleus is 
nearly equal to many of the germinal vesicles in which the chro- 
matic threads are fully developed. In Photos 18 and 20 we see 
the delicate chromatic threads just appearing, as yet very indis- 
tinct, and in the later stage shown in Photo 21 the threads are 
more distinct but too much tangled to warrant any precise in- 
terpretation. In later stages, however, we can often trace among 
the filaments, an unbroken thread, so long that it cannot possibly 
be interpreted as one of the fourteen univalents, it must represent 
at least two univalents attached end to end. Again some of 
these threads are long enough to justify the inference that they 
represent even more than a bivalent, indicating that in Euschistus 
the chromosomes emerge from the rest stage not as univalents, 
but as long threads representing at least one or more bivalents, 
thus supporting the observations of those who believe that bival- 
226 KATHARINE FOOT AND E. C. STROBELL. 
ents arise by the omission of a transverse division of a spireme, 
rather than by a conjugation of univalents. We believe that the 
evidence in Euschistus is opposed to the interpretation that the 
chromosomes first appear as univalents and later conjugate, either 
longitudinally or end to end. We believe, rather, that there is a 
tendency to form a continuous spireme, because we find cases in 
which very long, thin, unbroken threads are present, and a few 
examples of this kind outweigh as evidence, a large number of 
cases where we find the threads broken into small pieces. These 
delicate threads could easily be broken by the technique, for after 
pricking, the germinal vesicles flatten and dry on the slide^and 
this disturbance, combined with the shrinkage of drying, could 
easily account for the rarity of the cases in which the long thin 
threads remain unbroken. 
In Photo 24 we see this tendency to the formation of a spireme 
-a few of the threads can be traced certainly too far to justify 
any assumption that they are univalent or even bivalent chromo- 
somes. A long unbroken thread is also demonstrated in Photos 
23, 28 and 29. 
Tellyesniczky ('07) claims that the long thin chromatic 
threads do not become shorter and thicker by contraction. He 
thinks rather this is accomplished by a flowing together and 
change of place of the substance of which the thread is composed. 
He says, there is a continual disintegration and rebuilding of the 
threads and the new structure is formed at the expense of the old. 
We believe this is not the case in Ensc/iistiis. For example, it 
seems reasonable to interpret the chromosomes of Photo 30 as 
due to contraction of long thin bivalents (as shown in Photo 25), 
rather than to regard them as entirely new formations. Such a 
detail as this is interesting, but it seems to us only incidental to the 
all-important question, do the chromosomes retain their individu- 
ality throughout the rest stages do they differ in this particular 
from other structures of the cell the centrosome the aster - 
the spindle the nucleolus, etc. These elements return at each 
cell generation with as much regularity as to size and form as do 
the chromosomes, but those who believe in the individuality and 
continuity of the chromosomes attribute to the latter structures a 
causal significance not now ascribed to any other organ of the 
cell. 
NUCLEOLI IN EUSCHISTUS VARIOLARIUS. 22/ 
In the past, however, similar claims have been made for the 
centrosome--its individuality and continuity were asserted by 
many cytologists, some going so far as to believe the centrosome 
to be the cause rather than the expression of cell activity. 
Certain facts indicate that all the chromatic substance of the 
germinal vesicle of Enschistus is not used for the chromosomes. 
We find in many germinal vesicles various segregations of chro- 
matic substances which are present after the chromosome threads 
are formed and which often do not entirely disappear until the 
prophase stage. We sometimes find irregular granular or homo- 
geneous masses as in Photos 20 and 22, sometimes numerous large 
granules, so dense that they look like very small nucleoli (Photos 
19, 21, 23, 26, 27 and 29), and sometimes merely deeply staining 
areas, as shown in Photos 23, 24 and 27. We have no means of 
proving that this chromatic substance, not used for the chromo- 
somes, is chromatin, but for many forms it has been claimed that 
only a small part of the chromatin is used in the formation of the 
chromosomes and we are therefore perhaps justified in surmising 
that the chromatic substance in EnscJiistns (which often persists 
until after the chromosomes are formed) may be chromatin resi- 
due, which assumes various forms during its disintegration and 
disappearance. 
The early stages of the development of the germinal vesicles 
shown in Photos 1 1 to 1 7 we interpret as indicating that the chro- 
mosomes completely disintegrate --that any claim of their mor- 
phological persistence during these stages must be made as a 
pure assumption, as there is no morphological evidence whatever 
of their persistence. On the contrary there is every indication 
that they completely disintegrate. We believe we have no 
reason to assume that the reappearance of the chromosomes at 
certain stages differs essentially from the reappearance of the 
centrosome, aster, spindle, nucleolus and other cell structures. 
Pick ('07) in his critical review of the inconclusive evidence 
that has been offered as proof of the individuality and continuity 
of the chromosomes, and of their causal character, sounds a 
welcome note of protest against recent chromosome specula- 
tions. In addition to Pick such experienced cytologists as 
Hacker and Meves, have expressed their skepticism of the sex- 
228 KATHARINE FOOT AND E. C. STROBELL. 
determination hypothesis in no uncertain terms, and many of the 
investigators who have studied the accessory chromosomes in 
insects are equally incredulous, Montgomery, Schafer, Moore, 
Robinson, Arnold and others. 
We have reserved a description of the later stage of the ger- 
minal vesicle chromosomes of Euscliistus for a subsequent publi- 
cation, which will discuss only the chromosome groups in both 
male and female cells. In the male cells we have traced the 
idiochromosomes through both maturation divisions and are able 
to support, in the main, Wilson's results as to their division in 
these stages. He says, the idiochromosomes remain univalent 
in the first division, each dividing independently, and in the 
second division the two separate the small idiochromosome 
going to one pole and the large idiochromosome to the opposite 
pole. He gives no details however as to the method of the first 
division and as we find an interesting irregularity in this division, 
we believe it is worthy of consideration. As the idiochromo- 
somes separate in the second spindle, this is presumably the so- 
called reduction division for this unequal tetrad, and the first 
division should be therefore an equation division, and both idio- 
chromosomes should divide longitudinally, if the significance that 
has been attributed to the longitudinal and transverse divisions 
of the chromosomes holds true in this form. 
The value of the significance of a longitudinal or transverse 
division is based on the assumption that the so-called ids are 
arranged in a row, and a morphological demonstration of this 
assumption is claimed for those cases in which the chromosomes 
are rod- or thread-shaped these rods or threads often appear- 
ing as a single row of chromomeres. 
In our smear preparations of the testes of Euscliistus variolar- 
ius both the large and small idiochromosomes are distinctly rod- 
shaped, and at the late first prophase and metaphase it can be 
clearly demonstrated that the large rod divides longitudinally and 
the small rod transversely. There are rare (perhaps only appar- 
ent) exceptions to this, but as a rule one divides trans versely /?/-$/ 
as surely and as constantly as the other divides longitudinally, and 
this we have demonstrated in a number of photographs. 
The germ cells are being studied in order to obtain a morpho- 
NUCLEOLI IN EUSCHISTUS VARIOLARIUS. 22Q 
logical basis for conclusions or to put conclusions to a morpho- 
logical test, and the idiochromosomes seem to offer a fruitful test 
of the theoretical value of longitudinal and transverse divisions. 
In order to retain our faith in the special significance that has been 
attributed to the longitudinal and transverse divisions we are 
driven to the inconsistency of accepting the evidence in the case 
of oneidiochromosome, and disclaiming it for the other --though 
the evidence in both cases is equally strong. If, on the other 
hand, we accept the evidence and at the same time retain faith in 
the theory, we are forced to admit that the sister cells of the 
first division are also dimorphic and we have a so-called reduction 
division of one of the idiochromosomes of the first spindle as surely 
as we have a reduction division (the separation of the two idio- 
chromosomes) in the second spindle, we thus have a somewhat 
similar phenomenon occurring in both spindles. 
We have a number of photographs demonstrating both sper- 
matocyte divisions and also a large number showing the sperma- 
togonial, oogonial and embryonic chromosome groups several 
hundred in all, but the publication of these must be reserved for a 
separate paper, where we shall aim to make a very full compari- 
son of these chromosomes, giving especial attention to the groups 
that have not been presented in the papers which advocate the 
sex-determination hypothesis. We believe that these groups 
are of special value that we have no right to ignore exceptions 
because they do not fall in line with certain theories. The trite 
claim that exceptions prove the rule loses its force when enough 
exceptions are found to challenge the rule. A careful examin- 
tion of our preparations makes it possible to select chromosome 
groups which exactly fit a given theory, but many groups can 
also be found that are a serious menace to these theories, while 
on the other hand they present no difficulties to the conception 
of those who regard the number, size and form of the chromo- 
somes as inherited characters the expression of cell activities 
rather than the cause. 
January, 1909. 
230 KATHARINE FOOT AND E. C. STROBELL. 
APPENDIX. 
We have just received Wilson's latest " Study on Chromo- 
somes," No. V., 1 in which (in a footnote on page 159) he at- 
tempts to explain what he calls our "entirely mistaken conclu- 
sions " regarding the division of the accessory chromosome in 
Anasa tristis as follows : 
"The regrouping of the chromosomes in the second division, 
first described by Paulmier ('99) in Anasa tristis, is characteristic 
of the Coreidae generally, an eccentric position of the idiochromo- 
some being a nearly constant feature of the first division but not 
of the second. Failure to recognize this fact in the case of Anasa 
tristis seems to have been one of the main sources of error in the 
entirely mistaken conclusions of Foot and Strobell ('07^, '07$) 
regarding this species. (Cf. Lefevre and McGill, '08.) Demon- 
strative evidence on this point is given by polar views of rather 
late anaphases, in which every chromosome of each daughter 
plate may be seen, in the same section. Such views, of which I 
have studied many, both in Anasa and other genera, show that 
one of the chromosomes may indeed occupy an eccentric position, 
and may there divide ; but in such cases the odd chromosome is 
always found elsewhere in the group, lying either in or near one 
of the daughter groups and not in the other. When the odd 
chromosome is eccentric it is found in one of the daughter groups 
but not in the other." 
In order to speak with authority on the regrouping of the 
chromosomes, some accuracy at least is expected in the identifi- 
cation of the individual chromosomes. Paulmier's authority can 
be dismissed when we recall Wilson's interesting discovery that 
in Anasa tristis Paulmier erroneously identified the micro- 
chromosome as the accessory (undivided) chromosome of the 
second spindle. 
In Plate III. of our " Study of Chromosomes in the Spermato- 
genesis of Anasa tristis" ('07) we show in Photos 27, 28, 33, 35, 
36, 37, 38, 39, 42, 43, 44 and .45, late anaphases or early telo- 
phases of second spindles demonstrating the ring-like formation 
of the chromosomes, with one of the chromosomes eccentrically 
placed or lagging. In our smear preparations this grouping is 
Jour. Exp. Zool., Vol. VI., No. 2, 1909. 
NUCLEOLI IN EUSCHISTUS VARIOLARIUS. 23! 
characteristic of the second division as it is of the first, although, 
as we have already pointed out, there are exceptions to this rule. 
There can be no possible doubt of our identification of these 
groups as second spindles, because they are invariably found in 
regions of the testis surrounded by spermatids in different stages 
of development and, further, the form of the chromosomes is so 
beautifully demonstrated in our smear preparations, it is hardly 
possible to confound the first and second spindles, 1 a mistake 
easily made in sections. That the individual chromosomes may 
change their relative positions at this stage, while the form of 
the group remains unchanged, is a possibility that any observer 
would naturally consider. But as we find in our preparations, 
the ring-like grouping of the chromosomes of the first spindle 
so constantly duplicated in the second spindle, with the micro- 
chromosome (the one chromosome that can be identified beyond 
question) maintaining its characteristic position in the center of 
the ring, the identification of the eccentric chromosome of both 
spindles as probably the same chromosome, is certainly as legiti- 
mate a conclusion, as forcing an arbitrary contradiction of this 
assumption, in order to support a theory. 
In complete opposition to what Wilson calls his "demonstrative . 
evidence " on this point, we refer to our published photographs, 
Plate III. (in the paper quoted above), 29, 32, 33, 34, 35, 36, 
37, 40, 41, 43, showing polar views of late anaphases or telo- 
phases of second spindles, in which every chromosome of each 
daughter plate can be counted. In all these preparations there is 
an eccentric, lagging in most cases dyad - - chromosome, but as 
ten chromosomes can be clearly counted at each pole, we fail to 
see how the situation is improved, for Wilson to claim, that where 
the lagging chromosome divides in this spindle it is not to be 
identified with the " odd " chromosome which he says, " is, in 
such cases, akvays to be found elsewhere in the group'' We have 
given demonstrative evidence that this is not so in our preparations. 
The essential fact is not the relative position of the accessory 
chromosome in either the first or second spindles, but whether 
1 The ring-like grouping of the chromosomes is also characteristic of the first and 
second spindles of Euschistus variolarius, where the distribution of the unequal 
tetrad in the second spindle adds conclusive proof of the identity of the spindles. 
232 KATHARINE FOOT AND E. C. STROBELL. 
one of the chromosomes consistently fails to divide in the second 
spindle. It has been said of the accessory chromosome that in 
the spindle in which it is heterotropic, it tries to divide, but does 
not succeed. In the case of Anasa tristis, we have demonstrated 
that it is by no means always unsuccessful in these efforts. 
We regret that we cannot agree with Professor Wilson as to 
the value of the support he believes he has received from Lefevre 
and McGill's paper on Anasa tristis and Anax junius ('08). In 
this paper our results in Anasa are summarily disposed of in two 
pages of curt contradictions, supported by ten sketches made 
from sections. The futility of such methods as opposed to the 
very full demonstration we have given in 107 photographs to sup- 
port our conclusions in Anasa, must be apparent to any unpre- 
judiced observer. 
In this same paper, McGill corrects her earlier observations on 
Anax junius on three very important points : 
1. In her first work on this form ('04) (in which she ex- 
presses her indebtedness to Professor Lefevre) she found no 
evidence of a chromatin nucleolus in the resting spermatocyte. 
She and Lefevre now find, in t/ie same material, a chromatin 
nucleolus persisting through the resting spermatocyte. 
2. In her earlier work McGill identified the microchromo- 
some as the accessory chromosome. She and Lefevre now 
identify, in the same material, one of the larger spermatogonial 
chromosomes as the accessory chromosome. 
3. In her original count of the spermatogonial chromosomes, 
McGill found an even number, 28. She and Lefevre now find in 
the same material an odd number, 27 spermatogonial chromo- 
somes. 
In view of these contradictions we may justly hesitate to accept 
as definitive the recent conclusions reached by McGill and Le- 
fevre in Anasa tristis, believing a new point of view may give us 
still further variations in their very interesting observations. 
BIBLIOGRAPHY. 
Arnold, George. 
'08 The Nucleolus and Microchromosomes in the Spermatogenesis of Hydro- 
philus piceus (Linn.). Arch. f. Zellforschung, Bd. II., Hft. I. 
Deton, Willy. 
'08 L'etape Synaptique dans Povogenese du Thysanozoon Brocchii. La Cel- 
lule, t. XXV. 
NUCLEOLI IN EUSCHISTUS VARIOLAR1US. 233 
Fick, R. 
'07 Vererbungsfragen, Reduktions und Chromosomen-hypothesen. Ergebnisse 
der Anat. u. Entwick-geschichte, Bd. XVI., 1906. 
Foot and Strobell. 
'05 Prophases and Metaphase of the first Maturation Spindle of Allolobophora 
fcetida. Amer. Jour. Anat., Vol. IV., No. 2. 
'07 A Study of Chromosomes in the Spermatogenesis of Anasa tristis. Amer. 
Jour. Anat., Vol. VII., No. 2. 
Hacker, Valentin. 
'07 Die Chromosomen als angenommene Vererbungstrager. Ergebnisse u. Fort- 
schritte der Zoologie, Bd. I., Hft. I. 
McClung, C. E. 
'02 The Accessory Chromosome Sex Determinant? Biol. Bull., Vol. III., 
No. 2. 
Meves, Fr. 
'02 Struktur und Histogenese der Spermien. Ergeb. der Anat. u. Entw.-ge- 
schichte, Bd. XL, 1901. 
'08 Es gibt keine parallele Konjugation den Chromosomen. Arch. f. Zellf., Bd. 
I., Heft. 4. 
Montgomery, Th. H. 
'98 The Spermatogenesis in Pentatoma up to the Formation of the Spermatid. 
Zool. Jahrb., Bd. XII. 
'01 A Study of the Chromosomes of the Germ Cells of Metazoa. Trans. Amer. 
Philos. Soc., Vol. XX. 
'06 Chromosomes in the Spermatogenesis of the Hemiptera heteroptera. Trans. 
Amer. Philos. Soc., Vol. XXI. 
Moore, J. E. S., and Robinson, L. E. 
'05 On the Behavior of the Nucleolus in the Spermatogenesis of Periplanata 
Americana. Quart. Journ. Mic. Sci., Vol. XLVIII. 
Scbafer, Friedrich. 
'07 Spermatogenese von Dytiscus. Zool. Jahrb., Bd. XXIII., Hft. 4. 
Stevens, N. M. 
'os-'o6 Studies in Spermatogenesis. Publication No. 36, Parts I. and II., Car- 
negie Institution of Washington. 
v. Tellyesniczky, Koloman. 
'07 Die Entstehung der Chromosomen. Evolution oder Epigenese ? Urban & 
Schwarzenberg, Berlin . 
'05 Ruhekern und Mitose. Arch. f. mik. Anat., Bd. 66. 
Wassilieff, A. 
'07 Die Spermatogenese Von Blatta germanica. Arch. f. mik. Anat., Bd. 70. 
Wheeler, William Morton. 
'97 The Maturation, Fecundation and Early Cleavage of Mysostoma glabrum 
Leuckart. Archive de Biologic, Tome XV. 
Wilson, E. B. 
'05 Studies on Chromosomes, I. Journ. Exper. Zool., Vol. II., No. 2. 
'05 Studies on Chromosomes, II. Journ. Exper. Zool., Vol. II., No. 4. 
234 KATHARINE FOOT AND E. C. STROBELL. 
EXPLANATION OF PLATES. 
All the photographs were taken with a Zeiss Apo. 2 mm. immers. lens, 140 apr. 
and compensating ocular 4. Photos i to 31 are from smear preparations stained with 
a saturate solution of Bismarck brown. In all these preparations the whole nucleus 
is shown. 
Magnification of Photos i to 10 inclusive, 1,000 diameters. 
Magnification of Photos 1 1 to 32 inclusive, 600 diameters. 
As some of the germinal vesicles were too large to be included in the field at a 
magnification of 1,000 diameters, we used a magnification of 600 diameters for all 
the oocytes, in order to facilitate a comparison. 
The reproductions were made by the Holograph Company from our own negatives. 
PLATE I. 
Euschistus variolarius. 
PHOTO i . Two young resting spermatocytes, first order, one showing two nucleoli 
and the other one nucleolus. 
PHOTO 2. A little later stage than Photo i. Two nucleoli are shown, unequal in 
size, but this inequality is not so marked as in the two nucleoli of Photo 6. 
PHOTO 3. Resting first spermatocyte about the same stage of development as 
Photo 2. One nucleolus present. 
PHOTO 4. A resting first spermatocyte showing two nucleoli about equal in size. 
PHOTO 5. A resting first spermatocyte with two nucleoli, unequal in size, though 
the inequality here is not so conspicuous as in Photo 6. 
PHOTO 6. A later stage than Photo 5. Thechromatm shows the segregation 
which precedes the formation of the chromosomes. Two nucleoli are present, very 
unequal in size. 
PHOTO 7- First spermatocyte with the chromatin somewhat more closely segre- 
gated than in Photo 6. One nucleolus is present showing a typical vacuole. 
PHOTO 8. A larger first spermatocyte with one nucleolus, the disposition of the 
chromatin suggesting a network. 
PHOTO 9. A first spermatocyte showing an early stage in the formation of the 
chromosomes ; one nucleolus is present. 
PHOTO 10. A much later stage than Photo 9. The bivalent chromosomes are 
formed, but the number is not so clearly demonstrated as at later stages. (These later 
stages, as well as the transitional stages, between Photos 9 and IO, are reserved for 
a subsequent publication. ) The nucleolar-like thickening in the loop of the chromo- 
some on the lower periphery of the group is not a small nucleolus. In the prepara- 
tion it can be plainly recognized as a segregation of chromatin at the point where a 
loop of the chromosome brings two parts of the thread in contact. Similar, though 
less conspicuous thickenings are seen at other points where two threads cross. One 
clear nucleolus is present in this spermalocyte. 
PHOTO II. Nucleus of a young oocyte, with the chromatin granules quite evenly 
distributed throughout the nucleus. No nucleolus present. 
PHOTO 12. A slightly older germinal vesicle, with the chromatin segregated into 
clumps, too numerous to represent individuual chromosomes. 
Kuschistus variolarius. 
FOOT STROOELL PHOTOS. 
PLATE 1 
17 
BIOLOGICAL BULLETIN, VOL. XVI 
OCR text unavailable for this page.NUCLEOLI IN EUSCHISTUS VARIOLARIUS. 235 
PHOTO 13. A young germinal vesicle with thechromatin almost in a diffused con- 
dition. A small, clear, spherical space is present, which may represent the first ap- 
pearance of the achromatic nucleolus so conspicuous in the older germinal vesicles of 
Photos 1 8 to 30. 
PHOTO 14. A young germinal vesicle with the chromatin evenly distributed as in 
Photo II. 
PHOTO 15. An older germinal vesicle with the chromatin segregated into numerous 
clumps as in Photo 12. Many of the clumps in this germinal vesicle appear almost as 
dense as nucleolf, though in the preparation they are seen to be merely a close segre- 
gation of fine granules. 
PHOTO 1 6. An older germinal vesicle showing less pronounced segregations of 
chromatin, and a clear spherical space which can surely be interpreted as an early 
appearance of the achromatic nucleolus seen in the germinal vesicles of Photos 1 8 to 30. 
PHOTO 17. A much larger germinal vesicle with the chromatin quite evenly dis- 
tributed as granules. 
PHOTO 18. An older germinal vesicle with the first indication of the formation of 
delicate chromatic threads, which later develop into the chromosomes. At this stage 
the achromatic nucleolus is always present. The cracked spaces shown in this ger- 
minal vesicle and in Photos 21, 23, 24 and 29 are due to uneven shrinkage as the 
germinal vesicles rapidly dry on the slide. 
PHOTO 19. A much later stage than Photo 18 the chromatin threads being well 
formed. (The next stage to Photo 18 is shown in Photo 20, Plate II.) The ger- 
minal vesicle of this preparation is not well flattened and some of the chromatic 
threads are very much out of focus. 
236 KATHARINE FOOT AND E. C. STROBELL. 
PLATE II. 
Euschistus variolarius. 
PHOTO 20. A germinal vesicle a little further developed than the one shown in 
Photo 18. The delicate chromatic threads are a little more defined. The achromatic 
nucleolus is present near the periphery. 
PHOTO 21. A germinal vesicle with the delicate chromatic threads well formed. 
The achromatic nucleolus is seen on the periphery. 
PHOTO 22. A germinal vesicle with the chromosome threads further developed. 
The nucleolus near periphery. 
PHOTO 23. A germinal vesicle with long, thin chromosome threads. The nucleo- 
lus on periphery. 
PHOTO 24. A germinal vesicle at about the same stage of development as Photo 
23 The nucleolus on periphery. 
PHOTO 25. A germinal vesicle with the chromosome threads further developed. 
Kusrlnstus variohirius. 
FOOT STROOELL PHOTOS. 
PLATE II 
BIOLOGICAL BULLETIN, VOL xvi 
OCR text unavailable for this page.OCR text unavailable for this page.238 KATHARINE FOOT AND E. C. STROBELL. 
PLATE III. 
Euschistus vario/arius and Allolobophora fcetida, 
PHOTO 26. A germinal vesicle with chromosome threads well formed. Nucleolus 
on periphery. 
PHOTO 27. A germinal vesicle at about the same stage of development as Photo 
26. The nucleolus near the center. 
PHOTO 28. A germinal vesicle with long, thin, twisted chromosome threads. 
Nucleolus near the center. 
PHOTO 29. A germinal vesicle showing a later stage of development. The chromo- 
some threads have contracted into shorter, thicker pieces. Nucleolus still present. 
PHOTO 30. A germinal vesicle with the chromosomes almost at first prophase 
stage. Nucleolus still persisting. 
PHOTO 31. A germinal vesicle of Allolobophora fcetida. The same technique as 
that used for the Euschistus germinal vesicles (Photos 18 to 30). The chromosome 
threads are long, and thin, and longitudinally split. The large, less chromatic 
nucleolus is on the periphery, and the dense chromatin nucleolus is near the center. 
PHOTO 32. A section (2% /it thick) through the first maturation spindle of Allolo- 
bophora fci'tida. At the equator one chromosome is shown, the ten remaining chro- 
mosomes being clearly demonstrated in the adjoining sections. Near the upper pole 
we see the persisting chromatin nucleolus. A centrosome is shown at each pole 
Kiischistus variolarins 
FOOT Jv STHOI1ELL PHOTOS 
PLATE m 
.. 
: '. Yvv-\ 
V ;Vfc 
> z * ,' .vir<to < 
- , k ^.^ 
"^ -. *, 
3iOLOGICAL BULLETIN, VOL.XVI 
OCR text unavailable for this page.INHERITANCE IN THE "WALKING-STICK," 
APLOPUS MAYERI. 
CHARLES R. STOCKARD. 
The family Phasmatidae, as is well known, shows some of the 
most striking cases of " protective resemblance" found among 
the insects. Several of the genera are typically stick-like to a 
surprising degree while members of the genus Phyllium resemble 
in detail a leaf-like structure. These animals are no doubt pro- 
tected by their imitative forms provided they behave in a cer- 
tain manner. In fact the protection or concealment of such an 
animal depends as largely on its behavior as upon its resem- 
blance to surrounding objects. In order to ascertain whether 
these so-called protectively adapted insects really exhibited a 
"protective behavior" I 1 studied the habits of the "walking- 
stick," Aplopus mayeri, which is abundant on the Tortugas Islands, 
Florida. These large insects were found to behave in a manner 
almost ideal for their concealment among the twigs and stems 
of the plant on which they feed, Snriana inaritiina. 
My study was made during a season, June and July, when the 
enemies of Aplopus were extremely rare on these islands. In the 
spring and fall, however, the great numbers of migrating birds 
which stop here no doubt devour many of these large Orthoptera 
in spite of their almost perfect concealment. But for their pro- 
tective resemblance and habits birds might easily exterminate 
such slow-moving flightless insects within a few seasons, in fact 
the existence of creatures like Aplopi on these small islands is 
really dependent upon their ability to be passed unobserved by 
birds migrating between the eastern United States, West Indies 
and South America. 
The question arises whether the protective behavior in Aplopus 
is fully developed on hatching from the egg or whether it is at- 
tained with their large size and mature condition. In order to 
1 " Habits, Reactions and Mating Instincts of the ' Walking-Stick,' Aplopus may- 
eri," Science, N. S., XXVII., 1908 Publication No. 103, Carnegie Institution, 
Washington, 1908. 
239 
24O CHARLES R. STOCKARD. 
investigate this and the further question of first-leg form consid- 
ered below eggs were collected during the summer and brought 
to New York where the newly hatched individuals might be 
observed. 
These eggs were kept in small loosely stoppered bottles in the 
laboratory at ordinary room temperature. They began hatching 
about the first of December and during January a large number 
of the insects came out. 
BEHAVIOR OF NEWLY HATCHED APLOPI. 
The reactions of the small insect on the day it emerges from 
the egg are almost identical with those of the fully mature eight- 
inch females which I studied at the Tortugas. The young 
Aplopi are a light chocolate-brown in color with yellowish bands 
about the legs and the sexes are similar. The adult males, 
however, become greenish in color while the female retains her 
original brown. The adults also have rudimentary wings which 
are capable of being raised when the insect is excited, but the 
young are wingless. Their reactions will be referred to briefly 
at this time as they are given in some detail in my former paper. 
The insects when at rest among the twigs assume an attitude 
which in consequence of their stick-like shape makes them most 
difficult to detect. The first pair of legs are stretched directly 
forward enclosing the head between thin curved portions of the 
femora which fit perfectly against it. The antennae are brought 
B 
FIG. I. A. Lateral aspect of Aplopus, head and first two thoracic segments. The 
first pair of legs are stretched forward in the typical resting attitude. The femur fits 
perfectly about the ventro-lateral region of the head and leaves the eyes uncovered. 
B. Dorsal view of the same specimen showing the approximated antennae directed 
forward and enclosed between the first pair of legs. 
INHERITANCE IN THE "WALKING-STICK. 24! 
together between the first legs, Fig. I, B. (See also Fig. i, 
Plate I., and the photographs in my former paper.) The an- 
terior end of the insect thus closely resembles a more or less 
pointed stick. The newly hatched individual habitually assumes 
this position and the point of especial interest, which I shall re- 
turn to later, is that the thinned out curved part of the femora 
fit about the head as perfectly the first time the legs are stretched 
forward as they do in the adult. 
In walking from place to place the adult moves slowly and 
often exhibits a slow laterally swinging motion suggesting a twig 
swinging in a light breeze. The young also swings its body 
from side to side in a similar manner. When a number of young 
Aplopi are sitting motionless if the observer blows a current of air 
over them they all begin to swing very actively from side to side 
as if being swung by the breeze. This swinging motion no doubt 
serves to render them less conspicuous among the shrubs. 
The newly hatched individuals use the same methods to 
escape an enemy as those employed by the adult. When they 
are touched or pinched slightly they move away a short distance 
and immediately come to rest again, if the stimulus be repeated 
they begin to walk at a more rapid gait than before and move a 
greater distance away. If again touched they drop bodily to the 
floor and feign death just as the adult does. The death-feigning 
reaction is more readily induced in the young than in the adult, 
and no doubt serves to great advantage in enabling them to 
escape an enemy which fails to seize them securely in the first 
attempt. The chances of escape for this stick-like creature when 
it drops through the dense foliage and branches of the Suriana 
bush is most favorable. When in the death-feint the legs may 
be bent in any position and the body twisted without the least 
move on the part of the animal. They may actually be piled 
one on another and will remain as motionless as dead insects. 
The young walking-stick crawls upwards on any object that it 
may reach after emerging from the egg. As I previously re- 
corded the female Aplopus sits in the Suriana bushes and lays its 
eggs which fall to the ground where they later hatch. Thus the 
tendency of the young to crawl upwards on the first object with 
which it comes in contact serves to bring it up the Snriana bush 
242 CHARLES R. STOCKARD. 
to its leafy food. In crawling up the young insect waves its 
antennae to feel the way just as does the adult and reaches out 
with the first legs to grasp the object located by the antennae. 
Finally the young like the adult is more or less nocturnal in 
its movements. During the day they sit motionless with the 
first legs extended forward but at night they become active and 
move about to feed. The food of the adult is limited to the 
leaves of Suriana. I have made no attempt to feed the young 
since they may be kept alive for about one week after hatching 
without taking food. 
THE THIN CURVE OF THE FEMORA WHICH FITS AGAINST 
THE VENTRO-LATERAL SURFACES OF THE HEAD. 
When the first pair of legs are extended forward the femur of 
each is so curved near its proximal joint as to fit perfectly against 
the ventro-lateral parts of the head and at the same time leaves 
the eyes uncovered, Fig. I, A and B. The curved portion of 
the femur is also very thin in a lateral direction and thus when 
pressed closely to the head the legs go out as almost straight 
lines instead of bulging around the head to any great extent. It 
seems difficult to believe that the first pair of legs could through 
chance variations or mutations have come to fit so perfectly 
around the sides of the head and at the same time to have their 
dorsal line so curved as to leave the eyes uncovered. It must 
be remembered that when the first legs are in the extended posi- 
tion the head presses against the dorso-lateral surfaces of the 
femur and not straight against the inner lateral surface only. 
This arrangement may be better understood by a close examina- 
tion of the dorsal and lateral views given in Fig. I , A and B. 
The possibility suggests itself that the perfection of the fit is 
attained during the life of the individual since the legs are so 
habitually pressed against the head for about twelve hours daily. 
To test this it became desirable to study newly hatched indi- 
viduals in order to find whether the femur curve was as perfectly 
adjusted in them as in the adult. A careful examination of about 
one hundred Aplopi shortly after emerging from the egg has con- 
vinced me that the curve of the femur is as true to the head pat- 
tern in the newly hatched young as in the mature insect when 
several months old. 
INHERITANCE IN THE "WALKING-STICK. 
243 
Finally, is this adjustment between the structure of the femora 
and head due to the position of the insect when enclosed within 
the inelastic egg shell ? If the first legs were folded forward against 
the head the pressure during embryonic life might easily be suf- 
ficient to mold the femur curve into pattern for the head. Twenty 
eggs containing embryos at various stages shortly preceding 
hatching were dissected with this question in view. The ellipti- 
cal egg shell is of a rigid chitinous material with a circular oper- 
culum at one end and a hilum-like scar on one side to which the 
FIG. 2. An unhatched Aplopus with its egg membranes dissected away showing 
its folded position. The femora of the three right legs are marked with roman 
numerals and the tarsal ends of the same legs are indicated by figures. The part of 
the first femur which in later life fits against the head is shown between the points 
x-x, it is not molded against the head in the egg. A, antennae ; , compound eye ; 
M, mandible. 
inner egg membrane is attached. The size of the egg ranges 
from 2 mm. in narrow diameter by 4 mm. in long diameter to 3 
mm. by 4.5 mm. The length of the walking-stick on hatching 
is from 17 mm. to 23 mm. measured from the tip of abdomen to 
tip of the first pair of legs when extended forward, or from tip 
of head to tip of tail 9.5 mm. to 13 mm. The embryo within 
the egg is, therefore, necessarily much folded and bent. 
244 CHARLES R. STOCKARD. 
On dissecting the egg the embryo is found to be curled around 
in a rather constant manner, the head being usually, though not 
always, near the opercular end. The long legs are folded back 
and forth upon themselves in a very definite fashion as shown by 
the camera drawing, Fig. 2. The antennae (A) pass down the 
front of the head and then back along the ventro-lateral surface 
of the abdomen being sometimes bent around the first pair of 
legs. The point of most importance is that the femoral segments 
of the legs are all directed obliquely away from the head. The 
first pair of legs each of which is folded on itself four times does 
not touch the sides of the head at all. The head and large eyes 
are entirely uncovered and exposed. The femora of the first 
pair of legs not only fail to mold their curves against the head 
but the femora are so pressed against the thorax that the surfaces 
which will subsequently be concave (in Fig. 2 between x and x] 
are actually arched convexly. Thus it is seen that the mechanical 
arrangement of the embryo's parts within the egg is not respon- 
sible for the fit of the femur curve against the head. On the 
contrary the curve seems to develop in spite of these arrangements. 
When hatching the embryo's head and body come forth from 
the egg first, the antennas are then pulled out, the legs being the 
last parts liberated from the shell, Fig. 3. It often happens that 
the shell is carried around for some time dangling to the third 
pair of legs. In Fig. 3 the well developed curves of the femora 
are distinctly shown, x to x, and are being pulled in a direction 
away from the head, yet as soon as the legs are free from the 
shell the first pair may be straightened forward and their curved 
femora fit neatly against the sides of the head. We see, there- 
fore, that the curve of the femora to fit the sides of the head is a 
character transmitted to all of the young and perfectly formed at 
the time of hatching. It might seem that the origin of this 
character was most probably due to the habit of the insects to 
press the first pair of legs against the head. Gradually this 
pressure developed a thin concave region of the femur of the first 
leg which molded itself more and more perfectly to the contour 
of the head. If this curve arose in any other way the second and 
third pairs of legs might have developed at least a trace of such 
a character though this is not absolutely necessary. It must be 
INHERITANCE IN THE "WALKING-STICK.' 
245 
remembered that the curves fit the ventro-lateral contour of the 
head to a remarkable degree. 
When the first pair of legs are so stretched forward the insect's 
antennae are brought together. The legs have an irregular 
groove extending along the approximated surfaces and when 
complete approximation takes place a rather imperfect tube is 
formed enclosing the antennas. This is a case analogous to the 
above and it is difficult to imagine how chance variations could 
FIG. 3. Aplopus in the act of hatching from the egg. The body and head come 
out first, then the antennas and finally the legs free themselves from the shell. The 
parts of the first legs between xx are curved to fit the head when they are straight- 
ened forward although they have never touched the head up to this time. 
bring about such mechanical harmonies between organs only 
associated through an habitual attitude assumed by the animal 
when at rest. Yet it must not be forgotten that many other 
equally as nice morphological arrangements exist which have no 
habit or action connected with them. Indeed a crucial case of 
use inheritance is almost impossible to imagine from purely 
descriptive work. I would not be understood as advocating any 
principle of inheritance but merely bring forward the present case 
as being of interest in itself. 
CORNELL MEDICAL SCHOOL, NEW YORK CITY, 
February 6, 1909. 
THE CONNECTIONS OF THE GONADIAL BLOOD VES- 
SELS AND THE FORM OF THE NEPHRIDIA 
IN THE ARENICOLIDyE. 
ELLIOT ROWLAND DOWNING, PH.D., 
NORTHERN STATE NORMAL SCHOOL, MARQUETTE, MICHIGAN. 
The gonads of Arenicola occur on certain blood vessels that lie 
diagonally on the exterior of the glandular portion of the nephridia 
and that are consequently designated the gonadial vessels. In a 
study of the gonads, the results of which will be published shortly, 
it was found that the literature of the subject contained conflict- 
ing and inaccurate statements regarding the relations of these 
gonadial vessels to their connecting vessels. This paper is the 
result of an attempt to determine these relations. 
Gamble and Ashworth, in 1900, published an extended study 
of the anatomy of the several species which also reviewed the 
important literature on the family. They had previously pub- 
lished, 1898, a paper on the anatomy of Arenicola marina. In 
an article published in 1899, V. Willem disagreed with some of 
their statements as did R. Lillie in a more recent publication, 
1906. These contradictory statements I shall try to adjust : I 
must also take exception to some other statements of each. 
The general relations of the blood vessels may be easily un- 
derstood from the accompanying partly diagrammatic figure of a 
cross-section at about the level of the first nephridium of A. cris- 
tata (Fig. i). It will be noted that there are two main blood 
vessels, a dorsal and a ventral ; these run the entire length of the 
animal. There is a pair of neurals of much smaller caliber, also 
extending the full length of the body. The paired gastric lat- 
erals and subintestinals, as also the paired integumentary vessels 
known as the parietal (or dorsal longitudinal) and the longitudinal 
nephridial, are limited in their extent and vary in length and 
prominence in the several species. 
All previous authors agree in the arrangement of the nephridial 
vessels as follows : The afferent vessel, on approaching the 
246 
FORM OF NEPHKIDIA IN AREN1COLID/E. 
247 
nephridium, divides, one branch going to the setal sac and gill, 
if present, one to the integumentary vessels and one to the 
nephridium. The latter enters the nephrostome and forks, one 
branch traversing each lip. After reuniting, the vessel, reformed, 
runs out onto the wall of the nephridium as the gonadial vessel. 
Each of the three main branches of the afferent vessel gives off 
smaller vessels which, with more or less anastomosing with ad- 
jacent vessels, form capillary net-works in the organs supplied. 
FlG. I. Partly diagrammatic cross-section of Arenicola cristata at about the level 
of the first nephridium. D.V., dorsal blood vessel. H., heart. L.G.T., lateral 
gastric blood vessel. N., neural blood vessel. N.L. V., nephridial longitudinal 
blood vessel. P., parietal blood vessel. S. V., sub-intestinal blood vessel. V. t 
ventral blood vessel. 
The efferent vessels are formed by the union of capillaries in these 
same regions. 
The afferent vessel, as a rule, comes from the ventral vessel. 
This is true for all except the first two nephridia of A. Grnbii and 
of A. ecaudata, which nephridia are supplied, the first by a branch 
of the dorsal, the second by a branch of the parietal vessel (Fig. 
2, c/). In addition to the afferent vessel from the ventral vessel, 
the following nephridia also receive branches from the dorsal 
vessel : the first of A. cristata, the first two of A. Claparedii and 
the first three of A. marina (Fig. 2). 
The following nephridia return blood directly to the subintes- 
tinal vessels through efferent vessels whose numerous branches are 
adjacent to theirfunnels : the fourth, fifth and sixth of A. cristata, 
the fourth and fifth of A. Claparedii and the fifth and sixth of A. 
marina. All others must pour the blood into the parietal and 
nephridial longitudinal vessels which, in turn, pass it to some of 
the more posterior efferent vessels (Fig. 2). 
248 
ELLIOT ROWLAND DOWNING. 
These observations are directly antagonistic to those of other 
investigators on several points .of anatomical detail and of func- 
tion of some of the vessels. I have given above a statement of 
the general arrangement of the branching afferent vessels with 
reference to the nephridia ; it will be evident from text-figure 2 
that my results only agree in a general way with the statements 
FIG. 2. Diagrams of the blood vessels supplying the nephridia in the several 
species of Arenicola, a, A. Clafaredii, b, A. cristtita, c, A. marina, ti, A. Grubii ; 
d will also serve as a diagram of the blood vessels of A. ecaudata. The form of the 
nephridia and of the digestive organs in the latter species, would be somewhat differ- 
ent but the arrangement of the blood vessels is the same as in A. Grubii. The dorsal 
vessel is shown by a solid line, thus . Afferent vessels arising from the ventral 
vessel are shown in broken lines, . . . Efferent vessels connecting with the 
sub-intestinal blood vessel are shown in dotted lines, D. J-'., dorsal vessel. 
j\ r .L. V. t nephridial longitudinal vessel. P. V., parietal blood vessel. 
of previous authors. Thus in all nephridia of A. Claparedii, in 
some of A. marina and occasionally in A. Grubii, the afferent 
branch of the ventral vessel enters the apex of the funnel before it 
branches. In A. cristata one finds frequently, particularly in the 
somites containing the second and third nephridia. that the affer- 
FORM OF NEPHRIDIA IN ARENICOLID^. 249 
ent vessel gives off no branches to the nephridia, but the latter 
are supplied by a branch from the parietal. 
The branch of the dorsal vessel to the second nephridium is 
small in A. Claparedii, but functional still. Gamble and Ash- 
worth show and describe an afferent vessel to the first nephridium 
of A. marina ; with this V. Willem disagrees, showing for this 
nephridium only a branch from the dorsal. In the majority of 
cases I find a branch from the ventral vessel .also, thus agreeing 
with Gamble and Ash worth ; in a few animals it is apparently 
wanting. I do not find, however, the efferent vessel from the 
fourth nephridium, in A. marina, to the subintestinal, which 
Gamble and Ashworth show but which V. Willem claimed was not 
present. 
It is not true that the three main branches of the afferent vessel 
break up into capillaries ; the one to the setal sac and gill does. The 
one to the integumentary vessels unites with them. The branch 
which enters the funnel continues as the gonadial vessel, runs 
peripherad to the nephridium and connects with the nephridial 
longitudinal, except in A. Claparedii. 
The parietal and nephridial longitudinal vessels are distinct 
throughout the nephridial region in A. GmbH, A. ecaudata and A. 
marina. In A. cristata, the parietal is distinct but the nephrid- 
ial longitudinal, while large at the level of the first nephridium, 
tapers posteriorly, becoming obscure back of the third nephridium. 
The parietal vessel is distinct the entire length of the nephridial 
region in A. Claparedii. The statement of Gamble and Ashworth 
that it is absent in this form is only explicable because they 
worked on preserved material ; it is plainly evident in the living 
worm. The nephridial longitudinal, as a distinct vessel, is ab- 
sent in this form except in the region of the first nephridium: its 
place is taken by a series of small connecting vessels, as if the 
gonadial vessel branched and some of its branches ran back to 
connect with those of the next posterior gonadial vessel. 
In all cases the branches of the dorsal vessel running to the 
nephridia are afferent vessels. Gamble and Ashworth state that 
" The first nephridia of A. Grubii and A. ecaudata are supplied 
by a branch from the dorsal vessel " (the italics are mine), yet 
" The first three nephridia of A. marina, the first two of A. Clapa- 
25O ELLIOT ROWLAND DOWNING. 
redii and the first of A. cristata return blood to the dorsal 
vessel." On a priori grounds, it would be strange to find the 
homologous blood vessels in closely related species carrying blood 
in opposite directions. I am certain that in A. cristata the blood 
flows from the dorsal vessel out to the parietal, the nephridial 
longitudinal vessel and to the nephridium, in this branch of the 
dorsal vessel. In this species the individuals are so large that in 
chloretonized specimens the direction of the blood movement is 
seen with comparative ease. S. E. Keith who has worked care- 
fully for some time on the circulation of A. cristata gives me per- 
mission to quote her on this point as follows : " I have referred 
to these blood vessels in my own notes as the third pair of 
branches of the dorsal vessel. Gamble and Ashworth speak of 
the dorsal as receiving these branches, but the blood flows out- 
ward in them I am sure." 
I have watched the blood-flow in A. Claparcdii ; the skin is 
frequently so transparent it may be seen without dissection. The 
flow is certainly away from the dorsal vessel in this species also. 
I am reasonably certain that such is the case in A. marina, in 
chloretonized specimens of which I have watched the blood 
movement. In this opinion I am supported by Willem who says 
of marina, " II faut remarquer de plus, au point fonctionnel, que 
le sang circulant dans le tronc dorsal d'arriere en avant, le contenu 
des trois branches qui en emanent progresse dans une direction 
centrifuge." 
Lillie makes the statement that " The vessel (segmental) begins 
its formation at the junction with the subintestinal blood vessel. 
Near its junction with the body wall the main vessel (segmental) 
gives off a branch (the nephrostomial vessel) which curves back, 
passes inward and barkward along the dorsal lip of the nephros- 
tome." From this quotation it is evident that Lillie traces the 
afferent nephridial vessel to the subintestinal blood vessel, while, 
as stated above, I trace it, in agreement with Gamble and Ash- 
worth and other observers, to the ventral vessel. The contra- 
diction is only verbal, not real, for Lillie, in the explanation of 
his figures, labels this subintestinal vessel " the subintestinal or 
ventral vessel," a usage which he gets from oligochaet anatomy 
and which is incorrect here. Lillie finds that in Arenicola as in 
FORM OF NEPHRIDIA IN ARENICOL1D.E. 2$ I 
other Poiychata, (see Fraipont on Poygordius, for instance), as 
well as in the OligocJiceta (Wilson on Lumbricus), the first vessel 
to arise as a differentiation of the mesoderm, is the ventral vessel. 
Wilson says " The first vessel to appear (in Lumbricus] is the 
ventral or sub-intestinal." What the relation of this first vessel 
genenally is to the development of the circum-intestinal network 
in the polychaets is an unsettled point. (See Edward Myer's 
" Studien iiber den Korperbau der Anneliden," III., p. 464.) 
Recently Schiller writes as follows : " Nur bezuglich Blutsinus 
und Darmgefasznetz ist mann bei Arenicola GmbH noch nicht ins 
klare gekommen, welches von den beider das primare sei, da 
verscheidene Autoren ganz verscheidener Ansicht daruber sind. 
Die einer behaupten dasz zuerst der Sinus auftrete und sekundar 
sich in ein Netz auflose : die anderen bezeichnen den Blutsinus 
als ein Produkt des zusammengeschmolzen Darmgefasznetzes." 
He does not care to express an opinion on the subject although 
of the growing posterior region of A. GmbH he says : " Im 
Darmepithel keine besonderen Zellen vorkommen, die Antrit an 
der Bildung eines Sinus nehmen konnten." In considering der 
Blutsinus, Darmgefasznetz und Subintestinalgefasz he says : " Diese 
drei Gebilde werden hier zusammen als anatomisch und entwicke- 
lungsgeschichtlich zusammengehorende Elemente." 
I have taken pains to quote in order to show that in spite of 
the difference of opinion regarding the origin of the network of 
intestinal vessels there is no tendency to include the ventral vessel 
in Arenicola or other polychaets as a part of this network which 
does, however, include the subintestinals. In A. marina we 
know (Benham, 1893) that the subintestinal vessels and the 
gastric longitudinals develop as the result of a fusion of the walls 
of the circular vessels of the alimentary canal. While in the 
oligochaets the subintestinal vessels split off from the ventral 
vessel. (Beddard, F. E., "Monograph of the Oligochaeta," p. 
70.) The term subintestinal vessel is an incorrect term for the 
vessels so named in Arenicola. Lillie should not have used the 
term subintestinal when he meant ventral, as these two terms are 
not interchangeable in Arenicola. The ventral vessel or subin- 
testinal vessel of Lumbricus they remain united in this form - 
is not homologous to the ventral plus the subintestinal vessels of 
Arenicola. 
252 ELLIOT ROWLAND DOWNING. 
There is need of amendment in the description and figures of 
the form of the nephridia of the genus. The funnels, in par- 
ticular, are much more symmetrical, in some species, than they 
have heretofore been shown. The usual technique is at fault in 
distorting them. The funnels are attached to the oblique muscles 
in A. Claparedii, A. cristata and A. marina. When the worm is 
opened and pinned out for dissection, the strain on these oblique 
muscles or on the connecting blood vessels of the other species, 
pulls the funnels throughly out of shape. It is, therefore, well 
to stupefy the worm by adding, drop by drop, seventy per cent, 
alcohol to the smallest quantity of sea water that will cover the 
worm in a long narrow dish. Then when the muscular walls are 
relaxed, the body cavity is injected by means of a hypodermic 
syringe or a fine pipette, until the walls are well distended with 
C.P 
FIG. 3. Dorsal view of the second left nephridium of Arenicola cristata, X 3- 
C.P., ciliated plates. D.L., dorsal lip. G.V., gonadial blood vessel. O.M., 
oblique muscle. 
FIG. 4. Ventral view of the anterior portion of second left nephridium of Areni- 
cola cristata, ><3- D.L., dorsal lip. V.L., ventral lip. 
the fixing agent. (Kopsch fluid answers well.) The whole worm 
is then placed in the fixing fluid. The nephridia are thus har- 
dened without distortion and when dissected out later show their 
proper form. 
The figures of the funnels or entire nephridia of each of the 
species are shown in the accompanying figures. The funnel 
of A. cristata (Figs. 3 and 4) is the most complicated and at 
the same time the most symmetrical in the genus. It is flat- 
tened, with a broadly sagittate or hastate form. Its dorsal lip, 
FORM OF NEPHRIDIA IN ARENICOLID.E. 
253 
which is attached by its outer surface to the oblique muscle, 
forms the point of the arrow. The ventral edge of this lip is 
set with from thirty to sixty ciliated plates which stand nearly 
at right angles both to the plane of the lip and to its edge. A 
loop of the blood vessel which traverses the dorsal lip runs up 
into each plate. The ventral lip is bow-shaped and divided into 
three segments like the handle and ends of a bow ; the convex- 
ity of the bow turns toward the apex of the dorsal lip. The 
opening of the nephrostome is a long narrow slit between the 
base of the triangular dorsal lip and the bow-shaped ventral lip ; 
the opening leads into the rapidly narrowing throat. The has- 
tate funnel is held to the body by a short slender haft ; the body 
of the nephridium, the glandular portion, is club-shaped ; the 
funnel attaches to its side near the larger end, the axis of the 
FIG. 5. Dorsal view of a nephridium of A. marina, X *5- 
FIG. 6. Ventral view of the anterior end of the nephridium of Arenicola marina, 
< 15. N. L. V., nephridial longitudinal blood vessel. O.M., oblique muscle. 
funnel being nearly at right angles to the longitudinal axis of the 
body. The body of the nephridium joins the roughly spherical 
bladder at the smaller end of the body, the handle of the club. 
The bladder opens to the exterior on its peripheral side by a very 
short duct leading to the nephridiopore. 
The nephridium of A. marina approaches, at times, quite nea 
254 
ELLIOT ROWLAND DOWNING. 
to the form of that of A. cristata ; but usually it is quite different 
from it (Figs. 5 and 6). The general shape of the body and of 
the bladder is nearly the same as in the nephridium of A. cristata, 
although in A. marina the body is more frequently curved while 
that of A. cristata is straight. The funnel shows great variation 
and often departs widely from the type of A. cristata. It is a 
flattened flap having roughly the shape of an equilateral triangle. 
At one angle the funnel opens into the anterior end of the body 
of the nephridium ; along the opposite side lies the nephrostome. 
One of the sides adjacent to the neck is attached to the oblique 
muscle ; the other adheres to the margin of the body. The axis 
of the funnel forms, therefore, an angle of only thirty or so de- 
grees with the axis of the body. The dorsal lip is straight or 
slightly concave ; it is set with twenty five or thirty somewhat 
lanceolate ciliated plates, each of which is supplied with a loop of 
the blood vessel. The ventral lip, Fig. 6, is regularly concave. 
FIG. 7. Ventral view of the anterior portion of the third left nephridium of Areni- 
cola Claparedii, X 3- 
FIG. 8. Dorsal view of the nephridium of Arenicola Claparedii, X I S- 
The ciliated plates which run along the edge of the dorsal lip, 
tend to continue along the blood vessel past the angle which the 
dorsal lip makes with the side of the funnel that attaches to the 
oblique muscle. This blood vessel runs along the muscle nearly 
parallel with the edge of the funnel. There are from ten to twenty 
of these plates ; those along the blood vessel on the muscle 
FORM OF NEPHRIDIA IN ARENICOLID;E. 
255 
diminish in size. When occasionally there are an unusally large 
number of them along the blood vessel beyond the angle, the 
angle of the funnel adjacent to the muscle tends to become the 
apex of the funnel as in A. cristata, and the mouth shifts so as to 
more nearly face this angle instead of a side. In such cases the 
throat of the funnel opens into the body some distance back of 
the anterior end of the latter. 
The funnel of the nephridium of A. Claparedii (Figs. 7 and 8) 
is least complicated. If we imagine a simple funnel form with 
short stem to be flattened and to have one lip pulled out into a 
triangular protrusion, we may gain a clear idea of the funnel of 
this species. The apex of the dorsal lip is broadly obtuse ; the 
ventral lip is straight. The dorsal lip is set with the ciliated plates 
characteristic of the marina section of the genus which, as Gamble 
and Ashworth point out, includes the species Claparedii, cristata 
CT? 
c 
FIG. 9. Dorsal view of the first left nephridium of Arenicola ecandata, X 1 S 
C.P., ciliated processes. D.L., dorsal lip. G. F., gonadial blood vessel. 
FIG. 10. Ventral view of the anterior portion of nephridium (first left) of A. 
ecaudata, X J 5- V.L., ventral lip. 
and marina. There are ten or twelve of these plates in A. Clap- 
aredii. The funnel is at the anterior end of the body and its axis 
is also at right angles to the axis of the body of the nephridium. 
The funnel of the nephridium of A. ecaudata (Figs. 9 and 10) 
is reniform in outline with re volute ends as seen from the ventral 
face. It also is flattened ; the dorsal lip is slightly concave, the 
ventral lip is deeply indented. The dorsal lip is provided with 
ten to thirty blunt, at times much branched, finger-like processes ; 
256 
ELLIOT ROWLAND DOWNING. 
these are covered with cilia, and within each is a blood sinus in- 
stead of a loop of the blood vessel. The throat of the funnel 
is relatively wide ; the funnel attaches at the end of the body and 
usually has the customary position with its axis at right angles 
to the long axis of the nephridial body. Not infrequently how- 
ever, the axis of the funnel is continuous 
with the axis of the body and we have a 
simple, unbent, tubular nephridium (Fig. 
1 1). It is interesting to find such variation 
in this species for it makes evident how the 
more aberrant forms of the nephridium, such 
as occur in A. manna, are derived from a 
simple tubular type ; just as a paper tube 
may be bent with its end at right angles to 
its major axis. 
A. Grubii (Figs. 12 and 13) possesses the 
same type of funnel as A. ecandata. It is 
flattened ; the dorsal lip is semicircular, the 
ventral lip trilobate, with a small median 
lobe and large lateral ones, thus making 
this lip deeply notched. Ten or twelve 
blunt, digitate, ciliated processes attach to 
the dorsal lip ; these are often branched and 
are provided with the blood sinus. The 
funnel attaches to the body at some dis- 
tance from the anterior end and its axis is 
at right angles to the axis of the body. 
The bladder of the nephridium of this spe- 
cies is usually expanded : it is capable of equally wide expansion 
in A. ecaudata but is more often found contracted. In the other 
species, the bladder is not so distensible although it is relatively 
large at times, especially when filled with eggs or sperm about 
to be discharged through the nephridiopore. 
I have collected A. cristata and A. marina near Woods Hole, 
Mass., A. marina, A. Grubii and A. ecaudata in the bay at Ply- 
mouth, England, and have studied fresh A. Claparedii and A. 
Grubii at Naples. I wish to express my sincere thanks to the 
directors of the biological stations at these several places, who 
FIG. ii. An occa- 
sional form of the nephri- 
dium of Arenicola ecau- 
data, ;< 20. 
FORM OF NEPHRIDIA IN ARENICOLID.E. 
257 
have kindly placed at my disposal, material and facilities for the 
work, and to the Carnegie Institution whose table at Naples it 
was my pleasure to occupy. 
FIG. 12. The nephridium of Arenicola Gntbii, X J 5 dorsal view. The outline 
of the ventral lip of the funnel is shown in dotted line. 
FlG. 13. Ventral view of anterior portion of the third left nephridium of A. 
Grubii, X 2 - 
258 ELLIOT ROWLAND DOWNING. 
BIBLIOGRAPHY. 
Benham, W. B. 
'93 Postlarval Stages of Arenicola marina. Jour. Mar. Biol. Ass. Plymouth, 
N. S., Vol. III., pp. 48-52. 
Eisig, Hugo. 
'98 Zur Entwickelungsgeschichte der Capitelliden. Mitth. Z. Stat. Neapel, Vol. 
XIII., pp. 1-293. 
Fraipont, J. 
'87 Le Genre Polygordius. Fauna u. Flora d. Golfes von Neapel, Vol. XIV., 
Monographic. 
Freudweiler, Hedwig. 
'05 Studien iiber das Gefaszsystem niederer Oligochaeten. Jen. Zeitschr. f. 
Naturwiss., Vol. XL., pp. 1-28. 
Gamble, F. W., and Ashworth, J. H. 
'98 Habits and Structure of Arenicola marina. Quar. Jour. Micr. Sci., Vol. 
XLI., pp. 1-42. 
'oo The Anatomy and Classification of the Arenicolidse. Quar. Jour. Micr. 
Sci., Vol. XLIII., pp. 419-569. 
Lillie, Ralph. 
'06 Structure and Development of the Nephridia of Arenicola cristata, Stimpson. 
Mitth. Z. Stat. Neapel., Vol. XVII., pp. 348-405. 
Meyer, Ed. 
'01 Studien iiber den KSrperbau der Anneliden, III. Mitth. Z. Stat. Neapel., 
Vol. XIV., pp. 247-585. 
Schiller, Ignaz. 
'07 Ueber den feineren Bau der Blutgefasse bei den Arenicoliden. Doctor's 
Thesis, Univ. Zurich, Jena. 
Willem, Victor. 
'99 L' Excretion chez 1' Arenicola. Travaux de la Station Zoologique de Min- 
ereux, Vol. VII. 
Wilson, E. B. 
'89 The Embryology of the Earthworm. Jour, of Morph., Vol. III., pp. 388-462. 
Vejdovsky, Franz. 
'05 Zur HamocSltheorie. Zeit. f. wiss. Zool., Vol. LXXXII., pp. 80-170. 
SOME OBSERVATIONS ON THE HABITS OF 
PECTEN DISLOCATUS. 
B. H. GRAVE. 
With the purpose of studying the habits of the scallop, Pcctcn 
dislocatus, I collected many young specimens ranging from two 
to ten millimeters in length and placed them in small glass aqua- 
ria in the laboratory. 1 They were found in the harbor at Beau- 
fort, N. C., well above the muddy bottom, clinging to eel grass. 
They were usually attached by several strands of hyaline byssal 
threads, which were exceedingly strong and elastic. 
Although the Pcctcn is so generally known as to make a 
detailed description of its anatomy superfluous, yet a brief descrip- 
tion of certain parts is deemed necessary. For a more detailed 
study of the anatomy, reference can be made to a paper, by G. A. 
Drew, on " The Habits and Anatomy of the Giant Scallop." 
By reference to Figs. I and 2, it may be seen that the shell is 
rounded and eared. The ears make possible the long, straight 
hinge line; which extends along their upper borders to their ex- 
tremities. The right valve is slightly more convex than the left, 
and near the anterior 2 ear, it has a deep notch. This one feature 
mars the symmetry of the valves. Between the valves and just 
beneath the hinge ligament, there is a pad of cartilage-like sub- 
stance, which is compressed when the valves are closed. It 
serves to open them quickly when the adductor relaxes (Fig. 3). 
The form of the shell and the structure and arrangement of the 
soft parts within, adapt Pecten to the swimming habit. It swims 
by opening and closing the valves in rapid succession. By vary- 
ing the position of the mantle so as to control the direction of 
1 Through the courtesy of Hon. Geo. M. Bowers, U. S. Commissioner of Fish and 
Fisheries, I had the privilege of occupying a table in the Fisheries Laboratory, at 
Beaufort, N. C., for two months during the summer of 1908. For this privilege and 
for many kindnesses shown me by the Director, Henry D. Aller, I am glad to express 
appreciation. 
2 The hinge line is here considered dorsal for convenience in description, although 
it does not represent the true dorsal of the animal. 
259 
260 B. H. GRAVE. 
the currents of water expelled from the mantle chamber, it is 
enabled to swim either forwards or backwards, although it usually 
swims with the opening of the valves directed forwards. 
A perfectly symmetrical shell is the form best adapted to swim- 
ming, and the presence of any irregularity in it, such as that just 
mentioned in the pecten, is to be explained either as an adapta- 
tion to habits other than swimming, or as a structure, inherited 
from an ancestral form, and not as yet obliterated through adap- 
tation to the swimming habit. 
Although adult specimens were kept in aquaria all summer, no 
method of locomotion other than smimming was noted, and no 
clue was gained as to habits which would in any way explain the 
function of the notch in the right valve. They neither attached 
themselves by a byssus, nor used the foot for locomotion. Young 
specimens, however, showed much more activity than the adults, 
and some observations on their habits are recorded in the follow- 
ing pages. 
Concerning the function of the notch in a related species, Pecten 
tenuicostatns, Dr. Drew writes as follows : " I have been unable 
to satisfy myself as to the function performed by this notch. The 
sense tentacles on the mantle margin, opposite the notch, are 
somewhat longer than those adjacent, but I have been unable to 
determine that they have a special function or that they are espe- 
cially advantageously placed." 
THE SE^SE OF POSITION. 
The Pecten lies habitually upon the right valve and if placed 
upon the left, immediately turns over. When lying upon the 
left valve, it seems to feel the same sort of discomfort which a 
frog, or other animal with well developed balancing organs, feels 
when placed upon its back. However, after turning them over 
repeatedly, they sometimes remained resting on the left valve for 
several minutes. 
THE ASYMMETRY OF THE VALVES. 
When dropped through a considerable depth of water, Pectens 
settle about as frequently upon the left valve as upon the right. 
: The University of Maine Studies, No. 6, September, 1906, p. 7. 
OBSERVATIONS ON PECTEN DISLOCATUS. 26 1 
The slight flatness of the left valve does not serve to make them 
settle always upon the same side. 
THE FUNCTIONS OF THE FOOT. 
The foot lies just opposite the notch in the right valve. It 
appears to be functionless in adult specimens, or rarely used by 
them, but is made use of to good advantage by the young. 
Specimens were often seen to extend the foot anteriorly to a 
remarkable distance, attach it at the tip to the bottom and then, 
by a powerful contraction, draw the body forwards to the point 
of attachment. The foot is cylindrical and seems very small to 
carry such a load ; and frequently, after it has been extended 
and attached, the valves are opened and clapped together, at the 
same time the foot contracts, the body thus being drawn forward 
with much less strain upon that organ. This method of loco- 
motion is a combination of swimming and creeping. The force 
of the current of water expelled from the mantle chamber serves 
to raise the body and propel it forwards as far as the attached 
foot will permit. At the same time, the foot contracts and the 
body lands close to the point of attachment. When this method 
of locomotion is used, the foot, instead of being extended directly 
outwards, anteriorly, is usually directed more ventrally, so that 
the point of attachment is more nearly in line with the force ex- 
erted by the swimming movement. Except for the notch in the 
right valve, this sort of performance would not be possible, be- 
cause by the closure of the valves, the foot would be crushed. 
The foot is, also, frequently used in turning the body over, 
when placed upon the left valve ; it is extended anteriorly from 
the body and attached ; the valves are opened and clapped 
together vigorously ; the body, as a result, is raised and shot 
forwards, but the weight of the foot and the resisting pull from 
its attachment cause it to swing over upon the foot as a pivot, 
the scallop landing upon the other valve, having turned through 
an arc of 180 degrees. 
The above method of turning over, is usually, if not quite uni- 
versally, used by specimens when placed upon the left side for 
the first time. After a little handling, however, they become 
much more irritable, seeming to be excited, and at such times, they 
262 B. H. GRAVE. 
manage to right themselves by one of three methods : Sometimes 
without extending the foot, they open the valves and clap them 
together. After one or several trials, the body turns over upon 
the hinge line as a pivot. The mantle must have played a part 
in this by expelling currents of water in a direction such as to 
cause the body to turn over. 
At other times, a position on the right valve is gained by one 
or more short swims, the method being continued until the body 
comes to rest on the right side. They usually manage to alight 
upon the right valve after a few trials, and then they become 
quiet. 
THE BYSSUS. 
When specimens are allowed to lie undisturbed upon the right 
valve, they usually become attached by numerous strands of 
strong byssal threads. A short time only is required for this to 
take place. They frequently become firmly fixed in from two to 
five minutes and the threads are sufficiently strong to support a 
weight several times that of the body of the Pecten. The byssal 
threads pass through the notch in the right valve directly to the 
support below. They adhere, to some extent, to the shell where 
they come into contact with it. 
So long as specimens are kept lying upon the left valve, they 
cannot, or do not, attach themselves by the byssus. Since the 
byssal gland lies at the base of the foot, it is possible that the 
notch, in the shell opposite it, is a structural adaptation directly 
correlated with the function of the byssus. At any rate, because 
the byssal threads extend through the notch, in place of over the 
edge of the shell, the pull has less tendency to tilt the body than 
would be the case if no notch were present. 
In order for the byssus to become attached to the bottom, it 
is not necessary for the valves of the shell to be opened, since the 
attachment of the byssus is frequently accomplished while 
they are closed. The notch in the shell is sufficiently large to 
allow the extension of the foot to the support during the process 
of attaching the byssus. It seems that Dr. Drew has observed 
this process in individuals of Pecten irradians, to quote : 
" An individual of Pecten irradians placed in a glass dish of sea 
water will sometimes protrude its foot from the shell, apply it 
OBSERVATIONS ON PECTEN DISLOCATUS. 263 
closely to the bottom of the dish and after a short time, slowly 
withdraw it, leaving a rather broad band of slightly yellowish 
material attached to the glass and connected with the foot by the 
byssal gland. This is not composed of small threads as in the 
mussels mytihis and inodiola, but it may be sufficiently tough to 
support the weight of the animal, if, after a few minutes, the dish 
is carefully turned over." 1 
SUMMARY. 
By way of summary, therefore, it might be said concerning the 
function of the notch that it makes possible a much freer use of 
the foot and byssal gland, and is in some way connected with the 
function of these organs. Although many mollusks live in the 
mud, the fact that young Pectens do not is evidence that they do 
better out of it. The foot and byssus enable them to climb upon 
supports and maintain their position there. As they approach 
maturity, they assume more and more the swimming habit and 
the foot and byssus lose, to some extent or entirely, their func- 
tional activity. If these organs are not functional in full-grown 
Pectens, as seem probable, the notch is no longer of any value to 
them, although it is not obliterated. 
The Pecten has the sense of position well developed. 
EARLHAM COLLEGE, RICHMOND, INDIANA, 
March 6, 1909. 
1 The University of Maine Studies, No. 6, September, 1906, p. 18. 
264 B. H. GRAVE. 
EXPLANATION OF PLATE I. 
FIG. I. Shows the left valve as it appears from surface view. It is not quite 
symmetrical. 
FIG. 2. Shows the right valve as it appears from surface view. The prominent 
notch at the base of the anterior ear can be seen. 
FlG. 3. Is a view of the inner surface of the right valve. The dotted line shows 
the position of the cartilage pad which aids in opening the valves. 
BIOLOGICAL BULLETIN, VUL. X/l. B. H. GRAVE. 
PLATE I. 
OCR text unavailable for this page.THE DYNAMIC FACTOR IN REGENERATION. 
T. H. MORGAN. 
With the publication of the data here presented the series of 
experiments that I have carried out on Tubularia for several 
years may be considered as temporarily brought to a close. I 
take this opportunity therefore to sum up the evidence bearing 
on the problem of the formative factors of regeneration, as ex- 
hibited by this hydroid. In the course of my experiments ten- 
tative hypotheses have been proposed here and there that have 
at least served to suggest further experiments. The conflicting 
evidence sometimes inclined me towards one point of view, some- 
times towards another ; yet, all in all, the same general line of 
thought, if sometimes vague, can be traced through the attempts 
to analyze the results. It will be my endeavor here to bring 
more into the foreground those theoretical deductions that seem 
to me at present to be best in harmony with the experimental 
evidence. 
That the dynamic factor in regeneration is not primarily the 
outcome of physiological movements of the animal, or of its 
parts, is made probable not only by many facts familiar to every 
student of regeneration, facts that show that the new part often 
develops under conditions where movement or function in this 
sense is absent, but also by the important experiments of Zeleny 
and of Stockard with the jelly-fish Cassiope. They have shown 
that when one half of the disc is brought to rest by removing 
the sense organs (and scratching a barrier zone across the con- 
necting ectoderm) the quiescent half regenerates as well and as 
rapidly as the half left pulsating during the period of regeneration. 
Since every part of the stem of Tubnlaiia is capable of pro- 
ducing a hydranth the inhibition of basal development is obviously 
due to the presence of a hydranth or of a developing hydranth 
at the oral end. Two alternative chemico-materialistic explana- 
tions have been suggested for cases like this. (A) The oral 
hydranth may use up some materials necessary for the formation 
265 
266 T. H. MORGAN. 
of a basal hydranth. (7?) The hydranth may produce some 
materials that inhibit the development of other hydranths from 
the remainder of the piece ; hence the inhibition, as long as a 
hydranth is present or developing. On first thought these alterna- 
tives would seem to cover the only possible ways in which the 
problem of regeneration may be presented at least as long as 
the problem is confined to purely physiological actions of a chemi- 
cal order. There are, however, not a few considerations indicat- 
ing that the fundamental interpretation may lie in a different con- 
ception of the problem. I shall try here to emphasize this other 
point of view without attempting to develop it into a theory of 
regeneration. At most we may hope at the present time to find 
in the facts some indication of the nature of the problem if not 
its entire elucidation. 
A number of experiments have been made that seem to indi- 
cate that the temporary inhibition of the development of the basal 
hydranth in Titbnlaria is not the result either of the using up of 
materials by the oral hydranth, or of the setting free of inhibi- 
tory stuff. The simultaneous development of hydranths at both 
ends of a piece, which frequently occurs in short pieces, is a case 
in point. Both ends develop at the same rate as when a single 
hydranth develops, and not half as fast as the hypothesis demands. 
Again the development of a basal hydranth does not appear to 
inhibit the oral development as we should expect if the result 
were dependent simply on the presence of materials in the stem. 
Some experiments of MacCallum's with plants have an impor- 
tant bearing on this point. If the terminal bud of the bean is 
removed, the buds in the axils of the cotyledons develop. But 
if the activity of the terminal bud is simply lessened by inclosing 
that part in an atmosphere of hydrogen, the basal buds do not 
develop. Hence the result is due not to activity of the terminal 
end, but to its presence or absence. In a different way the same 
fact is brought out. A piece of willow stem is cut off, its middle 
third is inclosed in a tube filled with moist air, so that the buds 
in this part are encouraged to begin their development ; the dry 
air retarding the development of those outside. After the middle 
buds have unfolded, the entire piece is inclosed in a moist 
chamber, when the more apical buds sprout forth, while none 
THE DYNAMIC FACTOR IN REGENERATION. 267 
of the buds basal to the middle region develop. The presence 
of growing shoots in the middle of the piece does not inhibit the 
apical buds from developing, if external conditions are supplied 
favorable to their growth, but the basal buds are inhibited by the 
presence of shoots on the more distal parts. These facts are incom- 
patible with the assumption that the results are due to the pres- 
ence of materials used up by those parts that develop first to the 
exclusion of other parts. They also show that the alternative 
view is untenable, for, the presence of growing shoots in the middle 
of the piece is not antagonistic to the development of shoots in 
other regions provided those regions are situated more distally. 
In the case of Tubularia, it is more difficult to present con- 
vincing evidence that distal hydranths do not produce ma- 
terials inhibiting the development of basal hydranths, improbable 
as such an interpretation may now seem. But the fact that basal 
hydranths do develop after the oral hydranths have formed may 
seem to discredit this view. Here, however, an apparent para- 
dox is found. The experiments seem to show that when the 
oral hydranths develop, the basal hydranths are retarded in de- 
velopment, but they do develop later, and the results also show 
that if both start simultaneously both develop at the normal rate. 
The paradox is due, I think, to two antagonistic factors at work 
at the same time. Admitting that the oral development tends to 
inhibit the beginning of basal development, we also find that if 
other influences suffice to start both simultaneously, the on-rush, 
so to speak, of the process once begun changes the conditions 
that tended to prevent the starting. Strange as this seems it is 
little more than a statement of the facts. The same results may 
be put in a somewhat different way. A cut end being present, 
whether oral or basal the conditions that call forth hydranth 
formation are given. Experiments show that the oral end 
tends to develop first, its development acts as a partial inhibition 
of the basal hydranth-formation. If this influence is strong 
enough the basal development is temporarily held in abeyance, 
but if not the inhibition is overcome. Once overcome, the for- 
mative influences do not check the further action of the basal 
end. In this connection it is curious to note that small oral 
pieces produce simultaneous hydranths more often than larger 
268 T. H. MORGAN. 
pieces. The interpretation of this seems to be that the tendency 
to produce hydranths, both oral and basal, is stronger near the 
distal end and decreases basally. In short pieces the sensitive- 
ness of the two ends to those influences that call forth the hydranth 
is so great that both ends develop simultaneously or nearly so, 
hence the oral end has not time to get a sufficient start over the 
basal to stop its development. It should be noted in passing 
that it is probable that the influence preventing basal develop- 
ment is not only the oral development, but a direction-factor 
present in the stem at all times. This factor we call polarity. 
The interesting point is that this factor seems to be more capable 
of inhibiting basal hydranth formation when an oral end is 
developing than when such development has not yet begun. 
The basal development, however, does not appear to delay the 
oral process. It is acting against the polarization and its influ- 
ence is less felt throughout the stem, as experiments by Stevens 
and myself have shown. These considerations lead, I think, to 
the view that the essence of our problem lies in that peculiarity 
of the piece that we designate its polarity, and not in the absence 
or presence of formative substances in Sach's sense. 
If our analysis is correct, we are led to look upon living ma- 
terial as possessed of a certain formative principle that has so to 
speak a "sense of direction." The next step will be to study 
the nature of this principle and see what properties we are justi- 
fied in ascribing to it ; for while it may be beyond our powers at 
present to state precisely the nature of the directive principle, we 
may at least be enabled to work out its manifestations. Some 
of these manifestations become apparent in the study of the re- 
generation of Tubularia. One of its most striking modes of ac- 
tion is seen in the inhibition of basal-hydranth formation. Most 
interesting is the result that its action becomes intensified by de- 
velopmental processes going on at the oral end, as shown by 
the fact that if the oral development is suppressed by tying 
that end, the basal development is much accelerated. It is ac- 
celerated in the sense that basal hydranths more often develop 
at once than when both ends are open, but not in the sense that 
the basal development is faster than when this end also gets as 
early a start as the oral end. In other words, there is no speed- 
THE DYNAMIC FACTOR IN REGENERATION. 269 
ing up of hydranth formation as such, but the initial inhibition is 
overcome. 
The special problem with which this paper deals is the nature 
of what takes place at the basal ends when the oral end is kept 
open and when it is tied. Is the retardation of such a kind that 
a slower process of development is going tin at the basal end 
while the oral end is developing, or does the basal end not really 
begin to develop until the oral end has formed its polyps. If so, 
what gives it its start later ? The following experiments were 
devised to study these questions. 
Experiment I. - - The purpose of the experiment was to deter- 
mine whether when both oral and basal ends of a piece are left 
open constructive changes are slowly going on at the basal end. 
Some pieces were cut off and left open (A) ; later other pieces 
were cut off and the oral ends tied (Z>') and at the same time the 
oral ends of (A) were tied. It was found that the basal ends of 
the (A) pieces did not develop faster than those of the (>) pieces, 
showing that the changes at the basal end of (A) are not pro- 
gressing, but are held in check by the developing oral hydranth. 
Control I. - - In some pieces the old hydranths were left intact 
and the pieces cut off. No basal hydranths began to develop 
until the old heads began to be absorbed. The presence of the 
old heads inhibited the development of the basal hydranths until 
the heads had degenerated when the latter appeared. 
Experiment II. In order to find out whether, when the oral 
end is tied, changes take place throughout the piece that tend to 
make more rapid the development of basal hydranths, or whether 
these changes are localized at the basal end where the new hy- 
dranth develops, the following experiment was tried. The oral 
ends of many pieces of the same length were tied. Then after 
several hours' interval differing in several experiments, the basal 
end was cut off, (a] just inside of the area that would form the 
basal hydranth, (&) in the middle of the piece, (c) just below the 
ligature. In general the development of the basal hydranth was 
delayed as compared with control pieces tied but not cut off 
at the basal end ; the delay was the greater the further re- 
moved the cut from the basal end, despite the fact that oral levels 
tend to regenerate faster than more basal levels. The differences 
2/O T. H. MORGAN. 
are more apparent the longer the time that elapses before the 
basal pieces are removed. The differences are not very marked 
at the different levels indicating perhaps that changes take place 
throughout the piece and not only at the basal end although 
more pronounced in the latter. The different levels of the cuts 
make it difficult to ascribe the results solely to the general 
changes in the piece, for the more orally situated cut ends have 
an advantage in level as other experiments have shown. 
Experiment III. - - Pieces were cut off at the same oral levels. 
After 23 hours the hydranth region at the oral end was cut off 
of some pieces (^4), others were cut in two in the middle of the piece 
(.Z?), and for a control some pieces were left as before (C\ A 
slight retardation occurred after another 12 hours in (A}, less 
in (Z?) as compared with ((7). Removal of the hydranth forming 
region after 23 hours causes delay but the delay is not so much 
as though a new hydranth had developed at the new cut, show- 
ing that changes directed towards hydranth formation are going 
on not only in the region where the hydranth will develop but at 
more basal levels as well. 
Experiment IV. - - This experiment was like the last, except 
that the basal ends of all the pieces were tied, thus preventing the 
basal end from exerting any influence on the result. Other ex- 
periments had shown, however, that the basal development, even 
if it occurs, has apparently no retarding influence on the oral de- 
velopment. The results, as was to be expected, were the same 
as in the last experiment. It is interesting to note that in both 
the influence of the cutting causes a greater delay in the first ap- 
pearance of the primordia of the hydranth than on their later 
development ; for later the differences seem to be less than at 
first. This may be due to an acceleration extending through- 
out the whole time that is more effective after a beginning has 
been made than before the start. 
Experiment V. Some previous experiments had left undecided 
the question whether, when the oral end is left open for several 
hours and is then tied off, the basal development is more rapid 
than when the oral end is tied at once. If such an acceleration 
really occurs it might seem to indicate that changes take place 
in the oral end that produce accelerating materials even for the 
THE DYNAMIC FACTOR IN REGENERATION. 27! 
basal ends. I was particularly anxious to settle this point defi- 
nitely, for obviously, if such acceleration could be proved, it 
would furnish evidence in favor of a chemical process, especially 
since other experiments had seemed to show that the basal end 
does not begin its development when both ends are left open. I 
have carried out rather an extensive series of experiments that 
give, I think, a definite answer to the question. 
When pieces are left open at both ends from four to nine hours, 
and are then tied at the oral end, the basal development is slightly 
retarded as compared with its development in pieces tied at once. 
There is little evidence in favor of the view that the later tied 
pieces can make good the loss of four to nine hours, and of 
course they can not catch up if a longer time elapses. Whether 
they may do so in later stages is more difficult to decide, but 
this does not concern the main point here raised. 
Individual differences in rate, differences in stems, and uncon- 
trollable differences in level tend to obscure results that depend 
on only four, six, and nine hours differences in start. The above 
statement holds, therefore, only for average results. There was 
found no evidence in favor of actual acceleration, whether there is 
some relative acceleration is difficult to decide. If the hydranths 
do not develop promptly /. c., if a long time elapses between the 
tying and the appearance of hydranths, the initial differences of 
a few hours may be lost. 
Experiment VI. Another attempt was made to see whether 
changes take place in the piece as a whole, after it is cut off, that 
make more rapid the development of an oral hydranth when a 
new cut is made. 
Pieces were removed and after four hours somewhat more than 
the oral hydranth region was cut off. In some cases the newly 
cut ends developed as fast as did the hydranth in the small pieces 
cut off, but the latter may have been retarded by the operation or 
by the smallness of the pieces ; yet in some cases the development 
of the newly cut ends was as rapid as in control uncut pieces. 
This result indicates that changes take place in the pieces behind 
the actual region of hydranth formation that lead toward the 
development of a hydranth. 
Experiment VII. In this case pieces first cut off were after 
2/2 T. H. MORGAN. 
ten hours cut in two in the middle. Comparing the rates of de- 
velopment of the oral ends of the oral halves with that of the oral 
ends of the basal halves it appears that the latter are slower, 
but there is evidence that the retardation may be somewhat less 
than the ten hours difference in initial start. The basal halves in 
this experiment are also somewhat behind the basal halves in the 
last experiment, which seems to show that the general changes 
in the excised pieces that go towards oral development decrease 
from the cut ends inwards. 
It has not seemed necessary to give the details on which these 
general conclusions are based. The nature of the case makes it 
difficult to obtain results as definite as one might wish, despite 
the precautions that were undertaken to make the conditions as 
uniform as possible. The general conclusion that changes take 
place in the piece as a whole, after its removal, that are in the 
direction of hydranth formation, seems fairly certain. Less cer- 
tain perhaps is the evidence to show that when the oral end is 
tied similar changes take place in the pieces that accelerate basal 
development in regions beyond the hydranth forming region, but 
this conclusion too is, I think, quite probable. The nature of 
these changes is not revealed. 
THE DYNAMIC FACTOR IN EGG-DEVELOPMENT. 
Students of the processes of regeneration have without excep- 
tion made use of the term polarity to express a directive factor 
observable in their results, and to this factor is sometimes as- 
cribed an active role as a controlling influence, at other times the 
term is used descriptively merely as a statement that the new 
structures are directed in the same way as the part removed. In 
both respects the word has been useful, however vague our con- 
ception of what polarity may be. Our analysis of the process 
has now gone sufficiently far, I think, to justify us in an attempt 
to come to closer quarters with the term. Without reviewing 
the opinions that have been expressed as to the nature of polarity, 
I shall try to contrast two views of its nature that seem to me to 
represent the two main lines that speculative thought has followed. 
It should be noted that the term is used equally by students of 
embryology and by students of regeneration. The former 
THE DYNAMIC FACTOR IN REGENERATION. 2/3 
finds axial relations in the egg polarity, bilaterality, radial 
symmetry, etc. ; the latter finds the new organs regenerated in 
definite relations to the old. To some observers the distribution 
or the stratification of the materials of the egg, has seemed a 
sufficient basis for the results referred to under the term polarity ; 
to others it has seemed more probable that there exists in the 
egg an arrangement or structure that has axial relations from 
which result not only the depositions of the formed materials 
but also the nature of the action of the parts. Polarity is from 
this latter point of view not simply a passive structure, but a re- 
lation of the parts that directs the shifting series of changes that we 
call development. At one time one of these views has seemed 
more probable ; at other times the other. The history of modern 
experimental embryology and regeneration shows the influence 
that these views have had on those who have followed the 
new work. In a general way the two views may be classed as 
the materialistic or chemical and the dynamic or physical con- 
ceptions of the developmental process. At present it seems to 
the writer that the evidence has been steadily pointing to the second 
of these contrasting views as the more probable. As far as the 
egg is concerned, the recent experimental work goes to show 
that the visible inclusions of the protoplasm (yolk, oil and other 
granules perhaps) are not the fundamental causes of the forma- 
tive processes, although they may be needed in certain regions to 
carry out the future development of the structures that there 
appear. In regard to regeneration it has been evident for some 
time, that the specification or the differentiation (with its concomitant 
products), cannot be unreservedly utilized as a basis for an explana- 
tion of formative processes that take place. For example, if the 
gross materials or the differentiations of the head end of a plana- 
rian are the causes of that region being a head, it is inexplicable 
that when the head is removed it could regenerate a tail. There 
must be something else behind what we see that is responsible 
for the change that takes place. These and other considerations 
lead to the view that there exists a fundamental property of liv- 
ing matter that is the formative principle of development. On 
two former occasions, when attempting to analyze the results of 
regeneration in Tubularia, the author tried to account for the re- 
2/4 T. H. MORGAN. 
suits of polarity on the basis of a stratification of the materials. 
Influenced at the time by recent results in experimental embry- 
ology that seemed to show that visible substances of different 
kinds in the egg are really responsible for the development of 
its parts, the same idea was applied to the problem of regenera- 
tion, despite the fact that I had on more than one occasion re- 
jected the hypothesis of formative stuffs, in Sach's sense, as suf- 
ficient to account for the facts of regeneration. Yet a careful read- 
ing of the papers here referred to will show that I still held, 
though perhaps not always consistently, to the conception that 
back of these differentiated materials lay the real differentiating 
factors. 1 It now seems to me that the evidence, which at that 
time seemed so strongly to favor the idea of the importance of 
the grosser materials of the egg, is insufficient to establish its 
case, and that the important factors of development are dynamic 
properties of the bioplasm, rather than the formed products of 
the egg, or of the differentiated products of the adult animal. 
This statement does not mean that the visible products in the egg 
play no role in development. The evidence still shows that they 
may do so, but their role seems to be secondary, not primary. 
The interrelation of the parts seems to be one of the most 
evident expressions of the fundamental formative influences. 
Several years ago a consideration of a number of results in 
regeneration led me to state that this relation might be expressed 
as a sort of tension. This view has been objected to on the 
ground that it does not appear to explain the matter any better 
than before. In a moment of doubt and in order to give the 
l One further word of explanation. The rate of hydranth formation varies with 
the distance of the cut end from the original hydranth. I have spoken of this differ- 
ence in rate as explicable on the assumption of the hydranth-forming materials de- 
creasing toward the base, i. e., away from the hydranth. It was unfortunate to have 
used the term hydranth materials, although I made sufficiently clear in the text that 
I did not mean to invoke the stufi-hypothesis in this connection. It is not entirely 
clear on what the difference in rate depends ; most probably on the stem being less 
specialized as a store-house of food substance nearer the hydranth ; probably also 
on some difference connected with the thickness of the walls with which the speciali- 
zation may also be connected ; possibly neither of these but some more fundamental 
characteristic is responsible for differences in rate. In any case it is not obvious that there 
is any connection between this difference in rate and the polarity of the piece. The 
latter is the same for all levels the time it takes the piece to be remodelled seems to 
be referable to something else. 
THE DYNAMIC FACTOR IN REGENERATION. 275 
statement a meaning for those who believe no suggestion to be 
of value unless it refer the problem to ordinary properties of 
inorganic bodies, I suggested that osmotic pressure might be the 
cause of the tension differences in the parts. This was an 
unnecessary concession. The behavior of fluid crystals (accord- 
ing to Lehmann) shows that the formative changes can be 
accounted for on the basis of a tension exhibited by the mole- 
cules of the substance of the crystal. While the organism may 
not be put down as a fluid crystal, still we see that physical 
properties other than osmotic pressure and surface tension may 
play an all-important role in form-changes. It may be that a 
similar property is the cause of the formative changes in the 
organism. In any case, the facts that I had in mind suggested 
that tension of some sort is an important dynamic factor in 
development, perhaps the important factor. The facts still seem 
to me to indicate some such relation between the parts, and no 
one regrets more than I that we cannot "explain" the results 
even if my suggestion prove to be in the right direction. 
Still later a consideration of certain facts of development led to 
the suggestion that two known properties of the organism con- 
tractility and irritability also play a very important role in 
embryonic and regenerative development. I shall not attempt 
to review here the argument which led to this point of view. 
How far and in what sense contractility and irritability are better 
expressions of the tension hypothesis, it is not easy to state. 
So far as contractility is concerned Lehmann's recent important 
paper on " Scheinbarlebende Kristalle " l shows the possibility at 
least of referring this property also to a condition of molecular 
tension. We are still too ignorant of the physical basis of 
irritability to make speculations in this direction profitable, but it 
may be well not to lose sight of this property of living matter in 
our attempts to analyze further the problem of development. 
STEREOMETRY OF THE BIOPLASM. 
Polarity implies difference in one direction. Every student of 
regeneration knows that in all three dimensions of space the same 
factor is present. Polarity is therefore only a part of the problem, 
l io/. Centralb., XXVIII. , 1908. 
2/6 T. H. MORGAN. 
and so far as it draws attention away from the whole problem it 
seems best to substitute the term stereometry. 
Sufficient evidence has accumulated, I think, to show that 
stereometry has a dynamic side in so far as it is a result of the 
molecular factors that determine the relations of the parts to each 
other. A question of fundamental importance here presents it- 
self. If the formed substances at each level are the products of 
the bioplasm, must not the bioplasm itself be stratified in nearly 
the same sense ? It was this idea that I had in mind when I 
wrote in 1906 : " If we imagine a stereometric network as a part 
of the specialized structure, we must be prepared to admit that it 
changes at each level as the structure changes. Therefore it 
seems to me simpler to base our hypothesis of polarity on the 
difference in differentiation itself, and not on an imaginary polar- 
ized system associated with the living materials." But the point 
I overlooked was that there is no need to suppose that a hetero- 
geneous network of bioplasm exists because the visible structure 
formed by it is different. The relation of the polarized material 
to the ends of the material (indeed to all its directions) suffices to 
account for the difference of level. In fact if the stereometry 
rests on a dynamic and not a statical relation of the parts this is 
the logical standpoint. 
It has been suggested that irritability may be related to the 
dynamic factor of development. 
The effects of irritability at any level may be realised through 
the cliemical changes inaugurated. TJiese chemical changes once 
started may, if enzymatic, thenceforward continue (unless checked by 
other chemical processes], independently of the factor that set them 
going. 
Vol. XVI May, 1909. No. 6 
BIOLOGICAL BULLETIN 
THE REGULATORY CHANGE OF SHAPE IN PLAN- 
ARIA DOROTOCEPHALA. 
C. M. CHILD. 
The recent discovery of Planaria dorotocephala Woodworth 
(Woodvvorth, '97) in very large numbers, near Chicago, has made 
it possible for me to use this form for extensive series of experi- 
ments. The species is very similar to P. niacitlata in structure, 
behavior and regulation, but possesses some advantages over that 
species for experimental work. It attains a larger size, is more 
active, and can be obtained in unlimited numbers and all ages in 
this locality, while P. macnlata is much less abundant. I found 
the same species in California some years ago (Child, '06), but 
was unaware at that time that it had been described. 
In the present paper only certain experiments concerning the 
effect of anaesthetics on form regulation will be considered. It is 
possible by the use of dilute solutions of anaesthetics to control, 
modify and inhibit various regulatory processes almost at pleasure. 
For example, head-formation can be made a process of redifferentia- 
tion instead of regeneration in almost any desired degree (Figs. 14 
and 1 6) or can be completely inhibited, according to the condi- 
tions of the experiments, and the same is true concerning the 
formation of a new posterior end and a new pharynx, and the 
regulatory changes in the intestinal branches. Moreover, the use 
of anaesthetics permits, in greater or less degree, an analysis of 
some of the various factors concerned, and finally, it is possible 
by this means to produce individuals capable of continued exist- 
ence if returned to water which possess characteristics, or per- 
haps more properly, combinations of characteristics which do not 
occur in nature. 
The anaesthetics chiefly employed in my experiments thus far are 
2/8 C. M. CHILD. 
alcohol, ether and acetone-chloroform, commercially known as 
chloretone. The effects of all are essentially similar in kind, but 
of course differ quantitatively according to the substance and the 
concentration. Ether, for example, in a concentration of 0.4-0.5 
per cent, produces about the same effect as alcohol in a concen- 
tration of 1.5-1.6 percent. In solutions of these concentrations, 
individuals and pieces have been kept alive as long as three months, 
though the resistance differs greatly according to the condition of 
the animals and various other factors, most of which can be con- 
trolled experimentally either directly or indirectly. 
In order to avoid as far as possible decrease in concentration 
of the solution by evaporation the following method has been 
used. The animals or pieces are placed in Stender dishes of sev- 
eral hundred c.c. capacity with ground edges and a cover with 
ground groove exactly fitting the edge. The groove is filled 
with solution of the same concentration as that in the dish so 
that the dish is sealed so long as the fluid does not evaporate 
from the groove. After the dishes are thus closed they are 
placed in larger jars containing a liter or more of the same solu- 
tion and these are sealed with vaseline and the covers weighted 
o 
so that no escape of vapor or entrance of air is possible. And 
finally, all solutions are renewed every forty-eight hours and are 
made up immediately before using. In this way it is possible to 
compare the effect of the anaesthetic upon pieces of different size 
and from different regions of the body and also upon worms in 
different physiological condition. 
This method makes possible the control and modification of 
form regulation in this species to a greater extent than any other 
which has been described. At present, however, only certain 
points will be considered, a full account of the experiments being 
postponed to a future time. 
In several of my " Studies on Regulation " I discussed the 
changes in shape and proportion which occur in isolated pieces 
of various species of turbellaria and which, under the usual con- 
ditions, constitute an approach to the shape and proportions of 
the whole animal. These experiments \vith anaesthetics furnish 
new data which confirm my earlier conclusions, and it is some 
of these data which are to be presented here. 
CHANGE OF SHAPE IN PLANARIA. 
279 
a 
...b 
I. EXPERIMENTAL DATA. 
When whole individuals or pieces are placed in 1.5 per cent, 
alcohol or in 0.4-0.5 per cent, ether they lose the power of coor- 
dinated movement almost entirely for a time. After four to five 
days, however, they become in some degree acclimated to the 
new conditions and begin to move about very slowly, but with 
increasing vigor as time goes on, though they never attain the 
normal motor activity. Pieces including the old 
head begin to move about earlier than pieces with- 
out a head, for the latter must form a new head be- 
fore they can regain the usual degree of motor 
activity, and regulation is greatly delayed in the 
solution. The important point for the present pur- 
pose is that for some days movement, and particu- 
larly coordinated locomotion, is almost completely 
eliminated. It is of interest to determine to what 
extent regulation occurs under these conditions. 
The first experiment to be described concerns 
pieces including that part of the body anterior to the 
line b in Fig. I, /'. e., short pieces with the old heads. 
In Fig. 2 a piece of this kind after ten days in 1.5 
per cent, alcohol is shown : during this time the 
piece has moved about but little and that chiefly 
during the last few days. Fig. 6 represents a 
similar piece after five days in water, Fig. 7 the 
same piece after ten days. Comparison of Figs. 2 
and 7 shows that regulation in the alcohol is 
greatly delayed : a little new tissue has been formed 
at the posterior end, but, as a microscopic examina- 
tion under pressure shows, it is still a mass of cells 
without any marked visible differentiation, and it 
can readily be seen that it is not used as a tail and 
is not attached to the substratum as the animal 
creeps slowly about ; a small group of cells is 
present in the pharyngeal region, but these likewise show no 
marked differentiation. In water, on the other hand, a new 
tail has been formed which functions in the normal manner, con- 
tracting, extending and attaching to the substratum as the animal 
FIG. i. 
280 
C. M. CHILD. 
creeps ; the pharynx is well-developed and sections show that it 
possesses essentially the same structure as the pharynx in un- 
injured animals. 
But the difference between the two pieces is most marked as 
regards change of shape. The piece in alcohol has not elongated 
at all, in fact it has decreased in length and it may be noted 
incidentally that the "auricles" on the sides of the head are 
greatly reduced. The piece in water (Fig. 7) has in the same 
I f 
> t 
4 
length of time elongated to nearly twice its original length, has 
become much more slender and tapers posteriorly. This piece 
has moved about during regulation to an even greater extent 
than the uninjured animal, for short pieces with the old heads are 
usually more active than whole individuals. 
After ten days the piece in alcohol gradually becomes more 
CHANGE OF SHAPE IN PLANARIA. 28 1 
active, though it never attains anything like normal activity. At 
the end of twenty or twenty-five days it has acquired a shape 
like Fig. 3. The posterior end now functions as a tail to some 
extent and attaches itself to the substratum as the animal 
advances but the amount of new tissue has not increased. The 
piece may live for six weeks or more in alcohol but it never 
undergoes any appreciable further change in shape. The newly 
formed parts undergo some degree of differentiation but never 
attain the characteristic adult structure. Apparently the piece 
has attained approximately a condition of equilibrium for the 
conditions under which it is living. 
If the concentration of the solution is gradually increased after 
the first three or four days it is possible to inhibit practically all 
regulation beyond the closure of the wound : the pharynx does 
not appear, no further growth of new tissue at the posterior end 
occurs, and the piece undergoes no elongation (Fig. 4). Under 
these conditions the piece does not acquire the ability to move 
about. 
If now these pieces which have attained equilibrium in alcohol 
be returned to water they gradually resume the process of regu- 
lation, but with certain differences from pieces which have not 
been in alcohol. The chief difference for present purposes is 
that the outgrowth of new tissue at the posterior end does not 
proceed until it reaches the usual amount. The tail is formed 
almost entirely by a redifferentiation of the old tissue (Fig. 5). 
The pieces may undergo change of shape after their return to 
water until they attain practically the same shape as pieces which 
have not been in alcohol. Fig. 5 shows a later stage of Fig. 4 after 
its return to water and Fig. 8 will serve as regards shape for the 
late stages of either water or alcohol-water pieces. 
These results, which are merely illustrations of what I have 
observed in several hundred pieces, permit certain conclusions of 
interest. In the first place it is possible to inhibit entirely the 
change in shape without inhibiting entirely the processes of redif- 
ferentiation and regeneration, and the change in shape can be 
stopped at any point without stopping entirely other regulatory 
processes (Fig. 3). On the other hand, if the growth of new 
tissue from the cut surface is inhibited in earlier stages (Figs. 2 
282 
C. M. CHILD. 
and 4) the change in shape may occur in later stages (especially 
after return to water) without the formation of more new tissue 
(Figs. 3 and 5)- It follows that the factors determining the 
change in shape must be in greater or less degree different from 
those determining the localization andgrowth of new parts. More- 
over, the change in shape occurs only when the piece is capable 
of locomotion and it is in general proportional to the locomotor 
ability of the piece. 
v 
// 
Fir.s. 9-11. 
But pieces which do not possess the old head afford even more 
positive evidence for these conclusions. Fig. 9 represents a 
piece corresponding to the region between the lines a and c in 
Fig. i after fifteen days in 1.5 per cent, alcohol. A new head 
has appeared, a small new pharynx is present as a mass of undif- 
ferentiated cells, and some new tissue has formed at the posterior 
end, but the piece as a whole shows no approach to the normal 
shape : it has undergone no marked changes in proportion. Inci- 
dentally it may be noted that the pharynx in such pieces appears 
much further posteriorly than in similar pieces in water. After 
seven days more the piece has the shape shown in Fig. 10. It 
moves about slowly, but its movements are different in character 
from those of pieces possessing the old head : here the posterior 
half of the body is very evidently not under complete control, 
CHANGE OF SHAPE IN PLANARIA. 
283 
i. e., is not fully coordinated with the anterior region, and when 
the animal advances it is simply dragged along as a mass of 
inert material, its posterior end being only very rarely attached 
to the substratum. The shape of the piece suggests that the 
anterior part is being stretched by the strain upon it of the pos- 
terior portion, and I believe that is exactly what is occurring. 
Similar pieces in water attain in the same length of time the 
shape and structure shown in Fig. II. Pieces in alcohol of 1.5 
per cent, do not change in shape much beyond the condition 
shown in Fig. 10, but if they are returned to water they regain 
their normal locomotor activity and may finally reach a shape 
like that of Fig. 1 1. 
T r 
12 
14 
15 
FIGS. 1 2-1 6. 
In these cases a new head, a small new pharynx and some 
new tissue at the posterior end have appeared without any marked 
change of shape in the piece as a whole (Fig. 9). Evidently the 
change of shape and the localized formation of new tissue are not 
necessarily correlated. 
The same thing appears in Fig. 12, which shows a piece from 
the same region (a to c, Fig. i) after fifteen days in 0.5 per cent, 
ether. Head, pharynx and posterior end have formed but no 
change in shape has occurred. This piece was returned to water 
at this stage and after seven days more had attained the shape 
shown in Fig. 13. 
284 C. M. CHILD. 
In another series pieces comprising the whole post-pharyngeal 
region were used (posterior to d, Fig. i). These pieces, after 
eighteen days in 0.5 per cent, ether, had attained the condition 
shown in Fig. 14. A small head is forming, almost entirely by 
redifferentiation, at the anterior end, with one median eye, and a 
small pharynx is present. At this time half of the pieces were 
returned to water and half remained in ether. After nine days 
more the pieces in water had acquired the condition shown 
in Fig. 15 while those in ether were like Fig. 16. In the pieces 
returned to water the anterior half is greatly elongated but the 
posterior half remains much as before. In these pieces, as in the 
one described above, the posterior part was dragged about by the 
more active anterior portion. Gradually complete coordination 
returned and the posterior end began to attach itself to the sub- 
stratum as the animal advanced and after this the shape gradually 
approached that of the normal animal. 
Further data along this line could be presented but these cases 
are sufficient to show that it is possible to delay or inhibit the 
change in shape to any desired degree, and to induce its occur- 
rence at any time. Moreover, the manner in which it occurs 
can be controlled and altered indirectly by using pieces of dif- 
ferent sizes and from different regions of the body. 
To my mind the evidence indicates very clearly that the change 
in shape is primarily a mechanical deformation of the body in con- 
sequence of the altered direction of the strains to which it is sub- 
jected as the animal advances. To control the change experi- 
mentally we have only to control the locomotor activity. In 
several of my earlier papers (Child, '02, '03, '04^) I have de- 
scribed the method of locomotion in certain species of turbel- 
laria : in Planaria longitudinal strain arises in essentially the same 
manner as in the other forms discussed, the use of the posterior 
region as an organ of attachment being one of the chief factors. 
Moreover, there is considerable direct evidence that the tissues of 
a region undergoing change of shape are being stretched. In 
regions where the decrease in width and the elongation of the old 
parts begin, the chromatophores are always greatly elongated, 
and their elongation is greatest where the change in shape is most 
rapid. As the change goes on they become drawn out into long 
CHANGE OF SHAPE IN PLANARIA. 
28 5 
lines. Figs. 17 and 18 indicate this change in shape of the chro- 
matophores in the region posterior to a new head, and Figs. 19 
and 20 for a region anterior to a new tail. In the pieces in alco- 
hol and ether this change in shape of the chromatophores appears 
only when the change in shape occurs, not when the new tissue 
is formed. In cases where the change of shape is inhibited in the 
anaesthetic (Figs. 2, 4, 9, 12) it does not appear at all, but if such 
pieces are returned to water, and the change of shape occurs, the 
stretching of the chromatophores also appears. In Fig. 14, for 
example, it did not appear so long as the piece was kept in the 
ether, but after several days in water it was most conspicuous in 
the slender anterior region (Fig. I 5), this region appearing almost 
FIGS. 17-20. 
as if finely striped in the longitudinal direction. This change in 
shape of the chromatophores is actually a stretching, not a migra- 
tion, for it is possible to select some particular spot which happens 
to be conspicuous for some reason and to observe its change of 
shape from day to day : in such cases it can be seen clearly that 
merely elongation not migration occurs. 
A similar elongation is visible in the parenchyme cells in sec- 
tion. Stevens ('07) has recently described this elongation or 
orientation of the parenchyme cells and regards it as indicating 
migration, but Steinmann ('08) does not agree with her. As a 
matter of fact the specimens in which the change of shape is 
inhibited by anaesthetics show nothing of the sort even in regions 
adjoining those where new tissue is being formed, but if such 
pieces be returned to water the cells of the parenchyme become 
very distinctly elongated or oriented in the direction in which 
elongation of the body is occurring, even though no new tissue 
is being formed at the time. In short the change in shape or 
286 C. M. CHILD. 
arrangement of the parenchyme cells probably has nothing to do 
with active migration, but merely indicates the direction of the 
strain which produces deformation. 
These histological features then, support and confirm the other 
observations, and all appear to show that the change in shape is 
primarily a deformation in consequence of strain rather than a 
complex physiological process. 
There can be no doubt, however, as I have repeatedly pointed 
out that reactions of various kinds result from the strain and de- 
formation : muscles and other tissues undoubtedly " adapt " them- 
selves to the new relations of parts. In fact there is no apparent 
reason why the change of shape should not continue indefinitely, 
or at least until the elasticity of the tissues became involved, if 
nothing but the mechanical deformation took place. Undoubtedly 
"functional adaptation " to the altered strains occurs and this 
determines how far the change shall proceed. Sooner or later 
the tissues adjust themselves fully to the new conditions, /. <?., a 
condition of equilibrium is attained and change in shape ceases. 
Under the usual conditions this gives what we commonly call the 
" normal " shape, but I have shown above that under other con- 
ditions the shape may be different. No one would deny, I sup- 
pose, that the shape of cells is determined in many cases (e. g., 
the polyhedral shape of blastomeres in many eggs, the polygonal 
outline of many epithelial cells, etc.) to a greater or less extent 
by purely mechanical factors. If this is the case it seems scarcely 
possible to deny that purely mechanical factors may be concerned 
in determining the shape and arrangement of parts in animals 
without hard structures and with tissues of a high degree of 
physical plasticity. That they are the only factors in such cases, 
or that because they are factors in some cases, they must be in 
others, I certainly do not and never have maintained. 
In Planaria the width of the head is apparently an important 
factor in determining the width of the body behind it. The head 
itself is not involved to any appreciable extent in the change of 
shape. It furnishes, so to speak, a fixed point, or rather region 
at one end, and between this and the posterior end the change of 
shape occurs. Consequently, in short pieces, where the new 
head is small, the change of shape is very much greater than in 
CHANGE OF SHAPE IN PLANARIA. 287 
long pieces, where it is much broader. In short pieces possess- 
ing the old head and consequently capable of rapid and frequent 
locomotion, the change occurs much more rapidly than in other 
pieces and the body often becomes much more slender propor- 
tionally than that of the " normal " animal and tapers more toward 
the posterior end. In pieces from the posterior region of the 
animal the chief region of attachment forms the largest part of 
the piece. Since attachment occurs by the lateral margins as 
well as by the tip of this region, the change of shape occurs most 
rapidly in the regions just behind the new head, for these are the 
regions where the greatest change in the direction of the strains 
occurs. 
Summing up the data presented, the first point of importance 
is that no necessary connection exists between the change of 
shape and the redifferentiation or regeneration of parts : the two 
processes can be separated from each other almost completely 
in time, and are often separated in space. Secondly, it is possible 
to control the change of shape, to inhibit it, to stop it at any point, 
or to allow it to proceed, by controlling the locomotor activity. 
And finally, the tissues in regions undergoing change of shape 
show very distinct indications of physical deformation in the direc- 
tion of the strains. The conclusion which seems to me to accord 
most closely with the facts as they now stand is that the change of 
shape is primarily a mechanical deformation resulting, at least in 
large part, from the strains which arise in locomotion, these strains 
being altered in direction and amount in pieces as compared with 
the whole animal. The final shape, /. c., the cessation of change 
is probably determined by the physiological reaction of the dif- 
ferentiating or redifferentiating tissues to the altered conditions, 
and the consequent establishment of an equilibrium. 
II. FUNCTION, FORM AND REGULATION. 
Since the appearance of my earlier papers on regulation cer- 
tain reviews and criticisms of the work have appeared. From some 
of these, particularly from Driesch's review (Driesch, '05) I can 
only conclude either that I have failed to state my position clearly, 
or that the reviewers have not become sufficiently acquainted with 
my work to understand what that position is. Driesch, for example, 
288 C. M. CHILD. 
has imputed to me certain views which I have not only never 
held, but which I have expressly repudiated more than once. 
And recently, in a brief reference to my work, Morgan ('07, pp. 
3734) has cited certain conclusions as mine, which are very dif- 
ferent from those which I have reached. I have considered 
Driesch's criticisms elsewhere (Child, '07) and shall refer to these 
and other criticisms only incidentally. But a brief statement of 
my position with regard to certain features of the problems of 
form and regulation and with especial reference to mechanical 
factors, movement and use of parts and function in general seems 
desirable in connection with the new facts presented above. 
i. Tlic Relation between Function and Form. 1 
In a fully developed organ certain processes occur which are 
concerned with the maintenance of the organism as a whole. 
These processes may affect either the relations of parts to each 
other or the relations of the organism as a whole to the external 
world, or both. They are commonly designated as functions, 
i. e., the adult organism may be regarded as a complex machine, 
which works or functions in a characteristic manner. 
But the functions characteristic of the adult organs do not 
appear in development until a certain stage is reached and the 
organ possesses a certain structure. Apparently then, develop- 
ment up to the stage where function in the above sense begins is 
a process of machine-building. Roux's distinction between a 
formative and a functional stage in development expresses this 
idea. 
But how is the machine constructed and what is the agency 
which constructs it ? The material of which the machine is to be 
made must be acted upon, arranged, transformed, localized, dif- 
ferentiated, etc. Evidently this point of view necessitates the 
assumption of a special formative agent or agents of some kind. 
How shall we conceive this formative agent ? Pangenes, deter- 
minants, formative substances, entelechies are some of the answers 
which have been given to this question. All these answers are 
much alike in that they regard the construction of the organism 
as a process analogous in some sense to the construction of a 
1 Cf. Child, '06^, p. 402, '07. 
CHANGE OF SHAPE IN PLANARIA. 289 
machine by human agency. When the machine is constructed it 
begins to function. 
This point of view seems to me to be essentially naive and 
anthropomorphic : moreover it is responsible for certain " Schein- 
probleme " which have arisen from time to time, and for certain 
barren lines of research, particularly in zoology, which has been 
dominated by hypotheses of this general character to a much 
greater extent than botany. 
To my mind life is inconceivable without some sort of function 
in some sort of structure ; a living structure functions as long as 
it is alive ; its function is its life. If this view is correct, the 
organism is functioning in some manner at all stages of its exist- 
ence, from the earliest to the latest. Function in this sense is 
physiological activity of all kinds, all transformation and transfer- 
ence of energy. 
It is a well established fact that the special functional activity 
of organs is very generally a factor in their development and 
differentiation beyond a certain stage, and in the maintenance of 
their characteristic form and structure after development is com- 
pleted. In the absence of this functional activity the organ does 
not develop beyond a certain point and does not persist indefi- 
nitely. Functional hypertrophy, atrophy from disuse, functional 
adaptation, etc., are terms used to designate these relations between 
function and form. 
But the real difference between earlier and later stages of 
development consists, it seems to me, not in the absence of func- 
tion at one stage and its presence at another, but rather in the 
difference in character of function in the different stages. If a 
particular kind of function determines the structure and differentia- 
tion at one stage, and we know that it does this, is there not a 
logical basis for the belief that the functions which exist in other 
stages are also formative factors in those stages ? 
In other words, is there not good reason to believe that every 
stage of development is directly or indirectly the necessary con- 
sequence of the functional activity in the preceding stage ? Ac- 
cording to this view each stage of development is a machine in 
the broad sense and each is the product of the activity of a pre- 
existing machine. External factors play a part, particularly in 
2QO C. M. CHILD. 
plants and in the simpler animals, in determining the character 
of the machine and its activity, but this fact does not essentially 
alter the case. 
Viewed from this standpoint, development, from its earliest 
stages on, is just as strictly a functional process as functional 
differentiation or functional hypertrophy in the stricter sense. 
This view does not necessitate the assumption of any special 
formative factors different in character from functional processes 
as the factors behind or underlying development. The forma- 
tive factors of each stage are the functional activities of the 
immediately preceding stage (plus external factors). The proc- 
ess of development of the organism is not essentially different 
from the process of maintenance after development. Indeed 
strictly speaking, development ceases only when death occurs. 
The germ cell is an organism possessing a certain structure and 
function, and this forms the starting point. The functional acti- 
vity in this structure determines the next stage of development, 
i. e., a change in structure and therefore a change in function. 
This process continues and at the same time gradually approaches 
a condition of physiological equilibrium. 
This view of development is of course no more an " explana- 
tion " than is the assumption of an entelechy or that of determi- 
nants. As a point of view, however, and as a basis for attacking 
the problem of morphogenesis it possesses a certain value in that 
it does away with various assumptions and places the problem 
of morphogenesis on a strictly physiological basis. To say that 
all organic form and structure are functional in their origin is 
merely to say that the problem of morphogenesis is a physio- 
logical problem and nothing more. 
This is the position which I have held since the beginning of 
my work on regulation. I have used the word "function " with 
reference to any and all physiological processes and activities at 
any and all stages of development and have repeatedly called 
attention to this fact. 1 
1 It is somewhat surprising, therefore, to find Driesch ('05, p. 790; '07, p. 180) 
pointing out that I have put the cart before the horse in regarding function as a de- 
termining factor in form regulation, since, as he says, an organ develops "for func- 
tion" and does not function until it is developed, or at least until a certain stage of 
development has been reached. The difference between Driesch and myself on this 
CHANGE OF SHAPE IN PLANARIA. 29 1 
When we come to consider the process of development some- 
what more in detail we find that two complex groups of internal 
factors must be considered, viz., constitution and correlation. 
This, of course, is true of any machine : its function depends upon 
the constitution and correlations of its parts, provided, of course, 
that the necessary external conditions for function are present. 
In different organisms and in different features of development 
the two factors may of course play a very different part. 
2. Functional Regitlation and Form Regulation. 
It follows from what has been said that all form regulation is, 
according to my definition, either directly or indirectly functional 
regulation, /. t\, the physiological processess in the structure 
involved determine what the result of regulation shall be (Child, 
'064 
In certain cases among the turbellaria I have been able to con- 
trol, inhibit and modify the process of form regulation to a greater 
or less extent by controlling and modifying, in most cases indi- 
rectly, the movement and use of the parts most intimately involved 
in the regulatory processes (Child, '02, '03, 'o^a-c, 'o$a-cT). 
This work showed very clearly that movement and use of parts 
were important factors in certain cases and certain features of reg- 
ulation : they may even be primary factors in determining certain 
results. It is just as certain, however, that in many other cases 
they play no part at all. 1 
point is in reality merely one of definition ; he limits the term " function " to the proc- 
esses of which I spoke in the first paragraphs of this section, while I regard all proc- 
esses in the organism as functions. His criticism of my position is of course entirely 
beside the point since he substitutes his definition for mine. Physiological processes 
of some kind certainly occur even in the earliest stages of development, and that these 
are functions of an existing structure cannot be doubted. I believe that they are also 
the formative factors in development. 
Morgan ('07) calls attention to the fact that the functional idea is not new. This 
of course is true ; it is the position which every physiologist must hold in one form or 
another. I have never considered that I was formulating a new hypothesis of de- 
velopment ; I have merely attempted to apply certain physiological ideas to the phe- 
nomena of form regulation. Among botanists this view has been very generally held 
for a long time. 
1 Morgan ('07, p. 374) apparently believes that I regard movement as a universal 
factor in form regulation, for he calls attention to the fact that movement does not 
occur in many cases of regeneration. But I have repeatedly asserted that my con- 
clusions upon the effect of movement concerned only certain species and certain fea- 
tures of regulation (Child, '02, p. 218, p. 229; 'o^a, p. 131 ; '04/5, p. 467, etc.). 
292 C. M. CHILD. 
As I pointed out concerning Lcptoplana (Child, '04^) move- 
ment is undoubtedly not a factor of importance in determining the 
formation of new tissue at a cut surface, but experimental evidence 
indicates clearly enough that it is a factor in determining the 
rapidity, amount and direction of growth and to a greater or less 
extent the character of its differentiation. 
In the experiments on Planaria described above, the head and 
pharynx, for example, do not attain their " normal " shape and 
structure when movement is largely inhibited by anaesthetics. 
By the use of the higher concentrations it is possible to inhibit in 
almost any desired degree the process of form regulation so far 
as visible morphogenesis is concerned. Pieces which in water 
form heads very rapidly may be prevented entirely from forming 
heads by the proper use of the anaesthetic, or almost any inter- 
mediate condition between these two extremes may be attained by 
the formation of incomplete or partial heads. Pieces incapable 
of forming heads in the anaesthetic, regain their original power 
when returned to water. But it does not in the least follow from 
these facts that head formation is entirely the result of actual 
movement. My position is, and has been, that movement is 
merely one of the functional activities concerned in development, 
and which usually concerns later stages to a greater extent than 
earlier. But in regeneration in the turbellaria and in various, 
though by no means all, other forms a peculiar condition exists 
in that new tissue adjoins fully developed parts which may be in 
active movement. There can be no doubt that movement of these 
adjoining old parts influences conditions in the new tissue and so 
affects the result, either quantitatively or qualitatively or both. 
But movement and functional activity in the stricter sense, in 
which Driesch uses the word, are certainly far from being univer- 
sal factors in either form regulation or ontogeny. 
Whatever the importance of movement in a particular case, it 
is merely one of a great variety of functional factors. In many 
cases movement and motor use of parts are not concerned at all 
in regulation and in some other cases it is evident that while 
movement may affect the later stages this movement is possibly 
only in consequence of preceding regulation. In Cestoplaua 
(Child, '05/7), for example, where the posterior portion of an an- 
CHANGE OF SHAPE IN PLANARIA. 293 
terior piece becomes visibly "posterior" as regards the character 
of its movements before any marked structural changes occur, 
regulation must have occurred before such changes in movement 
could appear. 1 
3. Mechanical Regulation and Morphallaxis? 
In the first of the " Studies " the idea of mechanical regulation 
as the chief factor in the change of shape in pieces of Stenostoina was 
developed. According to my conception mechanical regulation is 
primarily a mechanical deformation of physically plastic tissues. 
Whether we consider regulation as a return or approach to the 
normal condition after a disturbance of this condition (Driesch, 
'01), or as a return or approach to a condition of physiological 
equilibrium after a preceding condition of equilibrium has been 
disturbed (my own definition), mechanical regulation in my sense 
1 In my paper I referred to these changesas functional regulation and called atten- 
tion to the altered character of the movement as indicating that they had occurred, 
but while the movement is undoubtedly a factor in what follows, it would be absurd 
to suppose that it is the primary factor in such a case. 
- Some years ago, in describing the course of regulation in Planaria macttlata, 
Morgan called attention to the peculiar changes in shape which the pieces undergo, 
these changes consisting mainly in a decrease in width and an increase in length. 
Concerning this process he says : " Thus the relative proportions of the planarian are 
attained by a remodelling of the old tissue. I would suggest that this process of 
transformation be called a process of morphallaxis " (Morgan, '98, p. 385). Later 
('oo) he applied the same term to the changes in shape and proportion in Hydra and 
other forms. So far as I am aware, Morgan has not stated at any time whether the 
term "morphallaxis" is to be applied to the whole process of form regulation in 
Planaria and other forms, including the regeneration and redifterentiation which 
occurs, or only to the changes in shape and proportions of the piece. From the quo- 
tation given above it would appear that he intended it to apply only to the change in 
shape and proportions, but in his latest statement ('07, p. 15) he apparently applies 
the term to the whole process of form regulation by redifterentiation. In my earlier 
"Studies " I used the term with reference only to the changes in proportion and 
shape : later, however, I substituted "change in proportions " for it as less ambigu- 
ous for my purposes (Child, '050, p. 253). Driesch ('oi) considered the term as 
synonymous with " Restitution durch Umdifferenzierung," a very different meaning 
from that which I had given it in my work. It is probable that Driesch's misunder- 
standing of my conclusions concerning form regulation is in part due to our different 
interpretations of this term. 
In several of my earlier papers the word "form" was used for " shape" and 
" outline." I agree with Driesch that this use of the word may be misleading, but I 
was careful to distinguish between form in this sense and structure, and pointed out 
that the factors concerned with change in form probably had nothing to do in many 
cases with change in structure (Child, '02, p. 218). 
2Q4 C. M. CHILD. 
is theoretically possible, provided certain characteristic mechan- 
ical conditions exist in the normal animal and are altered in char- 
acteristic manner in the piece. In several of the "Studies" the 
changes in shape in pieces of Stenostoma, Leptoplaua and Cesto- 
plana were shown to be apparently primarily mechanical regula- 
tions (Child, 02, '04*7, 'O5<2, '05^) and it was suggested that 
similar processes might occur elsewhere. The recognition of 
mechanical regulation is not in any sense an attempt to interpret 
regulation in general on a mechanical basis (cf. Child, '02, p. 
229) : it concerns merely changes in shape and outline. Me- 
chanical regulations are possible only under the conditions men- 
tioned above, and whether they occur or not can only be deter- 
mined by investigation of each particular case. It is not necessary 
to suppose that all cases of " morphallaxis " (in the sense of change 
in shape and proportions) are mechanical regulations. If they are 
they certainly depend on different mechanical conditions in dif- 
ferent cases. In Hydra, for example, the factors determining the 
change in shape are certainly not the same as in Planaria, and 
may not be mechanical at all. In the cases which I have con- 
sidered mechanical strains arise in connection with locomotion 
and these strains are altered in a characteristic manner in iso- 
lated pieces and, as I have shown, the changes in shape are 
exactly what might be expected in physically plastic material 
under these conditions. Mechanical regulation is to be expected 
only in organisms or parts possessing a considerable degree of 
physical plasticity and where skeletal structures are not con- 
cerned. As regards the shape of animals in general gross 
mechanical factors are certainly unimportant as compared with 
others, or not concerned at all, and I have never even suggested 
that such factors were of general significance (cf. Driesch, '05, p. 
790). On the other hand, it seems to me absurd to deny that 
characteristic strains or pressures existing in plastic material may 
play some part in determining shape. 
That the changes in shape in pieces of Stenostoma, Leptoplana, 
Ccstoplana, and Planaria are not primarily " functional adapta- 
tions " (cf. Driesch, '05, p. 766) is, I think, evident. A func- 
tional adaptation, as I understand it, is a change in structure 
which involves a change in functional capacity, either an increase 
CHANGE OF SHAPE IN PLANARIA. 2Q5 
or a decrease or a modification in kind according to the character 
of the factor inducing the change. The change in shape in the 
pieces of P/anaria, etc., does not in itself involve any such change 
in functional capacity, moreover, functional adaptations to me- 
chanical tension and pressure consist, usually if not always, in an 
altered resistance to the mechanical factor, not in a change of 
shape which accords with the laws of mechanics. The change 
in shape in these turbellaria is essentially similar to what would 
occur under the same conditions in any material of similar phys- 
ical constitution and there is not the slightest evidence that it 
results from a change in functional capacity. 
There is little doubt, however, that functional adaptations 
result from the change of shape : as was suggested above, it is 
probable that the cessation of the change and the final attain- 
ment of a more or less characteristic shape, i. e., a condition of 
equilibrium, is in part the consequence of the functional reaction 
of the differentiating and redifferentiating tissues to the altered 
mechanical conditions. In other words, the change of shape 
brings about the functional adaptation instead of resulting from it. 
But even though these changes in shape do not appear to be 
functional adaptations in Driesch's sense, they are functional 
regulations in my sense, and probably the simplest possible kind 
of functional regulation. 
HULL ZOOLOGICAL LABORATORY, 
UNIVERSITY OF CHICAGO, 
March, 1909. 
BIBLIOGRAPHY. 
Child, C. M. 
'02 Studies on Regulation. I., Fission and Regulation in Stenostoma. Arch. 
f. Entwickelungsmech., Bd. XV., H. 2 and 3, 1902. 
'03 Studies, etc. II., Experimental Control of Form- Regulation in Zoolds and 
Pieces of Stenostoma. Arch. f. Entwickelungsmech., Bd. XV., H. 4, 
1903. 
'o4a Studies, etc. IV., Some Experimental Modifications of Form-Regulation 
in Leptoplana. Journ. Exp. Zool., Vol. I., No. i, 1904. 
'o4b Studies, etc. V., The Relation between the Central Nervous System and 
Regeneration in Leptoplana : Posterior Regeneration. Journ. Exp. Zool., 
Vol. I., No. 3, 1904. 
'040 Studies, etc. VI., The Relation, etc. Anterior and Lateral Regeneration. 
Journ. Exp. Zool., Vol. I., No. 4, 1904. 
Studies, etc. VII., Further Experiments on Form-Regulation in Lepto- 
plana. Journ. Exp. Zool., Vol. II., No. 2, 1905. 
296 C. M. CHILD. 
Studies, etc. VIII., Functional Regulation and Regeneration in Cestoplana. 
Arch. f. Entwickelungsmech., Bd. XIX., II. 3, 1905. 
'050 Studies, etc. IX., The Positions and Proportions of Parts During Regula- 
tion in Cestoplana in the Presence of the Cephalic Ganglia. Arch. f. Ent- 
wickelungsmech., Bd. XX., II. I, 1905. 
'osd Studies, etc. X., The Positions and Proportions of Parts During Regula- 
tion in Cestoplana in the Absence of the Cephalic Ganglia. Arch. f. Ent- 
wickelungsmech., Bd. XX., H. 2, 1905. 
'o6a Contributions Toward a Theory of Regulation. I., The Significance of the 
Different Methods of Regulation in Turbellaria. Arch. f. Entwickelungs- 
mech., Bd. XX., H. 3, 1906. 
'o6b The Relation Between Regulation and Fission in Planaria. Biol. Bull., 
Vol. XL, No. 3, 1906. 
'o6c The Relation Between Functional Regulation and Form Regulation. Journ. 
Exp. Zool., Vol. III., No. 4, 1906. 
'07 Some Corrections and Criticisms. Arch. f. Entwickelungsmech., Bd. 
XXIV., H. i, 1907. 
Driesch, H. 
'01 Die Organischen Regulationen. Leipzig, 1901. 
'05 Die Entwickelungsphysiologie von 1902-1905. Ergebnisse der Anat. und 
Entwickelungsges., Bd. XIV. (1904), 1905. 
'08 The Science and Philosophy of the Organism, Vol. I., London, 1908. 
Morgan, T. H. 
'98 Experimental Studies of the Regeneration of Planaria maculata. Arch. f. 
Entwickelungsmech., Bd. VII. , H. 2-3, 1898. 
'oo Regeneration in Planarians. Arch. f. Entwickelungsmech., Bd. X., H. i, 
1900. 
'07 Regeneration. Uebersetzt von M. Moszkowski. Leipzig, 1907. 
Steinmann, P. 
'08 Untersuchungen iiber das Verhalten des Verdauungssystetns bei der Regen- 
eration der Tricladen. Arch. f. Entwickelungsmech., Bd. XXV., II. 3, 
1908. 
Stevens, N. M. 
'07 A Histological Study of Regeneration in Planaria simplicissima, Planaria 
iraculata and Planaria morgani. Arch. f. Entwickelungsmech., Bd. XXIV., 
H. 2, 1907. 
Woodworth, W. McM. 
'97 Contributions to the Morphology of the Turbellaria. II., On Some Turbel- 
laria from Illinois. Bull. Mus. Comp. Zool. Harvard Univ., Vol. XXXI., 
No. i, 1897. 
SOME ABNORMALITIES AND REGENERATION 
OF PLEIOPODS IN CAMBARUS AND 
OTHER DECAPODA. 1 
CHARLES ALBERT SHULL. 
INTRODUCTION. 
The two crayfishes whose abnormal appendages are described 
in this paper were found among a small number of specimens of 
Cauibarus virilis Hagen, which were captured near the city of 
Chicago in the autumn of 1904. The discovery of these abnormal 
pleiopods led to an experimental study of the regeneration of the 
abdominal appendages in the crayfish, the results of which are 
given herein. The earlier experiments were unsuccessful owing 
to imperfect methods of handling the material, and to the use o 
mature animals in which the moults were infrequent. The later 
experiments, with improved methods of caring for the material, 
and with small immature specimens, have yielded satisfactory 
results. 
These experiments were begun in the Hull Zoological Labora- 
tory of the University of Chicago, continued in the Biological 
Laboratory of Transylvania University, and completed in the 
Marine Biological Laboratory of the Brooklyn Institute of Arts 
and Sciences at Cold Spring Harbor, Long Island. I wish to 
express my thanks to Dr. Davenport, director of the Marine 
Laboratory, who kindly placed the facilities necessary for the 
completion of the work at my disposal ; also to Dr. A. E. Ort- 
mann, of the Carnegie Museum, Pittsburgh, to whom I am in- 
debted for identification of the species used in the experimental 
work ; and to Dr. Faxon, of Harvard, and to Dr. Hans Przibram, 
of Vienna, for examining the abnormal specimens. 
HISTORICAL SUMMARY. 
The regeneration of the abdominal appendages of decapod crus- 
taceans has been a subject of investigation for many years, but 
'Contributions from the Biological Laboratory of Transylvania University, No. 2. 
297 
298 CHARLES ALBERT SHULL. 
not until recently has a critical study of it been attempted. A 
review of the earlier literature relating to regeneration in the 
Decapoda has been given by Miss Steele, '04 ; it is necessary, 
therefore, to mention only a few of the later papers upon whose 
results my own observations have some bearing. 
Morgan, '98, desiring to test Weismann's hypothesis, that re- 
generation is an adaptation, experimented with Eupagnrus longi- 
carpns to determine what relation, if any, existed between the 
power of regeneration and liability to injury. Because the her- 
mit crab lives in shells where the abdominal appendages are pro- 
tected from injury, whereas the thoracic appendages are not so 
protected, these two series of organs were chosen for this experi- 
ment. He came to the general conclusion that no such relation 
existed, and further supported this view by a second paper in 
1900. The latter paper was based upon experiments in which 
the thoracic appendages were removed at unusual levels. 
He found in the course of his experiments that the ab- 
dominal appendages of Eupagurns did not regenerate readily, 
although slight regenerative power existed. He suggested that 
this rarity of regeneration "may be connected in some way with 
the amount of food supply brought to the region from which 
they arise." 
Miss Steele, '04, experimenting with Cambarus virilis and C. 
gracilis has obtained results similar to Morgan's. She says, 
speaking of the swimmerets : "I have found none to regenerate 
except the first pair in the male. ... In the case of the other 
abdominal appendages except the sixth pair, regeneration, if it 
does take place, is very slow in beginning." 
Emmel, '04, reported the observation of regeneration in the 
first appendages of the lobster. He experimented with the other 
appendages and says : " In experiments with the other four pairs 
of abdominal appendages or swimmerets, positive results were 
obtained in the second and third pairs, and it seems safe to say 
that all the swimmerets will regenerate." 
Haseman, '07, mentions the regeneration of swimmerets in C. 
propinquus, giving figures showing the progress of differentiation, 
but he does not mention the fact that the regeneration of these 
appendages was considered doubtful. 
REGENERATION IN CAMBARUS. 299 
A discussion of the results obtained by Morgan and Miss 
Steele will be found in connection with the discussion of my own 
observations in a later portion of this paper. 
ABNORMAL SWIMMERETS. 
Anomalous variations not of congenital origin are of little in- 
terest to the student of evolution ; for it has never been definitely 
demonstrated that any acquired somatic variation is inherited. 
Moreover, no very important laws of variation are likely to be 
discovered by a study of teratological specimens. Nevertheless, 
it is advisable to place interesting cases of abnormal growth and 
regeneration on record, for we may finally acquire a knowledge 
of the physiology and mechanics of development sufficient at 
least to explain the origin of such anomalous variations. 
The swimmeret shown in Fig. I is the right appendage of the 
second abdominal somite of a female C. I'irilis. The pleuron is 
much deformed, as is readily seen if compared with the normal 
pleuron of the opposite side of the same somite. The coxopo- 
dite of the abnormal appendage is much larger than normal. 
From this enlarged basal piece arise two appendages, one of 
which, near the posterior side of the coxopod, seems on casual 
observation to be a beautiful case of duplicity. The parts are 
perfectly doubled from the proximal portion of the basipodite 
outward. The two pleiopods thus united are normal in every 
particular beyond the point of union, the posterior member being 
somewhat larger than the anterior. From the anterior portion 
of the large coxopodite, completely separated in point of origin 
from the double pleiopod, arises another appendage which is 
uniramous, but of a size and structure typical of the endopodite 
of the second pleiopod. 
It is an interesting circumstance that the first pleiopod lying 
immediately anterior to the one just described is also abnormal. 
This swimmeret is shown in Fig. 2. The lower portion of the 
appendage is much enlarged, and at the proximal end of the 
basipodite, there is a posterior projection set with stiff hairs, 
reminding one somewhat of the structure of the coxopodite of 
the pereiopods. 
The third abnormal swimmeret is from a male C, virilis which 
3OO CHARLES ALBERT SHULL. 
was found to have three pairs of appendages instead of two, mod- 
ified for sexual purposes. . The right swimmeret of the pair is 
shown in Fig. 3, the left one being exactly like it. The charac- 
ter of the modification in this third pair is just the same as that of 
the second, except that the projection of the endopodite is smaller 
than on the second. The first and second pairs of appendages 
are perfectly normal. Moenkhaus, '03, has reported an exactly 
similar case in the same species, and so far as I know these are 
the only two cases on record. 
THE PROBLEM, MATERIALS, AND METHODS. 
Being convinced that the abnormal swimmerets of the female 
C. virilis described herein were the result of regeneration after a 
somewhat extensive injury, and having before me the work of 
Morgan on Eupagurus longicarpus, and of Miss Steele on C. virilis 
and C. gracilis, I determined to perform a series of experiments 
with the crayfish in order to ascertain whether or not the ab- 
dominal swimmerets regenerate, and if so, under what conditions. 
These experiments were begun in 1905 ; and after two years of 
rather unsuccessful work the choice of material and the methods 
of handling it were so improved that gratifying results have been 
obtained. 
At first attempts were made to keep the crayfishes in aquaria 
of running water. A large aquarium was divided by partitions 
of galvanized iron netting into a number of compartments, each 
about 30 cm. square. Each compartment was provided with 
gravel and flat stones, and fresh water from Lake Michigan was 
kept running constantly at a depth of about 5 cm. 
The aquarium was cleaned frequently, and all precautions were 
taken to keep the animals in sanitary conditions ; but in spite of 
the great care exercised, the crayfishes would die after several 
weeks of confinement. Since my first material was adult, and 
moulted therefore infrequently, death took place before any 
moults occurred. Many variations of these conditions were tried 
in an attempt to secure more favorable results, but without 
success. 
During the last two years a method has been employed, which 
has obviated all difficulties, and has given entirely satisfactory 
REGENERATION IN CAMBARUS. 30 1 
results. Young specimens of Cambarus (Bartonius) bartoni Fabr., 
probably at the beginning of their second and third years were 
taken on March 24, 1907, from a small stream which crosses the 
Bryan Station pike about three fourths of a mile northeast from 
the city limits of Lexington, Ky. The specimens used are not 
like the typical C. bartoni of the eastern United States, but have 
a narrower areola, less spiny carpus, and a shorter but broader ros- 
trum than the eastern form. Dr. Ortmann states that they agree 
with specimens described by W. P. Hay in the 2Oth Ann. Report 
of the Ind. Geol. Survey, 1895, after comparing the living, speci- 
mens with those in the Carnegie Museum from Mitchell, Lawrence 
Co., Ind. 
Each specimen was put into a tumbler with water not quite 
sufficient in amount to cover it. A small piece of mica schist 
coming not quite to the surface of the water, was placed in each 
tumbler so that the crayfishes could crawl upon it and expose 
themselves to the air at will. The tumblers were kept nearly 
covered by glass plates to prevent accidental desiccation. Care 
was taken not to cover the dishes entirely, as the air was found 
to become overcharged with carbon dioxide in a short time if so 
covered. The water was changed completely three or four times 
per day, and even oftener on very hot days ; for crayfishes seem 
unable to endure warm water for any great length of time. 
The water used in these experiments was supplied by the city 
water works of Lexington. This water is filtered through large 
Jewell filters before being pumped into the water mains, and is 
exceptionally pure. The complete change of such water every 
few hours rendered the multiplication of bacteria or growth of 
other fungous plants rather difficult, and tended to prevent the 
accumulation of any sediment. 
When it appeared that any fungous growth was forming upon 
the appendages and about the thoracic region, each crayfish was 
treated for a few minutes with a bath of copper sulphate. The 
vessels and stones kept in them were occasionally treated with the 
same solution. The copper sulphate was used with a strength of 
one part to a million, when the animals could be left in it for 
some time. Solutions much stronger than this, one part in ten 
thousand, can be used for a few minutes at a time with entire sue- 
3O2 CHARLES ALBERT SHULL. 
cess in destroying bacteria and fungi, and without injury to the 
crayfishes, provided they are washed thoroughly several times in 
plenty of pure water after the bath to cleanse their gills of the 
copper sulphate. 
Young crayfishes are voracious creatures, and need to be fed 
frequently. I have fed them every day with entire success, but 
they soon tire of a uniform diet. Several kinds of food were used ; 
and by rotation, so that they never received the same kind of food 
on successive days, their appetites were retained. Fresh raw beef 
was given them twice a week, and raw potatoes and pieces of Myr- 
iophyllum, which was kept growing in the laboratory, were used 
as vegetable foods. The Myriophyttum was covered with slime, 
which was found to contain unicellular algne, rotifers, nematode 
worms, annelids, such as sEolosoina kcmprickii, and other kinds of 
small animals. This slimy material was especially esteemed, and 
probably most nearly represents their normal diet. This food 
material was left in the glass with the animals for an hour or two, 
after which the remains were removed and fresh water placed in 
the vessels. 
Using these methods, I have kept them alive for months in 
perfect health, with rapid growth during the moulting season. 
Occasionally a death would occur among them, but these fatalities 
were always obviously due to special causes, not to any general 
defects in the methods employed. 
After ecdysis the cast off exoskeletons were used in the prepa- 
ration of the drawings. 
EXPERIMENTS. 
The experiments here described are a few typical ones selected 
from a series, all of which gave similar results. The numbers 
correspond to those used while recording notes on individual 
experiments. 
No. 4. C. (Bartonius) bartoni, ?. Fifth right abdominal ap- 
pendage removed March 24, 1907. The appendage was cut off, 
leaving a short stump attached to the body (Fig. 4). 
The first moult occurred in the afternoon of March 27, three 
days after the operation. No regeneration could be noticed, but 
the wound was perfectly healed (Fig. 5). The second moult 
REGENERATION IN CAMBARUS. 303 
occurred during the night following May 6, 1907. Between 
March 27 and May 6 no change could be detected even by use 
of the microscope. But that changes were taking place beneath 
the chitinous cuticle is evident from the expansion of the regen- 
erating limb, which occurred immediately after this second 
ecdysis (see Fig. 6). 
A third moult took place on June 19, 1907, at which time 
the appendage again expanded, reaching the size and condition 
shown in Fig. 7. The regenerated appendage measured at this 
time 3.3 mm. in length ; the one on the opposite side of the 
same somite 4.8 mm. (Fig. 8). The regeneration was equal to 
nearly 69 per cent, of the normal size from March 24 to June 
19, a rate which I consider by no means slow. 
Owing to an unfortunate accident this crayfish was killed on 
the morning of July 5, 1907. Had it lived until after its fourth 
moult it would probably have shown a much more nearly com- 
plete regeneration of the appendage. 
No. 7. C. (Bartonius) bartoni, ?. Fourth left pleiopod 
removed March 24. The first moult after the operation took 
place the night of April 22, 1907. The amount of regeneration 
while not extensive was plainly visible. The second moult 
occurred May 25, 1907. The regenerating appendage expanded 
to nearly half the natural size, immediately after the exoskeleton 
was cast (Fig. 9). 
No. 3. The same as No. 7 except that the fourth pleiopod 
was removed from the right side instead of the left. The first 
moult, on March 30, showed a small whitish papilla a little over 
one mm. in length, projecting from the base of the amputated 
limb. This projection increased slightly with age, but this indi- 
vidual was killed in July by the water in the vessel receiving 
direct sun-light through a partially open window blind, thus 
becoming much overheated. Fig. 10 reveals internal develop- 
ment. A later moult would probably have given results similar 
to No. 7. This specimen was probably a year, or perhaps even 
two years, older than No. 7, and consequently only one moult 
was recorded between March 24 and July 5, while some of the 
smaller, younger specimens moulted three times in a shorter 
period (cf. No. 4). 
304 CHARLES ALBERT SHULL. 
No. 2. C. (Bartoniits] bartoni, d\ The third right appendage 
was cut off March 24. On May 9, no moult having occurred, 
the stump of the appendage was cut longitudinally by means of 
sterilized scissors. The moult occurred on June 17, but no 
abnormal growth was noted. The appendage was perfect, and 
showed regeneration of more than 50 per cent, in size (Fig. 1 1). 
No. 5. C. (Bartonius] bartoni, 9. The second left appendage 
was removed March 24, 1907. The first moult occurred on 
May 8, 1907, and the regeneration amounted to about 30 per 
cent. The second moult took place on June 26, 1907, when the 
appendage showed about 65 per cent, of complete regeneration. 
Fig. 1 2 shows the point of amputation, and Fig. 13 the appendage 
after the second moult. 
No. 6. C. (Bartonins] bartoni, d\ Sixth right appendage was 
amputated just beyond the basal joint March 27, 1907, as shown 
in Fig. 14. It moulted the following day. A very narrow edge 
of white tissue was visible along the cut edges, but it is not 
probable that any regeneration had taken place. During the 
time preceding the next moult which occurred on May 27, the 
basal portions of the rami increased somewhat in size (Fig. 15), 
and after moulting the appendage was about one half natural size, 
and perfectly formed (Fig. 16). 
A series of experiments on the pleiopods of Paltzmonetcs vnl- 
garis was carried on at Cold Spring Harbor, and regeneration of 
all the abdominal swimmerets takes place rapidly in young 
specimens. 
At present a series of experiments to test the regeneration of 
the antennse from various levels and to compare the rate of 
regeneration of pleiopods and pereiopods in C. (Bartonius] bartoni 
is being carried on ; but the data are insufficient as yet to permit 
a general statement regarding either phase of the series. 
DISCUSSION. 
A. The Abnormalities : As far as I have been able to ascer- 
tain, only one abnormal abdominal appendage has been recorded 
for any of the decapod crustaceans. In as much as all abnormal 
pleiopods described herein or elsewhere have been discovered 
accidentally, the rarity of such records is due in part, perhaps, to 
REGENERATION IN CAMBARUS. 305 
lack of careful observation ; but the examinations which have 
been made show that they occur infrequently. 
The abnormalities presented here are of interest for several 
reasons. The abnormal pleiopod is evidently the result of a 
regeneration after extensive injury in one instance, and just as 
evidently the result of undisturbed growth processes in the other 
instance. That is, the condition is congenital and may be con- 
sidered as having arisen by mutation. 
It would appear that the swimmerets shown in Figs, i and 2 
probably resulted from an injury which removed part of the 
pleuron, and tore the first and second pleiopods near the base of 
each, as represented diagrammatically in Fig. 17. The coxop- 
odite in Fig. I is abnormally large, which condition may pos- 
sibly be explained by the fact that a large area was exposed 
when the mutilation occurred. 
Very few genuine cases of duplicity have been described. 
Bateson mentions only four cases, all chelae ; Herrick has figured 
a double chela of a female lobster, and Zeleny has described a 
double chela of Gelasiinns pugilator which regenerated instead of 
a normal single one during the course of his experiments. 
Bateson states emphatically that " in arthropods and verte- 
brates such a phenomenon as the representation of one of the 
appendages by two identical appendages standing in succession 
is unknown. No right arm is ever succeeded on the same side 
of the body by another arm properly formed as right, and no 
crustacean has two right legs in succession where one should be." 
While this supernumerary appendage may be regarded as a 
complementary image of the normal one, and therefore a left 
appendage instead of a right, there is nothing in the structure of 
either member to indicate that such a relation exists. The 
members are not imperfect, and are placed in succession, differ- 
ing in these two respects from the cases admitted by Bateson. 
At first appearance the coxopod of the abnormal first append- 
age is very much like that of the last pereiopod. And the an- 
terior uniramous part of the second pleiopod is like the first 
pleiopod in being uniramous, but is a typical endopodite. There 
is a mere suggestion here of a shifting backward of the series of 
organs, a condition to which Bateson applied the term backward 
homceosis. 
306 CHARLES ALBERT SHULL. 
However if the injury actually took place as indicated in the 
diagram referred to above, the most plausible explanation which 
can be offered is in harmony with Bateson's statement. The 
regeneration should lead theoretically to triplication, that is, to 
the production of three appendages, two supernumerary ones on 
each cut appendage ; and these on the right side of the body 
should stand as a left between two rights. The abnormal 
pleiopod shows defective regeneration only in the suppression of 
the exopodite of the anterior supernumerary appendage. The 
rounded projection of the first accessory pleiopod may be ex- 
plained similarly as the fused basal portions of the two super- 
numeraries, whose tapering jointed ends have been completely 
suppressed. 
The abnormalities have probably arisen after an injury caused 
by some force acting in the direction of the arrow in the diagram, 
Fig. 17, which produced two breaking surfaces, from which the 
new pleiopods arose, following the laws of symmetry for super- 
numerary appendages as stated by Bateson. 
The abnormal appendage shown in Fig. 3, is a case of back- 
ward homoeosis. It would be of interest to know the hereditary 
behavior of such unisexual characters if bred in confinement. 
B. Experiments. - - Few investigations of the regenerative power 
of the abdominal appendages of the Decapoda have been made, 
and the results obtained have not been entirely satisfactory. 
Morgan's experiments on Eupagunts longicarpus in 1898 showed 
that a slight power of regeneration existed in the appendages of 
two or three of the individuals he used ; but his experiments ex- 
tended over too brief a period of time to secure any marked 
regeneration. The first experiment was continued only twenty 
days, and the second for twenty-eight days. And the conditions 
under which the material was kept were possibly not the most 
sanitary, as the fact that over 40 per cent, of the individuals died 
during the twenty-eight days would seem to indicate. One 
would hardly expect regeneration to be rapid under conditions in 
which life itself could scarcely be maintained. My experience 
has shown that although regeneration may occur after an opera- 
tion, and become visible without an intervening moult, it usually 
does not do so in the pleiopods. None of the individuals Morgan 
used was kept until a moult had occurred. 
REGENERATION IN CAMBARUS. 307 
Miss Steele's experiments on C. virilis and gracilis were carried 
on for a sufficient length of time, but she found it difficult to pro- 
vide the sanitary conditions necessary to such prolonged experi- 
mentation. The specimens used in her work were probably too 
old to give the best results. They measured not more than three 
inches in length, a size which would indicate that they were 
several years old at least. The younger the specimens used, the 
more frequently the moults take place. Those used in my ex- 
periments measured from one and six tenths to two and five 
tenths cm. in length, and were probably at the beginning of the 
second or third year. Two individuals measuring five cm. in 
length were also used. These did not yield as satisfactory 
results although some regeneration occurred. The difference 
was probably more in the frequency of moulting than in any- 
thing else. While one moult took place in these older ones, I 
could secure three in the younger. And to have kept the older 
individuals until the same number of moults occurred as in the 
younger ones, would have required two years instead of four or 
five months. The use of more nearly adult material, and the 
fact that slight attention was paid to the swimmerets may account 
for the slight regenerative power which Miss Steele was able to 
report. 
My experiments show that the swimmerets of young specimens 
of C. (Bartonius] bartoni regenerate rapidly, and that the regen- 
eration of any of the appendages may be practically completed 
in a single season of growth. The regeneration is not particularly 
slow in beginning, having become visible in one instance only six 
days after the operation was performed. It is usual, however, to 
find that the regeneration begins and takes place under the old 
exoskeleton, without showing any visible indications that the new 
parts are forming. For instance, No. 9 moulted on March 28, 
four days after the operation, and no regeneration could be noticed. 
The next moult took place April 30, but during that time no in- 
dication whatever that regeneration was occurring could be seen. 
Nevertheless, when the moult occurred it was evident that regen- 
eration had taken place. If the experiment had been continued for 
only thirty days, the conclusion naturally drawn would have been 
that regeneration either did not occur, or was " very slow in begin- 
3O8 CHARLES ALBERT SHULL. 
ning." As a matter of fact there is noway to tell how soon after 
March 28 the regeneration did begin, if indeed it had not already 
begun at that time. 
To explain the slight regenerative power which he found in the 
abdominal appendages of Enpagurns, Morgan suggested that the 
food supply of these organs might be considerably less than that 
of the thoracic legs. It seems to me quite unnecessary to make 
this assumption, especially since I have shown that there is a rapid 
and complete regeneration of the swimmerets in young specimens 
of C, (Bartonius) bartoni. In this connection Emmel, '04, ob- 
served that swimmerets in the lobster will regenerate more rapidly 
than the pereiopods if the latter are cut " only a relatively short 
distance below the breaking plane." And he questions whether 
the supply of food material can explain the comparative difference 
in the regeneration of pereiopods and swimmerets. Moreover, 
if the limited food supply is responsible for lack of regenerative 
power how must we regard the regeneration of two supernumer- 
ary appendages in the abnormal swimmeret figured ? It seems 
to me that some other explanation must be offered for the differ- 
ence in regenerative power. 
It has been a common experience with those who experiment 
with regenerative tissues, that the regeneration is always more 
rapid and complete in young individuals than in old ones. This 
fact is due probably to the greater plasticity, the more active 
and mobile condition of young tissues. They are more nearly 
embryonic in character, differentiation is not so complete, nor so 
fixed as in the older tissues. Emmel's ('08) recent work on the 
reversal of asymmetry in the lobster lends emphasis to this state- 
ment. He says : " In the first four stages of the lobster's develop- 
ment, a crusher may be produced on either the right or the left 
side of the body by the autotomous amputation of the chela on 
the opposite side the regenerated chela becoming a nipper. 
During the fifth stage, although the chelae are still symmetrical, 
the possibility for such experimental control disappears." 
Since differentiation is known to proceed at different rates in 
different parts of the body, we may suppose that one part re- 
tains its primitive condition longer than another. If any portion 
of the adult regenerates less rapidly than another portion, may 
REGENERATION IN CAMBARUS. 309 
it not be an index of the comparative plasticity of the two parts? 
The more rapid regeneration of the thoracic appendages than 
of the swimmerets in the adult crayfish, may be due to a longer 
retention of the plastic embryonic condition of the cells in the 
region of the breaking joint by the former, certainly not to any pro- 
portionate difference in the food supply of the two series of organs. 
The important point in Morgan's results which he properly 
emphasized, was that regeneration did take place, however slight 
it might be. While we cannot say that the abdominal append- 
ages of crayfishes are never injured, yet injuries to them are rare 
in nature. Examination of hundreds of individuals has shown 
that there is a remarkable uniformity of size and structure in 
these organs, a condition which could not exist if mutilations 
were frequent. Yet these appendages possess as high a power 
of regeneration in youthful stages as any of the appendages 
which are so frequently torn away. My experiments strengthen 
very much the evidence in favor of Morgan's statement that 
there is no relation between power of regeneration and liability 
to injury. 
CONCLUSIONS. 
The following conclusions may be drawn from the study of 
abnormalities, and the results of the experiments described in 
this paper. 
1 . The abdominal swimmerets of C. (Bartonius) bartoni Fabr. 
all possess a high power of regeneration in immature specimens. 
Since regeneration of swimmerets has been noted in Cambarus 
propingnus, Palcsmonetes vn/garns, Homarus americanns and Eu- 
pagurus longicarpus, it is probable that the decapods generally 
possess this power, especially in young individuals. 
2. This regeneration usually cannot be seen till after one, and 
sometimes two, moults have occurred, due to masking of the 
regeneration by the exoskeleton. 
3. Slow regenerative processes in the swimmerets of.the older 
individuals are due probably to a lower degree of plasticity in the 
protoplasm rather than to insufficiency of the food supply. 
4. Injuries may occur, but they are rare in swimmerets, and 
the power of regeneration and liability of the parts to injury are 
apparently independent. 
3IO CHARLES ALBERT SHULL. 
LITERATURE CITED. 
Bateson, W. 
'94 Materials for the Study of Variation. London, 1894. 
Emmel, V. E. 
'04 The Regeneration of Lost Parts in the Lobster. Thirty-fifth Annual Report 
of the Commissioners of Inland Fisheries of Rhode Island, pp. 81-117, 1904. 
'08 The Experimental Control of Asymmetry at Different Stages in the Develop- 
ment of the Lobster. Jour, of Exp. Zool., pp. 471-484, 1908. 
Haseman, J. D. 
'07 The Direction of Differentiation in Regenerating Crustacean Appendages. 
Archiv fiir Entwicklungsmechanik der Organismen, pp. 617-637, 1907. 
Herrick, F. H. 
'95 The American Lobster. Bull, of U. S. Fish Comm., 1895. 
Hay, W. P. 
'95 The Crawfishes of the State of Indiana. Twentieth Ann. Report of the 
Indiana Dept. of Geol. and Natural Resources, pp. 475-507, 1895. 
Moenkhaus, W. J. 
'03 An Extra Pair of Appendages Modified for Copulatory Purposes in Cambarus 
virilis. Indiana Univ. Bull., Vol. I., No. 4, pp. 17-18, 1903. 
Morgan, T. H. 
'98 Regeneration and Liability to Injury. Zool. Bull., I., 1898. 
'oo P'urther Experiments on the Regeneration of the Appendages of the Hermit 
Crab. Anat. Anz., XVII., 1900. 
Steele, M. I. 
'04 Regeneration of Crayfish Appendages. Univ. of Mo. Studies, Vol. II., 
No. 4, 1904. 
Zeleny, C. 
'05 The Regeneration of a Double Chela in the Fiddler Crab (Gelasimus 
pugilator) in Place of a Normal Single One. Biol. Bull., Vol. IX., pp. 152- 
155. I905- 
OCR text unavailable for this page.312 CHARLES ALBERT SHULL. 
EXPLANATION OF PLATE I. 
All figures except Fig. 17 are X 5- 
FIG. I. Second right pleiopod of C. virilis Hagen, 9 showing two supernumerary 
appendages. 
FIG. 2. First right pleiopod of the same specimen. 
FIG. 3. Third right pleiopod of C. virilis, $ , modified for sexual purposes. 
FIGS. 4-7. Successive steps in the regeneration of the fifth right pleiopod in 6. 
(Bartonius) bartoni, 9- 
FIG. 8. Left pleiopod from same somite as Fig. 7, without regeneration, drawn to 
same scale. 
FIG. 9. Fourth left pleiopod of young C. (Bartonius) bartoni, 9> a ft er two 
moults. 
FIG. 10. Fourth left pleiopod of older C. (Bartonius} bartoni, 9, after three and 
one half months. 
FIG. ii. Third right pleiopod of C. (Bartonius} bartoni, $ , after one moult. 
FIGS. 12-13. Second left pleiopod of C. (Bartonius} bartoni, 9, condition im- 
mediately after amputation, and after the second moult. 
FIGS. 14-16. Right uropod of C. (Barlonius} bartoni, $, conditional amputa- 
tion, before, and after, second ecdysis. 
FIG. 17. Diagram illustrating the way in which the abnormal appendages shown 
in Figs. I and 2 may have been produced. 
BIOLOGICAL BULLETIN, VOL. XVI. CHARLES ALBERT SHULL. 
PLATE I. 
Ci % 
12 
' " 
16 
OCR text unavailable for this page.SEX RECOGNITION IN CYCLOPS. 
S. J. HOLMES. 
The sexual behavior of copepods presents several points of 
similarity with that of the amphipods which was described by the 
writer in a previous paper. 1 In both groups the males clasp and 
swim about with the females for a long time previous to copula- 
tion ; and in both groups the behavior of the female is much the 
same while being clasped by the male. Having an opportunity 
to study a thriving culture of Cyclops fimbriatus in which pairing 
was actively going on the endeavor was made to ascertain if the 
method of the sex recognition employed in the amphipods occurs 
also in this species of a quite distantly related group. 
Male Cyclops, as is well known, have the first antennae en- 
larged and modified to form a clasping organ. In Cyclops fim- 
briatus the male usually clasps the female just in front of an 
enlargement at the base of the abdomen. Females carrying eggs 
are sometimes seized, and also females not more than half grown. 
Males show great eagerness in grasping the females, and they 
can be compelled to release their hold only with difficulty. They 
may be poked about roughly with a needle and the posterior 
part of the body may be cut off without causing them to leave 
the female. I have often picked up pairs in a fine pipette and 
forcibly squirted them out several times without succeeding in 
separating the two sexes. 
As the pairs of Cyclops swim through the water the males are 
usually the more active. Frequently the female remains entirely 
quiet with the appendages drawn close to the body, and the body 
flexed vertrally, allowing herself to be passively carried about 
by her mate. At other times the female may swim as actively 
as the male. In general the behavior of the females and their 
attitude while being carried closely resemble what is found among 
the Amphipoda. So also does their behavior when the males 
come in contact with them and attempt to seize them. The 
1 BIOLOGICAL BULLETIN, Vol. 5, 1903. 
313 
3H S. J. HOLMES. 
female during the efforts of the male to clasp her around the 
base of the abdomen usually lies quiet with the appendages 
drawn close to the body. She may be seized by the legs, tip 
of the abdomen, or any other part of the body, but the male 
works around until he gets into his normal position which he 
sometimes attains only after much labor. Females vary greatly 
however in respect to their willingness to be clasped by the male, 
certain individuals resisting seizure for a long time. 
So far as could be detected the males do not seek or follow 
the females at a distance as Parker concluded they did in Labido- 
cera. The association of the sexes seems to be due merely to 
chance collisions. Males often attempt to seize other copepods 
with which they collide regardless of their sex. The males resist 
such attempts at seizure and dart quickly away, while the females 
often stop and submit readily to the clasping propensities of their 
companions. Several males were injured so that they could not 
resist seizure, and in many cases they were seized by other males 
who worked industriously until they got their burden clasped 
around the base of the abdomen in the usual way. These asso- 
ciations did not last long however ; the active males apparently 
appreciating that something was wrong soon swam away. Re- 
cently killed females were often seized and in some cases carried 
about for a while, but they were finally dropped. Males seem 
rather more prone to seize dead females than members of their 
own sex. In one case I saw three males tugging away at a 
dead female, and they were soon joined by a fourth male who 
participated in the same effort. 
It is possible that the odor of the female determines to a cer- 
tain extent the sexual behavior of the males, but my experiments 
yielded no evidence of this. Several females were put into a 
tube one end of which was covered with fine gauze and the tube 
was then placed obliquely in water in which were numerous 
males. The males showed no tendency to congregate around 
the end of the tube where the females were confined. In another 
experiment several females were placed in a glass tube in which 
a small plug of loose cotton was inserted a short distance from 
one end. This end was laid obliquely in the water. The males 
showed no tendency to enter the open mouth of the tube as they 
SEX RECOGNITION IN CVCLOPS. 315 
might be expected to do if they were attracted by the odor of 
the females. The experiment of removing the organ of smell 
which was performed in the case of the amphipods would be 
a fruitless one in Cyclops* as the seat of smell is located to a con- 
siderable degree at least in all probability in the same organs that 
are used for clasping. 
It is evident that mating in Cyclops is brought about much as 
it is in the Amphipoda. The males have a strong tendency to 
clasp other copepods ; the females tend to remain quiet in a 
condition somewhat resembling the death feint while being seized 
by the males. It is not improbable that olfactory stimuli may 
cause the males to remain with the females longer than they 
otherwise would, and they may render the males rather more 
prone to seize females than other males, but so far as could be 
determined by watching the behavior of the animals the specific 
reaction of the two sexes to certain kinds of contact stimuli is 
the main factor in bringing about their association. 
OUR KNOWLEDGE OF MELANIN COLOR FORMA- 
TION AND ITS BEARING ON THE MENDE- 
LIAN DESCRIPTION OF HEREDITY. 1 
OSCAR RIDDLE. 
Hardly a year has passed since the rediscovery of Mendel's 
Law without several additions to its descriptive terminology. 
This may signify either one of two things : a very healthy and 
vigorous growth, or the onset of senescence. Some of these 
newly introduced features are plainly justifiable; but there is 
reason to believe that the rather long series of extensions which 
has been made in recent times carries as its result not so much 
description of fact, as of deduction and far-reaching theory. 
The facts and phenomena discovered by Mendel, and the array 
of facts of high importance which later workers in this field have 
brought before biologists, have already proved their value. The 
proved value of these facts is, however, no proof of the correct- 
ness of Mendelian interpretations of the processes of inheritance. 
I shall here present some facts which seem to indicate that these 
Mendelian interpretations are not sound ; and further, that these 
unsound interpretations now stand as a formidable block in the 
path of progress to a better knowledge of the mechanism of 
inheritance and development. 
Mendelian workers think that they have discovered, and cer- 
tainly they have named and labelled, many " factors " 2 as neces- 
sary for the production of some single characters. These 
workers tie all of these factors together, and for them to- 
gether they go into the germ cells, and whatever appears or 
fails to appear in the zygote is interpreted in terms of the 
1 Read January 19 before the Biological Club of the University of Chicago. 
2 The word factor as used here in a purely Mendelian sense represents quite a 
different thing from the physiological sense of the same word. The whole series of 
ordinary environmental factors temperature, light, reaction of medium, concentra- 
tion, moisture, etc., all these would not constitute even one Mendelian "factor." 
The Mendelian " factor " is often a rather unidentifiable thing, but it is conceived of 
as something capable of residence in, and of segregation by, the germ cells. 
316 
MELANIN COLOR FORMATION. 317 
presence or absence (dominance, inhibition, contamination, etc., 
made use of by some workers) of particular factors in the gamete, 
But facts at hand, despite an opposite contention, will go very 
far towards showing that the method of analysis of the Men- 
delian worker has not permitted him to decide the question as to 
the real number and scparateness of the factors ; nor yet to de- 
termine as to whether certain of the factors were at all repre- 
sented in the germ cells, or whether they may not have arisen 
during the ontogeny as a direct result of tissue differentiations, 
through regulatory processes, or otherwise, and quite indepen- 
dently of the existence of a definite determiner in the gamete or 
germ cells. 
As Mendelism has developed, it has lent support to the doc- 
trines of preformation, unit characters, and discontinuous varia- 
tion. The facts and interpretations here brought forward disclose, 
on the other hand, no small amount of epigenesis, and strongly 
support the proposition that present and new knowledge will 
lessen, not widen, the apparent gap between discontinuous and 
continuous variability. There is, too, at present a marked tend- 
ency in some quarters to further elaborate and extend the 
" factor " hypothesis, which furnishes an additional and specific 
reason for my calling attention to some facts from my province 
of study which indicate that already we have represented too many 
factors in the germ cells ; that qnite certainly some factors which 
have by Mendelian interpretation been made to circulate through 
the germ cells are never represented (in the Mendelian sense) in 
these cells at all ; and finally that many factors considered most 
separate and discreet by Mendelians, can now be proved to be 
but points in lines of perfect continuity. 
It will no doubt be urged by some Mendelians that the obser- 
vations recorded here are quite wide of the mark because the 
writer has no experience in animal breeding. It is very true that 
I have not personally carried through any breeding experiments 
whatever. For information in this field I have depended upon 
what I have been able to see of the breeding and hybridization 
experiments conducted by others, and upon the literature of the 
subject. My own work for several years has been largely in the 
field of developmental and color physiology ; its aim being to get 
318 OSCAR RIDDLE. 
at the basis of the color characters of organisms. It must rest 
with biologists generally, however, to decide whether the facts 
here presented have, or have not, to do with the Mendelian inter- 
pretation and description of the processes involved in heredity 
and development. 
The basis, then, of my objections to much of the Mendelian 
interpretation rests upon chemical and physiological facts regard- 
ing the origin and development of melanin pigments. It is neces- 
sary to anticipate the query as to how, or by what right, has 
melanin color formation anything to do with the essential points 
of Mendelism ? I realize fully that the line of contact between 
these two provinces of activity is apparently not a line of contact 
at all, and so new and untrammeled is the territory that one 
would almost hesitate, to enter, had not a pair of such good Men- 
delians as Cuenot and Bateson already knocked importuningly 
at the gateway which leads into it. 
It should be recognized at the outset that, in thus presenting 
a body of facts from one field, as having important bearing on 
facts and theoretical deductions in another field, there is every 
risk that a short presentation will be incomplete, inaccurate, 
and at the same time may fail to properly or sufficiently orient 
the reader with respect to the writer's point of view. It is here 
impossible entirely to avoid incompleteness, inaccuracy, and but 
partial explanation of an opposed interpretation of the facts of 
color development and inheritance ; it is hoped, however, that a 
presentation, and rather general though cursory discussion, of a 
limited number of facts facts with which most biologists are 
not familiar, and which have never before been treated in this 
connection - - will make it possible to recognize that some points 
of present biological theory are involved. 
And, though many of my statements concerning such points 
of theory may seem dogmatic, I should like to make it clear that 
I am not deceived or blind to the fact that my' present function 
is merely to introduce subject matter for a chapter, not to con- 
clude a volume ; to propose, not to decide. These discussions 
of theory would have been omitted entirely from this paper if it 
had been thought that the facts here brought for the first time 
into the field of heredity and developmental physiology would 
MELANIN COLOR FORMATION. 319 
receive the attention which they deserve without such a setting. 
I am, of course, rather confident of the correctness of the point 
of view set forth. I am absolutely convinced that the facts here 
presented will prove valuable assets to the student of develop- 
ment and inheritance. 
There are three reasons why melanin color formation, better 
than any other process or group of processes, may furnish the 
starting point for certain inquiries and criticisms regarding the 
way Mendelian inheritance is construed and described : 
1. Color characters have been more extensively studied and 
described from the Mendelian standpoint than have any others. 
A very considerable share of the color investigated all mam- 
malian color, for example is due to melanin pigment. 
2. It was to recognize a fact in melanin color development that 
Cuenot ('03) introduced the idea of presence and absence of a 
character, or character determiner, etc.; an idea which is now 
made by many workers to support practically the whole struc- 
ture of Mendelian description and interpretation. Again, the 
now rather elaborate terminology introduced by Castle is based 
almost wholly upon the behavior of melanin colors. A few 
paragraphs of the paper by Cuenot furnish practically all there 
is of a tangible basis for representing cJiromogens, enzymes, and 
activators in gametic formulae. 
3. There is already at hand a certain amount of definite chem- 
ical knowledge, and some reasonably safe physiological informa- 
tion, which can be brought to bear on some points of the color 
philosophy of Mendelism. There is, moreover, something which 
though apparently less substantial, is none the less important - 
namely, the assurance of further, definite light from these same 
sources. There can be no doubt that we can use biochemical 
and physiological methods and data to give us what is now more 
needed than all else, perhaps, in the study of evolution and de- 
velopment namely, the intimate developmental history, and na- 
ture of some one character ; I mean the proximate history, the 
mechanics of what some would call the "late stages" of the 
development, or the " differentiation " of a character. 
It may help to keep the reader oriented throughout this dis- 
32O OSCAR RIDDLE. 
cussion to state that I shall first describe some of the facts of 
color development as they are known at present from chemical, 
pathological, and physiological experience ; and afterward sketch 
very briefly the nature of the Mendelian terminology ; this to be 
followed by some discussion of my point of view. Such facts 
of color will be considered as have bearing on the following 
points : 
Do the known facts of the genesis, nature and history of color 
characters harmonize with, supplement, modify, or radically differ 
from, the demands of present Mendelian interpretation ? Do 
they enable us to decide as to whether color characters are 
qualitative or quantitative in nature ? Are color differences cases 
of continuous or discontinuous variability ? Can these facts 
throw light upon the existence or nature of unit characters ? 
What of the purity of gametes ? Do these facts indicate a dif- 
ferent or sounder basis for the interpretation of Mendelian, or 
other inheritance? What justification or light, if any, is thrown 
upon the present practices of (a) adding " factors " in order to 
account for the inheritance phenomena exhibited by a character ; 
($) of tying all these " factors " together and postulating that all 
pass (by means of their representatives) through the germ cells ? 
SOME FACTS OF MELANIN COLOR FORMATION. 
In a consideration of the facts of the origin of melanin color- 
ation, one might deal at some length with the distribution and 
histogenesis of melanin. Though several interesting and illumi- 
nating facts lie in each of these directions, I shall dismiss these 
two phases of the origin of melanin colors with the single state- 
ment that the melanins are usually dark, amorphous or granular 
pigments, chiefly of intracellular, animal origin ; extending within 
this kingdom from the trypanosomes (Protozoa) to man. There is 
no vertebrate species (unless we may think of pure albinos as such), 
but has this coloring matter in one or several parts of its body. 
It is, however, the chemical and physiological phases of the 
origin of color that it is most desirable to discuss, and it is 
from this angle of approach that we find most of the facts which 
bear on the Mendelian description of heredity. 
Our knowledge of what has been called the " mechanics of 
MELANIN COLOR FORMATION. 321 
melanogenesis " may be thought of as having begun with studies 
in the production of artificial melanins, and the accompanying 
search for the (melanin) chromogen in the albumen molecule. 
This work was shared by many workers : Stadelmann ('90), 
Gmelin ('94), Nencki ('95), Schmiedeberg ('97), Chittenden and 
Albro ('99), Hofmeister (see v. Fiirth, '04), v. Fiirth ('99, '01, 
'04), Hopkins and Cole ('oi, '03), Schneider ('01), Samuely 
('02) and others. Through these workers it was made known, 
first, that melanins artificially produced are essentially the equiva- 
lents of natural melanins ; and second, that tyrosin and related 
aromatic compounds are the chromogens concerned. 
The second step in the progress of this knowledge was con- 
cerned with the nature of the process by which the melanin is 
formed from the chromogen. Hlasiwetz and Habermann ('73) 
had first recognized oxidative processes as necessary for the 
formation of the artificial melanins. Landolt ('99) extended this 
fact to the natural pigment of the choroid. 
Bertrand then discovered ('96) an oxidizing enzyme tyrosi- 
nase which was able to transform tyrosin into melanin-like 
bodies. Bertrand found the enzyme in certain plants. It has 
since been found to be of wide distribution, having been found 
by Biedermann ('98) in the contents of the alimentary canal of 
meal worms ; by Leptnois ('99) and Gessard ('oi) in the adrenal 
glands ; by Gonnermann ('oo) in beet roots ; by v. Fiirth and 
Schneider ('oi) in the haemolymph of insects; by Przibram (see 
preceding, 'oi) in the ink-sacs of cephalopods (Sepia); by 
Ducceschi ('oi) in the blood of Bombyx ; by Gessard (o2a, 'o$a, 
'03^) in the ink-sacs of Sepia, in the integuments of insects, and 
in melanotic tumors of horses ; by Dewitz ('02) in the blood of 
certain insects ; by Durham ('04) in the skins of mammals and 
birds ; by Weindl ('07) in the skin, eyes, ink-sacs and eggs of 
Loligo ; and by Bertrand and Mutermilch ('07) in wheat bran, 
v. Fiirth and Schneider ('oi) concluded that " tyrosinase-like 
ferments are widely distributed in the animal organism, and prob- 
ably always appear wherever and whenever a physiological or 
pathological formation of melanin occurs." 
Meanwhile, another advance in our knowledge of melano- 
genesis was made when Dewitz ('02) demonstrated the role of an 
322 OSCAR RIDDLE. 
oxidizing enzyme (tyrosinase) in the normal development of the 
dark pigment of the integuments of living, growing animals (fly- 
larvae Lucilia Ccesar}. At the same time he was able to prove 
that, in the forms with which he worked, free oxygen is also an 
indispensable factor in the development of the color. This work, 
important and suggestive as it was then, is now made still more 
valuable by new knowledge of the chromogen that is, the other 
factor involved in the pigment formation. Without knowing 
just what this chromogen might be, Dewitz was able to conclude 
(p. 45), " We cannot doubt that we have here in the blood of the 
larvae an enzyme under the influence of which a chromogen is 
oxidized and forms a brown or black pigment." 
A year later Gessard ('03^) was able to show that in the 
melanotic tumors of white horses not only tyrosinase but free 
tyrosin is present. He concludes (p. 1088) : " Tyrosin is then 
the chromogen, the oxidation of which by tyrosinase determines 
the formation of the black pigment which is common to many 
physiological and pathological products of the animal economy ; 
and it can be said that the color of the negro is due to the same 
reaction that produces the ink of the squid, or the black color of 
some mushrooms." Gessard states, too, that when tyrosin is 
oxidized with tyrosinase it gives a series of colors "rose, 
rouge-grenat et brune." In a later work ('O3</) he made a 
closer study of the color reactions of tyrosin in which he showed 
that the presence of acids, alkalis and salts have marked effects 
on the colors produced. 
The recent work (May, 1908) of Bertrand, is, however, of the 
highest interest. He has been able to determine (i) the type of 
substance of w/iic/t there are many representatives which can 
by the use of tyrosinase be oxidised to melanin compounds ; (2) he 
has shown that each one of these compounds passes through a series 
of colors before arriving at the final stage of oxidation ; (3) that 
this series varies somewhat as to the exact tint of the initial and 
final colors, but that (4) the early stages of oxidation uniformly 
give lighter colors than the later ones, the series usually running 
from yellow to orange, through darker tints to brown or black. 
Bertrand's studies make it clear that any benzene nucleus with an 
attached hydroxyl can be acted upon by tyrosinase and converted 
MELANIN COLOR FORMATION. 323 
into melanin pigment. Thus the whole series of compounds in the 
table given below (and many others besides) can be oxidized to 
colored compounds. On the other hand, phenylalanine, phenyl- 
methylamine, phenylominoacetic, phenylpropionic and phenyl- 
acetic acids, alanin, glycocoll, etc., give no coloration whatever. 
t 
The size, complexity, and nature of the lateral chain has only a 
subordinate influence ; if it is not very strongly acid or basic it will 
not interfere with the oxidation. Thus for example, ethyltyrosin, 
chloracetyltyrosin, and glycyltyrosin were oxidized with ease. 
The nature of these side chains does, however, considerably modify 
the colors produced by the oxidation. Neither of the three last 
named bodies (tyrosin compounds) give a final black color ; they 
begin with orange or yellow and end with red or mahogany 
(chocolate ?). 
In order to further appreciate something of the variety of color 
which may arise from a single cliromogen, and to get an introduc- 
tory idea of the number and variety of chromogens to be found 
in the animal body, careful reference to Table I. should be made. 
And here it is of the higlicst importance to see that a single chro- 
mogcn acted upon by a single enzyme (so far as all chemical ex- 
perience has detected) produces several colors depending upon the 
degree of oxidation involved. 
In regard to the rate at which these colors appeared the 
author's statement may be cited that in a 20 per cent, tyrosinase 
(80 per cent, strong tyrosin) solution, tyrosin developed a rose 
color in ten minutes and its black color in four to five hours. 
TABLE I. (From Bertrand, '08.) 
Name of Body. Colors Produced by Oxidation. 
Tyrosin red grenadine, then inky black, 
p-oxyphenylethylamine red grenadine, then black olivaceous, 
p-oxyphenylmethylamine orange yellow, orange red, maroon, 
p-oxyphenylamine orange, mahogany red, then brown, 
p-oxyphenylpropionic acid orange yellow, grenadine red, brown, 
p-oxyphenylacetic yellow, orange yellow, then brown, 
p-oxybenzoic " (weak) rose, orange, then yellow, 
p-cresol yellow, orange, then red. 
Phenol yellow, orange, red, then brown. 
These are, from our present point of view, the more notable 
results of Bertrand's investigations of what he calls " the mechan- 
324 OSCAR RIDDLE. 
ism of melanogenesis." This author does not in any way con- 
sider the bearing of his findings on Mendelian descriptions of the 
mechanism of the origin of melanin color characters (neither on any 
aspect of inheritance). Nor as previously stated has any- 
one, at any time, inquired as to whether the facts obtained in the 
former sphere are compatible with the assumptions made in the 
latter. Bertrand's studies had other and quite different objects ; 
these were : First, to learn the degree of specificity of the enzyme 
tyrosinase ; the conclusion here being (p. 387) " that the results 
speak once more against the principle of very absolute specificity 
which one nowadays often hears applied to enzyme actions." It 
will be very well for us to bear in mind this result, since, as we 
shall see, Mendelian description demands a still higer degree of 
enzyme specificity than the philosophy of biochemists has yet 
dreamed of. A second object of his work seems to have been 
to consider the possibility of identifying certain of these tyrosin 
bodies by the color reactions they give upon oxidation with 
tyrosinase. A third purpose of the study was concerned with 
the causes of the rather wide variations in the elementary com- 
position of different melanins ; he believes that the results give 
some reason for believing that the simplest melanins arise from 
the oxidation of tyrosin itself, while the more complex ones 
those containing sulphur or iron are formed by the oxidation 
of less complete products of protein hydrolysis namely, tyrosin- 
containing di- or polypeptids. 1 A fourth and final phase of Bert- 
rand's investigation touched upon a hitherto unrecognized, but 
seemingly possible, mode of union between tyrosin and other 
amino acids. 
It will now be well to briefly sketch a few facts obtained from 
the study of abnormal tyrosin metabolism, and from pathological 
pigmentations (melanin) of the human body, as supplementary 
to the account of melanogenesis which is given above. These 
facts will also serve to illustrate the dependence of tyrosin oxida- 
tions upon somatic conditions which may be of such a temporary, 
intermittent, quantitative, or reversible character as to preclude 
1 In the formation of melanins, condensations as well as oxidations occur, but the 
former process need not concern us in treating the present theme. 
MELANIN COLOR FORMATION. 325 
the possibility of accounting for them on the basis of specific, 
independent transmissions, once for all segregated by the germ 
cells. 
Rather fortunately for the completer view of our present 
theme, pathologists and clinicians have been frequently confronted 
with cases of incomplete tyrosin oxidation (alkaptonuria), and 
unusual and pathological melanin pigmentations (melanotic 
tumors, Addison's disease, ochronosis) in the human body. 
These subjects because of their medical bearings have received 
a very great amount of attention, of accurate and searching study, 
at the hands of investigators. It seems self-evident that the 
student of melanogenesis should here find much data to interest 
him. 
In the condition known as " alkaptonuria " l the alkapton acids 
uroleuic and homogentisic appear in the urine. The last- 
named compound represents a stage in the oxidation of tyrosin. 
The intermediate stages and the chemical structure of these sev- 
eral compounds may be best understood by reference to Table II. 
Our interest in these early stages of tyrosin oxidation is very 
great since we know that the same, or similar steps, lead in special 
cases to the formation of melanin. 2 Neubauer ('08) has very 
recently determined the exact course of the first four steps of 
tyrosin oxidation as they occur in the living (human) body. The 
chemical expression of these stages is given in the table. 
Garrod ('02) found that certain individuals, who in their youth 
excreted urine containing homogentisic and urolencic 3 acids, 
produced only the former during adult life. Other cases of tem- 
porary and intermittent alkaptonuria are known. Here certainly 
the evidence of our senses is simply that the pow er of the organ- 
ism to oxidise tyrosin compounds is not dependent primarily upon 
germinal segregations, but rather upon tissue activities, relations 
Resume and literature by Falta, Biochem. Centralb., 3, p. 174, 1904. See also 
Abderhalden, " Lehrbuch der Physioiogischen Chemie," Berlin, 1906, pp. 294-298; 
and Neubauer ('08), loc. cit. 
2 A chemical research directed to the determination of the reasons for some tyrosin 
oxidation leading to melanin formation, instead of to the usual end-products of oxida- 
tion NH 3 , CO 2 , H 2 O, etc., would be of the greatest interest and help in studies on 
the physiology, development and heredity of color. 
3 According to Neubauer ('08) this body is not an intermediate step in the oxi- 
dation of tyrosin to homogentisic acid. 
326 OSCAR RIDDLE. 
TABLE II. 
SHOWING SOME OF THE KNOWN FACTS OF CHEMICAL EXPERIENCE CONCERNING 
THE SUCCESSION OF OXIDATIONS OF TYROSIN AND CLOSELY RELATED 
BODIES TO MELANIN PIGMENTS. 
Tyrosin = HO- -CH 2 CH(NH 2 )_ COOH 1 Colorless. 
P-oxyphenyl-pyrotartaric acid= HO c^ J> CH 2 CO COOH 1 " 
CH 2 CO COOH 1 
>H 
Hydrochinon-pyrotartaric acid = / \ CH 2 CO COOH 1 " 
Homogentisic acid 4 = ... ( \ CH 2 COOH 1 " 
HO 
,OH 
Gentisic acid = N COOH " 
HO 
Melanogen ( ) " 
Melanin ( ) (White)? 2 
Melanin ( ) Pale yellow. 3 
Melanin ( ) Deep yellow. 3 
Melanin ( ) Red. 3 
Melanin ( ) Brown. 3 
Melanin (C 50 H 58 N g SO J2 ) Black. 3 
Melanin (C 45 H 78 N 10 SO 20 ) (White)? 2 
and conditions ; these may vary or change from year to year 
a power of oxidation not possessed by the individual during 
many early years of life being attained at manhood, or vice versa. 
Bateson ('02) has seen fit to claim that (p. 133, note) alkapto- 
nuria is the result in inheritance of the union of " two recessives." 
Garrod has concurred in the view. But in the light of the in- 
constant and intermittent character of the phenomenon, it seems 
necessary to draw a directly opposite conclusion. 
1 This according to Neubauer ('08). 
2 See discussion of Spiegler's work ('03), etc., p. 328 of this paper. 
3 These from Gessard ('03) and from Bertrand ('08). 
4 Concerning the formation of homogentisic acid from tyrosin in plant tissues, see 
Schulze and Castoro, Zeit. f. Physiol. Chem., Bd. 48, p. 396, 1906. 
MELANIN COLOR FORMATION. 327 
Examples of other temporary or intermittent oxidative powers 
might be much extended to include cases of glycosuria, cys- 
tinuria, purin metabolism, etc. I shall not discuss these cases 
which have only an indirect bearing on our question of tyrosin 
oxidation. It is, however, of some interest to state that it has 
become evident through the work of Abderhalden and Schitten- 
helm ('05), Garrod and Hurtley ('06), and others, that the body 
may possess a low oxidizing power for several different protei)i 
constituents at the same time ; as, for example, in some cases of 
cystinu'ria, when diamines, tyrosin, lysin, etc., in addition to cys- 
tin, pass through the body unoxidized and appear as such in the 
urine. 
The known facts of abnormal pigmentations deserve a larger 
share of attention than they receive here. They are mentioned 
chiefly to direct attention to a field of facts that are quite com- 
pletely ignored in our theories of the heredity of color. 
In the condition known as ochronosis, certain cartilages (e. g., 
those of the ear) and connective tissues become pigmented. The 
work of Albrecht ('02), Osier ('04), Pick ('06), and others, make 
it certain that ochronosis is a form of melanotic pigmentation, 
and that it is not uncommonly associated with melanuria, or 
alkaptonuria and even with the pigmentation of the sclerotics 
and skin (Osier). Similarly, in Addisori 's disease there is de- 
posited in the skin a pigment which, according to Pforringer 
('oo) differs from that produced normally only in quantity and 
not in origin or composition. It is well known too that nerve 
lesions are often accompanied by pathological pigmentation of 
the skin. 1 
A word in regard to melanotic tumors. These are known to 
occur particularly in white horses. The amount of melanin 
produced is often very great. Abel and Davis ('96) estimate the 
melanin of the entire skin of a negro at I gram, whereas Nencki 
and Berdez ('85) found 300 grams of melanin in a sarcomatous 
liver, and estimate that the entire body contained 500 grams. 
These several facts from pathology are significant in that they 
indicate that/<?r the building of any melanin at all, the actual local 
conditions of the organism, or the organ, have a role to play that 
1 See resume by Schmidt, Ergeb. Jer Pathol., Bd. III., Abt. I, p. 551, 1896. 
328 OSCAR RIDDLE. 
is quite out of keeping with any " once for all determination " by 
the shuffling of color "factors" through the germs. 
In this connection attention should be called to the fact that 
Spiegler ('03) has reported the finding of a white melanin in the 
white hair of horses, and in sheep's wool. It would seem that 
the melanin isolated by him represents a more advanced stage 
of oxidation than does black melanin. If this is true it is obvi- 
ously an important fact. The isolation of a similar pigment 
from the white hair of albinos seems, however, neither to have been 
sought for, nor found ; but it is quite possible that such a pig- 
ment also exists in mammalian albinos. There are nevertheless 
some reasons --chiefly biological rather than chemical for be- 
lieving that among birds, at least, a white color exists which rep- 
resents a less advanced stage of oxidation than that in any other 
of the melanin color series; in fact, here the "white" seems to 
be a purely "physical " color. The chromatophores do not de- 
velop, and no white pigment is histologically discoverable. Ap- 
parently, therefore, the oxidation of tyrosin, etc., is here not 
carried far enough to produce any color whatever. 
The actual facts regarding the "white" pigment or color of 
birds and mammals are not yet clear. " White " forms at present 
a most awkward, and at the same time a most interesting gap in 
our knowledge of the melanin series. The elementary formulas 
for the black and white pigments given in the table are Spiegler's 
findings for the white and black hair of the horse. Certain 
phases of Spiegler's research have been confirmed by Wolff ('04), 
but the particular facts which we have referred to above have not 
been reviewed or confirmed. 
We are fortunate in having, from physiological experiments 
with melanin pigments in living animals, some facts which con- 
firm the data from chemistry and pathology regarding the 
mechanism of melanin production. On this point special attention 
may be called to the recent work of Gustav Tornier. His work 
furnishes a splendid view of the control of the color of the in- 
tegument. The colors of Amphibia, from larval stages to old 
age, were determined at will by controlling the physiological 
state, particularly the nutrition, of the animals. When the color 
MELANIN COLOR FORMATION. 329 
phenomena observed by Tornier are viewed in the light of the 
work of Bertrand, it seems certain that in these two sets of phe- 
nomena we are really dealing with the same facts. 
Tornier found that tadpoles divided into lots for differential 
feedings gave (i) little or no pigment in the ones fed the mini- 
mum, and progressively more pigment as the maximum is ap- 
proached ; (2) a series of colors : u'/iite, yellow, red, gray, black. 
Tornier concludes ('07^, p. 288) : It is possible, therefore, by 
adjusting the dosage of fleshy food to force the epidermal color- 
ation of Pelobates larvae (he elsewhere describes this as true for 
other amphibia) into white, yellow (see p. 285), red, gray, black. 1 
It will be seen that the types of color and the order of their ap- 
pearance in the organism - - when we put these organisms into 
such conditions as will force them to do the work of pigment 
formation (oxidation) in stages closely follow the lines of our 
purely chemical experience. Tornier produces entirely compar- 
able effects upon the coloration by two other means, viz.: by re- 
moving more or less yolk from the vegetative pole of the egg 
through an opening made by a needle ; or by coagulating in situ 
a part of yolk proteid by the introduction of water ; such coag- 
ulations of yolk proteid cannot be digested by the develop- 
ing embryo. The three methods employed all reduce the nutri- 
tion of the animals, and produced albinism, erythrinism, black- 
ness, or melanism, depending upon the state of nutrition. 
It was further found ('07) that the experiment could be carried 
out in the opposite direction as well ; that is to say, the highly 
fed, black-containing, and black-producing larvae of large or 
small size could be made, through a diminution of feeding, to 
produce a series of colors in the order of: black, brown, red, gray, 
i^'Jiite. Certain observations by Powers ('08) are confirmations 
of Tornier' s L results. 
Without extending these illustrations it can be stated that if 
these facts and experiments mean anything they mean that in an 
animal that produces melaninic color, there exists all the machinery 
necessary to produce a series or scale of these colors. And that 
1 The proof that all these shades of color in Tornier' s tadpoles were melanins is 
hardly as complete as is desirable, since lipochromes and traces of guanin are also 
known to develop normally in late larval and adult stages of these amphibia. Many 
facts, however, indicate that the colors here described by Tornier are true melanins. 
33O OSCAR RIDDLE. 
what is actually produced is, in several demonstrated instances, 
dependent upon the physiological state of the organism. Or, per- 
haps, in certain cases it may be possible to say more definitely 
that the limiting factor is none other than the available oxygen 
or food-supply. I have been able to prove ('oS) definitely that 
in many birds the daily nutritive changes which accompany the 
low blood-pressures occurring at night influence the quantity of 
melanin produced. 
The specific color of an animal then is an index, not of the 
presence in the germ from which this animal arose, of certain 
chromogens and specific zymogens, and the absence of a wide 
series of others ; but, this specific color means that a process with 
a wide range of possibilities, because of a particular physiological 
state and environmental conditions has struck this particular equi- 
librium. One and the same organism has within it all tJiat is 
necessary to move that equilibrium up or down-- taking the red 
form for example, we can in the words of Tornier " force it to 
black or to white." 
Tornier did not consider the relation of his results to any of 
the facts of the chemistry of melanin, nor did he consider their 
bearing on color inheritance. His chief concern has been appar- 
ently to establish two points in color physiology ; first, the effect 
of varying degrees of nutrition on the size, shape, color-produc- 
tion, etc., of chromatophores ; and, secondly, a defense of the 
thesis that the pigment granules of these chromatophores act as 
reserve food-materials in cases of inanition, etc. 
MENDELIAN DESCRIPTION. 
It is now possible to consider whether Mendelian interpreta- 
tion and description is in accord with the facts of color formation. 
The Mendelian position can be best presented in the words which 
Cuenot ('03) used in the original formulation and statement of 
it. I quote practically the whole of his discussion, and ask that 
it be remembered that it is almost entirely upon this slender basis 
that the "presence, absence" hypothesis, and consequently a 
great share of Mendelian nomenclature, rests : 
" Again one learns that the authors whp have recently studied 
the origin of melanin pigments, Biedermann, v. Furth, Schneider 
MELANIN COLOR FORMATION. 33! 
and Gessard, state that these pigments result from the action of 
an oxidizing enzyme (tyrosinase) upon a chromogenic substance ; 
there are good reasons for supposing that things happen simi- 
larly in the pigmentation of the skin ; there should be, however, 
in this case, either two different chromogens and only one enzyme, 
or only one chromogen and two enzymes, the one for the blackish 
pigment and the other for the yellow pigment. We adopt pro- 
visionally, for convenience of language, this latter hypothesis. 
" The germ plasma of a gray mouse should contain potentially 
the three substances which, by their reciprocal reactions later 
produce the deposition of pigment in the hair ; and doubtless 
these three substances are contained in the potential state within 
many of the material particles of the germ plasma (representa- 
tive particles or qualitative substances of the egg mnemons). 
In a gray mouse (black and yellow pigmented) there are three 
mnemons, one for the chromogen and two for the two ferments ; 
in a black mouse there are only two mnemons, one for the 
chromogen and another for the formative enzyme of black 
pigment. 
" In regard to albinos, all is explained if we admit that their 
germ plasma contains only the mnemons of the enzymes, that of 
the chromogen being totally absent. With these conditions, 
colored hair cannot be formed in albinos, since one of the sub- 
stances indispensable to the reaction is absent, but one easily 
understands that the albino will transmit to its progeny either the 
mnemons for the two enzymes, or one mnemon only, if it pos- 
sesses but one." 
The Mendelians have one further point to confirm the faith 
that is in them. Soon after the appearance of the paper by 
Cuenot, Durham undertook to find whether in the skins of black, 
chocolate, yellow and albino mammals there is the appropriate 
enzyme in each for the production of its particular color --when 
this acts upon a tyrosin solution. Only a short preliminary 
statement ('04) of the results has appeared ; and although posi- 
tive results were reported for the black, chocolate and yellow 
pigments, it is evident that from no point of view can these re- 
sults be regarded as satisfactory ; particularly because the ex- 
tracts used by her are stated to have had a reddish color before 
332 OSCAR RIDDLE. 
adding the tyrosin ; an adequate account of the history or 
nature of color changes in such a solution seems hardly possible 
since as Table I. shows much of melanin production results in 
colors which are paler than red and the origin of all these, and to 
some extent the final color, would be obscured by the initial 
presence of red. Miss Durham was unable to state exactly 
what the extracts of the albino skins were capable of doing, but 
thought they probably contained no such enzymes. 
The factor hypothesis of Castle avoids some of the pitfalls of 
the earlier theory, but seems to rest on essentially the same base. 
It is necessary to examine in detail the statements and conclusions 
in Cuenot's paper. 
Cuenot says : " There should be, however, in this case either 
two different chromogens and only one enzyme, or only one 
chromogen and two enzymes, the one for the blackish, the other 
for the yellow pigment." The facts of the origin of melanin do 
not substantiate Cuenot's hypothesis because the colors do not 
form on this plan ; one and the same chromogen is known to 
form yellow and black ; one and the same ferment tyrosinase 
-is known to produce both yellow and black from the same 
chromogen. 
According to Cuenot : "The germ plasma of the gray mouse 
(black and yellow pigment) should contain potentially the three 
substances which, by their reciprocal reactions, later produce the 
deposition of pigment in the hair ; and doubtless these three sub- 
stances are contained in the potential state within many of the 
material particles of the germ plasma (representative particles, 
mnemons, etc.)." But three such substances are not required 
for the production of black and yellow ; these two color com- 
pounds are known to be but different stages of oxidation of the 
same substance. Furthermore, the locating of all the factors which 
determine the development of these colors in the germ plasm 
does not reckon with the facts of color physiology already cited. 
Continuing, Cuenot says : " In a gray mouse (black and yellow 
pigmented) there are three mnemons, one for the chromogen, 
and two for the two enzymes ; in a black mouse there are only 
two mnemons, one for the chromogen and another for the forma- 
tive enzyme of black pigment." Statements made above furnish 
a sufficient refutation of this conception of Cuenot's. 
MELANIN COLOR FORMATION. 333 
Cuenot's concluding statement : " In regard to albinos all is 
explained if we admit that the germ-plasma contains only 
mnemons of the enzymes, that of the chromogen being totally 
absent. With these conditions, colored hair cannot be formed 
in albinos, since one of the substances indispensable to the reac- 
tion is absent, but one easily understands that the albino will 
transmit to its progeny either the mnemons for the two enzymes, 
or one mnemon only if it possesses but one." To this statement 
must be opposed first the opinion of Durham that the skins of 
the albino mammals studied by her contained no tyrosin-oxidizing 
enzymes ; a second much more weighty and conclusive objection 
is that the absence of melanin chromogens in albinos is practically 
inconceivable. It is now certain that tyrosin and its many related 
compounds are such chromogens, and that these compounds 
have a distribution in the universe almost or quite co-extensive 
with protoplasm itself. The postulation of the formation of chro- 
mogen-containing, and non-chromogen-containing gametes is there- 
fore reduced to an absurdity. It is moreover quite certain that the 
food of tJie albino mouse must daily bring it quantities of chro- 
mogens, even if such could have been excluded from the germ cells. 
There is no doubt and no middle ground here. Cuenot's concep- 
tion fails completely. 
Space does not permit a discussion of the facts and interpreta- 
tions of the inheritance of coat colors of mice since the work of 
Cuenot ; a subject which has been investigated or discussed by 
many workers, notably by Bateson, Allen, Cuenot, Morgan, 
Wilson, Castle and Durham. A detailed consideration of these 
results is omitted also because the facts are well known and do 
not belong here. The behavior of the color yellow first reported 
by Cuenot ('05) is, however, of such unusual interest as to de- 
serve special mention. It was found that yellow is " dominant " 
in mice, whereas elsewhere in animals it is usually " recessive " 
to black ; this is an unexpected result from the Mendelian stand- 
point, and the difficulties which it presents have called forth 
several highly complex, supplementary Mendelian hypotheses. 
It seems not to have occurred to anyone that yellow may be a 
blend formed from the union of other colors, e. g., between 
albino and black. A glance at our scale of colors (p. 326) sug- 
334 OSCAR RIDDLE. 
gests that such may be a true, and, indeed, the truest conception 
of a color blend ; and an examination of the experimental results 
gives considerable support to this view. Castle ('06) reports 
that he secured yellow mice from three sources, viz. : Y<^ x Ch $, 
A x B, A x Ch. It will be observed that each of these crosses 
has one color (yellow, chocolate, black, albino) more oxidized 
than yellow, and one less oxidized than yellow (see Table II.) ; 
that is, the yellow produced in these cases is apparently a blend. 1 
The general fact of the unstable (heterozygous) character of all 
yellow mice is, quite possibly, evidence of the same kind. 
Heretofore a blend, say between white and black, has been 
considered a mixture of these two colors, a spotted animal, a 
form in which black was diminished, etc., but a little reflection 
upon facts already stated concerning the nature of these color 
characters reveals very distinct colors as none the less very distinct 
blends ! 
At present the biological data are wanting to quantitatively 
seriate all of the several colors ; but there is apparently enough 
data to warrant the definite statement that yellow mice are forms 
with the power to oxidize tyrosin compounds to an intermediate 
point. Thus the biological data again parallel chemical experience. 
Cuenot, Bateson and others " explain " color inheritance on the 
assumption that "recessives " lack altogether the factors of color 
production ; but Castle has convincingly argued that this cannot 
be true, because in such forms small amounts of pigment actually 
form, etc. Castle tried to explain these phenomena upon the 
supposition that all of the factors may be present, but that " the 
presence of one character often inhibits the activity of another," 
i. e., upon grounds of activity and latency. I would urge that 
we are now quite ready to take the next step, which seems to lie 
in the opposite direction, and say that we have to do neither with 
absent factors nor with the inhibition of present factors ; that in 
gametic unions we deal not at all with " factor " particles, but 
merely mix, and amalgamate to various degrees, powers of tyrosin 
oxidation ; and the conditions supplied by the differentiation of 
1 Cuenot says of his yellow mice that they contain numerous unfixable variations, 
not hereditary, ranging from a clear orange yellow to a sooty or grayish yellow, not 
very different from the color of gray mice. Does this look like purity of gametes, or 
a wide range of blending, which? 
MELANIN COLOR FORMATION. 335 
tissues and organs, together with environmental conditions ex- 
ternal and internal, supply whatever else is concerned in color 
production. 1 
In his work on mice Castle ('06) shows conclusively that in 
these forms he is not dealing with the total absence of a character, 
even in some cases where this seems apparent ; he there states 
that the purity of gametes does not exist, and further, " no more 
does the purity of factors exist." It would seem that this paper 
by Castle is one of the best, if not the best document extant to 
convince that no such things as factors exist. Yet, quite recently, 2 
Castle has carried the factor hypothesis to its highest state of 
complexity, to what is apparently its logical conclusion, if any 
"particle" basis whatever could be granted for hereditary proc- 
esses. Castle pictures his conception of the factors in melanin 
color inheritance and their relations in the following diagram, 
modelled after a chemical formula : 
S^U 3 
I C = Color, or chromogen factor. E = Extension factor. 
A C^ R F A = Agouti factor. I = = Intensity " 
B =. Black U = Uniformity " 
D ^p D= Dilution " S = Spotted 
This visualizes for us the body of factors which are to be 
shuffled in the germs and " determine " the colors of the progeny. 
From what has been said it is obvious how far this conception 
leads us astray ; because this complex Medelian interpretation 
and description of color inheritance leads us away from a simple 
series of color developments due to the difference in degree to 
which one substance may be oxidized. 
The placing of the " uniformity-spotted," "intensity-dilution," 
and " extension " factors in these germs is a virtual surrender of 
the whole theory of discontinuous variation ; and in reality puts 
'The many and accumulating additions, qualifications, " contaminations," " laten- 
cies," etc., that have been attached to Mendel's ('65) original conception of domi- 
nance and recessiveness, or to Cuenot's ('03) presence and absence of a factor 
hypothesis, are but so many direct admissions that the "purity of gametes" concep- 
tion is an error ; they are but so many secondary and tertiary hypotheses of completer 
preformation made to bolster up a primal preformation hypothesis. Incidentally, they 
give present-day students the opportunity to see the child of Weismannism recapitulate 
the developmental history of the parent. 
2 Darwinian lecture, Baltimore, January I, 1909. 
3 Either one of the pair may be present (or active). 
336 OSCAR RIDDLE. 
the Mendelian who accepts tKis terminology in no position to 
deny the development of the melanin series on a basis of closely 
graduated powers of tyrosin oxidation, and so without any basis 
on particles or factors, whatever. It is, moreover, as plain as it 
is certain, that this degree of oxidising poiver covers several of 
these factors (A, B, E, I, D) which are thus reduced to one ; 
while the kernel of this formula the C, or chromogen, goes out 
entirely as a factor, /. c., something capable of being shuffled in 
the germs since as we have pointed out, such a chromogen is 
universal in protoplasm. 
DISCUSSION. 
If the later oxidations those which produce color of the 
tyrosin compounds are each individually controlled by a separate, 
specific enzyme, why are not the several earlier oxidations of the 
same compounds similarly conditioned ? If assumption will give 
us the whole series of tyrosin oxidations only on condition of 
their production by means of separate and distinct enzymes, why 
should it hesitate to put the whole vast array of oxidations of all 
aromatic, or even of all organic compounds on a similar basis - 
a separate and specific enzyme, separately heritable, for each 
step in oxidation ? This is certainly not true. 
It is impossible at present to announce the limits to the speci- 
ficity of enzymes, and of the oxidases generally ; but it can be 
said that it is pretty generally conceded that the oxidases present 
less specificity than do the digestive enzymes (see Wilcock, '06). 
That which weighs most heavily against the Mendelian assump- 
tion of a high specificity of tyrosin oxidizing enzymes is, however, 
the result of Bertrand's special study of tyrosinase which indicates 
no such specificity (quoted p. 324 of this paper). 
But granted the greatest possible specificity of these enzymes, 
the Mendelian description of color inheritance becomes even 
more untenable ; for Gessard and Bertrand have shown that 
"black" is the end-result of a series of successive oxidations 
and this final color can be attained only by having all of the 
intermediate stages actually attained. This means that the 
animal that transmits the enzyme for black, /. e., produces black 
colored offspring, must at the same time transmit also the enzyme 
MELANIN COLOR FORMATION. 337 
for brffivn, chocolate, red, yellow, etc. (more accurately, an enzyme 
for each step of oxidation from tyrosin to black melanin), without 
the absence of a single one. If there be introduced here the primal 
Mendelian conception of the freedom of each of these factors to 
be distributed in gametogenesis according to the laws of chance, 
how often then may we expect pure-bred black parents to produce 
black offspring? 1 
There are reasons, derived from our general knowledge of 
oxidizing enzymes, why this assumption of high tyrosinase 
specificity is highly improbable and some evidence against this 
position has been cited from Gessard and Bertrand ; a direct 
refutation of it is furnished by the experiments of Tornier. He 
showed that he could take animals which would have, according 
to Mendelian assumption, the enzyme for " black," and make 
them produce any one of three or four of the less oxidized mem- 
bers of the color series ; these same forms could again under other 
conditions be made to produce the more highly oxidized black, 
etc. Obviously, the presence of a black-producing enzyme did 
not determine the color here ; but conditions of life did so deter- 
mine. The further assumption of inhibiting factors ready at all 
points of color-production to account for lower grades of color 
formation, and all other secondary assumptions to support the 
primary one are clearly unnecessary, and since a clearer, saner 
interpretation is possible they need not be considered. 
'This would be the usual or expected type of specificity if such thing should exist. 
I call attention to its implications merely to forestall any further thought regard- 
ing its possibility. The sort of specificity of enzymes that has thus far been 
assumed by the Mendelians has, however, been of a different sort ; namely, that for 
the production of each color only one enzyme is necessary, but the enzyme which 
produces any particular color is specifically different from those which produce other 
colors; the case is not really different from an assumption that pepsin is specifically 
different in different, but closely related races and varieties. Unlike the case cited 
above, here each germ possesses only one or two zymogens, each capable of super- 
vising the complete production of some one color, and therefore all the offspring 
could (in contrast to above) be provided with color. The only evidence that has been 
adduced for this sort of specificity is the rather incomplete and unsatisfactory results of 
Miss Durham already cited. All else has been mere assumption on the part of the 
Mendelians. But this type of specificity becomes a very unusual and extraordinary 
thing as soon as we find that all of the colors form in a continuous oxidation series. 
To cover this fact the assumption is obligatory that in melanin production, six, ten or 
a dozen tyrosin oxidizing enzymes are concerned ; that these are all able to take the 
first steps of tyrosin oxidation ; that they are differentiated merely by their " strength," 
that is, the extent to which they can carry the oxidations. 
338 OSCAR RIDDLE. 
The doctrine of numerous specific enzymes, 1 then, goes the 
way of the doctrine of specific cliromogens, which is so decisively 
settled by the work of Bertrand. 2 Remembering that the Men- 
delian description (by implication) of what is happening on the 
(color-producing) surface of the body in late stages of develop- 
ment is thus completely awry, we may not be surprised to find 
that their assumptions regarding conditions in the germ are in a 
similarly contradictory tangle. 
Our present knowledge permits neither the realization nor the 
imagination of a " color factor " in the germ ; not even in a 
simple form ; much less does it grant us the very composite and 
elaborate picture presented by Castle. 
The " production of color " is a special manifestation, in rather 
restricted regions of an organism, of a general po^ver to oxidise 
organic compounds, possessed, presumably, by all parts of the germ 
cell from which the organism arose. With the development of 
the body, the specialization of tissues, there arise very new en- 
vironments for the oxidative processes, producing localized 
changes and variations in this power. That this is so is evi- 
denced by the fact that the living substance of the various body 
regions oxidizes fats, sugars, and proteids with unequal ease. 
There can be, moreover, scarcely a doubt that certain of these 
regions, owing to new structure, new environment, new condi- 
tions, are able to oxidize different protein substances with variable 
ease, and to a variable extent, and even in a different way. 
When one has grasped the nature of the process by which 
melanin color characters are formed, there is as little necessity or 
truth in assuming that the germ cells contain representatives or 
determiners to correspond to a particular color, as there is in 
assuming that the vapors of the Gulf contain determiners for the 
depth, or distribution, of the mantle of snow which they are to 
1 Mendelians must, however, accept the specificity or non-specificity of enzymes 
in the production of color (they must have some sort of representative particle in the 
germ) and either choice leads, when critically examined, to the unqualified refuta- 
tion of some of their fundamental conceptions and interpretations of heredity. The 
established fact that each of the melanin colors represents but a point in a line a 
line which records the continuity of a continuous oxidation process is the fact that 
strikes hard at the very basis of Mendelian interpretation and description. 
2 Table I. shows, for example, that at least seven of the nine compounds repre- 
sented are capable of producing the colors yellow and red. 
MELANIN COLOR FORMATION. 339 
form on the northern hills. There exists, to be sure, a relation 
between the vapors and the heaps of snow, between the egg and 
the definitive character but the one is not the other, does not 
contain the other. 
In accounting for melanin color characters, I would maintain 
that -- granted the continuance of other life and developmental 
processes we can account for all the major happenings of color 
development and inheritance with extremely little of assumption. 
It is knoivn that one oxidase is concerned. I think we need 
make use of but one. It is knoivn that the germ possesses 
actively the power to oxidize amino acids. I think we need 
make use of nothing in the germ than just this power. It is 
known that the protoplasm of different species, of different tissues, 
of different parts of a cell, possess different powers of oxidizing 
protein bodies. It is no tremendous assumption that germ cells 
are not freaks of nature in this respect, and that they too have 
different powers of tyrosin oxidation. The fate of color charac- 
ters then is bound up (i) in the union of these particular powers 
of the two germ cells ; (2) in the origin (through other outside 
developmental processess) of favorable and unfavorable regions 
for tyrosin oxidations ; and (3) in environmental conditions. 
Let us now for a moment return to the matter of color-blends. 
The data given furnish a nearer view practically a new concep- 
tion of the nature of color-blends in inheritance ; if the position 
stated in regard to these color-blends is correct a little thought 
will convince that at the same time new light is furnished on alter- 
native color inheritance ; and particularly on what is happening 
in the case of the so-called alternative (Mendelian) inheritance of 
a color character. It can be said definitely in such a cross that 
it is the pcnver to oxidize tyrosin compounds (a power which I be- 
lieve, by no stretch of imagination, needs to be, or can be, repre- 
sented by a particle in the germ, but by a general property of 
the protoplasm of germ cells and of tissue cells) that is transmitted 
and that here this power of one of the gametes is continued into * 
1 Tyrosinase has recently been found in the eggs of cephalopods by Weindl ('07). 
Similarly, several enzymes have been isolated from germ cells in recent years ; in 
none of these cases can we for a moment suppose that these enzymes or zymogens 
existed as a particle in any one, or in each, of the chromosomes. I would, however, 
by no means have it inferred that I consider the presence, quantity, etc., of tyrosinase 
34O OSCAR RIDDLE. 
the zygote without being very appreciably increased or diminished. 
It may be, however, that more than one (almost certainly several 
for yellow) oxidation stage- of a tyrosin compound presents only 
the one color black. It could conceivably (see scale of colors, 
p. 326) happen, therefore, that a certain black (low stage of oxi- 
dation) x light yellow = yellow (dominant ?) ; but another black 
(highly oxidized) x yellow = black (dominant ?) ; and yet each 
of these ntigJit be true blends, i. c., attain to an exact intermediate 
stage of oxidation to the two forms crossed. Some cases of 
supposed dominance of color may be therefore in reality true 
examples of blended inheritance. 
In describing color inheritance I believe that less violence is 
done to the known facts of color formation, and at the same time 
a sounder view of developmental and hereditary processes is main- 
tained, if it be said - -without delimiting terminology, and with- 
out putting a single thing into the germ except what every one 
knows is there, and there in the form which is stated for it that 
in the union of germ cells derived from two pure color varieties, 
each cell brings with it a power of oxidizing tyrosin compounds, 
and that the union of the pair of cells may give one or more of 
the following results : 
1. The tyrosin oxidizing power of the male cell is established 
(a) at once, or (b] in the next generation or later, throughout the 
fertilized ovum and its derivatives. 
2. The tyrosin oxidizing power of the female cell is so estab- 
lished. (In neither of these cases do we need to postulate the 
continued existence of a subdued recessive factor or repre- 
sentative.) These would be so-called dominants. 
3. The oxidizing power resulting from the union is somewhat 
more, or somewhat less, than that of either of the gametes. 
in the egg as an index, measure, or determiner of the tyrosin oxidation powers of the 
adult. Secrets of protoplasmic differentiation, zymogenesis, stereochemistry, and 
catalytic action, all block for the present the tracing or predication of any such 
relations. 
Undoubtedly enzymes have a very considerable importance in development. By 
focusing attention too closely upon them it is possible, however, to underestimate 
the importance of other, and even of related phenomena in which the possibility of 
"inherited specificity" is entirely eliminated. I refer particularly to the "auto- 
catalysis" of such substances as oils studied by Guenthe ('07), Mathews, Walker and 
the writer ('o8<$), and others. 
MELANIN COLOR FORMATION. 34! 
4. The result of the union is a blend ; /. e., an oxidizing 
power intermediate to that of the two gam'etes. 
Numbers I and 2 represent colors at points of fairly fixed color 
equilibrium, as proved by the fact that individuals, varieties and 
species tend to stop color formation at those points ; and most of 
the offspring of such hybrids may reasonably be expected to 
breed true with reference to this character, because of such stable 
equilibrium. Categories 3 and 4 often represent, on the other 
hand, colors at points of unfixed equilibrium ; stages in the oxi- 
dation of tyrosin are not easily held at these points ; that such 
points of unstable equilibrium arise in the chemical building of 
melanin, as elsewhere, is practically certain ; this unstable condi- 
tion is followed by an immediate tendency - - in the second (next) 
generation usually --to shift to one or both of the stable points 
represented by the male and female condition, or to a new point. 1 
The above considerations seem to be of far-reaching applica- 
bility. They are, I think, rigorously consistent with what we 
know of the oxidation process, and with the various facts of 
melanogenesis ; whereas Mendelian interpretation is consistent 
with neither. How much of the totality of color-inheritance is 
thus brought under one point of view will be at once appreciated 
by naturalists ; while in striking contrast is the very small frac- 
tion of such inheritance that can be brought into the Mendelian 
system, even with all its elaborations. 
In following out the implications of our conception of the state 
in which these color characters exist in the germ, it may be said 
that hybrid offspring possessing a color of easy, fixed equilibrium, 
mated with similar forms may usually be expected to breed true 
(that is, to continue this oxidizing power into their germ cells) 
with respect to this character. If mated, however, with another 
variety possessing some very different character, let us say size, 
which is also in very stable, fixed equilibrium, it seems quite 
1 These four types give nothing of qualitativeness nor of discontinuity ; we have to 
deal absolutely in these initial stages with quantity, degree or pitch of oxidation 
power, and the gaps which we find in the end-result of the development of the color 
characters are but the cumulative, final expressions of different degrees of oxidation 
power, and of the fact that certain stages of oxidation are more stable in firmer 
equilibrium than others. 
342 OSCAR RIDDLE. 
conceivable almost a necessity that each variety should at 
least sometimes be able to force its stable character (if these do 
not both rest directly upon the same process, e. g., ty rosin oxi- 
dation) practically unchanged into a new combination and pro- 
duce a new form --an animal with the color of one parent and 
the size of the other. 
This clearly brings us face to face with something resembling 
unit characters and particulate inheritance. I can see no reason, 
from studies on the nature of color formation, and from the 
necessary deductions as to the way in which color must be trans- 
mitted, to doubt the possibility or the probability of the formation 
of races with new characters, or rather a race with a new combi- 
nation of old characters ; and with some such newly combined 
characters in very stable equilibrium, i. e., breeding true. In 
fact, it is the recognition of the state in which our color character 
exists in the germ, that is, as a given pitch of power for oxidizing 
tyrosin compounds, and this not latent but active in the germ as 
later, that enables us to view the mechanism by which such a 
result is brought about. 
The demonstration of the existence of such combinations of 
characters is, I believe, the supreme contribution of Mendelism to 
our knowledge of heredity. Phenomena of dominance, of segre- 
gation (?) and proportion, are but minor and special manifesta- 
tions of a process much more important, general and inclusive ; 
and which general process is, in color inheritance at any rate, 
the propagation and occasional quantitative modification (four 
types above) of oxidizing powers, with their more or less con- 
stancy of expression through settling into points of easy or fixed 
oxidizing equilibrium. 
Our view does not, however, allow the acceptance of the unit 
character hypothesis without very considerable and rather radical 
modification. The prevailing idea among Mendelian workers 
has been essentially that each character, each recognizable differ- 
entiation, each member of a group of factors that forms a char- 
acter, is quite separate and capable of being shuffled in the germs, 
and of independent appearance in the zygote. Now, as already 
noted, experience with the melanins drives home the point that 
a long series of very distinct characters have not each even one 
MELANIN COLOR FORMATION. 343 
representative, but all together have one basis in the germ 
a power to oxidize tyrosin compounds, and this capable of close 
and continuous gradations. Many other characters, moreover, 
may also be closely connected genetically with the tyrosin oxi- 
dizing power of the organism. This being true, it is easily seen 
that what has been quite generally taken to be a " unit char- 
acter " among colors cannot be in fact a " unit " of modification. 
All tyrosin oxidations, and some others as well, may be, and 
probably are, modified simultaneously and in corresponding 
manner. 
The conception of " units of modification," then, which we 
adopt as an apparent logical necessity cannot regard as " units " 
the things that have been called unit characters in the past ; quite 
certainly some unsound criteria, a false interpretation, a mis- 
leading nomenclature, and some observed facts are all back of 
the old conception as it has been applied to color behavior. , We 
need not be surprised if it be found to embrace very little truth. 
To determine what a real " unit of modification " is, would seem 
to be no easy matter. Present Mendelian methods will doubtless 
contribute something, though not nearly all, to the complete 
story. For if and there is little doubt other definitive char- 
acters of the soma are likewise present in the germ only as vary- 
ing "strengths" of rather general powers or processes, these 
same powers will exercise influence directly and indirectly, and 
to greater or less extent, in many and diverging directions during 
development ; so that a " unit of modification " in inheritance 
would, in the broadest sense, include all such effects (there would 
probably be found all gradations of such effects). Obviously, there- 
fore, individual characters arc by no means units of inheritance or 
modification. And in any exhaustive search for such influence 
or modification it is quite possible that vision will sometimes be 
forced to cover the whole field from antibodies and immunity to 
size and color ; from the grossest structural modification to the 
most delicate functional idiosyncrasy. 
The question will be asked how may those who reject the 
Mendelian interpretation based on representative particles, account 
for the segregation and proportions observed in Mendelian 
behavior? To undertake a discussion of this point is obviously 
344 OSCAR RIDDLE. 
to leave the province of fact, with which the body of this paper 
deals to enter the realm of assumption and hypothesis. The 
fact, that in these pages we have been able to get what we are 
inclined to consider the clearest view that we have at present as 
to the state in which a character or set of characters exists in the 
germ, perhaps furnishes a reason why some statement may here 
be made in regard to the possible mechanism of segregation. 
It has already been stated that to us the phenomena of domi- 
nance and segregation are only minor, surface, and incidental 
phenomena of heredity j 1 the really important Mendelian contri- 
bution being that certain different characters (such as have, ac- 
cording to my belief, different rather general processes as a basis) 
of different races may be combined to form new fixed races. The 
establishment of this last-named fact has been most commonly con- 
sidered by Mendelians as, on the one hand, a consequence of the 
laws of dominance and segregation, and, on the other hand, as a 
strong argument for a " representative particle" basis for these 
two sets of phenomena. When, however, we learn that a certain 
character has no other existence in the germ than a rather 
general protoplasmic power, the " mnemon " conception fails 
completely and with it the supposed mechanism of its segrega- 
tion /. e., the shuffling of mnernons in the reduction divisions 
of the chromosomes. 
How then on my view of the basis of color inheritance may 
the segregation and proportions which result in Mendelian 
behavior be accounted for ? I may say at once that I do not 
know ; but in this respect I consider myself hardly worse off than 
the wisest Mendelian. My supposition, however, would put less 
faith than is theirs in the behavior of chromosomes during the 
1 It would seem that instead of viewing the real and entire sea of heredity, learn- 
ing of its intimate nature, searching its boundaries, sounding its depths, too often 
Mendelians have to indulge a figure focused their visual instruments upon an 
optical section lying perhaps a few meters above sea-level. Here, occasionally, 
beautiful and regular phenomena come into sight; but for the most part the field is 
blank. At times great ocean swells pass in fine order and precision, permitting the 
observer to predict quite well some of the attributes of the next undulation and the 
time of its appearance ; again, bits of spray or foam or mist may sometimes come into 
view, and the seeming disconnectedness of it all permits the be-focused onlooker to 
name and classify and wonder. All while the great ocean of heredity with its 
perfect continuities, its essential oneness, its inclusiveness, lies in unseen constancy 
and majesty beneath. 
MELANIN COLOR FORMATION. 345 
maturation divisions. It may for the present be assumed, and 
this is, I think, all that is really demanded in accounting for the 
differences in end-result in color development and inheritance - 
that the germ cells vary in "strength," i. e., in such general 
powers as assimilation, growth, oxidation, etc. (and this proposi- 
tion is not all assumption) ; this general difference of " strength" 
of germ cells may (or may not) arise during the maturation 
divisions ; one or more of the four cells receiving in some species, 
though not in all, a type of protoplasm in better or worse condi- 
tion than the others --influenced, for example, by more or less 
yolk ; yolk more or less affected by previous nuclear and cyto- 
plasmic contact ; by variable distribution of the protoplasm of the 
astrosphere ; by variable admixtures of nuclear (not necessarily 
chromatin) and cytoplasmic matter ; by receiving more or fewer 
entire chromosomes, etc. (Chromatin is well known to be a very 
reactive substance and we may very well believe that wide varia- 
tions of it in amount influence the vigor or intensity of vital proc- 
esses occurring in a cell and in its derivatives ; but in this respect 
chromatin is not unique, the other variables just mentioned doubt- 
less do the same.) Any or all of these things will influence the 
development of a character only by serving to strengthen or 
weaken some process which underlies its formation. 1 
Such general differences of germ cells as arise from the several 
possible causes mentioned above would conceivably tend to affect 
the strength of such a general process as oxidation such as 
produces scales of color. From the nature of these differences 
it will be seen that the growth and maturation stages may occa- 
sionally among organisms constantly perhaps in some forms - 
furnish the conditions for the production of four sister sperm 
cells of unequal strength, one or two may be especially favor- 
1 It is perhaps well in our estimation of the basis for segregation of color characters, 
which rest directly on an oxidation process, to call attention to the special significance 
of the centrosome and cytaster. Considerable evidence is adduced by Mathews ('07) 
to show that the centrosome is the reduction center of the cell. Jf this should prove 
true, very obviously our estimation of the oxidation powers of the cell would be better 
treated in relation to this body than to any other in the cell. It cannot be too much 
emphasized, "however, that upon this view the centrosome and the sphere is but a 
region where intense reduction occurs ; the intensity grading oft from centriole to 
(presumably) the periphery of the cell ; and this region should be thought of as an 
expression of general powers of the whole cell. 
34-6 OSCAR RIDDLE. 
ably equipped by the distribution of such materials, while one 
or two are left poorer than the others. 1 A similar conception 
may be applied to the ova. ' Then upon the union of male and 
female cell, two oxidizing powers of equal or unequal rank, of 
higher or of lower degree, meet. A stable (pure breed) high 
equilibrium results in, say one of four ; a stable (pure breed) 
lower equilibrium results in another ; while in the other two per- 
fect equilibrium is not at once attained. 
Here is then a possible picture of the basis of Mendelian segre- 
gation and proportion, but without recourse to hypothetical 
"particles" or to immutable and immortal factors. An appar- 
ently very specific end-result of an oxidation would be traceable 
in the germ only in the strength or pitch of a general vital process, 
and not at all in mnemons or representative particles packed with 
unthinkable precision, order and potentiality into (presumably) 
the chromosomes. But the above is a possible picture only ; and 
it is not here my purpose to furnish nor to defend at length a pos- 
sible or probable theory of the mechanism of heredity. The 
material in hand lends itself first of all to the demonstration of 
the impossibility of many Mendelian views. 
The nature of present Mendelian interpretation and description 
inextricably commits to the " doctrine of particles " in the germ 
and elsewhere. It demands a " morphological basis " in the 
germ for the minutest pha^e (factor) of a definitive character. It 
is essentially a morphological conception with but a trace of func- 
tional feature. Although heredity is quite surely a functional 
process of major complexity, it may be recalled that the primary 
and fundamental Mendelian conception of this process utilizes 
not a single finding of the science of biochemistry ; that the only 
physiological fact utilized is the one of certain occasionally ob- 
served segregation behavior which is exhibited in the end-results 
of varietal 2 or specific character formation ; such segregation, by 
1 We know that in the corresponding divisions of ovogenesis that extremely dispro- 
portionate distribution of yolk and cytoplasm occurs quite constantly. We may be- 
lieve that the laws which there give rise to the extreme differences between polar body 
and oocyte are not everywhere else, and completely, unoperative ; they may be opera- 
tive though in a much less pronounced degree in providing different mature ova, 
and different sperms with varying amounts of the materials mentioned above. 
2 De Vries ('oi, '05) has asserted that only varietal, not specific, differences ex- 
hibit Mendelian heredity; this statement is not accepted by Bateson ('06), and is 
controverted by still other workers. 
MELANIN COLOR FORMATION. 347 
inference, arising from the temporary mixture (heterozygote), or 
failure to mix (homozygote) in the gamete, of something, no one 
knew what but which has been generally conceived of as some 
sort of " particles." (In later additions to, and special applications 
of the Mendelian conception, certain other biochemical and 
physiological facts have, of course, been considered.) 
This is precisely why present Mendelian interpretation and de- 
scription of heredity is a bar to the progress of studies in inherit- 
ance and 'development ; with an eye seeing only particles, and a 
speech only symbolizing them, there is no such thing as the study 
of a process possible. 
The conception that organic color has at its basis not rigid, 
immortal particles, but yielding, equilibrium-seeking powers, or 
strengths of processes, makes the infinite variety of colors in 
organisms intelligible. If, on the other hand, particles, and a 
mechanism for their continual segregation and propagation pure 
were in reality at the basis of color inheritance, we should rather 
expect uniformity, not the actual diversity, to be the dominant fea- 
ture of organic coloration. Indeed, a modification of the strength 
of many organic processes (and so of color formation) would be 
a necessary accompaniment a result, even if not a cause of 
that " transformation of organs " which has been the very labor 
of phylogenetic development. The atrophy, superior develop- 
ment and transformation of organs are certainly efficient factors 
in such modification, for it is a physiological fact of common ex- 
perience that many, or most, of the vital processes are not equally 
strong or pronounced in all of the organs of the same organism ; 
and that many metabolic processes of the body are dependent 
upon special organs for their highest expression, for the com- 
pletest manifestation of their power. 
Let me not be understood to say that our knowledge of the 
development of melanin color characters is complete. There is 
much yet to be learned. But the significant thing about it is 
that we now know so much of the mechanism of the building 
of these characters in comparison with what we know of a sim- 
ilar nature in non-color characters. It has been possible, I think, 
to show by means of what we know of the genesis of these color 
characters that the Mendelian description of color inheritance 
t < 
348 OSCAR RIDDLE. 
at least has strayed very wide of the facts ; it has put factors 
in the germ cells that it is now quite certainly our privilege to 
remove ; it has declared discontinuity where there is now proved 
continuity ; it has postulated preformation where there is now 
evident epigenesis. 
Is it too much to expect that the further application of such 
tests as the one here presented in outline for the melanin colors 
will in the end remove many of the Mendelian " factors " from 
the germ cells? That many of their " characters " will come to 
rest on a more proximate basis ; will be known to have their 
"determination " and origin in very general germinal powers, and 
in somatic conditions obtaining previous to, or at the time of, 
their development? Will not other Mendelian discontinuities 
then begin to disclose gradations, and other qualitative differences 
then appear more and more as quantitative sequences ? 
LABORATORIES OF EXPERIMENTAL THERAPEUTICS AND ZOOLOGY, 
THE UNIVERSITY OF CHICAGO, 
January, 1909. 
LITERATURE CITED. 
Abel, J., and Davis, W. 
'96 The Pigments of the Negro's Skin and Hair. Jour. Exp. Med., I., p. 361. 
Abderhalden, E., and Schittenhelm, A. 
'05 Ausscheidung von tyrosin und leucin in einem falle von cystinurie. Zeit. f. 
Physiol. Chem., XLV., p. 468. 
Albrecht, H. 
'02 Ueber ochronose. Zeit. f. Heilk. Path., Abt. 23, p. 366. 
Bateson, W., and Sanders, Miss E. R. 
'02 Report to the Evolution Committee. 
Bateson, W. 
'06 Progress of Genetics. 
Bertrand, G. 
'96 Sur une nouvelle oxydase ou ferment soluble oxydant, d'origine vegetale. 
Bull. Soc. Chim., 3d Ser., T. XV., p. 791, and C. R. Ac. Sc., T. 122, p. 
1215. 
'03 Sur les relations du chromogene surrenal avec la tyrosine. Compt. rend. Ac. 
Sc., T. 138. 
Bertrand, G., and Mutermilch, W. 
'07 Recherches sur le mode de coloration du pain bis. Ann. Inst. Pasteur, T. 
21, p. 833; also C. R. Ac. Sc. and Bull. Soc. Chim., '07. 
Bertrand, G. 
'08 Recherches sur la melanogenesis : Action de la Tyrosinase sur la Tyrosine. 
Ann. de Inst. Pasteur, Vol. 22, No. 5, p. 381. 
Biedermann, W. 
'98 Beitrage zur vergleichenden physiologic der verdauung. I. Die verdauung 
der larve von Tenebrio molitor. Pfl tiger's Archiv, 72, p. 105. 
MELANIN COLOR FORMATION. 349 
Castle, W. E. 
'06 Mice and Gametic Purity. Science, N. S., 24, No. 609, p. 275. 
Chittenden, R. H., and Albro, Alice H. 
'gg Formation of Melanins or Melanin-like Pigments from Proteid Substances. 
Amer. Jour. Phys., IT., p. 291. 
Cuenot, L. 
'03 L'heredite de la pigmentation chez les souris. Arch. Zool. Exp., 4, No. 3, 
P- 33- 
'05 Les races pures et leurs combinaisons chez les souris. Ibid., 4, T. 3, Notes 
et revue, p. 123. 
DeVries, H. 
'01 Die mutationstheorie. Leipzig ('oi and '03). 
'05 Species and Varieties, Their Origin by Mutation. Chicago. 
Dewitz, J. 
'02 Recherches experimentales sur la metamorphose des insectes. Compt. rend., 
T. 54, p. 44. 
t Duccheschi, V. 
'oi Sulla nature delle melanine e di alcune sostanze ad esse affini. Rendic. 
della R. Acad. dei Lincei, Cl. di Scienzi fisiche (5), Vol. X., Fasc. 5, 
Seduta del 3 Marzo. 
Durham, F. 
'04 On the Presence of Tyrosinase in the Skins of Some Pigmented Vertebrates. 
Prelim. Note. Proc. Roy. Soc. London, Vol. 74. 
Garrod, A. E. 
'02 The Incidence of Alkaptonuria : A Study in Chemical Individuality. Lan- 
cet (2), p. 1616. 
Garrod, A. E., and Hartley, W. H. 
'06 Concerning Cystinuria. Jour. Physiol., Vol. 34, p. 217. 
Gessard, M. C. 
'oi Etudes sur la tyrosinase. Ann. Inst. Pasteur, T. XV., p. 593. 
'O2a Tyrosinase animale. Compt. rend. soc. biol., T. 54, p. 1304. 
'o2b Antityrosinase animale. Ibid., p. 1398. 
Sur lesoxydases des Seiches. Compt. rend. Ac. Sc., T. 136, p. 631. 
Sur la formation du pigment melanique dans les tumeursdu cheval. Compt. 
rend. soc. biol., T. 136, p. 1086. 
'030 Ueber das Pigment des Nebennieren. Compt. rend.,T. 138, p. 586. 
'O3d Sur les reactions colorees consecutives a 1' action de la tyrosinase. Compt. 
rend., T. 138, p. 774. 
Gmelin, L. 
'g4 Ueber das schwarze Pigment des Auges. Beitrage z. Chemie u. Physik von 
Schweigger, Bd. X. 
Gonnermann, M. 
'oo Homogentisinsaure die farbebedingende substanz dunkler Riibensafte. 
Pfluger's Archiv, Bd. 82. 
Hlasiwetz, H., and Habermann, J. 
'73 Ueber die proteinstoffe. Liebig's Annalen, 169, p. 150. 
Hopkins, F. G., and Cole, S. W. 
'oi, '03 A Contribution to the Chemistry of Proteids. Part I. and II. Journ. 
Phys., 27. 
35O OSCAR RIDDLE. 
Landolt, H. 
'99 Ueber das Melanin der Augenhaute. Zeit. f. phys. Chem., Bd. 28. 
Lepinois, M. E. 
'99 Etudes sur le chromogene des capsules surrenales et sur 1'origene de la colora- 
tion rouge que ces glandes prennant au contact de 1'air. Compt. rend. Soc. 
Biol., Vol. 51, p. 315. 
Mathews, A. P. 
'07 A Contribution to the Chemistry of Cell Division, Maturation, and Fertiliza- 
tion. Am. Jour. Phys., Vol. 18, p. 89. 
Mendel, G. 
'65 Versuche ueber Pflanzen-Hybriden. Abhandl. Naturf. Ver. in Brunn, IV. 
Also English translation by Bateson, Jour. Roy. Hort. Soc., XXVI., 1901. 
Nencki, M. 
'95 Zur Kenntniss der pankreatischen Verdauungsprodukte der Enweisskorper. 
Ber. Dtsch. Chem. Ges., Bd. 28. 
Nencki, M., and Berdez, J. 
'85 Ueber die Farbstoffe der melanotischen Sarkome. Arch. Exp. Path, und 
Pharm., Vol. XX., p. 346. 
Neubauer, 0. 
'08 Ueber den Abbau der Animosauren im Gesunden und kranken Organismus. 
Habitationsschrift-Vogel, Leipzig. 
Osier, W. 
'04 Ochronosis : The Pigmentation of Cartilages, Sclerotics and Skin in Alkap- 
tonuria. Lancet (i), p. 10. 
Pforringer. 
'oo Zur entstehung des hautpigments bei Morbus Addison. Cent. f. Allg. 
Path. (u). 
Powers, J. H. 
'08 Morphological Variation and its Causes in Amblystoma tigrinum. Studies 
from Zool. Lab. of Nebraska, No. 71. 
Pick, L. 
'06 Ueber die ochronosis. Ber. klin. Wochenschr. (43), p. 478. 
Ravenna, E. 
'03 Beitrage zur Histogenese der Melanotischen Hautgeschwiilste. Virchow's 
Arch., Bd. 171. 
Riddle, 0. 
>o8a The Genesis of Fault-bars in Feathers and the Cause of Alternation of Light 
and Dark Fundamental Bars. Biol. Bull., XIV., p. 328. 
>o8b (With A. P. Mathews and S. Walker. ) The Spontaneous Oxidation of Some 
Cell Constituents. Abst. in Jour. Biol. Chem., IV., June, p. xx. 
Samuely, F. 
'02 Ueber die aus Ei weiss hervorgehenden Melanine. Hofmeister' s Beitrage, II. 
Schmiedeberg, 0. 
'97 Ueber Elementarformeln einiger Eiweisskorper und ueber die Zusammen- 
setzung und die Natur der Melanine. Arch. Path. u. Pharm., Bd. 39. 
Spiegler, E. 
'03 Ueber das Haarpigment. Hofmeister's Beitrage, Vol. IV. 
Stadelmann, E. 
'90 Ueber das beim Tiefen Zerfall der EiweisskSrper entstehende Proteinochro- 
mogen den die Bromreaction gebenden Korper. Zeit. f. Biol., Vol. 26. 
MELANIN COLOR FORMATION. 35! 
Tornier, G. 
'073 Experimentelles ueber Erythrose und Albinismus der Kriechtierhaut. Sitz'- 
her. Ges. Nat. Freunde, Berlin, p. 71. 
*O7b Nachweis ueber das Entstehen von Albinismus, Melanismus und Neotemie 
bei Froschen. Zool. Anz., Bd. 32, p. 284. 
'o8a Ueber eine Albinotische Ringelnatter und ihre Entstehen. Sitz'ber. Ges. 
Nat. Freunde, Berlin, No. 8, p. 196. 
'o8b Vorlaufiges ueber experimentell erzielten Hautalbinismus bei Axolotl-Larven. 
Sitz'ber. Ges. Nat. Freunde, Berlin, No. 3. 
'o8c Vorlaufiges ueber das Entstehen der Goldfischvassen. Ibid., No. 3. 
von Fiirth, 0. 
'99 Ueber die Einwirkung von Salpetersaure auf Eiweisstoffe. Habilitato- 
Schrift., Strassburg. 
'04 Physiologische und chemische Untersuchungen ueber melanotische Pigmente. 
Centralb. f. All. Path., Bd. XV. (older literature). 
von Fiirth, 0., und Schneider, H. 
'01 Ueber tierische Tyrosinasen und ihre Beziehungen zur Pigmentbildung. 
Hofmeister's BeitrSge, Bd. I. 
Weindl, Theod. 
'07 Pigmententstehung auf Grund vorgebildeter Tyrosinasen. Arch. Ent'mech., 
Bd. 23, pp. 632. 
Wilcock, E. G. 
'06 The Action of Radium Rays on Tyrosinase. Journ. Physiol., Vol. 34, p. 
207. 
Wolff, H. 
'04 Zur kenntnis der Melanotischen Pigmente. Hofmeister's Beitrage, V., p. 
476. 
OCR text unavailable for this page.OCR text unavailable for this page.OCR text unavailable for this page.OCR text unavailable for this page.OCR text unavailable for this page.