RE C A PITU LA TI ON 
OF THE 
.EMBRYOLOGY OF THE TURTLE, 
AS GIVEN IN 
PROFESSOR AGASSIZ’S “CONTRIBUTIONS TO THE NAT- 
URAL HISTORY OF THE UNITED STATES OF 
NORTH AMERICA.”—Vol. II, Part III. 
HY 
H. JAMES CLAHK, of Cambridge, Mass. 
EXTRACTED FROM THE AM. JOURNAL OF SCIENCE AND ARTS, VOL. XXV, MAY, 1858. 
NEW HAVEN: 
P KIN TED BY E. HAYES. 
1 8 5 87 [From the American Journal of Science and Arts, Vol. 25, p. 342.] 
RECAPITULATION OF THE “EMBRYOLOGY OF THE TURTLE,” 
AS GIVEN IN PROF. AGASSIZ’S CONTRIBUTIONS TO THE NATURAL 
HISTORY OF THE UNITED STATES, VOL. II, PART III. 
H. JAMES CLARK, of Cambridge, Mass. 
The following summary, of the “ Embryology of the Turtle,” 
was undertaken at the request of Professor J. D. Dana, one of 
the editors of this Journal. For certain reasons, given below, 
it will be seen that neither comment nor criticism upon the work 
is in place here. The method adopted in writing out this sub- 
ject, as it stands in the original work, is so far from the aphoristic 
style that I find it will be next to impossible to make extracts, 
which will give the pith of the matter, without inserting here a 
great portion of the whole. On this account I shall be compelled 
to rewrite, whatever may be presented, in a condensed form, 
with perhaps here and there a short extract. 
Had it not fallen to my lot, as Professor Agassiz’s assistant, 
to write out the embryological portion of these “Contributions” 
I would not have dared here to take the liberty of reconstruct- 
ing the fabric of the story of the development of the Turtle. 
As it is, it will not be possible, for want of room, nor really 
necessary, to render account of the whole history as originally 
written, but I will confine myself mostly to what is new in this 
department of science. 
Whenever anything is presented to the scientific world as 
new and especially controvertive of old and perhaps apparently 
well established views, every caution is necessary in proving 
that all liability to error has been guarded against whilst pursu- 
ing the investigations. 
On this account considerable care has been taken, in the open- 
ing part of the “ Embryology,” to state clearly, and how that 
the egg and embryo, so called, were kept in as natural a condition 
as possible. In the study upon the origin and development of 
the egg, “ a young animal was resorted to on account of the 
greater abundance of the smallest sized eggs and also because 
the ovary is less opaque than in the adult." 
The origin of the egg.—The ovary was cut out entire and 
floated in serum, so that the eggs might not be distorted by pres- 
sure nor by pulling and tearing in order to get them under the 
microscope. In the first place a comparative view was taken of 
the whole mass of the eggs, from the youngest to the oldest, 
with a low magnifying power, and in this manner a rapid survey 2 
Agassiz on the Embryology of the Turtle. 
was made of the respective condition of each egg, and the mind 
prepared in part, by anticipation as it were, to more readily 
comprehend the relation of the phase of any one to that of any 
other or of the whole. 
The physiognomy of the egg in its younger stages is so pecu- 
liar, with its thick dark outline, brilliant, strongly refractive, 
homogeneous, yellowish contents, and the lateral nucleus, that 
it cannot be mistaken for one of the cells, of the corpus graafia- 
nurn, which press upon it from all sides. “The initial form of 
an egg is a dark, oily looking, granule-like, spherical body, situ- 
ated among the interstices* of the cells of the corpus graafianum. 
As the latter not only, but even their nuclei surpass such an 
egg in size by several diameters, it is superfluous to debate the 
question, whether the egg may not be the nucleus of a cell of 
the generating organ.” 
At first the egg has no wall about it, but finally by a differen- 
tiation of the superficial particles a consistent envelop is elabo- 
rated and answers to the name of the vitelline sac. 
The origin of the Purkinjean vesicle.—About this time, or soon 
after, the Purkinjean vesicle—germinal vesicle oftentimes called 
—becomes visible, by a condensation of a portion of the homo- 
geneous yolk against the inner surface of the vitelline sac. The 
mode of origin of this vesicle is very important to notice, be- 
cause it bears reference, in a very pointed manner, to the theory 
of the origin of free cells. Here it is evident that the egg-cell 
did not originate, as is usually stated of cells, around its nucleus, 
but that its nucleus is the offspring of the cell which encloses it. 
The Purkinjean vesicle always remains close to the parieties of 
the egg, even to the time when it disappears, and in no way 
points to a falsely claimed relation to fecundation and the origin 
of the embryo. The wall of this vesicle originates like that of 
the vitelline sac, about a previously conglomerated mass of 
particles. 
The appearance of numerous vesicular bodies, known as the 
Wagnerian vesicles, upon the inner surface of the wall of the 
Purkinjean vesicle, completes the operations, which have here 
been going on, necessary to the perfecting of the egg-cell, and 
the rendering of it, although a cell, totally different in properties 
from all other apparently similar cells. 
The development of the yolk.—It will be more convenient after 
this, on account of the complicity of the contents of the egg, to 
* “ The first blood corpuscles are yolk-cell nuclei which have undergone changes 
identical with those of the whole ‘ embryo,’ and they alone remain free, circulating 
in the channels hollowed out in a mass of cells identical with themselves. These 
are the first cells originating interstitially, but yet, after all, not essentially so, as is 
the case with the egg; for each blood corpuscle is a segment of an original yolk-cell 
nucleus, which has gone through the process of self division; whilst the egg originates 
just as the primary yolk cell does, by conglomeration of particles, and the formation 
of a membrane around the parieties of this concretion.” Agassiz on the Embryology of the Turtle. 
3 
treat of its different constituents separately. We will commence 
with the yolk first, as that is the fostering substance from which 
all the other components essentially originate. Previously to 
the development of distinct yolk cells, the yolk passes through 
some peculiar granulated phases, the most remarkable of which 
is the gradual encroachment of granules, commencing at one 
side of the egg, upon the hitherto homogeneous contents, till 
they pass across the whole bulk of the vitellus. At one time, 
during this progressive incursion of the granules, the egg appears 
as if a half of two different eggs had been stuck together, one- 
half being homogeneous and the other thickly granulated. After 
the granular stages, at the time the egg is about one-sixteenth of 
an inch in diameter, the oily looking granules gradually disap- 
pear and at the same time minute hyaline, albuminous, vesicular 
bodies begin to develop. These again, as is the case with the 
egg-cell and Purkinjean vesicle, originate without the interven- 
tion of a so-called nucleus,* and each one grows for some time 
without the least sign of a second body within its wall. 
“Here then we have essentially, nay in every sense, a cell, a 
hollow layer of spherical surface derived from the lateral adhe- 
rence of the superficial particles of a homogeneous globule. It 
is not a cell formation by the hollowing out of a solid substance, 
forming at first a very thick wall, which would stretch by the 
increase of the contents, as it gradually surrounds a larger space, 
till it thins out to the ordinary crassitude of such envelops. 
Never, throughout the whole range of cell development in the 
egg, is there the merest hint of this mode of genesis. From the 
beginning to the end of the growth of the ectoblast it ever pre- 
serves the same thin stratum, apparently of a single layer of 
corpuscles, and moreover the same tenderness and the same re- 
fractory power. Nor can we compare this process to the re- 
* “Thus far we have employed, in our descriptions of the egg and its contents’ 
the nomenclature generally in use to designate its different parts, and those of the 
cell. But this nomenclature, framed to express particular views respecting the 
mode of formation and the functions of these parts, is completely theoretical in its 
meaning. It appears desirable, therefore, now that we are about to consider more 
fully the origin and successive growth of the yolk cells, to discard every technical ex- 
pression which may imply a theory, and to adopt such only as designate the natural 
relations of the objects under consideration, especially since the views to which we 
have arrived cannot be reconciled with the theories which the current nomenclature 
is intended to express. These parts are therefore designated in the sequel by the 
following names: ectoblast is applied to the outer envelop; mesoblast to the so-called 
nucleus; entobla.it to the so-called nucleolus; and, when this contains a still smaller 
body, this is called entost/iobla.it. 
In the nomenclature of the egg. similar objections may be raised against the use 
of germinal or germinative vesicle and dot, as neither of these parts has the slight- 
est reference to the formation of the germ. We shall therefore designate them, 
henceforth, as some embryologists do, by the names of the Purkinjean and Wagne- 
rian vesicles. Applying our nomenclature to a comparison of the egg with the cell, 
the yolk-membrane is to be considered as an ectoblast, the Purkinjean vesicle as a 
mesoblast, the Wagnerian vesicle as an entoblast, and the Valentinian vesicle as an 
entosthoblaxt.” Agassiz on the Embryology of the Turtle. 
4 
ceived mode of cell origin according to winch a wall is con- 
densed around and upon a ‘nucleus,’ for the mesoblast is often 
absent in quite large cells; in fact, an egg little more than one- 
sixteenth of an inch in mean diameter contains numerous cells 
of considerable size, no one of which contains a mesoblast. 
Nor can it by any possibility be advocated, that these cells are 
the contents of other cells, for no others exist; even in a much 
larger egg, up to full grown ones, this holds good just as un- 
doubtedly, for in such a mass of yolk as larger eggs contain, 
the mesoblasts and ectoblasts have respectively very peculiar 
and unmistakable properties, not to be confounded with any 
other cell contents.” 
No ectoblast, as these albuminous hyaline vesicles are called, 
which has attained to a diameter of jo-th of an inch, is with- 
out a mesoblast. This latter body, like the Purkinjean vesicle, 
originates as a conglomeration against the wall of the cell,—the 
ectoblast,—which directly encloses it, but as soon as it has be- 
come well defined in outline it breaks away from its attachment 
and takes up a position in the centre of the parent cell. 
Notwithstanding that the mesoblast soon loses its globular 
shape, which free cells usually assume, and passes through some 
rather sharply angular phases, and finally ends with becoming 
more or less irregularly spheroidal, it appears to have a superfi- 
cial layer of its mass so consistent as to constitute a wall. Inas- 
much as a cell wall has been found, here, in these investigations, 
to have no definite density and in some cases to be hardly differ- 
entiated from the substance which it contains, it has been thought 
best to extend the definition hitherto used, and characterize it as 
“ a hollow, more or less spherical layer, of indefinite density, 
tenacity, and refraction, which surrounds the field of some defi- 
nite though isolated and homogeneous function.” 
By the time the egg has attained to its ultimate size, within 
the ovary, the mesoblast has grown so large as to almost com- 
pletely fill the ectoblast. The size of the mesoblast, at this pe- 
riod, is enormous, far exceeding in this respect the mesoblast of 
any other free cell known, being about TIoth °f an inch in 
diameter. 
Perhaps the most remarkable of all the constituents of the 
egg are the entoblasts of the yolk cells. These bodies appear 
about the time the egg has reached from one-eighth to one-sixth 
of an inch in diameter; a single one at first originates in the 
centre of the mesoblast, and not at the side as we have seen in 
other cases, but soon others take their places around the first. 
The marked peculiarity of these bodies is their sharp angularity, 
from the time of their origin, which increases till they resemble 
spiculse; and eventually fill, to surfeiting, the mesoblast. Then, 
when the egg has reached a diameter of one quarter of an inch, 
these angular, crystalloid entoblasts begin to lose their corners, Agassiz on the Embryology of the Turtle. 
5 
and many to disappear altogether, whilst new ones, with rounded 
angles and more equilateral, develop in place of these. 
The Purkinjean vesicle.—We have already described the mode 
of origin of the Purkinjean vesicle, and mentioned the appear- 
ance of the Wagnerian vesicles. These last are developed at 
first few in number, but subsequently they cover, like drops of 
dew, the whole inner surface of the wall of the Purkinjean 
vesicle. Like all cell development heretofore described, the 
Wagnerian vesicle begins with a faintly visible conglomeration 
of particles, which eventually obtains a well defined outline. 
The Valentinian vesicle is remarkable for having very little re- 
fractive power, so as to appear like a flat disc in the midst of 
the parent cell, the Wagnerian vesicle. 
By the time the egg has become one quarter of an inch in 
diameter, the Wagnerian vesicles disappear and give place to an 
almost total homogeneity of contents in the Purkinjean vesicle. 
From this time forward the latter grows more and more albu- 
minous, and may be hardened, by boiling the egg, so as to be 
taken up on the point of a knife. This albuminousness is so 
marked in the mature ovarian egg that boiling contracts the ve- 
sicle by nearly one-half, and the side next the yolk sac becomes 
collapsed.* 
The Zona pellucida.—The zona pellucida, which eventually 
takes the place of the yolk sac, when the latter becomes resorbed, 
* “ The clear space, observable in the egg of various animals just previous to 
segmentation, to which the name of “ embryo-cell” has been given, from its supposed 
intimate connection with the formation of the germ, may be identical with the white 
area about the Purkinjean vesicle observed in Testudinata. We would take this 
opportunity to express the opinion, that very probably too much stress has thus far 
been laid upon the assumption that the Purkinjean vesicle performs a peculiar and 
exclusive function in reference to the formation of the so-called embryo-cells; and, 
moreover, that the Purkinjean vesicle is not to be so definitely separated, as regards 
its essential element, from the immediately juxtaposed substance of similar appear- 
ance, but should rather be looked upon as the crowning point of albuminous con- 
centration, to which the opposite side of the egg stands in the reverse extreme of a 
highly oleaginous nature. A reference to the mode of origin of this vesicle shows 
this conclusively; for it is developed as a phase of secondary accession in the egg- 
evolution, and not as the primary basis to a succeeding structure ever after retain- 
ing a significance of superior import, and leading, as some would have it, to its be- 
coming in the end the essential element in the genesis of the embryo. This mode 
of origin alone, we maintain, is sufficient to show that the very foundation upon 
which its importance is laid cannot be tenable, in this light. The Purkinjean vesicle, 
therefore, loses all its advocated claims to preponderance over the rest of the egg 
constituents; to say nothing of the fact that it takes no part in the building up of 
the blastoderm, excepting that its discharged contents may become absorbed in the 
endosmotic and exosmotic interchanges of substances between the oily yolk cells, 
and the albuminous matter in which they float. True enough, the region about this 
vesicle exhibits a specialized nature; it is there that the embryo first develops cer- 
tain of its characteristics, previous to its further extension; but it does not follow, 
that because the Purkinjean vesicle is situated thereabout, it is the basis of this evo- 
lution, or in any way causatively connected with it. On the contrary, its presence 
is itself rather the result of certain tendencies, for instance, the concentration of 
albumen in that direction; and its disappearance also is the consequence of the 
consummation of these tendencies.” 6 
Agassiz on the Embryology of the Turtle. 
is formed by a stratification of the inner cells of the Graafian 
follicle, and therefore is not developed inside of the yolk sac, as 
has been asserted of it by some authors when treating of its 
origin in the eggs of other animals. Eventually the zona ap- 
pears as if made up of columnar cells, and thus remains during 
the rest of its existence. It may be detected even as late as 
when the amnios has begun to form.* 
The Embryonal Membrane.—The embryonal membrane is a 
layer, originating by a transformation of the superficial part of 
the yolk when the egg is hardly visible to the naked eye, and 
the formation of an excessively transparent cellular tissue, which 
envelops the whole vitelline mass. In its earliest stages it can- 
not be compared nor homologized with anything heretofore no- 
ticed in the eggs of animals, but when the embryo has begun 
to develop, it occupies the same relative position as the “Keim- 
blase” of Bischoff, or the “ Umhullungshaut” of Keichert, and the 
“Investing membrane'1'1 of English authors. Yet, insomuch as the 
“Investing membrane'1'1 of these authors, according to them, does 
not develop till the embryo begins to form, it cannot, for a cer- 
tainty, be homologized with the embryonal membrane of the tur- 
tle’s egg. But it is remarkable that this membrane like the 
“investing membrane” follows all the curvatures of the dorsal 
region of the embryo and even becomes a lining to the spinal 
tube. At the time the turtle is hatched this membrane may be 
detected as readily as in any of its earlier stages. 
Fecundation.—Professor Agassiz’s observations are so remark- 
able, and the facts in regard to fecundation so unprecedented, 
that I think it worth while to quote here the whole of the sec- 
tion he has written on this subject. 
“ Ever since I have known that our Turtles lay only once a year, I 
have been struck with the fact that the ovary nevertheless contains eggs 
* “Thompson (article Ovum in Cyclop. Anat. p. 78) compares the early yolk-sac 
of Birds (which he hardly admits as a true membrana vitelli, notwithstanding Meckel’9 
researches) to the zona pellucida of Mammals, (the true primary vitelline sac of 
these animals, interior to the zona, being totally ignored by him; see also p. 60, 
where he describes the zona as the original yolk-sac and the only one existing in 
Mammals) and the secondary yolk-sac (the true zona) to the tunica granulosa of vi- 
viparous vertebrates. The secondary yolk-sac, he infers, is derived from the cellular 
lining of the Graafian follicle; but, since at the same time he makes it merely the 
exterior stratum of a concentric series, the inner of which he insists become the true 
yellow yolk granules, (the primary yolk-sac, zona pellucida, as he calls it, having 
disappeared by deliquescence) it looks very much as if he had mistaken the devel- 
opment of the ‘membrana investiens’ for that of the membrana vitelli. Again he 
says (p. 78), ‘ the external edge of the layer of prismatic cells, the length of which is 
considerably increased, is now surrounded by a narrow pellucid space inclosed by a 
double line, presenting the appearance as if a small part of the bases of these cells 
had been fused together in a homogeneous film.’ This probably is the true zona 
pellucida of Birds; he having failed to see the membrana vitelli (already disappeared 
as he thinks), situated between it and the layer of prismatic cells, from which latter 
he supposes, but without direct research, that the ‘ pellucid space ’ because of its 
traces of hexagonal markings, is an immediate development.” Agassiz on the Embryology of the Turtle. 
7 
of very different sizes. I was led by this observation to inquire into the 
duration of the growth of the ovarian eggs, when I further noticed that 
these eggs appear in well-marked sets of different sizes, each set being 
equal in number to the average number of eggs laid by the species under 
observation. It thus became evident that the eggs require more than 
one year for their full development. Once upon this track, it appeared 
practicable to determine how long a period this growth embraces; for, 
as soon as it could be ascertained how many eggs different species of 
Turtles lay, there was a standard of comparison obtained for the investi- 
gation of the ovaries; and, as I early learned that the species most com- 
mon about Cambridge exhibit marked differences in that respect, I se- 
lected these species for my first studies. Chrysemys picta lays always 
between five and seven eggs. I have never observed as few as four, and 
only occasionally eight. Nanemys guttata lays generally two or three; 
I have only once or twice found four eggs in its nest, and three times in 
its ovary. There was therefore no chance of making any mistake, when 
comparing the number of their ovarian eggs with that of the eggs they 
lay, after I had ascertained that a few weeks before the breeding season 
there are the same numbers of mature eggs to be found in the ovary as 
these species usually lay in the spring. I felt still greater confidence in 
the possibility of coming to precise results, after I had found again and 
again the very same number of eggs in the oviduct, and noticed that at 
that time another set of eggs could be readily distinguished, of the same 
number as the larger eggs left in the ovary. Indeed, the difference be- 
tween this largest set of ovarian eggs and the smaller ones is so great, 
even at the time when the eggs about to be laid are still in the oviduct, 
that they are distinguished at the first glance; for, though they have un- 
questionably to remain another year in the ovary, they are already nearly 
as large in diameter as those which have just left it. 
“ With a knowledge of these facts, it was easy to arrive at a full un- 
derstanding of the normal periodicity in the growth of the ovarian eggs. 
It soon became plain, that shortly before the period of laying there were 
not only two, but as many as four, distinct sets of eggs in every ovary; 
and that, after the largest set had been laid, a new small set was started 
from among the innumerable smallest eggs of variable size. It now 
seemed that a single question remained to be answered. What is the 
age at which the Turtle discloses for the first time such differences be- 
tween its eggs? Upon opening large numbers of young Chrysemys 
picta it was ascertained, that, up to their seventh year, the ovary contains 
only eggs of very small size, not distinguishable into sets; but that with 
every succeeding year there appears in that organ a larger and larger set 
of eggs, each set made up of the usual average number of eggs which 
this species lays, so that specimens eleven years old, for the first time, 
contain mature eggs, ready to be laid in the spring. 
“Now another question arose, When are the eggs fecundated? Field 
observations soon taught me that this species copulates before it is eleven 
years old; I have even seen those that were not over seven years old 
already performing the act, though I have never seen any in copulation 
younger than these. Thus it appears that the first copulation coincides 
with a new development of the eggs, in consequence of which, a certain 8 
Agassiz on the Embryology of the Turtle. 
number of them, equal to that which the species lays, acquire a larger 
size, and go on growing for four successive years before they are laid, 
whilst a new set is started every year, at the period of copulation in the 
spring, enabling this species to lay annually from five to seven eggs, after 
it has reached its eleventh year. 
“ The question was then naturally suggested, whether fecundation is 
the result of the first act of copulation, or of the second, the third, or 
the last; or whether the first copulation only determines the further 
growth of a certain number of eggs, which require a series of successive 
fecundations to undergo their final development. The second alternative 
appears the more probable when it is remembered that Turtles were ob- 
served which did not lay their eggs as usual, though the yolk had under- 
gone all the regular changes through which it passes, up to the time the 
egg has entered the oviducts. This is another fact which tends to prove 
that fecundation is a successive act. Though Turtles lay only once every 
year, soon after the period of copulation in the spring, copulation itself does 
not take place once merely, every year, as in all the animals known to 
bring forth young once annually ; it is repeated a second time, every year, 
in the autumn, shortly before the Turtles retire to their winter quarters; 
and this takes place without apparent connection with any marked change 
in the growth of the egg at that season. So, in Turtles, fecundation does 
not appear to be an instantaneous act, resulting from one successful con- 
nection of the sexes, as it is with most animals. The facts related above 
show, on the contrary, that, in Turtles, a repetition of the act, twice every 
year, for four successive years, is necessary to determine the final develop- 
ment of a new individual, which may be accomplished in other animals 
by a single copulation. 
“ It may be suggested, that, by an investigation of the spermatic parti- 
cles, additional light would be thrown upon these remarkable circumstan- 
ces. But such investigations present greater difficulties in these animals 
than could be supposed at first; and notwithstanding the most diligent 
search, my efforts to trace the spermatic particles through the oviduct, as 
high up as the ovarian eggs, have been unsuccessful. Turtles do not 
copulate in confinement; and those which I could catch in coitu in their 
native haunts have only exhibited spermatic particles in the oviduct. I 
have, still less, been able to trace the sperm into the egg itself. Indeed, 
there is no micropyle in the egg of Turtles; and I must confess that I 
have not yet seen the first fact which could lead me to admit that the 
spermatic particles penetrate into the egg. I am therefore obliged to 
abstain from expressing any decided opinion upon the question of the 
penetration of the spermatic particles into the egg, which has of late at- 
tracted so much attention among embryologists. I can only say, that, 
notwithstanding the high authority upon which it is asserted as a fact 
that the spermatic particles do pass into the substance of the egg through 
a definite aperture of its envelope, I am still rather inclined to doubt it. 
“The aperture observed in the outer membrane of the egg, which has 
been called micropyle, has always appeared to me to be the result of the 
separation of the sac in which the egg is developed, and by no means to 
pass through the vitelline sac. Without the most careful examination it 
is not possible to perceive how complicated the sac is, in which the egg Agassiz on the Embryology of the Turtle. 
9 
is inclosed; and I suspect that a kind of Graafian follicle, which in many 
animals drops from the ovary with the egg, has frequently been mistaken 
for a vitelline membrane. I believe, further, that the scar resulting from 
the separation of that follicle forms the opening called micropyle, and 
that this opening does not traverse the vitelline membrane. In Turtles 
the perforated appearance of the yolk sac arises from the presence of the 
Purkinjean vesicle near the surface of the yolk, and not from the exist- 
ence of a real hole. After what has been said above of the lateral origin 
of the Purkinjean vesicle, it is superfluous to insist upon the incorrectness 
of the view of those who wrould ascribe its superficial position to the in- 
fluence of fecundation. It is formed in that position, and preserves it as 
long as it exists.”—pp. 489-492. 
Egg Laying.—By careful comparisons, it has been possible to 
ascertain at what age at least one species begins to lay its eggs 
and reproduce its kind. Chrysemys (Emys) picta does not lay 
its eggs before the eleventh year. With other turtles the data 
were not so complete, but in all probability they may be said to 
lay their eggs from the eleventh to the fourteenth year, accord- 
ing to the species. They all do this too in the spring, in the 
month of June, both at the north and south; physical differ- 
ences not seeming to have the least effect upon this particular 
function. 
The formation of Albumen.—The mode of formation of the 
albumen is very different from what it is among birds, there be- 
ing no chalazas nor spiral arrangement of the layers. The shell 
and albumen are both deposited at the posterior end of the ovi- 
duct, and the albumen is not applied against the surface of the 
yolk, by direct contact of the inner surface of the oviduct, but 
in a great measure infiltrates through the previously formed 
shell membrane. 
The albumen is composed of innumerable layers of amorphous 
substance, in which are imbedded a multitude of elongate, oval, 
coarse, granular bodies pointing in a particular direction in each 
layer, some along the longer axis of the egg, some across the 
same, and others in intermediate directions between these two. 
The shell membrane.—This arrangement is carried out in a 
beautiful manner in the shell membrane whose inner layers are 
composed of the same sort of oval bodies, arranged end to end, 
and forming a beaded thread. Going outwardly these oval 
bodies become more and more blended with each other till 
finally they form a mesh of uniform, even fibres, of excessive 
thinness and nacreous aspect. 
The shell.—The shell is deposited in what appear to be ex- 
ceedingly tender layers of albuminous substance originating 
after the manner of the albumen and the shell membrane. The 
shell is composed of characteristic groups of carbonate of lime, 
more or less intimately soldered together side by side, each 10 
Agassiz on the Embryology of the Turtle. 
group or nodule consisting of concentric layers of columnar 
crystals. 
The absorption of Albumen.—The albumen is absorbed into the 
yolk sac in a very peculiar manner; the outer layers are re- 
moved first, at a point always above the embryonic disc, so that 
in the end an inverted conical hole is formed. From this hole 
the absorption spreads centrifugally till all the albumen is drawn 
into the yolk sac, or zona pellucida now, and the latter has dis- 
tended so as to completely fill the shell. At or near the time 
that the absorption of the albumen commences, the segmentation 
of the yolk begins; and the embryonic disc is formed, in some 
species, before the clear hyaline space appears under the embryo. 
Self-division of the Mesoblast.—Before we describe the segmen- 
tation of the yolk, it is proper to mention another kind of seg- 
mentation never before observed in the eggs of any animal. 
We refer to the self-division of the mesoblast of the yolk cell. 
This takes place, at least to a certain extent, without the influ- 
ence of fecundation within a year, but at the same time has been 
seen only in those eggs which have been expelled from the ovary. 
Moreover this happens before the so-called segmentation of the 
yolk mass, and in fact is a forerunner, as if to prepare the way 
for the latter change. 
The mesoblast, which by this time so completely fills the cell 
that the ectoblast appears like a bright and very narrow halo 
about it, usually divides at first into two equal portions, rarely 
into three, each of these into two more, and they each again and 
again into two more till the ectoblast is filled by an innumerable 
host of mesoblasts. This process goes on slowly, and lasts as 
long as the young turtle is within the shell. It is completed 
first in those ectoblasts which enter into the formation of the 
embryonic disc, and afterwards in those which are added to the 
later formed parts of the embryo. After the segmentation of 
the yolk mass, the ectoblasts become resorbed and the innumer- 
able mesoblasts are left in heaps, which are at first quite distinct, 
but gradually they blend with each other till finally they form 
a uniform layer all over the disc and the whole extent of the 
germinal layer. 
The great point gained by these last observations, is to prove 
that these self divided mesoblasts become the original cells, “ the 
primitive embryonic cells,” in the composition of the different or- 
gans of the body. 
The segmentation of the yolk.—The segmentation of the yolk 
mass has been seen in the eggs of only one species of turtle, viz: 
Glyptemys (Emys) insculpta. This was observed for the first 
time on the 27th of May, 1854, and afterwards during the two 
succeeding days, in a series of eggs taken from three different 
individuals. The process of segmentation is not so regular, and Agassiz on the Embryology of the Turtle. 
11 
there does not seem to be always, in the beginning, a symmetri- 
cal halving of the embryonic area, as has been observed among 
birds; but in other respects it resembles what takes place within 
the eggs of the latter animals, and finally results in shaping out 
the embryonic disc. The most remarkable features to be noticed 
in the egg at this time, are certain phenomena which tend very 
strongly to show that segmentation is not confined to the embry- 
onic area, at the upperside of the egg, but that, as has never 
before been noticed in oviparous allantoidian animals, the whole 
mass of the yolk becomes segmented. 
The whole egg is the embryo.—“Since we have shown in for- 
mer pages, that the embryonic disc, and its extension, the germi- 
nal layer, are formed by the original apposition of yolk-cell 
mesoblasts minutely subdivided, and that these yolk-cells are all 
the same through the whole yolk mass from centre to surface, 
even to the very walls of the superficial Purkinjean vesicle; and, 
moreover, since it is proved that segmentation obtains beyond 
the embryonic disc, and very probably all over and throughout 
the whole yolk, it is evident, that, in the egg of the Testudinata 
at least, the region around the Purkinjean vesicle cannot be sep- 
arated from the more exterior or inferior mass which constitutes 
the greater bulk of the vitelline substance, and that the last 
cannot be homologous to the contents of the Graafian follicle, 
which bears no part whatever in the formation of the embryo, 
but is totally exterior to the Mammalian egg. Again, as will 
be shown hereafter, that portion of the yolk which is originally 
excluded from the primary circumscription of the outlines of the 
embryonic disc cannot be separated from the animal as an ap- 
pendage, for it very soon afterward becomes an essential part of 
the ‘embryo,’ as the latter extends itself in the form of a germi- 
nal layer and a vascular area, not only all over the surface of the 
yolk, but in the case of the area vasculosa, throughout the whole 
vitelline mass, the latter becoming a great spongy network of 
blood-vessels, formed by the lateral apposition of the cells com- 
posing this large body. This vascular mass is finally drawn 
into the body, and though gradually disappearing by resorption, 
remains for nearly six months after birth as one of the essential 
portions of the organization of the freely moving animal.” 
The formation of organs.—The first step taken that indicates 
the formation of organs, is the originating of a furrow around 
and close to the edge of the embryonic disc. The furrow is the 
folding of the germinal layer as it begins to turn upward to form 
the amnios. At this time there are two layers all over the sur- 
face of the yolk mass; the outer one, the germinal layer, is very 
thin, excepting where it enters into the composition of the em- 
bryonic disc; the inner layer, the subsidiary layer, is quite thick, 
and forms the basis of all the organs, except the cerebro-spinal 12 
Agassiz on the\Embryology of the Turtle. 
marrow. The germinal layer performs a triple office; in the 
median line of the embryo, it forms the basis of the “primitive 
furrow” which is the incipient spinal marrow; exterior to this it 
constitutes the musculo-cutaneous layer; and more exteriorly 
still, it folds upwards over the back of the embryo, and becomes 
the amniotic sac. 
The vertebral lamina is formed by the separation of a broad 
band, about equal in length with the embryonic disc, of loose, 
unconnected cells, from the upper surface of the subsidiary layer. 
From this the vertebrae are developed after the usual manner 
among vertebrates. The cells in the median line of the verte- 
bral lamina become changed very soon, and more compact and 
united, forming the chorda dorsalis, or the axis of the young 
vertebrae, which are not as yet apparent. By the development 
of the chorda dorsalis, the vertebral lamina becomes equally 
divided into two laminae lying right and left of the axis of the 
body. In about twelve days from this time, when four or five 
vertebrae have appeared, and the spinal marrow is closed over in 
the anterior of the cerebral region, the subsidiary layer has de- 
veloped a thick annular ridge on its under side all around the 
embryo, at a short distance beyond the confines of the original 
embryonic disc. This ridge eventually becomes a hollow mesh, 
like a sponge, and then constitutes the vena terminalis. Cotem- 
poraneously with the last, another change occurs in the subsidi- 
ary layer by which it becomes separated along the median line 
of the ventral side of the body, from the vertebral laminae above 
it, and the heart and dorsal artery are hollowmd out in its upper 
side. Soon after this the omphalo-meseraic arteries are also hol- 
lowed out in the upper surface of the same layer from which the 
heart and dorsal artery originated. In these vessels a granular 
albuminous fluid surges backward and forward; thus evincing 
at this early period a beginning of blood circulation in vessels 
which do not form a complete circle with each other. It is very 
easily seen that the heart is the impelling power in this case. 
The mode of formation and development of the Wolffian 
bodies is so simple that it is a wonder there has been so much 
dispute among embryologists about their genesis. They originate 
as mere thickenings of the subsidiary layer, around and oblique- 
ly above each abdominal vein, and leaning over toward the dor- 
sal artery. Being in the same layer with the arteries and veins, 
the blood vessels of the two systems meet in these organs through 
very simple channels, running as yet direct from the principal 
artery to the principal vein. About this time, when the heart 
has become three-chambered, the vertebrae reach to the root of 
the tail, and the eyes have become entirely enclosed in complete 
orbits, the allantois begins to grow. Its formation is as simple Agassiz on the Embryology of the Turtle. 
13 
as that of the Wolffian bodies, and developed in the same layer 
with these. 
The subsidiary layer approximates its opposite sides and uni- 
ting them, arches over a certain space so as to form a sac. At first 
this sac does not project beyond the body. Soon after or almost 
synchronously with its formation the allantoidian arteries and 
veins are developed by simply hollowing out channels direct 
from the dorsal artery into the allantois, and thence, all in the 
same layer, into the allantoidian veins which course along the 
edge of the abdominal opening. Now too the embryo turns 
upon its axis, and rests always upon its left side. Why it rests 
upon this side and not upon the other, sometimes at least, is 
simply because this position belongs to its plan of development. 
A little later the omphalo-meseraic arteries have joined their nu- 
merous currents into one single vessel for a short distance from 
the dorsal artery. The nostrils have begun to develop at this 
period, and may be recognized as two simple indentations at the 
end of the head. They have no communication with the mouth 
at this time. 
The vena terminalis merits attention at this time, inasmuch as 
it is as concentrated in its course as it ever will be. It never 
has, from the beginning, been a single vessel, as occurs with 
birds, but is composed of numerous closely set anastomosing 
currents, running parallel wise to each other. Not long after 
this time the vena terminalis, to one half of its circuit, sinks a 
short distance below the surface of the yolk, carrying along with 
it the afferent vein. The nostrils have grown deeper and broad- 
er, and communicate with the mouth, but only through a very 
shallow furrow, nothing like the deep channel which has been 
described in Mammals, but just deep enough to show that there 
is a tendency to follow the type of formation prevalent among 
higher Yertebrata. In the next step the nostrils become nar- 
rowed at the aperture, and the shallow furrow, mentioned above, 
is closed over, leaving a very narrow fissure to indicate its for- 
mer position. The degree of communication between the mouth 
and nostrils in this case, is intermediate between that which ob- 
tains among Mammals where the mouth and nostrils are all one 
cavity during the early periods of development, and that of 
Fishes, where these two openings are always independent of each 
other. The shield begins to develop by a budding out laterally 
of the musculo-cutaneous layer along the sides of the body, and 
the development of narrow ribs extending to the very edge of 
the roof-like covering. 
The feet, or rather paddles of the lower forms of turtles, the 
Chelonioidae, do not remain in a partially undeveloped state, as 
might be expected from what is observed among other verte- 
brates, but undergo what may be called an excess of develop- 14 
Agassiz on the Embryology of the Turtle. 
ment; the bones of the toes becoming very much elongated, and 
the web,—which remains soft among some turtles with moder- 
ately elongated toes,—is hardened by the development of densely 
packed scales, so that the whole foot is almost as rigid as the 
blade of an oar. About two months before the time they were 
hatched, some young of Chelydra (Chelonura) serpentina, were 
observed to move the head, feet, tail, lower jaw, tongue, and also 
the toes separately, and to roll the eyes. In a turtle a little older 
than these, the omphalo-meseraic vein runs in a direct line 
through the yolk mass, and joins the exterior boundaries of the 
vascular area at the lower side of the egg. 
By the time the shield has become broadly oval the embryo 
assumes again an erect position in the egg, and thus remains 
till it is hatched. The embryo of Chelydra serpentina already 
shows its predaceous propensities, by snapping at everything 
which touches it. Just young were hatched the edges 
of the jaws were cut open longitudinally, and disclosed a series 
of small cavities, into each of which, a branch from the maxil- 
lary nerve ran. No doubt these indicate a typical tendency to 
the formation of teeth. When the young is hatching a large 
mass of the yolk sac is still hanging outside of the umbilical 
opening; but within a few hours it is altogether drawn into the 
body and occupies a large space in the abdominal cavity. The 
circulation in the yolk sac at this time is in full tide of operation. 
Externally the allantois withers and falls away, but, within the 
young turtle, its neck is persistent and becomes the urinary 
bladder. 
The brain at this time does not extend in nearly a straight 
line, as is the case in the adult, but is still considerably bent 
downward at its anterior end. The eye presents some hitherto 
unknown peculiarities at this time; the capsule of the lens is a 
triple membrane, excessively thin, very tough, glassy, and elas- 
tic. It is this membrane which supports the “membrana pupil- 
laris” and not the hyaloid membrane as has been claimed for 
some of the Mammalia. The “ membrana pupillaris is a thick 
double membrane and does not seem to be resorbing, but on the 
contrary, would appear to be persistent through life, inasmuch 
as it was found in a turtle,—Trachemys scabra,—twenty years 
of age. 
Histology.—A great part of the histology relates to the young 
turtle about the time it was hatched. In the formation of nerve 
fibres or tubuli, the olfactory nerve furnished the most conclusive 
proof that these tubuli originate by the soldering end to end of 
the nervous cells, and the obliteration of the intervening walls 
at the point of coutact. All stages of development were repre- 
sented here, from cells arranged in a line end to end, with par- 
tially absorbed walls, up to those where the only indication of Agassiz on the Embryology of the Turtle. 
15 
the position of the cells which compose the tubuli, were the 
mesoblasts (nuclei) arranged in a line at regular intervals. In 
the bones, the granular ossification presents nothing very remark- 
able ; but, in the ribs, the relation of the outermost fibrous layers 
to the fibres of the corium of the shield, is very important to 
notice. The only difference observable between the fibres of the 
growing and widening edge of the ribs and those of the corium, 
was that the former were hardened by ossified granules deposited 
in lines, and the latter were perfectly soft. In all probability 
the wings of the ribs, which unite eventually and form in some 
turtles, a solid shield, are developed by the hardening of the 
component fibres, by calcification. In fact it would be difficult 
to determine how much of the shield is composed of the fibrous 
layers of the ribs and how much of calcified fibres of the skin. 
In the eye, the nervous fibres, which run in lines from the en- 
trance of the optic nerve along the inner surface of the retina to 
its anterior border, do not spring directly from the optic nerve, 
as has been usually asserted in regard to the eye of Vertebrata, 
but are tail-like prolongations from each retinal cell, which lies 
abutting against the hyaloid membrane. 
The whole retina, including the membrana Jacobi, is composed 
of five apparently distinct layers of cells. When examined 
closely these layers are seen to be only different varieties of cells. 
The cells of one layer send tail-like prolongations into the layer 
above and below, but in no instance have these prolongations 
been seen united so as to form a continuous string of cells reach- 
ing from the outer to the inner surface of the retina. In this 
respect the retina is lower in degree of development than that of 
Mammals, where H. Muller and Kolliker have seen the string of 
fibres in the greatest perfection. The epithelial cells of the 
stomach and esophagus are endowed with an unheard of ar- 
rangement of vibratile cilia. Instead of being scattered over 
the whole free surface, the cilia are placed in a circle forming a 
crown to the end of each cell. So that when seen face-wise the 
interior of the esophageal and stomachial cavity appears as if 
lined with a net-work of vibratile cilia. The surface of the 
lungs is covered by a thin layer of very pale round cells, with 
numerous dark granules intervening. Over the course of the 
blood vessels these pale cells are almost hidden by densely 
packed dark granules, arranged in two, three, four, or five rayed 
star-like heaps: 
The blood corpuscles at first are very minute globular bodies, 
in fact, nothing more nor less than some of the cells of the sub- 
sidiary layer, loosened from the sides of the original blood chan- 
nels. The cells increase in size and grow transparent; then they 
become oval, and the mesoblast leaves its lateral position and 
takes a central one, and finally in the last month of incubation 16 
Agassiz on the Embryology of the Turtle. 
they begin to be flattened, and assume the discoid shape proper 
to the adult state. A short time before the turtle is hatched, 
the muscles attached to the dorsal arch exhibit very clearly how 
the fibres originate from the soldering of the ends of the spindle- 
shaped cells to each other, and the obliteration of the interven- 
ing walls. The granular contents of these cells is partially ar- 
ranged in lines parallel with the sides of the developing fibres. 
In the muscles of the fore-leg the granules may be recognised as 
components of the fibrillae, and as such giving the fibres a trans- 
verse striation.