512 EMBRYOLOGY OF THE TURTLE. Part III. 
we infer that the earliest layers of albumen were at once liquefied and drawn 
within the yolk sac. This view of the case seems to be warranted by the fact, 
that the layers of albumen still outside of the yolk sac, are, to all appearances, 
undisturbed and perfectly symmetrical all around the egg, even after the yolk 
sac has assumed an oval form. No further alteration of the yolk mass con- 
nected with the absorption of albumen was noticed in much older eggs of this 
species, (Pl 10, fig. lla, 11b,) up to the time when the embryonic area became 
a distinct disc (fig. 15a). 
The mode of absorption of the albumen mentioned above, and the consequent 
change in the shape of the yolk sac, have been observed in the oval eggs of 
several other genera of the family of Emydoide, namely, in Nanemys guttata, 
(Pl. 9b, fig. la,) Chrysemys picta, and Cistudo virgmea (which had the same 
aspect as Pl. 10, fig. 15a). In these the embryonic area was already a distinct 
disc, (Pl. 10, fig. 12, 13, 14,) and the albumen and shell were complete. 
The yolk of the oval eggs of at least two other genera (Ozotheca and Cinos- 
ternum) belonging to another family, the Cinosternoid, does not assume an oval 
form so early as in the Emydoide. It does not appear even that the albumen 
mixes at all with the yolk in the beginning, as is the case in the eggs of 
Emydoidx. At least, in all the younger eggs of the family of Cimosternoide which 
had already a shell, the albumen was arranged in perfect symmetry around the 
yolk mass; and the latter was perfectly globular, and to all appearances not 
larger than when it left the ovary (Pl. 9d, fig. 4). In the eggs of the Testudi- 
nina, Trionychide, Chelydroide, and Chelonioid, which have a globular shell and 
a globular albuminous deposit, neither an oval form nor an increase in the size 
of the yolk mass has been observed as long as the yolk remained homogeneous. 
In fact, the earliest period at which we have known the albumen to enter the yolk 
sac of the eggs of the families just mentioned was when the cephalic hood had 
already began to form (Pl. 11, fig. 1, la; Pl. 9b, fi 
albumen bore a very different relation to the yolk mass from that in the cases 
e. 5, 7, Ta); and then the 
pointed out before, as will presently be shown. 
Notwithstanding the infiltration of a small portion of albumen in some of 
them, the eggs of all Testudinata, whether their shell be oval or round, retain 
a homogeneous aspect till the embryonic disc has assumed a sharply defined out- 
line (Pl. 9b, fig. la, e, 4b, e; Pl. 10, fig. 12, 13, 14, 15, lda). But it is a sig 
nificant fact, that, at this period, the oval egg shell of the family of Emydoide 
should contain an enlarged oval yolk mass, whilst the oval egg shell of the 
family of Cinosternoide contains a yolk mass which is perfectly globular, and 
not larger than when in the ovary. 
We have already mentioned, that all eggs with globular shells retain, until 
