524 EMBRYOLOGY OF THE TURTLE. Part III. 
nature but only about half as long extended laterally and obliquely outward, 
one on each side of the several points of departure, so as to include between 
themselves and the terminal thirds of the main furrow, equilateral, triangular 
spaces, and on each side of the median third a truncated isosceles triangle, each 
of which was partially split, as it were, by a short depression, originating from the 
central part of the principal furrow, and terminating sharply at its ends. Thus, 
on the whole, six segment masses, two of which were again partially divided, 
were included in the formation of the embryonic area at this early stage. The 
depth of these furrows was not ascertamed in a definite manner; yet, judging 
from appearances and the known thickness of this portion of the germinal layer, 
they must have penetrated very nearly, if not fully, to the inner surface of 
the latter. 
In another egg, a little older, and the anterior of four in the left oviduct, 
the main furrow was not parallel to the longer axis of the egg, running more 
in a zigzag line, (PI. 10, fig. 2,) and the lateral oblique furrows trending so as to 
be more nearly perpendicular to the longer axis of the yolk. Those which ter- 
minated sharply in the last egg were here represented by much more lengthened 
forms, and not exactly symmetrical as to their pomt of origin as in the former 
egos, being nearer to the end of the principal furrow on one side than on the 
other. So, even at this early period, there is here a considerable want of bilateral 
symmetry, which still further justifies the doubt, already expressed, as to the 
constancy of a single furrow, in the beginning of the segmentation. It will also 
be noticed, that there is some difference in the two figures which are given of this 
stage from the same egg (PI. 10, fig. 1 and 2); but this apparent disparity is 
explained by the circumstance that one view is more superficial than the other ; 
the one (Pl 10, fig. 1 and la) representing only the surface of the embryonic 
area where the outermost edges of the segment masses more or less overlap 
each other, and the other (PI. 10, fig. 2) a deeper view, through the overlap- 
ping edges just mentioned: thus showing that the furrows are not perpendicular 
chasms, but bend, some in one direction and some in another, opening below, 
in one instance, (Pl. 10, fig. 2, 6,) in a line at right angles to the main fur- 
row, but gaping above, (Pl 10, fig. 1, 6,) with edges runnimg quite obliquely to 
it, or, as in the case of the median transverse furrows, opening above (Pl. 10, 
fig. 1) in one and the same line, and terminating differently one from the other 
below (PL 10, fig. 2). This obliquity becomes more and more evident as the 
number of segment masses increases, and their contours attain a more rounded 
outline, just as would happen were a collection of plastic, rounded bodies pressed 
against each other as they laid spread out upon a convex surface. The embry- 
onic area of the egg next behind the one first mentioned (Pl. 10, fig. 3) was 
