546 EMBRYOLOGY OF THE TURTLE. Part III. 
the eye the same appearance.) Upon closer examination of the layers of cells 
within this space, we find that the subsidiary layer (Pl. 9d, fig. 1, 01; Pl. Qe, 
fig. 5, o') has undergone a change, both in the closer aggregation and further 
selfdivision of its cells, so that it approaches in intimate structure that portion 
of its expansion (Pl. 9d, fig. 1, 1; Pl. 9e, fig. 5, ) which lines the lower arch 
of the embryo. The exterior edge of this layer is thickened below, (Pl. Qe, 
fig. 5, 7.) so as to present a projecting annular ridge all round. Beyond this, 
again, the subsidiary layer remains as _ heretofore. 
Another embryo, (Pl. 12, fig. 2,) although two days younger than the last, 
is considerably more advanced in its development. The amnios is much more 
closed over, (a, a’,) and the head more sunk towards the centre of the yolk 
mass. The spinal marrow, for some distance behind the head, has become a 
closed tube by the uniting of its upward folding edges, (Pl. 9e, fig. 6, ¢,) and 
its wall (e") has increased in thickness. At the posterior third of the body it 
still remains open, and gradually loses its distinctness from the portion of the 
germinal layer which extends beyond. That portion of the germinal layer which 
lies on each side of the part of the spinal marrow that is closed over rests at 
a lower level than in the last, younger, stage which we have just described, and 
is considerably increased in thickness, (p,) but thins out towards its periphery, till, 
at its second duplicature (a*) in the amniotic portion, it suddenly becomes exceed- 
ingly tenuous (a); and so it remaims wherever it may be found beyond the 
embryonic region. 
The figure we have last referred to (Pl. 9e, fig. 6) represents a transverse 
section of the body at the anterior edge of its posterior third, along a line just 
behind the point (Pl. 12, fig. 2, a’) where the amnios is still open. It will 
be noticed here that the vertebral layer (Pl. 9e, fig. 6, /’, f°) is much thinner 
than in the section of a younger embryo made at the middle region. In the 
latter case the dorsal vertebrae were already marked out, (Pl. 9d, fig. 1, 7; Pl. 9e, 
fiz. 5, f; Pl. 12, fig. 1, f, fig. la, /;) and so they are in this embryo at the same 
place, and also much farther backwards; but, as we have shown in very young 
stages that the vertebral layer grows thinner backwards, so here the same obtains. 
The posterior ends of the spmal and vertebral layers appear to expand into broad, 
spatulate figures (Pl. 12, fig. 2); but this is not so much a peculiarity of these 
strata alone, as a feature arising from the manner of their partial connection 
with the respective layers from which they take their origin. In both cases, as 
development defines the position and shape of each, the posterior expansions pass 
gradually farther and farther backwards (Pl. 12, fig. 3, 3a, 3b, 4, 7, 11, 12, 13). 
The chorda dorsalis (Pl. 9e, fig. 6, y) is large and well marked, appearing darker 
than the vertebral layer on each side of it, on account of the increased transpar- 
