556 EMBRYOLOGY OF THE TURTLE. Part III. 
cavity into which the branchial fissures open; its right and left sides are so 
folded inwardly and longitudinally as to shape the whole into a double channel, 
one of which (J, I’) lies next to the back, and the other, which is much nar- 
rower, (I’”,) next to the ventral side of the body. These chamels communicate 
with each other by means of a small aperture, (5,) which is situated near the 
posterior end of the longitudinal fold. From the posterior end of the lower 
channel, (1’”,) two short blind sacs, (J, I,) one on each side, protrude horizontally 
and in a backward direction, pressing against that portion of the subsidiary layer 
(I’) which remains a single tube. 
The larger of these channels occupies by far the greater part of the cavity 
in front of the heart; but, behind this organ, it grows narrow, giving place to 
a larger, globular, dark body, (Pl. 24, fig. 9, 7, 9a, 7,) which occupies nearly the 
whole breadth of the body next to its ventral side, and close behind the heart ; 
and finally, at the abdominal opening, it flares broadly open, but not so widely 
as in the last phase, forming a direct communication with the yolk mass below. 
This is the beginning of the intestinal canal, the broadest part of which corre- 
sponds to the cesophagus, and the part that follows behind, to the stomach; the 
long, thin intestine of the adult being at this age a broad, open layer, except- 
ing at the extreme posterior end, where the allantois springs from it. 
The smaller of these channels commences in front, close to the angle where 
the head is bent upon the neck, (PL 14, fig. 5,) and extends backwards as far 
as the heart, and there, as we have already mentioned, divides imto two blind 
sacs. Every relation which this channel bears to the oesophagus points out its 
identity with the respiratory system, and therefore the single part of the channel 
must be the windpipe, and the two blind sacs, the lungs. We have not traced 
the origin of the large, dark body (Pl 24, fig. 9, 7, 9a, 7) behind the heart; 
but from its size, position, relation, and dark color, it must be the liver. It will 
be noticed that the subsidiary layer, which composes these two channels, is sep- 
arated into two strata. The interior of these strata, both im the intestine (PI. 
24, fig. 1, 2’, 2”) and in the lungs and windpipe, (2, 2”;) is no doubt the epi- 
thelial layer. The allantois (Pl. 14, fig. 5) is not larger than m the last-men- 
tioned phase, but it shows traces of bloodvessels. The Wolffian bodies, blending 
closely with the vena abdominalis, (Pl. 24, fig. 9a, 7,) render the latter appar- 
ently larger than they really are. The abdominal opening is now contracted 
to a much diminished space, lying between the liver (Pl. 24, fig. 9a, 7) in front, 
and the allantois behind, and narrowed to half the width of the body. The feet 
(Pl. 24, fig. 9a, w) are not further developed than before. The caudal portion 
of the body, the tail, (Pl. 14, fig. 5,) beyond the allantois, is much longer and 
more slender than in the last stage; and at its base, close behind the allantois, 
