1889. ] THE BODY-CAVITY IN LIZARDS, ETC. 461 
Figs. 8, 9, and 10 show this right-hand sac at the beginning of the 
9th day, and we see (fig. 10, 7. W.) that the foramen of Winslow is 
still open. But on the 9th day this foramen becomes closed (that 
is, I could find no aperture in a complete series of sections of a chick 
of 9 days 1 hour incubation), and moreover that part of the recess 
which surrounds the gizzard, as opposed to the cesophagus, becomes 
obliterated ' (compare fig. 12, 2, with fig. 10, 2), 
This space (2), which can be followed through the sections of the 
12th day (ef. figs. 14-17 and 26-28) and seen in those of the adult 
fowl (figs. 44-46), corresponds, as stated, to rather more than the 
“‘recessus superior sacci omenti’’ of Mammals, since its posterior 
part represents a portion of the omental sac. For this reason, 
and because there is a similar space on the left side (where there is 
no omental sac to have a recess), and because these spaces are in fact 
no mere diverticula of the omental sac, but have a distinct origin, I 
have spoken of them as the pulmohepatic recesses (cf. ante, p. 454). 
This name expresses their relation to the liver and respiratory 
organs, and to the pulmohepatic ligaments which form part of their 
outer wall. 
In the fowl the foramen of Winslow up to the 9th day, when it 
closes, has precisely the same relations as in Mammals, being 
bounded by the inferior vena cava antero-dorsally, and the duodeno- 
hepatic omentum, or ventral mesentery that bears the bile-ducts 
and portal vein, postero-ventrally. 
With regard to the corresponding recess on the left side (2' in 
the figures previously referred to), a reference to the figures shows 
that it is from the first in much freer communication with the rest 
of the peritoneal cavity than that on the right, and the alimentary 
canal, instead of (as in the case of the other recess) bending round 
to form, together with the median vertical membrane that supports 
it (m, figs. 9, 12, 14, 15), a posterior or omental wall, appears 
rather on the contrary to hinder the development of its outer wall 
by leaning over on the left side between the lung and liver. 
III. (d). On the Homology of the Avian and Mammalian 
Diaphragms. 
On considering the relations of the recesses and ligaments (2, 2’, 
a, @) in the two above-named groups, we see that the whole of the 
Mammalian diaphragm lies laterally or centrifugally to the attach- 
ments (a, a) of the liver to the mediastinal tissues in front of it, while 
the avian diaphragm lies practically entirely within or centripetally 
to these attachments. So that the condition in birds is expressed by 
saying that the middle mediastinal tissue of the two sides, instead of 
coming to wrap round the pericardium, diverges posteriorly and 
ventrally to become attached to the lateral body-walls*. 
1 Apparently by constriction off from the rest and the adhesion of its walls. 
2 We may, for illustration, compare the middle mediastinal tissue of birds in 
its relation to the pericardium to a coat which, instead of being buttoned 
across the chest (the pericardium), is extended like a wing on either side by 
laying hold of the front bottom corners. 
