1892.] BODY-CAVITY IN SNAKES. 495 



in Crocodiles \cf. (5) figs. 42 and 43, in which a small unpaired 

 space is seen in the region of the stomach ']. 



(iv.) The omental space is, as described above [§ IV. (iv.) note], 

 well represented in all the other groups of Sauropsida. 



(v.) As to the liver-sacs, the task of comparison is not so simple. 

 The ventral liver-sacs of Crocodiles are no more comparable with 

 those of Snakes than with those of Birds. We also soon see that 

 there is a considerable difference between the body-cavity in the 

 pulmohepatic region of Snakes and that in the corresponding region 

 of the Bird. Thus, if we start from the median ventral ligament 

 of the liver and proceed outwards, we find in the case of the 

 Fowl [(5) fig. 45] that the spaces 1, 1 are bounded laterally by 

 the body-wall, or rather by the " oblique septum " ; but in Snakes 

 (fig. 3^) we pass all round the liver and reach the median dorsal 

 attachment of that organ. This difference is clearly correlated with 

 the face that while in Snakes the " pulmohepatic recesses " are totally 

 or almost totally absent [see above, § VI. (i.)], in Birds they are 

 remarkably well developed [(5) fig. 45, 2, 2']. 



In short, the "avian diaphragm" is developed out of fibrous 

 tissue on the ventro-mesial face of the lungs, which tissue bounds 

 these spaces (2, 2') externally. But while the fibrous tissue, which 

 forms a sort of diaphragm ventral to the lung in Snakes, and 

 bounds the right liver-sac dorsally, appears to have a somewhat 

 similar relation to the lung, its other relations are different. In fact, 

 but for its adherence to the lung, one might, from its topographical 

 relations, rather compare the diaphragmatic tissue in Snakes with 

 that in Mammals. 



It would seem hardly more possible to closely homologize the 

 " diaphragm " and liver-sacs of Snakes with those of Birds than the 

 diaphragm of Mammals with that of Birds. In the case of animals 

 which have descended from common ancestors along more or less 

 divergent lines, we ought not of course to expect to be always able 

 to compare directly corresponding parts, but rather to have to 

 content ourselves sometimes with tracing the condition of things in 

 each back to their common origin. 



It may be remarked that the lungs of Varanus {Monitor) are 

 excluded from the peritoneal cavity by a sort of " membranous 

 diaphragm "^ and this is less markedly the case with the lungs of 

 certain Chelonia. It may well be, therefore, that the relations of the 

 lung in Snakes can be explained by a reference to one or both of 

 these groups ; but this embryology alone can decide. 



^ It is marked in these figures with a number 1, like the paired ventral liver- 

 sacs, because I was led to regard both the ventral part of the postliepatic septum 

 (/3) which closes the liver-sacs posteriorly, and the fibrous tissue (fS) which 

 closes this gastric sac posteriorly, as but lateral expansions of the median ventral 

 attachment of the stomach and liver, from which point of view the sac in the 

 region of the stomach and the paired ventral liver-sacs might be grouped 

 together. The embryology of Snakes seems to be in favour of the essential 

 similarity between the septa 13(5 {see § VI.) ; but as in Snakes this gastric space 

 is so distinctly posterior to the two liver-sacs, it may with advantage be 

 considered apart from them in both those animals and Crocodiles. 



2 Martin, P. Z. S. 1831, p. 138. See also P. Z. S. 1889, p. 608. 



34* 



