1892.] BODY-CAVITY IN SNAKES. 491 



does not extend forwards into the region of the lung ^ ; it is confined 

 to the hepatic region— the pancreas, which forms the posterior wall of 

 the space, being, together with the gall-bladder, at this stage and for 

 some time longer in contact with the hver. But as development 

 proceeds, and as the lung extends back to and beyond the hinder 

 end of the liver, and as the pancreas and gall-bladder come to lie, as 

 they almost invariably do in Snakes, a considerable distance behind 

 the liver", this space (if not obliterated) comes to be, with rare 

 exceptions, entirely posthepatic in position. 



§ VI. (ii.). Embryos o/Elaphis quadrilineatus, 11 cm. long. 



The next stage that I have, an Elaphis embryo 11 cm. long 

 (Plate XXVIII. fig. A), (all allowance being made for Elaphis being 

 a larger Snake than Tropidonotus or Zatnenis, with larger eggs), is 

 considerably more advanced than the stage just described, and yet for 

 our present purpose there is no important gap between them. There 

 is in fact, as far as the pleuroperitoneal cavity is concerned, at first 

 sight as yet nothing to suggest the characteristic Ophidian condition. 

 The liver-lobe of either side and the lung for the greater part of its 

 length project freely into the common pleuroperitoneal cavity. 



The only definite change that we have to note is the closing of the 

 " Foramen of "Winslow " ; this, however, not only occurs in Birds 

 (Gallus), Crocodiles, and many Chelonians, but also in the snake-like 

 but truly lacertilian Amphisbsenidse. In fact an Elaphis embryo of 

 1 1 cm. long is still lacertilian as to its pleuroperitoneal cavity ; but, 

 nevertJieless, the changes that are shortly to supervene are fore- 

 shadowed. 



Lizards, and pei-haps little importance is to be attached to tlie difference. 

 Whichever is the more primitive state of things, the one may easily be derived 

 from the other. The condition in Snakes and the Scincoids first mentioned is 

 probably associated with the elongation of form, and with the origin of the 

 liver at some distance behind the point of origin of the lungs. In fact, in these 

 Lizards, as in all the Snakes I have examined (with the exception of the species 

 of Vipera, Hydro^ihls, Felamis, and less markedly of Typhlops), there is even in 

 the adult a distinct gap between the anterior end of the liver and the heart. 



^ Some of the Snakes examined (see list, p. 4S1), viz. the Pythonidas [Eryx, 

 Enygrus, Python) and Xenopeltis, have two well-developed limgs, the right, 

 however, being the larger. Others, viz. Ehinophis, Cylindrophis, Aspidura, 

 Elcqjs, have a more or less distinct rudiment of a left lung. Others again — 

 Tropidonottis, Elaphis, Bipsas — have the merest trace of this, only to be found 

 by careful seai'ch near the posterior corner of the heart. In some, again, 

 Vipera {bents and aspis), Crotalus, Lamprophis, and others, I did not find any 

 trace of a left lung. In Vipera aspis I find no trace of a left lung even in early 

 embryos. Though I have n o embryonic stages of the Pythonidis or Xenopeltid^, 

 a comparison of their anatomy with that of the more usual one-lunged forms 

 seems to assure us that, for our present purpose, there is no noteworthy difference 

 between them. The left lung, when present, lies between the dorso-lateral wall 

 of the left liver-sac and the oesophagus, in a position, in fact, corresponding to 

 that of the right lung of the other side, and has not, any more than its fellow, 

 any trace of pleural cavity round it. 



^ However, in B.hinophis and certain specimens of Aspidura the gall-bladder 

 is close to the liver, and it is not far removed in the Common Viper {Vipera 

 berus). 



