730 GENERAL CONCLUSIONS. 



typically formed of (i) a peritoneal funnel opening into (2) a 

 Malpighian body, from which there proceeds (3) a coiled gland- 

 ular tube, finally opening by (4) a collecting tube into the 

 segmental duct, which constitutes the primitive duct for the 

 mesonephros as well as for the pronephros. 



The development of the mesonephridian tubules is subject to 

 considerable variations. 



(1) They may be formed as differentiations of the inter- 

 mediate cell mass, and be from the first provided with a lumen, 

 opening into the body-cavity, and directly derived from the 

 section of the body-cavity present in the intermediate cell 

 mass; the peritoneal funnels often persisting for life (Elasmo- 

 branchii). 



(2) They may be formed as solid cords either attached to 

 or independent of the peritoneal epithelium, which after first 

 becoming independent of the peritoneal epithelium subsequently 

 send downwards a process, which unites with it and forms a 

 peritoneal funnel, which may or may not persist (Acipenser, 

 Amphibia). 



(3) They may be formed as in the last case, but acquire no 

 secondary connection with the peritoneal epithelium (Teleostei, 

 Amniota). In connection with the original attachment to the 

 peritoneal epithelium, a true peritoneal funnel may however be 

 developed (Aves, Lacertilia). 



Physiological considerations appear to shew that of these 

 three methods of development the first is the most primitive. 

 The development of the tubes as solid cords can hardly be 

 primary. 



A question which has to be answered in reference to the segmental tubes 

 is that of the homology of the secondarily developed peritoneal openings of 

 Amphibia, with the primary openings of the Elasmobranchii. It is on the 

 one hand difficult to understand why, if the openings are homologous in the 

 two types, the original peritoneal attachment should be obliterated in 

 Amphibia, only to be shortly afterwards reacquired. On the other hand 

 it is still more difficult to understand what physiological gain there could be, 

 on the assumption of the non-homology of the openings, in the replacement 

 of the primary opening by a secondary opening exactly similar to it. 

 Considering the great variations in development which occur in undoubtedly 

 homologous parts I incline to the view that the openings in the two types 

 are homologous. 



