MUSEUM OF COMPARATIVE ZOOLOGY. 265 



it remains for me to consider whether the theory of Boveri be better 

 grounded. 



The arguments which are adduced in favor of the homology of the 

 gonadial pouches and the mesonephros may be reduced to the following 

 points of similarity. The gonadial pouches of Amphioxus are metameric 

 diverticula of the dorso-pharyngeal coelom, in accordance with the estab- 

 lished views of Kowalewsky and Rolph, as confirmed by Boveri, who finds 

 in the adult a continuity of the epithelia belonging to the two tracts ; the 

 mesonephric tubules likewise are primitively metameric diverticula from 

 the dorsal portion of the body cavity (see Sedgwick, '80*, et al.). The 

 generative cells develop in the walls of the gonadial diverticula ; the 

 early occurrence of germinal cells at the proximal ends of the forming 

 mesonephric tubes has also been described by Riickert ('88, p. 257) for 

 Selachii. Finally, the canal by which the gonadial pouches primitively 

 communicated with the coelom arches over the dorsal angle of the atrial 

 cavity in a way that is very similar to that in which the mesonephric 

 tubules curve outward to join the duct. The only reason — save those 

 that require the prior assumption thatlhe gonadia represent mesonephric 

 tubules — which I can see for identifying the atrium with the segmental 

 duct is the fact that uephridial (pronephric V) tubes open into it. This 

 argument seems to me of very little weight. Boveri himself believes that 

 the pronephros primitively opened directly to the exterior. Unless other 

 evidence can be adduced, I see no adequate reason for regarding the 

 formation of the atrial cavity as a step in the development of the seg- 

 mental duct. On the other hand, that interpretation seems to me quite 

 opposed to all that is known of the development of the segmental duct. 

 As I have shown in the preceding pages, there can be no doubt that, 

 in Amphibia at least, the duct develops solely from the mesoderm. 

 According to the opposed view — the ectodermal origin of the duct — the 

 development always proceeds from a pair of narrow rod-like thickenings 

 of ectoderm, one on each side of the body, which ai'e very different from 

 the unpaired ventral groove from which, accoi'ding to the most recent 

 account (Lankester and Willey, '90) the atrium develops. If, now, we 

 deny the homology of the atrium with the segmental duct, the outward 

 arching of the gonadia becomes a most insignificant topographical resem- 

 blance. It seems to me that it would be manifestly unfair to base so 

 far reaching a homology on the remaining points of resemblance, viz. the 

 early occurrence of germinal cells in the mesonephric tubules, and the 

 circumstance that the gonadia are metameric diverticula of the dorso- 

 pharyngeal coelom. 



