264 BULLETIN OF THE 



tubules ; for while the metamerism of the gill bars does not correspond 

 in the adult to that of the myotomes, yet we should not lose sight of the 

 fact that according to Kowalewsky ('67, see his Figs. 3G and 39) such a 

 correspondence exists in the embryo. At such a stage, then, there would 

 be present a single excretory tubule for each myotome. 



In a recent lecture before the Gesellschaft fiir Morphologic und Physi- 

 ologie in Miinchen, Boveri ('90) has endeavored to show the existence in 

 Amphioxus of homologues of the pronephros, the mesonephros, and the 

 segmental duct. The tubules which Boveri regards as pronephric are 

 probably the same structures as the excretory tubules of Weiss ; and I 

 infer that the same have been seen by Spengel ('90, p. 282), though this 

 writer makes no suggestion as to their significance. Both Weiss and 

 Boveri claim to have proved by feeding the animals with carmine that the 

 tubes are actually excretory. According to Boveri, also, they open into 

 the atrium at the upper margin of each secondary gill bar ; but their 

 course is somewhat differently described by the two authors. Boveri 

 maintains that each tube communicates by means of several openings 

 with the dorso-pharyngeal coelom. As confirmatory of his position that 

 these canals represent the pronephric tubules of Craniota, he describes 

 the relations they bear to the gill vessels, which he identifies with the seg- 

 mental vessels described by Paul Mayer ('87, p. 343) in Selachii. Accord- 

 ing to Ptiickert ('88, pp. 239-242), the glomus of Elasmobranchs consists 

 of a rete mirabile in connection with these segmental vessels. Adjacent 

 to the excretory tubules, Boveri finds that the gills display an increase in 

 vascularity, and that anastomoses are formed between the branchial ves- 

 sels. This condition does not seem to have been noticed by Weiss. 

 Spengel, who made a special study of the gill vessels, describes a longi- 

 tudinal vessel at a corresponding level (longitudinal trunk of the liga- 

 mentum dcnticulatum), but does not discuss its significance. It seems 

 to me that Boveri's observations, provided they be confirmed, aff'ord fairly 

 satisfactory evidence of the existence of true nephridia in Amphioxus ; 

 and, as I shall endeavor to show in the sequel, that these are constructed 

 on a type which may be assumed to represent a primitive condition of the 

 Vertebrate kidney. 



The starting point of Boveri's researches was the hypothesis that the 

 atrial cavity and gonadial pouches of Amphioxus correspond to the seg- 

 mental duct and mesonephros respectively of Craniota. The attempts 

 of Haeckel ('74», p. 37, and '74^ p. 305) and of Huxley ('76, pp. 221, 

 222) to discover a homologue of the segmental duct in Amphioxus must, 

 in my opinion, be held to have at present merely an historical interest; 



