EVOLUTION OF THE VERTEBRATA. 



323 



cestral Tunicata did not undergo this degenerative metamorphosis 

 for it is to such types that we must look for the ancestors of the 

 other Vertebrata, the Acrania and the Craniata. And here pale- 

 ontology steps in and throws new light on the question. I have 

 pointed out briefly, in the " American Naturalist/' * that a second 

 order must be added to the Urochorda, viz., the Antiarcha, in 

 which the anus presents the same position as in the Acrania, at 

 the posterior end of the body, while an orifice of the upper surface 

 represents the mouth of the Tunicata. To this order is to be re- 

 ferred the family of the Pterichthyidag, of which the typical genus, 



Fig. 63. 



Fig. 62. 



Fig. 62. Bothriolepis canadensis Whiteaves, from above, half size of a small speci- 

 men. The valve of the dorsal mouth, or notostome, is broken. Fig. 63. Chelyosoma 

 maclomanum Brod. & Sow., Y3 natural size, from Point Barrow, Alaska. 



Pterichthys, is a well-known form of the Devonian period. This 

 genus retained its tail, which was the cause, in connection with 

 the i^resence of lateral fin-like appendages, of its having been sup- 

 posed to be a fish, by Agassiz, Hugh Miller, and others. It is pos- 

 sible that the American Bothriolepis canadensis lost its tail, as in 

 the majority of Urochorda. The tunicate which approaches near- 

 est to the Antiarcha is the Arctic Chelyosoma. 



From the Antiarcha to the Acrania and Craniata, then, the 

 line is an ascending one. 



* March, 1885, p. 289. 



