THE ANCESTRY OF THE VERTEBRATES 535 



formed into four longitudinal cords, slightly differentiated 

 from the rest, a dorsal, a ventral, and two lateral, all of which 

 run the entire length of the animal. Of these the dorsal re- 

 ceives slightly more emphasis than the others, since it continues 

 forward to the base of the proboscis, where it divides into 

 two diverging branches, which encircle it in the form of a 

 ring. As for the notochord, this has been doubtfully identified 

 with a small diverticulum, which arises from the dorsal wall 

 of the pharynx, and extends some distance forward into the 

 proboscis, and this supposition has been greatly strengthened 

 through the recent discovery of an allied form belonging to 

 a new genus (Harrimania) in which the diverticulum is much 



es.JSoc 



es - bro 



FIG. 148. Harrimania maculosa. [After RITTER.] 



Schematic representation of dissection, including collar and small portion of 

 the anterior pharyngeal region. The anterior and posterior aspects are designated 

 as A and P, respectively, es, oesophagus; es. No-c, cesophageal notochord; d. n. c, 

 dorsal nerve cord; SK, C, skeletal crura; br. o, branchial orifices. 



larger, and in its mode of origin is strikingly similar to that 

 of the true vertebrate notochord, and is thus without much 

 doubt homologous with this organ. 



From the testimony afforded by the structure of Balanoglos- 

 sus and its allied genera (the group Enteropneusta) it may be 

 quite confidently asserted that these forms lie nearly in the line 

 of vertebrate descent, and represent an earlier stage than that 

 of the tunicates. But here the chain seems to end, for Balano- 

 glossus is itself unusually isolated and shows no close affinity 

 to any other invertebrate types. There is, in such cases, but 



