348 ARTHUR WILLEY. 



intestine, bifurcates at its distal extremit}'^ this being the 

 introduction to an extremely complicated system of branching, 

 which occurs at first to a certain extent dichotoraously, but 

 the mode of branching is marked at a later stage by the 

 occurrence of numerous anastomoses between the various 

 branches. The epithelium of the csecum, or, as it is called 

 after it has commenced to ramify, pyloric gland, soon loses its 

 original columnar character, and the wall of the gland then 

 consists of a thin transparent membrane which is presumably 

 lined by a flat epithelium. The branches of the gland grow 

 in such a way as to embrace in a finger-like fashion the intes- 

 tine throughout its whole course from the stomach to near the 

 level of the anus. 



The topographical relations of the caecum or pyloric gland 

 of Ascidians are different from those of the caecum of Amphi- 

 oxus, — with the point of agreement, however, that they both 

 lie on the left side and are directed forwards. 



If now the intestine of Amphioxus be imagined to be 

 doubled upon itself, the topographical relations of the caeca in 

 the two cases would be essentially identical. 



In view of these considerations, and remembering what has 

 been said above as to the curvature of the intestine in the 

 Ascidians, I think it may be stated categorically that the 

 pyloric gland of the Ascidians is homologous with the hepatic 

 csecum of Amphioxus. 



7. Scheme of the Relations of the Protochordata, 

 with Remarks on Appendicularia. 



The following diagram will give a bird^s-eye view of the 

 relations which I imagine to exist between the various groups 

 of the Protochordata. 



