28o Susarina Phelps Gage 



over the cephalic end of the myel. In transection it has a 

 considerable lateral extent. The lamprey may be considered 

 to have a very small sac in a corresponding position (Fig. iii) 

 as, just cephalad of the closure of the myel, the walls ap- 

 proximate closely, the endyma being in contact except at the 

 dorsal limit where it swells out into a minute sac. In neither 

 lamprey nor amia was anj^ lack of continuity in the endyma 

 seen. Rex (43) in this region of the elasmobranchs shows 

 that the vessels leave a little space at this point, though all 

 around they form a close mesh work, and Burckhardt (8) shows 

 a sac similar to amia in protopterus. 



In diemyctylus such a pocket was not found nor, in the 

 earliest stages, any lack of continuity of the end5'ma. In a 

 half grown aquatic larva the last section before the closure of 

 the myel shows that the endyma of the plexus is not complete 

 and the pia does not cross the meson, though in the section on 

 either side these conditions do not exist. This same condi- 

 tion was found in various series up to a full sized red form. 

 In a large red form and all the adults examined which were 

 perfect in this region the opening was much more pronounced, 

 and was observed in sections cut in different planes (Fig. 54- 

 56). Here the pia with its vessels extends toward the open- 

 ing but not across it, and the endyma is recurved at the lateral 

 and caudal margins, but suddenly ceases, leaving a free com- 

 munication of the cavity with a subarachnoid space. It may 

 be objected that the cells lining the pigmented dura are en- 

 dymal cells, but they do not have the size nor appearance of 

 endymal cells and furthermore form a continuous layer lining 

 the dura in its whole circumference and are united by con- 

 nective-tissue like cells with the pia at frequent intervals as 

 in figure 55, ar. The recurved endyma may represent the 

 remnant of a stage like the amia. 



After the most careful study of the sections, which show no 

 indication of tearing, delicate connective tissue cells and pro- 

 cesses retaining their position, the conclusion is unavoidable 

 that a true metapore exists in the adult diemyctylus, and that 

 its beginning or initial form arises in the larva. 



