M. M. METCALF ON THE EYES AND SUBNEURAL GLAND OF SALPA. 355 



distance further as a very fine canal within the basement membrane 

 (see Plate LIV, Figs. 3-8). Unfortunately, the only two specimens of 

 the chain Salpa africana-maxima I had for study were injured, having 

 been brought up from some depth on thermometer lines and having 

 been in this way badly torn. In one of these the ganglion was missing ; 

 in the other the ciliated funnel ; so I have been unable to certainly 

 establish the relation of this rod of cells to the ciliated funnel. I am 

 convinced, however, that in this rod of cells we have the homologue of 

 the tube which in doliolum connects the ciliated funnel with the wart- 

 like antero-ventral process from the ganglion. If this be so, we have in 

 Salpa africana-maxima a condition that is found in doliolum and in the 

 young chain Cyclosalpa pinnata. 1 The nervous system of Salpa africana- 

 maxima is then, in this respect, more primitive than that of any other 

 adult salpa I have studied. 



What is the relation of the ganglion, the subneural gland and its 

 accessory structures in salpa to the organs described in the first part of 

 this section for doliolum, pyrosoma and ascidians? The ganglion of 

 salpa has been regarded by Salensky [14] and others as homologous 

 with the whole central nervous system of the ascidian larva. He has 

 described three slight enlargements of the neural canal of the embryonic 

 solitary salpa which he says correspond to the three primary vesicles of 

 the vertebrate brain and to the three regions of the larval ascidian neural 

 canal the sense vesicle, the visceral portion and the caudal portion 

 respectively. A careful -comparison of the conditions in the different 

 ascidians and in pyrosoma, doliolum and salpa, seems to me, on the con- 

 trary, to indicate that the ganglion of salpa is homologous with only 

 the visceral portion of the larval ascidian nervous system. In favor 

 of this homology there are the following points : 



1. It is only in the visceral portion of the larval ascidian nervous 

 system that the ventral wall of the neural canal is thickened. Only 

 that portion of the embryonic salpa's neural canal which has a thick- 

 ened ventral wall persists to form the adult ganglion. 



2. The adult ascidian ganglion is formed from the dorsal cells of the 

 visceral portion of the larval nervous system. The main portion (dorsal 

 two-thirds or three-fourths) of the adult salpa ganglion is derived from 



1 A similar tube is found in the embryo of the solitary form, connecting the 

 ganglion with the ciliated funnel, Plate LI, Fig. 4. 



