I'^O"] Bitter. — California Coast Ascidians. 31 



From this it appears that Benthascidia michaelseni is nearer 

 akin to Corynascidia than to any other genns. But even so the 

 differences in the branchial sac of the two are too great to permit 

 them to be placed in the same genns. Even were w^e to imagine 

 the retrogressive modification of the interstigmatic vessels of 

 Corijnascidia already so far advanced, continued to their com- 

 plete obliteration, the internal vessels with their long triangular 

 processes (see Herdman '82, pi. XXV, fig. 6) would still have 

 no counterpart in the sac of Benthascidia nvicliaelseni. I am not 

 disposed to attach as much importance to the internal vessels and 

 papillae of the sac as some writers have done, yet considering all 

 the groups of Ascidians in which they occur, they certainly pre- 

 sent a high degree of constancy, and to place together in the same 

 genus two species one of which possesses true stigmata, internal 

 vessels, and papillae, while the other has the vessels alone, would 

 be to grossly violate the most reasonable view we have been able 

 thus far to attain relative to Ascidian morphology and classifi- 

 cation. But it is probable that more knowledge will find other 

 good generic differences between Corynascidia and Benthascidia. 

 The minute processes on the inner surface of the pharyngeal area 

 of the latter are likely to be such. And in this connection atten- 

 tion may be again called to the irregular club form of the branch- 

 ial tentacles. I have thought that the enlargements and irregu- 

 larities might possibly be either the beginnings or the remnants of 

 branches of compound tentacles. The conjecture rests however 

 on meagre evidence. Benthascidia michaelseni I place then, 

 pending more light on its structure, near Corynascidia, and with- 

 out taking a definite stand on the merits of the somewhat diverg- 

 ent views recently expressed by Michaelsen and Sluiter relative 

 to just how this and the several related genera should be grouped 

 into families, should hold it to belong to the order Ascidiidae as 

 recognized by Herdman. However, before dropping the question 

 of the relationship of our species, I deem it worth while to call 

 attention to the particulars in which it inclines toward Clavelina, 

 or perhaps better the Clavelinidae. The large open, lobeless 

 branchial orifice may first be noted ; but should my interpretation 

 of the branchial web as being without true stigmata (which I 

 do not believe will happen) prove to be at fault, then we should 



