482 PROCEEDINGS OF THE NATIONAL MUSEUM. tol. 45. 



Hartmeyer, 1903, points out that most of the characters which I 

 reUed upon chiefly for distinguishing tuberculata from suhpedunculata, 

 and both these from aggregata, occur within the range of individual 

 variation of aggregata. He, however, finally decides not to include 

 my species in the older species, but considerately leaves to me to 

 determine what weight in this direction should be attached to his 

 observation. After deliberating long on the considerable evidence 

 now before me I am strongly of the opinion that, in spite of the 

 undoubted great range of variation of both aggregata and the Bering 

 Sea animals, the latter are sufficiently distinct from the former to 

 deserve being considered as specifically distinct. The pomts which 

 I rely upon most in support of this conclusion are the greater degree 

 of tuberculation and of hardness of the test of tuberculata, the darker 

 coloration of tuberculata, and particularly the character of the stomach. 

 This member is, judging from my examination of specimens of aggre- 

 gata sent me by Hartmeyer, as well as by the figures and descriptions 

 given by various authors, relatively longer, more sharply set off from 

 the intestine, and, perhaps most important of all, more distinctly 

 folded as viewed from the external surface than in any of the north 

 Pacific animals. It is probable, furthermore, that the longer rectal 

 portion of the mtestine in aggregata and the absence there of anal 

 lobes are constant differential marks. Although, as indicated by the 

 tabulation, there is great variety in the lobulation of the anal rim in 

 the specimens I have examined, it is doubtful if they are wholly 

 absent in any case, as they seem to be regularly in aggregata. 



The almost complete invisibility of the folds of the stomach wall 

 as seen from the external surface, though very prominent inside in 

 aU the Bering Sea specimens examined, is in striking contrast with 

 their distinctness in aggregata. The difference is due seemingly to 

 the greater thickness and closer adherence of the peritoneal layer in 

 the tuberculata group. I am quite confident, therefore, that future, 

 more thoroughgoing study will increase rather than diminish the 

 distinctness of a group that will fall under the designation tuberculata. 



Concerning the specimens from Albatross 3262 I have more doubt 

 than about any others which I am calling tuberculata. Externally 

 they strongly resemble many individuals of Styela rustica. The ten- 

 tacles, furthermore, in some of these are remarkably short and stubby. 

 Also the great number, seven, of branches of the ovary in some indi- 

 viduals arouses suspicion when considered in comiection with the other 

 differences noted that we may be dealing here with representatives 

 of a different group. In view, however, of the great variation to 

 which the whole genus is obviously subject I have not deemed it 

 best to describe these as a distinct species. For example, the large 

 number of ovarian branches (fig. 27) would have been considered, 

 according to the earlier practices in dealing with the Dendrodoas, in 

 itseK a sufficient characterization of a distinct species. But, as the 



