W. G. VanN'ame — Bermuda Ascidians. 395 



Avere not luucli smaller. They "svere growing attaclied to stones- 

 along the shores of Coney Island and Long Bird Island. 



From these, though the material is too scanty to give a satisfactory 

 idea of the individual variations Avhich specimens of this species are 

 likely to exhibit, I believe that the Bermuda form is sufficiently dif- 

 ferent from the West Indian one. to justify its description, provi- 

 sionally at least, as a new variety. 



The body is ovate, slightly longer than deep, and decidedly com- 

 pressed laterally. The test is not thick; it is soft and flexible, light 

 colored, and Avould be translucent were it not for the dense coating 

 of sand and shell fragments which cover not only the surface, but 

 are more or less buried in the test substance. The area of attach- 

 ment is small. The siphons are Avide apart in two specimens, in the 

 other they are rather near together. They are rather short in all 

 cases. The appearance of the animals is rather that of a Molgida 

 than one of the famih^ to which they really belong. 



The mantle is thin and more or less transparent with Aveak mus- 

 culature. In one specimen the tips of the siphons are pink. None 

 of the other specimens show any red color on any part of the body. 



In all these particulars the examples differ from Traustedt's de- 

 scription, in which the test is described as leatherly Avith a wrinkled 

 surface, and the mantle musculature as A^ery strong. 



There are about a dozen large tentacles beside some smaller ones. 

 They differ greatly from those of JI. ruhrilnhki, the largest ones 

 being bipinnate (fig. 84). The dorsal tubercle, in the specimen in 

 Avhich I examined it, had a U-shaped aperture with one horn incurved, 

 but not sufficiently to form a spiral. The dorsal lamina is provided 

 with numerous tentacular languets. They begin a little Avay back 

 from the anterior end, the lamina being plain for a little distance. 



As in the last described form, there are six branchial folds on each 

 side. There are, however, fewer internal longitudinal bars (I counted 

 only ten or elcA^en on one side of one of the longest folds) and they 

 ai'e separated hy 7 or 8 or even 9 stigmata in the meshes on the 

 interspaces between the folds, instead of 4 or 5 as in ^ rubrilabia. 

 The stigmata are also longer and narrower than in that species, but 

 in other respects the branchial sac resembles that of H. rubrilabici. 



The intestinal loop is rather narrower than in that species and the 

 reproductiA'e organs differ, the gonads being spherical though 

 arranged in a similar manner along each side of the genital ducts, 

 Avith which they communicate by short branch ducts. Thei'e is only 

 one series of reproductive organs on each side. On the left side it 

 lies within the intestinal loop. 



