350 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



(Vosmaer, 1885, pLl). Doubtless the shape of the individual animal 

 is subject to some change, but it seems improbable that the same 

 (oscular) face could at one time be nearly flat and at another strongly 

 convex. In some other points, however, the Philippine specimen 

 agrees with T. hemisphericum. Thus the oscular face in the latter is 

 covered with a furze of short tylostyles, and there are radial ascend- 

 ing skeletal bundles passing from the center of the lower cortex to the 

 upper cortex (Vosmaer, 1885, pi. 2, figs. 17, 18). In Topsent's (1913 

 b, p. 20) specimens (from 72° 37' north, 20° 0' east) of this species 

 the general anatomy is as in the Albatross specimens. But in the 

 latter the stratification of the peripheral body appears to be much 

 more conspicuous and indeed different in details from that of the 

 Atlantic species. 



T. sarsii Ridley and Dendy (1887, p. 218), which the Philippine 

 sponge resembles in general shape, as do some specimens of T. hem- 

 isphericum (see Vosmaer, 1885, pi. 1), has one comparatively long 

 oscular tube about in the center of the flat face. Round it are a num- 

 ber of "small round, translucent patches," the nature of which is 

 doubtful (Ridley and Dendy, 1887). In the figure given, pi. 43, they 

 look much like the oscular papillae of my form. The cortical skele- 

 ton of both surfaces is considerably simpler than in the Philippine 

 species (Ridley and Dendy, 1887, p. 217, fig. 8) ; the choanosomal 

 spicules form radiating groups; and ascending radial bundles are 

 apparently absent. Thus there are differences, although minute ones. 

 T. sarsii occurs in Australian waters, off the Azores, and in the In- 

 dian Ocean (Dendy, 1921 b, p. 151). 



Another species of the genus is T. irregularis Ridley and Dendy 

 (1887, p. 220), from off Valparaiso. Radiella sol Schmidt (1870, 

 p. 48), from near Cuba is regarded by Hansen (see Ridley and 

 Dendy, 1887, p. 217) and Topsent (1898) as synonymous with T. 

 hemisphericum. It has, however, only one oscular protuberance in 

 the center of the upper face, and the fringe spicules project upward 

 (Schmidt, 1870, pi. 4, fig. 6). Halicnemia patera Bowerbank (see 

 Ridley and Dendy, 1887, p. 218) is probably yet another species, 

 although Ridley and Dendy think that, in spite of the similarity 

 of external form, this sponge does not belong in Trichostemma. 

 The peculiar spicules, however (Bowerbank, 1864, pi. 10, figs. 228- 

 233), on which Ridley and Dendy lay stress seem to be only modified 

 tylostyles, in which case they can scarcely be thought of as a reason 

 for excluding this species from the genus. Topsent (19136, p. 21) 

 has recently described a new and beautiful species, T. grimaldii, 

 from the North Atlantic. 



