32 THE VOYAGE OF H.M.S. CHALLENGER. 



of a wing, so that an accurately transverse section of the trilobate filament exhibits 

 the mesoglcea in form of a cross, the arms of which are broad and wing-shaped. 



The generative organs lie in the thin septum wliich is intercalated between the 

 retractor and the mesenterial filament, and were mnle in one specimen investigated, in 

 the other female. The testes are r5 cm. long, 0-2 cm. broad, composed of separate 

 follicles which are arranged in about thirteen transverse swellings. At the edge of the 

 oro-an occur small bodies, recognisable only in transverse sections, which I take to be 

 the first commencements of the follicles ; the supporting lamina widens out, enclosing 

 a space in which are included roundish cells (spermatoblasts?), fewer (five) or more 

 numerous according to the size of the cavity. The latter always opens towards the 

 epithelium by a small but obvious pore. The latter would argue, if there were any 

 question here of stages of development of the testis, for its derivation from endodcrm ; 

 unfortunately, however, the mesenteries were not sufficiently well preserved for a close 

 histological investigation. 



In the female organ the conditions were similar ; the ova are irregularly scattered 

 in the mesentery as larger or smaller grains ; those of fair size project above the surface, 

 while the largest of all stand out markedly beyond its plane, and are connected with 

 the mesentery only by means of a fine pedicle. The pedicle passes into a chorion which 

 surrounds the ovum on all sides, the latter being about 1 mm. in diameter. In this 

 condition the ovum appears to be already in segmentation. 



In the mesenteries occur, finally, external stomata ; they are oval, about 0-5 cm. 

 lono-, and occur rather to the outer side of the great mesenterial muscles, on a level 

 with the wTcath of tentacles. Whether also internal mesenterial stomata exist just 

 below the oral lip, remains doubtful. 



From the type of the true Hakampce, this Actinian diverges in exhiljiting a 

 commencement of additional mesenterial cycles, although these are extremely weakly 

 developed. The accessory mesenteries are small projections, which, in the upper part 

 of the body alone, emerge from the angle between body-wall and oral disc ; here 

 there occur pairs of mesenteries both of the second and third orders, readily dis- 

 tino-uishable by difference of size. Since, as we have seen above, the number of the 

 tentacles also is larger than in the true Hcdcamjjce, the genus Hcdcampdla leads up 

 to the remaining Ilyanthidte, and through them to the true Actiniae. 



Halcampella, sp. (?)* 



Habitat. — Shallow water; St. Vincent, Cape Verde Islands, July 1873. 



To the genus Halcampdla doubtless belongs another Ilyanthidan with numerous 

 tentacles, although too much mutilated for close investigation or systematic determina- 

 tion. It is to be distinguished from Halcampella maxima at once, by the absence of 

 incrustation on the body-wall. 



