NOliTH AMERICAN PORIFERiE. PART II. 541 



bles Phyllospongia in its general aspect. On the other hand, an eqvially close approxima- 

 tion to true Spongia is to be observed in the softer and more pliable fronds of C. perforata, 

 which might readily be mistaken for a true Spongian form. 



Carteriospongia Otahitica Hyatt. 



Spongia Otahitica Espcr, Pflanz., 2, 270, \i\. 44. 

 " " Ellis and Solander, pi. 59. 



This curious species is singularly imiform in the aspect of the surface, and the specimens, 

 all of which are from Zanzibar, have the leaf-like or palmate form. This seems rarely to 

 be single ; in most cases the leaves of neighboring colonies seem to anastomose, so that 

 nearly every colony is supported by several stems. Both surfaces appear to be similar at 

 first sight, but they are quite distinct. The outer side in j^erfect specimens is ribljed, like 

 the inner, but the ribs, or longitudinal ridges, are thinner and longer than those on the 

 inner or upper side. The latter have a more luxuriant growth and a more tufted aspect. 

 Between these minute ridges the wall of fibres is thin, and pierced by rows of holes. 



The primary fibres are in fascicles, and overloaded witli debris. The mesh, where it can 

 be seen, is regular, as in other species of this genus. The preparation of the fibres of the 

 original of Caccospongia lamellosa Ehler has been compared, and shows great differences. 

 The primary fibres are single and loaded with debris, but the mesh is quite similar to that 

 which is usual in this genus, esjiecially as regards its regular form and the thickness of the 

 secondary fibres. These differences, though considerable, might occur in the same species, 

 especially if one lived in clear waters, and the other in a place where there was much float- 

 ing matter. It is very evident, however, that there are no holes through the fronds in 

 lamellosa, and that the whole surface is not cut up by discontinuous ridges as it is in this 

 form. 



Loc, Zanzibar, in Soc. Coll., and in Peabody Academj^, and Yale Coll. 



Variety aplanata. This precisely resembles the preceding, but the ribs or ridges are 

 depressed and thinner, and the surface is finely channelled. 



Loc, Madagascar ? m Soc. Coll. 



Carteriospongia radiata Hyatt. 



This species closely resembles G. Otahitica in general appearance and mode of growth, 

 but is easily distinguished from it by the smoothness of the surface on both sides. The 

 surface, also, instead of being evenly hirsute on the upper side, as in C. mystica, has a 

 veil composed of minute fibres, which bridge over all inequalities of the interior, and con- 

 ceal the longitudinal ridges which are formed internally. The primary fibres are single 

 and pointed, but project slightly above the veil and are evenly distributed. Upon the 

 lower or outer side no veil is found, and it differs from the upper or inner side in its 

 loose texture, which allows the primary filjres to be seen, and in the figures made by the 

 superficial canals. These radiate from a central, large pore, which represents the cloacal 

 aperture, and form numerous star-like or dendritic figures in the surface of the skeleton. 

 The interior is completely honeycombed by canals throughout the upper half of some 

 leaves, while the lower part, the stems, and the younger sponges, are solid. Whether this 

 is due to parasites or not, is uncertain. The microscopical structure is similar to 

 that of 8p. lamellosa Esper, but the fibres are much larger. There is also a notable 



MEMOIRS B08T. SOC. NAT. HIST. VOL. II. 



