GEODIA AGASSIZII. 117 



by chonal sphincters, generally found open and usually 120-500 ii wide. 

 Many of these efferent chonal canals are destitute of dermal sieves and open 

 out freely on the surface (Plate 26, fig. 13, those to the right). In some of 

 the specimens nearly all of them are thus naked; in most, however, some 

 of these efferents are covered by dermal sieves, composed of nets of threads, 50- 

 120 fj. broad, with round meshes of very variable size (Plate 26, fig. 13, those to 

 the left, figs. 14b, 15b). Thus, at first sight, it appears as if there were, in this 

 sponge, two different kinds of efferents, cribriporal and uniporal ones. A closer 

 examination, however, reveals remnants of dermal sieves in most of the aper- 

 tures appearing at first sight uniporal. I think it therefore highly probable 

 that all the efferents are, like the afferents, provided with sieve-membranes 

 (cribriporal) in the living sponge, and that, where they are now missing, they 

 have been lost post mortem. 



The skeleton consists chiefly of large choanosomal and small dermal amphi- 

 oxes, orthoplagiotriaenes, mesoprotriaenes, anatriaenes, large oxyasters, large 

 oxysphaerasters, small strongylosphaerasters, and sterrasters. To these spicules, 

 which occur in all the specimens, a few large ampliistrongyles, slender and 

 cylindrical or thick and club-shaped styles, mesoclade or amphiclade ortho- 

 plagiotriaene-derivates, anadiaenes, sterroids, and other irregular forms may 

 be added. In the specimens attached to hexactinellid skeletons, particularly 

 in the young specimen from Station 4228, small hexactinellid spicules, hexac- 

 tines, and scopules are also found imbedded in the choanosome. These foreign 

 spicules are by no means restricted to the base of the sponge which is attached 

 to the dictyonine network of the hexactinellid, but are found in all parts. 



The large choanosomal amphioxes are arranged radially and form loose, 

 conical bundles extending from the centre or base to the surface of the sponge. 

 Some of the outermost of these spicules protrude beyond the surface and thus 

 take part in the formation of the fur (Plate 27, figs. 1, 2). The rare amphi- 

 strongyles, which I have observed only in the young specimen from Station 

 4228, are scattered in small numbers between them. The rare, large styles, 

 both the thinner cylindrical and the thicker club-shaped ones, are arranged 

 radially like the large amphioxes among which they occur, their rounded end 

 being situated distally, their pointed end proximally. These spicules are more 

 numerous in the distal than in the proximal parts of the bundles and often pro- 

 trude their rounded end beyond the surface. I have found the thick club-shaped 

 styles only in the specimens from Stations 3168 and 4193, the thin cylindrical 

 ones in all the specimens with the exception of those from Stations 2887, 2978, 



