On Bactylocalyx pumiceus (ShUchhury). By W. J. SolJas. 127 



that the great majority of these spicules are incomplete at one end, 

 and thus incapable of exact measurement. In some cases the end 

 has been apparently broken off, in others it appears to have yielded 

 to some solvent action, either after the death of the sponge, or 

 quite as possibly during its hfe ; for the sponge appears to have 

 been alive when first procured, and the eroded umbones of Anodon 

 and Unio shells show that such contemporaneous solution is not an 

 unknown phenomenon in the animal kingdom. 



The ends of the acerates are roughened by minute spines, 

 which give them a ragged appearance, and their tenuous extre- 

 mities are pointed. Associated with them are other acerate 

 spicules (Figs. 3 and 18) which differ in a number of minor 

 characters ; thus the latter are usually smaller than the former, 

 more often curved, and though sometimes pointed, yet very 

 frequently capitate clavately at one or both ends. The larger 

 acerates are excavated by a well-defined axial canal which, however, 

 never exhibits any trace of a sexradiate cross in any part of its course. 

 I have repeatedly examined a large number of perfect acerate 

 spicules with a view to making sure on this point, and 1 am able to 

 state therefore with full confidence that none of them show the 

 least signs of a sexradiate character.* Instead of being aborted 

 sexradiate spicules, they are from my point of view the least modi- 

 fied descendants of the simple acerate spicules of which the early 

 ancestral sponge was composed ; the sexradiates on the other hand 

 having departed the most widely from the original type. 



The coarse meshes where they open at the surface of the 

 sponge, appear as the circular mouths of minute tubes, walled in 

 with the large fibre, and reminding one somewhat of the structure 

 of Aplirocallistes. It is into the large fibre surrounding these tubes, 

 but not into that forming their floor, that the acerate spicules are 

 inserted, which thus leave the tubes unencumbered within, but form 

 a beautiful fringe to them externally. 



Sexradiate sincules of the dermal layer. — These are remarkably 

 variable in all their characters ; the most typical form being that of 

 Fig. 5, Plate VII. This possesses the full complement of six rays, 

 four lying on the surface of the sponge, one descending into its 

 network, and the sixth projecting distally : the distal ray is short, 

 straight, and rounded off" at the end, the other five rays are much 

 longer, more or less curved, and attenuated to very fine pointed 

 extremities. All are minutely microspined for the whole or a 

 portion of their length. The greatest breadth of the rays is 

 0-0003". 



In other instances we find the distal ray becoming much 

 shorter, frequently capitate (Fig. 16, Plate VII.), and often disap- 



* On referring to Mr. Carter's paper (Joe. cit.) I find that his examination of 

 the acerate fusiform spicules oi Dact;jlocalyx subglobosa led him to the same results. 



