102 J. G. WALLER ON AN UNDESCRIBED 



from the other. R. lingua is the most robust of the genus, growing 

 to a larger size than any other ; its skeleton spicules are corres- 

 pondent ; we lose the clavate form, and they are simply acuate. It 

 differs mostly in the larger and more compound character of its 

 fasciculi, in which there is no pretence of the birch-like form 

 or of the bases or apices being coincident. The skeleton spicules 

 ally it to R. simplissima, which, however, is wanting in all the other 

 forms found in the genus.* 



I have thus shown the natural alliance of the gi'oup, and how 

 they diverge the one from the other. Whether they must not 

 hereafter, be brought closer to each other, is a matter we cannot at 

 present decide. I confess the tendency of my mind is to reject as 

 evidence of distinction of species very minute details ; at best it is 

 arbitrary, and the result of ouic ignorance rather than of our know- 

 ledge. It is, therefore, with something of a pang or twinge of 

 conscience I feel compelled to add a new name, but I sincerely hope 

 it may hereafter be disposed of. I propose that of Raphiodesma 

 minima, which expresses the smallness of its size as found. 



In the Sponge under consideration, the form of the anchorate 

 palmate spicule does not differ materially from that common to the 

 genus (Fig. 7). In its smaller size, it is very abundant on the 

 membranes, intermingled with its fine hair-like spicules. These latter 

 may be analogous to the extremely small acuate spicule of R. lingua. 

 It is well to note that spicules which belong to the membranes, and 

 generally of a minute size, are sometimes found nearly obsolete. In 

 few sections of R. sordida have I found in situ the tricurvate 

 figure given by Dr. Bowerbank, but in those of H. macilenta^ they 

 are extremely abundant, interweaving with the bihamates. These 

 latter also are often sparsely distributed in R. sordida, and vary in 

 size ; but this must naturally be looked for in examples not fully 

 developed. Dimensions of spicula, when given, must be taken with 

 some little reserve, for they are not constant in the same species, 

 and can never be depended upon as a character to distinguish a 

 species. 



There is still, however, something further to be said on analogies, 

 and the harmonious alliance between divergent forms. It is inter- 

 esting to trace the connection of species, still more so, when through 



* Refer to Dr. Bowerbank's " SpongiadaB," Vol. iii,, Plates xxxvii, xlvii, 

 Ixxvii, Ixxvi, xc. 



t I nsc this term as rofcrring to examples in my collection which agree 

 with Dr. B.'s description, ))nt my argument is that it is not a species. 



