254 J. G. WALLER ON AN UNDESCllIBED BRITISH SPONGE. 



Classifications are all artificial ; few genera or species are so 

 ■absolutely distinct but that they may be found occasionally to 

 overlap one another. It is this indeed that adds so much interest 

 to the study of natural history in all its departments. The 

 example, about to be considered, appears to me to have this 

 interest. 



The genus Hymeniacidon, of the: author referred to, is cliarac 

 terised by a very irregular distribution of the spicules in the 

 formation of the skeleton. Out of this genus he made another, 

 which he called Rapkiodesma, on account of some of the spicules 

 assuming a fasicular arrangement, in parts at least. Besides 

 this, there were other special forms : notably, minute spicules of 

 an anchorate shape, making rosettes upon the membranes, as well 

 as otherwise dispersed. Foreign observers had already noted these 

 characters, and given a name to a large cognate class — viz., 

 Esjyeria, from Esper the naturalist. Another feature is also 

 pretty constant — viz., that the skeleton spicules are chiefly acuate, 

 varying to the fusiform in the shaft with a pin-like knob. 



Now, this example differs from others in having two kinds of 

 spicules in its skeleton — the acerate, of the variety called fusiform, 

 and the acuate, with variations as above given. The skeleton 

 conforms to the characters expressed by Dr. Bowerbank under 

 his genus Hymeniacidon. The somewhat stout dermal mem- 

 brane, however, has, besides the spicules similar to the skeleton, 

 fine hairlike forms whicb, here and there, make up fasciculi, 

 though slightly compacted together. Intermingled with these, 

 though sparsely, are very fine tricurvate spicules, and occasional 

 bihamate ones, with no or v-ery slight variations in the thickness 

 of the shaft. Now these forms are common to the genus Rapliio- 

 desma and its allies, though the bihamate spicule is generally 

 contort — that is, has its terminals twisted in different planes. 

 The examples in this instance appear to keep to the same plane, 

 though it is not easy to determine this point when they are 

 seen in situ ; still this is not without a parallel, to which I 

 shall refer. 



I now come to the form of spicule which is almost invariably 

 associated with the genus referred to — viz., the anchorate, and 

 that specialised as inequipalmate anchorate, tridentate, and 

 bidentate. These, which are now generally called " flesh spicules," 

 have a special interest in their varying shapes, and also in the 



