J. G. WALLER ON A NEWLY DISCOVERED BRITISH SPONGE. 221 



The forcepiform is equally so, and they should not be classed 

 together. I do not think there can be anything gained by it, and 

 though one must feel respect for the opinion of such an eminent 

 worker in this department as Schmidt, it is a question of so simple 

 a character that all observers stand on the same ground. 



We have now to consider the analogies, which are not many, and 

 do not belong to the genus. First amongst them is Raphiodesma 

 sordida, a coating sponge also, but of a higher or more complex 

 type. In the dermal membrane of this sponge we get a great 

 number of bihamate spicules, also numerous anchorate ones, 

 and the skeleton spicule has resemblances in its arcuate and 

 clavate character. But the anchorate spicules differ in shape, and 

 many other important divergencies are to be found. I have before 

 alluded to the association of forms of spicules together as bearing 

 on the affinities of one class of sponges with another. But I do 

 not believe that any good classification of sponges will ever be 

 generally accepted if based on the forms of spicules only. They 

 are very important adjuncts, but of themselves are certainly not 

 sufficient. 



The nearly allied genus Hymedesmia furnishes an analogy in the 

 numerous complication of the bihamate spicules associated with 

 anchorate forms, in the species H. Zetlandica, and the fasci- 

 culated arrangement of skeleton spicules, there seen, bear some 

 resemblance. Hymeniacidon variantia is also remarkable in this 

 particular. 



In some exotic sponges also the membranes have often an abun- 

 dance of bihamate spicules, which form usually occurs indeed in 

 numbers, and rarely otherwise. But in this instance they are ex- 

 ceptionally numerous, and evidently play an important part in the 

 structure. 



In examining a filmy sponge, one must be very careful to assure 

 ourselves that it is fully developed, and not merely a beginning, or 

 you may very easily be mistaken, and call that a new species, which 

 indeed is but a known form in an early state. Some of the early 

 conditions of Halichondria incrustans might easily seem to belong 

 to Dr. Bowerbank's genus Hymedesmia, but a careful examination 

 of the spicules in this, as in other cases, may be sufficient to warn you 

 from falling into error. 



Before leaving the subject, I think I must call attention, for the 

 second time, to an erroneous conception of Dr. Bowerbank, if not 



