Ixviii THE VOYAGE OF H.M.S. CHALLENGER. 



matter, the fibre thus constituted being echiuated by special spicules which project 

 from its surface more or less at right angles, or the spicular core may be absent, 

 leaving only the horny fibre and the echinating spicules. Yet here again, we have at 

 least one species, viz., Myxilla frondosa, nobis, which seems to indicate a transition 

 between these two types of skeleton arrangement. In the universal absence of chelae, 

 however, the Axinellidse are sharply defined from the Ectyoninae and other Desma- 

 cidonidae. 



We come now to the second of our suborders, viz., the Clavulina. We have already 

 indicated the probable relations of this group to the Halichondrina. The four most 

 prominent features of the group (none of which, however, will probably be found to be 

 quite constant throughout) arc (1) the presence of radiately arranged tylostylote 

 spicules ; (2) the cork-like, granular ground substance ; (3) the presence of a distinct 

 fibrous cortex ; and (4) the absence of spongin from the skeleton. As we have already 

 hinted, we have no doubt whatever that this group is closely related, through the 

 Tethyadse, to the so-called Tetractinellida. We shall not further discuss this question, 

 but refer the reader to Vosmaer's latest remarks on the subject.^ 



According to the presence or absence of microsclera we divide the Clavulina into 

 two families ; (l) Suberitidse, in which there are no microsclera, and (2) Spirastrellidie, 

 in which microsclera are present. This appears to us to be a much more natural 

 arrangement than that proposed by Dr. Vosmaer,' viz., according to the presence or absence 

 of mammiform projections on the surface of the sponge.^ This difference in ideas is 

 perhaps partly accounted for by the fact that Vosmaer has omitted from his scheme the 

 two very important genera Spirastrella and Latrunculia, both of which are well repre- 

 sented in the Challenger collection. The Clioniadse, of which group the members, save 

 that they ajjpear to be mostly Clavulina, have only the boring habit in common, will have 

 to be distributed chiefly between the Suberitid^ and Spirastrellidse, as there is no reason 

 at all for supposing that species of both these families (and of others also) may not have 

 independently acquired the boring habit ; and the spiculation of the difierent boring 

 sponges seems to indicate that this is the case. 



1 Bronn's Klass. u. Ordnung. d. Thierreichs, Porifera, p. 474 (translated by Dendy, Ann. Mid Mag. Nat. Hist., 

 ser. 5, vol. xix. p. 249). 



2 Bronn's Klass. u. Ordnung. d. Tliierrichs, Porifera, p. 328, et seq. 



3 Vosmaer also gives in the diagnosis (loc. cit.) of his Polyniastida; " Faserrinde meist sehr deutlich," and of his 

 Suberitidje "eine echte Faserrinde fehlt immer;" hut, inasmuch as he includes the genera Siiherites and Stijlocordyla in 

 his Suberitidae, we must emphatically deny the correctness of the latter statement ((/. our own researches on the minute 

 anatomy of these genera, supra, and PI. L.). 



