250 Dr. G. C. J. Vosmaer on the 



adopted in the systematic portion of the work could not clearly 

 elucidate questions of relationship, and hence the necessity 

 for a more careful consideration of the question in this place. 

 Many people have placed the calcareous sponges in too close 

 connexion with the siliceous ones, largely owing to Fritz 

 Miiller's vain attempt to derive the calcareous and siliceous 

 structures from horny fibres. I have here followed the 

 main division of Gray, and have accepted two classes — 

 Calcarea and Non-calcarea — a proceeding as to the correct- 

 ness of which people seem to be more and more agreed. The 

 first spongologist of the present day, F. E. Schulze, accepts 

 this classification*. There are absolutely no transitions be- 

 tween the two classes ; and since the spicules appear at a very 

 early date in the larva, it can be only the very earliest develop- 

 mental phases which are common to the two. This primary 

 division thus appears to be a natural one. 



The Porifera Non-calcarea appear to me to be divisible into 

 three orders : — Hyalospongias, Spiculispongite, and Cornacu- 

 spoHgige. The Hyalospongise t all have this in common : — 

 their skeleton is composed of spicules based upon the triax- 

 onid type. The Cornacuspongise are distinguished by a new 

 element, spongin ; and in the Spiculispongia3 the " spi- 

 cula " are the chief distinguishing feature. It appears to me 

 that the genera within each order are more nearly related to 

 each other than to the genera of other orders ; and if this be 

 so, as I shall immediately endeavour to show, then the classi- 

 fication is a natural one. These three orders nevertheless are 

 not nearly so sharply separated from one another as are the 

 two classes. We do not, it is true, know of any direct transi- 

 tions from the Hyalospongife to the other orders. Still there 

 are certain facts which perhaps indicate a possible connexion. 

 Schulze appears to accept no connexion at all when he says, 

 "and however plausible, indeed almost self-evident is the hypo- 

 thesis that the latter (six-rayed spicules) may also atrophy and 

 give rise to spicules with fewer axes, so that the spicules 

 might even all become monaxonid, we know as yet no 

 Monaxonia in whose spicules we can detect any indication 

 (such as through crossing canals) of a descent from triaxonid 

 spicules " \. But we do find in the literature statements to 



* ' Ueber deu Bau und das System der Hexactinelliden ' (Berliu, 1886), 

 p. 82. 



t Hexactinellidee, anctorum ; but I prefer on principle to give names 

 ■with similar endings to equivalent divisions, rather than to abide rigor- 

 ously by priority. In the case of genera and species, on the contrary, I 

 keep as much as possible to the laws of priority. 



X Loc, cit. p. 34. 



