182 CEYLON PEARL OYSTER REPORT. 



Owing to their loss of microscleres and want of other well-marked characters, this 

 family is one of the most unsatisfactory with which we have to deal. The plumose 

 arrangement of the megascleres in the main skeleton cannot be regarded as exclusively 

 diagnostic, for it is met with also in Ectyoninse and even in Chalininse, and too great 

 reliance upon this character has led to the inclusion of forms amongst the Axinellidse 

 which certainly should not be included in that family. I have already indicated that 

 those so-called Axinellids in which true asters have been observed (e.g., Vibulinus) 

 should be placed amongst the Astromonaxonellida. The genus Raspailia, similarly, 

 has been removed from the Axinellidae to the Ectyoninse, but we are still left with a 

 somewhat heterogeneous collection of sponges which it is extremely difficult to define, 

 and which will probably be subjected to considerable re-arrangement in the future. 

 A curiously constant feature in this group is the irregularity exhibited by the ends of 

 the megascleres and the manner in which the stylote and oxeote forms tend to pass 

 into one another. 



It is highly probable that the group, even as here restricted, is of polyphyletic 

 origin. 



Spongosorites, Topsent (emend.). 



Axinellidse with the main skeleton composed of an irregular but dense reticulation of 

 large oxea or styli, and dermal skeleton composed of a thin layer of very much 

 smaller oxea lying tangentially, and in close contact with the main skeleton ; 

 oxea typically biangulate. 



Topsent (14, &c.) places the genus Spongosorites in his family Ooppatiidse, assuming 

 that it has lost the characteristic microscleres, and calling attention to the apparent 

 tetractinellid affinity indicated by the biangulate oxea. It appears to me, on the 

 other hand, to come more naturally amongst the Axinellids, being not distantly 

 removed from Leueophlceus and Ciocalypta, as indicated by the new species described 

 below, in one of which we perhaps find a clue to the evolution of the curious finger- 

 shaped " processes " of the latter genus. In Ciocalypta tyleri, var. aberrant, we also 

 sometimes find biangulate oxea. 



The genus Spongosorites may also be nearly related to Thiele's Dactylella (39). 



Spongosorites topsenti, n. sp. Plate XII., fig. 1. 



There are six specimens of this curious sponge in the collection, differing so mucn 

 amongst themselves in external form that it seems desirable to give a short account 

 of each. 



R.N. 152 is a depressed cake-shaped sponge of nearly circular outline, with convex 

 upper and almost flat lower surface, the latter having been evidently attached to the 

 substratum all over. Greatest diameter 36 millims., thickness in the middle 

 13 millims. The upper surface is somewhat corrugated, with shallow grooves 

 (indicating underlying canals) radiating in a stellate manner from low, mound-like 



