SPONGES. 107 



that author sub-divides into two sections, according to the prevailing form of 

 megasclere, viz., " (Jlavulida " and " Aciculida." This sub-division, however, appears 

 to me to be both unnecessary and unnatural, especially when its originator places the 

 genus Tethya, with its stylote megascleres, in the division " Aciculida," which is 

 characterized by oxeote megascleres.'" I therefore propose to divide the sub-order 

 immediately into families. Of these TorsENT (26) makes nine, of which five are 

 represented in the present collection. 



In addition to these, I here include the family Ghondrosidse, which have completely 

 lost their megascleres, and, in the genus Choadrosia itself, their microscleres also. 

 Some of the old group Axinellidaa (e.g., Vibulinus, with astrose microscleres) must 

 likewise be included in this sub-order, and it may prove necessary to institute a new 

 family Astraxinellidaa for their reception. 



On the other hand, I am inclined to think that some of Topsent's families will 

 prove to be superfluous. Thus I propose to abandon his " Coppatiidse " and 

 " Streptasteridse " in favour of the older " Epipolasidse " of Sollas. 



Family: EPIPOLASIDSE. 



Astromonaxonellida with oxeote megascleres, and usually euasters for microscleres. 



This family, founded by Sollas (15) for the reception of the genera Amphius, 

 Coppatias, and Asteropus, is, at any rate as regards what we may consider to be its 

 typical representatives, evidently very closely related to the Stellettidaa, a fact which 

 has already been recognised by previous writers. In short, we may safely regard 

 these forms as reduced Stellettids which have lost their tetraxon megascleres, and 

 Sollas himself placed the family next to the Stellettidse as an appendix to the 

 Euastrosa, though apparently with some doubt. 



Topsent divided the genera of Sollas's Epipolasidas between his own families 

 Coppatiidse and Streptasteridse. In the Coppathdae he also includes certain genera, 

 such as Spongosorites, which have no microscleres at all. Spongosorites is well 

 represented in the Sponge-Fauna of Ceylon, but, for reasons which will appear 

 subsequently, I find it desirable to remove it to the Axinellidse. 



Coppatias, Sollas (15). 



Epipolasidse with the skeleton composed of an irregular interlacement of oxea, and 

 microscleres in the form of euasters only. 



This genus is almost synonymous with the genus Stelletthiopsis as employed by 

 Carter. 



* It should be pointed out, however, that Sollas and ToPSENT regard the megascleres of Tethya as 

 modified oxea, which they term strongyloxea, but it is impossible to distinguish these from styli. 



P 2 



