DENDY— HOMOSCLEROPHORA AND ASTROTETRAXONIDA 



245 



is evidently closely related to a corresponding species of Aurora which still retains its 

 trisenes. Aurora (Rhabdastrella) distincta (Thiele), with large smooth spherasters but 

 without trisenes, is closely related to Aurora globostellata (Carter) with large smooth 

 spherasters and trisenes. Aurora crihriporosa, n. sp., with large rough spherasters but 

 without trisenes, is closely related to Aurora providentice, n. sp., with large rough spher- 

 asters and trisenes ; and Aurora (Diastra) sterrastrcea (Row), with sterrospherasters but 

 without trisenes, is closely related to Aurora roiui, n. sp., with sterrospherasters and 

 trisenes. 



Aurora (Rhahdastrella) distincta was first described by Thiele [1900] as a Coppatias, 

 but he subsequently [1903] separated it from Coppatias under the new generic name 

 Rhabdastrella. 



Aurora sterrastrcea was described by Row in 1911 and also recognised as an 

 epipolasid, for which the new genus Diastra was instituted. If these genera were to 

 be retained it would, I think, be necessary also to propose a new genus for Aurora 

 cribriporosa, but all three genera would have to be based upon very inadequate characters. 

 We cannot include all three species in a single genus distinct from Aurora unless we are 

 willing to retain that single genus merely as an artificial one of polyphyletic origin, and 

 any argument in favour of doing this would apply equally to the case of the so-called 

 family Epipolasidse. There seenis nothing for it, then, but to abandon the genera 

 Rhabdastrella and Diastra and place all three species in Aurora. It may be said that 

 we ought also to abandon the genera Amphius, Asteropus and Jaspis (Coppatias), and 

 this I shall be quite prepared to do when I feel as certain about their origin from 

 particular tetractinellid genera as I do in the case of Diastra and Rhabdastrella. 



In this connection it is also necessary to say something about SoUas's genus Magog 

 [1888], which was proposed for the reception of Carter's Ghondrilla sacciformis [1879 b]. 

 This species was described by Carter as having " acerates " (oxea) and " globostellates " 

 (spherasters). SoUas accepted this as correct after examination of one of Carter's slides, 

 and diagnosed the genus Magog thus : — " Tethyidse in which the rhabdus spicule is an 

 oxea, which is confined to the choanosome." 



Were the spiculation of Chondrilla sacciformis really as described I think we should be 

 obliged to accept that species also as an epipolasid Aurora. I have in my possession, how- 

 ever, in Mr Carter's cabinet, two microscopical preparations of his Chondrilla sacciformis, 

 one consisting of teased fragments mounted in balsam and the other of a number of fair-sized 

 fragments put up dry in a cell. I have examined sections and boiled-out spicules from 

 one of these fragments, but neither in these preparations nor in Carter's own balsam slide 

 can I find any oxea at all. The species is a genuine Chondrilla, identical with Thiele's 

 Chondrilla grandistellata [1900]. As to how the mistake arose I can only surmise that 

 the preparation from which Carter's original description was taken, and which is now 

 presumably lost, must have contained, as accidental inclusions, oxeote megascleres of 

 Rhap>hidhistia spectahilis, which Carter described as a thinly encrusting sponge growing 

 on the same mass as Chondrilla sacciformis. That the preparations now in Mr Carter's 

 cabinet, and labelled by himself " Chondrilla sacciformis" were made at a later date than 

 the original description seems certain, as they bear the date 1881. That they were taken 



