60 CEYLON PEARL OYSTER REPORT. 



SYSTEMATIC DESCRIPTION OF THE GENERA AND SPECIES 

 OF SPONGES IN PROFESSOR HERDMAN'S COLLECTION. 



Phylum PORIFERA, 



Class: NON-CALCAREA. 



Porifera without calcareous spicules. 



The non-calcareous sponges are divisible into four natural groups, for which 

 I propose to use the names Myxospongida, Triaxonida, Tetraxonida, and 

 Euceratosa respectively, and which may be conveniently regarded as of ordinal rank. 



Order: MYXOSPONGIDA. 



Non-calcarea which are primitively destitute of spicules and horny fibre ; with simple 

 canal system and usually large flagellate chambers. 



In this order I include the genera Halisarca, Bajalus, Hexaddla, and Oscarella, 

 all of which appear to me to be nearly related to one another, so that it is hardly 

 necessary to divide them, as is usually done, into two families, Halisarcidse and 

 Oscarellidse. All of them are, I believe, primitive forms, and should therefore stand 

 at the bottom of the series of non-calcareous sponges. 



It is cpiite possible that Lendenfeld's " Hexaceratina " may be closely related to 

 the " Halisarcidse," as supposed by that author ; but, if so, it appears to me that the 

 Halisarcida? are the more primitive forms, from which both Hexactinellida and 

 Hexaceratina have been derived. Lendenfeld, on the other hand, appears to regard 

 the Halisarcidse as being derived from the Hexaceratina by loss of horny fibres, and 

 in this opinion he is followed by Topsent (11), who observes " Depourvu a la fois de 

 spicules et de fibres conies, le genre Hexadella est vraiment le dernier chainon de la 

 chaine des Hexaceratina et doit occuper la place qu'on a parfois assignee, sans raison 

 valable, aux Halisarca." 



The Myxospongida may therefore be regarded as representing a common starting- 

 point, from which have originated the Triaxonida, the Tetraxonida, and the 

 Euceratosa, and I cannot agree with Professor Minchin (12) in placing this order 

 at the end of the siliceous series instead of at the beginning. The primitive 

 character of the canal system argues strongly against the belief that they are forms 

 in which the skeleton has been lost by gradual reduction, and the case is obviously 

 quite different from that of Chondrosia, whose highly specialized canal-system, 

 strongly developed cortex, and evident relationship to Choiulrilla, afford good 

 grounds for regarding it as a reduced siliceous sponge in which the absence of 

 skeleton is a secondary and not a primary character. 



