igii.] N. Annandai^e : Spoiigcs and Vcrmetid Molluscs. 51 



extent. Northwards the fauna they represent gives place to one 

 consisting largely of sedentary organisms such as Alcyonaria and 

 Antipatharia fixed to the recent conglomerate to which reference 

 has already been made. vStill further northwards, all round the head 

 of the Bay, only those animals can exist which can endure muddy 

 water and can live without a solid surface of attachment. South- 

 wards the Siliquaria-heds are replaced, in the more sheltered 

 and probably Salter waters of the Gulf of Manaar, by coral 

 reefs. 



From a strictly biological point of view it is interesting to 

 notice that neither of the two species of sponges found associated 

 with the three species of Vermetidae is peculiar to these shells or, 

 indeed, to a habitat or manner of life similar to that implied by 

 the mollusc's peculiar method of growth. In two cases out of 

 three, however, the sponge appears to be modified to some extent 

 by the peculiarities of the shell with which it is associated, or at 

 any rate in accordance with these peculiarities. 



Several other organisms were found in large or considerable 

 numbers and in some degree associated with the shells and sponges. 

 Dead shells (of which there were a considerable number in some 

 masses) of both Siliquaria coclilearis and Spiroglyphtis cummingi 

 were often inhabited by the peculiar little hermit-crab Troglo- 

 pagurus manaarensis ,^ hitherto only known to frequent holes in 

 corals in the Gulf of Manaar. A small bivalve mollusc {Area 

 domingensis var. divaricata) was also found in considerable num- 

 bers in dead shells of the two species, anchored to the inner sur- 

 face by a byssus of horny consistency, while specimens of a larger 

 species of the same genus {A. adamsiana) were found in inter- 

 stices of the sponges and between the shells. Sedentary organisms 

 were not so numerous on the external surface of the masses as 

 might perhaps have been expected, but a considerable number of 

 small monaxon sponges and a few polyzoa occurred in this posi- 

 tion and shells of the bivalve Chama ruppcllii were common. 



II.— DESCRIPTION OF THE SPONGES. 



The sponges here described belong to two grades of the order 

 Tetraxonida, namely the Lithistida and the Monaxonellida, if we 

 adopt the nomenclature proposed by Prof. Dendy in his account 

 of the sponges in part iii of Prof. Herdman's report on the pearl- 

 fisheries of Ceylon ; for two of the three forms found in associa- 

 tion with Siliquaria shells off the Madras coast represent varieties 

 of a species of Racodiscula (Lithistida) originally described by 

 Carter from the Gulf of Manaar, while the third represents a 

 species of Spongosorites (Monaxonellida) also described from the 

 Gulf of Manaar, by Prof. Dendy in the report to which reference 

 has just been made. 



1 Dr. J. R. Henderson, the author of the species, has been kind enough to 

 identify specimens. 



