54 Records of the Indian Museum. [Vol. VI, 



have found in different preparations of a considerable number of 

 specimens, single spicules or small groups of spicules belonging to 

 seven different types, the commonest of which consists of am- 

 phioxi that have clearly been derived from the skeletons of mon- 

 axons growing on the surface of the Lithistid. In several instances 

 I have found the sponge from which they clearly originated. These 

 adventitious spicules are scattered, together with small grains of 

 sand, in the collenchyma of the efferent canals and in the outer 

 parts of the sponge. In one preparation I found several rhabdi 

 like those of the var. spiroglyphi. Ihey apparently formed a small 

 vertical strand in the collenchyma of an efferent canal, but it is 

 impossible to be quite sure that they were not adventitious. The 

 external surface has been rubbed off most of my specimens of the 

 var. siliquariae , and with it the phyllotriaenes and amphiasters 

 have disappeared, but I have found them both in one preparation, 

 having the same forms and arrangement as in var spiroglyphi , 

 except that the amphiasters were perhaps a little shorter and more 

 regular in shape. 



Lying loose in the efferent channels near the osculum of a 

 dried specimen of this variety I found several little siliceous bodies 

 (pi. ix, fig. i8) that are probably the skeletons of embryo sponges. 

 They are formed of closely welded spicules resembling the desmas of 

 the adult sponge but smaller and more slender. Each body has the 

 form of a figure of eight somewhat attenuated, and measures 

 between 2 and 3 mm. in length ; the proportions differ in different 

 specimens, but one of the loops is usually rather larger than the 

 other One surface is flat, the other distinctly convex. Each loop 

 contains three relativeh^ large apertures, one on the convex sur- 

 face, a corresponding but smaller aperture on the flat surface, and 

 one (still smaller) at the free extremity. The last is surrounded by 

 projecting cladi of desmas. There is no channel, so far as can 

 be seen from a bare skeleton, between the two loops. 



Habitat. — Off the coast of Ganjam and Vizagapatam in 15 — 

 30 fathoms ; associated with. Siliquaria cochlearis , Morch. 



It is unfortunate that the data supplied with the specimens of 

 these two varieties are not sufficiently precise to enable me to state 

 whether their peculiarities are correlated with any difference in 

 habitat or environment. The colour of the two is of course strik- 

 ingly diff'erent, while the absence or paucity of rhabdi in the 

 var. siliquariae enables sections or other preparations of this form 

 to be distinguished at a glance from those of the var. spiroglyphi. 

 The former variety, therefore, seems to agree as regards spicula- 

 tion (except in the absence of discoidal phyllotriaenes) with the 

 typical form of the species, but to differ both from it and from 

 the var. spiroglyphi in colour. It differs from both in its more 

 massive structure. Colour is perhaps a more important character 

 as regards the forms of R. sceptrellifera than it is in most 

 sponges, for in the case of all the specimens examined I have found 

 it to some extent persistent. In the var. siliquariae, however, 

 the purple patches are probably due to the sporulation of some 



