50 Memoirs of the Indian Museum. [Voi<. V, 



several layers, and are held in this position by vertical spicules which 

 transfix them. Lichenoid coherent masses of gemmules, which can be 

 detached as a whole, are thus formed, instead of a single adherent layer as 

 in L. aquae-dulcioris . 



In general structure the two sponges resemble L. rugosus (Schmidt) ' , except that 

 they are much less vigorous in their growth and that their main exhalent channels 

 and the main component parts of their skeleton exhibit a greater tendency to be 

 horizontal. The two facts are probably correlated. The spicules of both species 

 differ from those of Schmidt's in having the heads practically always differentiated, 

 though often irregular. They are also more variable in size. In these respects 

 they come nearer ths spicules of L. ectyoninus, Topsent, from which they differ in 

 that by no means all of them are directed 'towards the periphery of the body ' 

 (Topsent, op. cit., p. 189, pi. vii, figs. 11, 12). As regards the form of the spicules 

 both species agree closely with the variable Australian L. proteus, Hentschel : , but 

 from all varieties of this sponge they are separated by the structure of their skele- 

 ton and the general smoothness of their surface. 



L. lacustris was always found on stones or rocks except in one instance in which it 

 was on a dead bamboo. On rocks it grows on vertical faces and on the lower surfaces of 

 overhanging projections ; on stones it occupies the lower side only, unless the stone 

 is protected by others above it. This seems to be not so much due to avoidance of 

 light as to the fact that its comparatively flat surface renders it liable to be com- 

 pletely smothered by the settling of silt if it spreads out in an exposed position. 

 If sponges of the species are placed alive and surface uppermost in an aquarium full 

 of lake-water they rapidly become covered with fine mud and débris, through which 

 their oscular collars project upwards. The convexity of the roof of the superficial 

 exhalent channels, combined doubtless with the steady movements of the water in 

 the canals, keeps the roofs free of extraneous matter for some time and the plan of 

 the canals is mapped out in a very conspicuous manner by clear hyaline streaks in 

 the general area of mud; but the dermal pores are soon choked, and the sponge dies. 



Larvae were found in April ready to be liberated in the canals of sponges which 

 also contained gemmules. Gemmules are produced at all times of the year but 

 particularly at the approach of the hot weather. At this season most of the rocks 

 on which the sponge flourishes are gradually exposed by the retreat of the water. 

 As it dries up it naturally dies. Sponges that suffer thus before producing gem- 

 mules, as is not infrequently the case, cling tightly as dried skeletons to the stone, 

 their horizontal fibres being pressed against their adherent basal membrane (pi. v, 

 fig. 2) ; but no fibres persist in the gemmule-masses and the basal membrane has 

 practically disappeared below them. When these masses are thoroughly dry, there- 

 fore, they begin to curl up round the edges owing to the unequal contraction of the 



1 Topsent, Arch. Zool. expérim (3) V, p. 185, pi. v, figs. 1-4. 



2 " Tetraxonida " in Michaelsen and Eïartmeyer's Faun. Sildw. Australiens II, pp. 389, 391 

 (figs. 20, 21) , 392 (6gs. 22, 23), pi. xxii, figs. 1-3 (1909). 



