^88 N. Annandalk : Xotiw on Freslnvafer Spoil (res. [VOL. I, 



tremely fragile, spongin being present in exceedingly small 

 quantities. Skeleton spicules smooth, moderately stout, com- 

 paratively large, ampioxous, gradually pointed ; flesh spicules 

 numerous both in the dermal membrane and in the paren- 

 chyma, slender, abruptly pointed or blunt, curved in a wide 

 ' arc or nearl}^ straight, covered irregularly with relatively large 

 spines, which tend, especially towards the ends of the spicule, 

 to be bent backwards and inwards ; gemmule spicules closely 

 similar but stouter. Gemmules large, spherical, 3^ellow, 

 abundant, both in the basal layer and in the branches, 

 covered with a thick la3^er of granular substance, which is 

 confined externally by a definite chitinous coat ; the gemmule 

 spicules arranged horizontally in the latter and tangentially 

 on the former ; the single aperture infundibular, not provided 

 with a chitinous tube. 



This vSponge is closely related to the very variable species 

 Spongilla alba, Carter, from which it may be distinguished by its 

 external form, b}^ the presence of green bodies in the cells of its 

 parenchyma, and by its soft consistency and fragile skeleton. 



Spongilla alba ' is, again, very closel}' allied to S. lacustris,^ of 

 which S. reticulata may be no more than a specialized race. An 

 examination of a considerable number of specimens from different 

 parts of Bengal convinces me that the only constant differences 

 between S. alba and S. lacustris are the following : — 



Spongilla alba. Spongilla lacustris. 



Branches frequently absent, Branches rarely absent, when 



when present, laterally com- present, cylindrical. Colour, in 



pressed. Colour even in a a bright light, leaf-green owing 



bright hght, white or grey, oc to the presence of chlorophyl 



casionally dark green owing to corpuscles in cells of the paren- 



the presence in the tissues of chyma. 

 extracellular algse. 



The skeleton is also stouter in S. alba than in 5. lacustris, and 

 this is perhaps the most important difference. 



Differences in external form and in colour are by no means 

 satisfactory foundations for the creating of species in the Spong- 

 illinse as a rule. The latter is liable to change from a variety of 

 causes, e.g., leaden-grey examples of Ephydatia indica become 

 white if kept alive in an aquarium, and it is well known that the 

 chlorophyl corpuscles, which probably start life as independent 

 organisms, become colourless if kept in the dark or even in a 

 dull light. As regards the presence of such bodies in 5. lacustris, 

 however, and their absence from 5. alba, it is not sufficient to 

 suppose that the free-living organism does not occur in the 



1 Petr differentiates between the two forms (in Bohemian) in Abh. Bohmisch 

 Ges., viii, p. 27, pi. i. Unfortunately I am unable to read what he says. His 

 figures of the gemmules are clear, if somewhat diagrammatic, but do not, of 

 course, illustrate their range of variation. (Lately I have found the typical 

 5. lacustris in W. India. Dec, 1907.) 



