1919'] N. Annandale : Bombay Streams Fauna. 159 



corpuscles of many freshwater sponges in structure but of a deep 

 purple colour, which is not soluble in spirit. 



The spinosit}' of the megascleres is so faint that it is apt to 

 escape notice altogether unless they are examined under a very 

 high power of the microscope. The tips are smooth. 



The regularity of the skeleton is due mainly to the compact 

 formation and large number of the primary radiating or vertical 

 spicule-fibres, which run upwards through the sponge for consider- 

 able distances without branching and are joined together by much 

 less regular cross-fibres or by groups of spicules. The vertical 

 fibres contain a considerable amount of binding substance. 



The gemmules are small and very numerous in all the speci- 

 mens I have examined. They have a thick pneumatic layer in 

 which the air-spaces are very minute, and a slender, straight, pro- 

 jecting foraminal tubule. 



In February. 1918 Dr. Gravely and I found specimens at two 

 localities in the Singbhum district of Chota Nagpur, in a rocky 

 stream close to Chakradharpur and in a pool of muddy water, 

 evidently part of a sluggish stream in the rains, on the road be- 

 tween that place and Chaibassa. 



At Chakradharpur the sponge was growing on the rocky bed 

 of the stream in clear running water. It was essentially of 

 encrusting habit but in little pockets in the rock showed a ten- 

 dency to adopt a cushion-like form and was then a centimetre or 

 more thick. The oscula were small on the flat rock but in the 

 pockets become larger (about 5 mm. in diameter). In all cases 

 they opened into wide vertical exhalent canals and horizontal 

 subdermal exhalent canals were absent. The sponge was very 

 soft and of a deep purple-brown colour. Our specimens from a 

 pool on the Chaibassa road were attached to the lower surface of 

 bricks at the base of the piers of a bridge. They had a cushion- 

 like form and were harder and blacker than those from the stream. 

 Their oscula were small and branching horizontal exhalent chan- 

 nels were conspicuous on the surface of the parenchyma. 



Spongilla (Euspongilla) perviridis, sp. nov. 

 (Plate VI, figs. 2, 2a, zb). 



1912. Spo)igilla cinerea, Annandale (nee Carton, jRec. Imi. Mits. \'II, 

 PP- 137. 387- 



This sponge, though closely related to S. cinerea, can always 

 be distinguished by the following characters : — 



(i) Purple corpuscles are absent from both parenchyma 

 and dermal membrane and are replaced in the former 

 only by green corpuscles. 



(2) The skeleton-spicules are more distincth' spiny, but also 



have smooth extremities. 



(3) The skeleton is much less regular. 



(4; The oscula are always small and surrounded b}" radiat- 

 ing exhalent channels. 



