248 DK. N. ANNANDALE ON FRESHWATEK 



a little light reached them between the stones. The spicules and gemmules, 

 as perhaps was natural, were smaller than usual ; the gemmules measuring 

 about 0*334 mm. in diameter, and the skeleton-spicules being 0*279 mm. 

 long and proportionately rather slender. Distinct spicule-fibres projected 

 vertically from the surface. 



On the whole, this form of S. lacustris from Mull is nearly as worthy 

 of specific rank as my S. reticulata from Eastern Bengal, although its 

 peculiarities lie in a different direction, the Indian form * being distinguished 

 by the great development of its branches, which are laterally compressed 

 and anastomose to form a reticulated structure. The Mull sponge, however, 

 is linked to the typical S. lacustris by one found in America and named by 

 Bowerbank f S, lacustris var. abortiva, which has few spicules on the gemmule 

 and an apparently aspiculous dermal membrane. This form, however, 

 is described by Potts X as " coating and ])ranching," and has the skeleton- 

 spicules rather stout ; there is no crust on the gemmules, a character in 

 which it agrees with the Hebridean form. I do not propose, for reasons 

 stated above, to give the latter a name, but it is a form of considerable 

 interest, possibly a distinct local race. -5'. jyroliferens from Betigal resembles 

 it in its small size and has no true branches, but is remarkable for its 

 prolific reproduction by means of external buds and has a tubular outgrowth 

 attached to the foramen of the gemmule. 



There is one other point to which it may be interesting to refer before 

 leaving Spongilla lacustris (on which there is already far more literature than 

 on any other species), namely, its occasional association with a Phylactolse- 

 matous Polyzoon. Growing in the substance of the specimens given me 

 by Mr. Evans I found a Plumatella identical with AUman's P. coralloides, 

 which was originally found in similar circumstances. This Polyzoon is 

 a temporary phase of the same author^s P. fruticosa, not of Linnets 

 P. repens as I formerly § thought. The same phase is common in S. carteri 

 in Calcutta and Bombay, and occurs occasionally in the former locality 

 in at least one other species of sponge, viz., S. crassissima, the hardness 

 of which, however, is evidently less favourable to its groM^th. 



TUBELLA PENNSYLVANICA, Potts. 

 Tubella pennsylvanica, Potts, Proc. Acad. Nat. Sci. Philadelphia, 1887, p. 14 ; 

 Hanitsch, Irish Naturalist, iv. p. 129, 1895. 

 This species was originally described from several localities in the 

 United States, and was later recorded by Hanitsch from the West of 



* ' Records of the Indian Museum,' i. p. 387. f Proc. Zool. Soc. 1863, p. 470. 



X Potts, Proc. Acad. Nat. Sci. Philadelphia, 1887, p. 189. 



§ Journ. Asiat. Soc. Bengal, 1907, p. 88. Allman says that the statoblasts are broad, 

 but figures them as somewhat elongated (Monograph of the Freshwater Polyzoa, p. 103, 

 pi. vii. fig. 4). 



