DENDY— NOZSr-CALCAREOUS SPONGES 137 



and Quasillina as at present understood, the soft, pulpy internal structure and thin 

 cortex being suggestive of the latter, and the hollow fistular processes of the former. 

 It seems quite possible that, after all, the distinction between these two genera may 

 have to be abandoned/ 



The process of bud-formation by gemmation from the filiform ends of tapering 

 fistulse appears to be identical with that described, with admirable illustrations, by 

 Merejkowsky [1878] in Polymastia (Rinalda) arctica, a species from the White Sea. 

 A somewhat similar process is, of course, familiar in certain other genera, such as 

 Donatia. 



In Polymastia arctica there are one or more osculum-bearing fistulse between the 

 gemmiparous ones. This may also be the case with the present species, but the 

 specimen is not in a fit condition to enable me to determine the point with certainty. 



Register Number, Locality, dec. IV. 7, dredged off S.W. coast of Beyt Island. 



53. Megalopastas retiaria n. sp.— (Plate IV., Fig. 27.) 



This is certainly one of the most interesting, and at the same time one of the most 



beautiful species in the collection, in which it is well represented by three good spirit 



specimens and a couple of washed-out skeletons. The general appearance of the sponge 



is well shown in Fig. 27, which represents the best preserved specimen twice the natural 



size. The sponge is irregularly lobose, or simply massive. Apparently it is sessile, 



with several points of attachment to relatively small foreign objects, as though it had 



grown on a gravelly or shelly bottom. The surface is produced into moderate-sized, 



acute conuli, scattered at rather wide and irregular intervals, the height of the conuli 



being about 1-2 mm. Internally the sponge is cavernous and the wide vestibules 



open on the surface by large pseudoscula. These pseudoscula, of various shapes and sizes, 



are frequently (perhaps always in life) covered over by a very beautiful network, which 



is a continuation of the delicately reticulate dermal membrane, but with much coarser 



meshes. The dermal membrane in general covers the whole sponge like a gauzy veil. 



It is strengthened by a close reticulation of delicate, deeply staining, fibrillar bands (not 



horny). Under a pocket-lens the principal lines of this network appear (in the spirit 



specimens) as fine white lines radiating from the apices of the conuli into the hollows 



between, where they lose themselves in a network of finer fines. The meshes of this 



network are occupied by the pore-bearing dermal membrane, which is itself reduced 



to a secondary, quite microscopic network by the very numerous inhalant pores, each 



only about 0-03 mm. in diameter. The dermal membrane is interrupted here and 



there, pretty frequently, by small circular apertures about 1 mm. in diameter. These 



apertures have well-defined margins, being bounded each by an unusually large, circular 



1 Cf. my remarks on Quasillina hrevis in Journ. Linn. Soc. Zool., Vol. XXXII. (1914), p. 271. 



