I915.] N. ANNANDALE: Notes on Freshwater Sponges. 173 
genus Euspongilla typified by S. crateriformis (Potts), and dis- 
tinguished from other members of the subgenus by the erect or 
semi-erect posture of all or most of their gemmule-spicules and 
by the fact that the terminations of these spicules are clearly 
specialized. The specialization may, however, take one or other 
of two directions, for the ends of the spicule may (as in S. 
cratertformis and the closely allied S. biseriata, Weltner') bear an 
imperfect horizontal rotule of large recurved spines, or they may 
(as in my own S. hemephydatia, in S. sansibarica, Weltner!-and 
apparently in Haswell’s imperfectly known S. botryoides) be in- 
flated, so that the spicule is technically tornote. The group is, 
therefore, of particular interest as representing the ancestral 
form, at any rate so far as the gemmule-spicule is concerned, 
of both Ephydatia and Pectispongilla. 
I give figures here, in both cases from the type-specimens, 
of S. hemephydatia (fig. 1) and S. sanstbarica (fig. 2). The 
tetraxon spicules that occur not uncommonly among the macros- 
cleres of the type-specimen of the former are of course abnormal, 
but they have some interest as possible examples of reversion of 
the type by no means uncommon in the Spongillidae. 
The general structure of the skeleton in Pectispongilla is 
identical in the different species and does not differ in any im- 
portant feature from that found as a rule in Euspongilla. In 
particular it agrees closely with that which can be readily demon- 
strated by the use of pyrogallic acid in S. hemephydatia, S. crateri- 
formis and S. sanstbarica. Weltner (op. cit., p. 127), indeed, 
states that the skeleton of the last species corresponds precisely 
with that of the Chalininae, in that the fibres are enclosed in a 
sheath of horny substance. That this substance is present in 
amount much greater than can be seen in unstained preparations 
or might be argued from the thinness of the fibres, is certainly a 
fact ; but its arrangement seems to me to be quite different from 
that I recently demonstrated in Lubomirskia*, for although it 
permeates the fibre, cementing together the component spicules 
and occupying the spaces between them, I can detect no external 
fibre-sheath. Where, as is often the case, it forms veil-like films 
at the nodes of the skeleton, it has the appearance of a perfectly 
homogeneous film. 
The geographical distribution of Pecsispongilla is peculiar. 
It is apparently the only genus of the Spongillinae that has so 
limited a range, for even Astevomeyenia,’ which is confined, so 
far as we know, to the southern part of the United States of 
America, considerably surpasses it in this respect. 
The two species of Spongilla most closely allied to Pectispon- 
gilla (S. hemephydatia and S. sansibarica) occur in the main area 
| Mitt. Naturh. Museum Hamburg XV, pp. 121, 127, pl—, figs. 1-5, 
13-17 (1897). 
2 Rec. Ind. Mus., X, p. 144, pl. ix, fig. 1a (1914). 
5 Annandale, Proc. U.S. Nat. Mus., XL, p. 593 (1911). 
