22 Kecords of the Indian Museum. [Vor Xals 
supposed generic character, in having the head of the tylostyle 
oval (instead of usually spherical) and with a group of minute 
vacuoles in its centre. 
In the collection of the Indian Museum there are two shells 
from the Andamans, one of a Tvidacna and one of a Maileus,! that 
contain the burrows of a Clionid which agrees well with Topsent’s 
description of 7. hancocct so far as the general structure and the 
colour are concerned. In the Tvidacna-shell the papillae of the 
sponge have been destroyed, but they are well preserved in that of 
the Malleus. In neither specimen have I been able to find a single 
nodular amphiaster, although there has been no difficulty in re- 
moving the papillae for microscopic examination from one of them. 
The slender amphiasters are abundant in both specimens, scattered 
in the galleries of the sponge, and the majority of the tylostyles in 
the galleries have spherical heads, but those in the papillae are 
variable in shape. In no single spicule can I detect a group of 
vacuoles in this part. 
The question naturally arises, Is Cliothosa a distinct genus or 
merely a phase of Thoosa? In considering this question the facts 
known in reference to other species of the family must be noted. 
In the first place, it is known that Cliona celata* may lose two 
-ypes of spicules in the course of its latter development and that 
Thoosa armata*® does the same at an earlier stage. Secondly, we 
know that the nodular amphiasters are sometimes scarce in 
T. hancocet itself and, apparently, may be either confined to the 
papillae* or scattered throughout the sponge.’ Thirdly, in the 
type-specimens of T. investigatoris (antea, p.18) and T. laeviaster 
(p. 23, postea) these spicules were not found in the fully formed 
papillae but in what were apparently papillae in the process of 
formation. Furthermore, in the case of the former species, they 
sometimes exhibited distinct traces of wear in that position. All 
these facts seem to me to point to the possibility of there being a 
stage, perhaps but seldom attained, in the life-cycle of Thoosa at 
which the characteristic spicules of the genus disappear and the 
sponge gains nominal generic distinction undex the title Cliothosa. 
If Iam right, there can, I think, be no doubt that at least one of 
my specimens from the Andamans has reached this stage. 
Thoosa hancocci was originally described from a Tridacna-shell 
from the Indian Ocean. It is apparently common in coral from 
shallow water in the neighbourhood of Java and was taken by 
Prof. Stanley Gardiner, also in coral, in the Maldives (fide Topsent, 
1905, Pp. 94). 
Thoosa laeviaster, sp. nov. 
Spicules and fragments of the sponge of this species were 
found in the piece of dead coral referred to by Carter, whose 
1 One valve of the individual in the other valve of which Thoosa armata was 
found intermingled with Cliona vastifica. 
2 Topsent, 1900, p. 42, etc. 
8 Topsent, 1904, p. 111: see synonomy of 7. armata, p. 21. 
4 Topsent, 1905, p. 94. 5 Lindgren, 1898, p. 321. 
