4 Records of the Indian Museum. [VoL. XI, 
This grouping is convenient for the purposes of a provisional 
classification, which is all that is possible until the life-histories 
of the species are known; but it must be remembered that in at 
least one species (Cliona celata, Grant) phases occur in the life of 
an individual sponge that would fall respectively into groups I, II 
and V. The sponge in its younger stage possesses tylostyles, 
diactinial spicules and microscleres, but as it grows it loses first 
the diactinial spicules and then, sometimes, the microscleres, so that 
in its mature form it has only tylostyles. It is possible, and indeed 
probable, that other species resemble it in this respect, so that 
groups V and VI may actually consist of species whose earlier 
stages are unknown and if known would fall into other groups, or 
even in some cases of species known by other names and assigned 
to other groups at different phases of development. 
Taking the groups as they stand, we find that among the 
species known from Indian seas all but group VI are represented. 
Group V, so far as hitherto described species are concerned, has 
not withstood recent criticism and research', but a new species 
belonging to it is described here on p. 14. In the following key 
to the species found in the Indian Ocean (including the Red 
Sea, the Bay of Bengal with its appurtenances and the western 
part of the Malay Archipelago) I have found it more convenient 
to make the primary division between species that possess and 
those that do not possess microscleres. Even so, it is necessary 
to include C. celata under three separate headings in accordance ~ 
with its three phases of development. 
Of the sixteen species now known from the Indian Ocean at 
least twelve have been found in the Bay of Bengal or the Gulf of 
Manaar. Of these, four are of very wide distribution (C. celata, 
C. vastifica, C. carpentert, C. viridis): C. carpenter is essentially 
a circumtropical sponge, but the other three are cosmopolitan. 
Three species have a wide distribution in the Indo-Pacific Region, 
namely C. margaritiferae, C. mucronata and C. orientalis; while 
five (C. annulifera, C. indica, C. enstfera, C. acustella and C. 
warrent) have been definitely recorded only from the Bay of 
Bengal and Ceylon. Of the four species not known from these 
seas, two were originally described, or are only known definitely, 
from the “‘Indian Ocean,” namely C. michelini and C. mille- 
punctata, but the original specimen of the latter was doubtfully 
ascribed to the N. Atlantic. One species (C. mussae) has been 
found only in the Red Sea, and one (C. patera) in the western 
part of the Malay Archipelago. 
I have not included C. gracilis, Hancock, among the species 
known from the Indian Ocean, although Topsent (1887, p. 77) has 
done so; because the latter author’s brief description of his speci- 
men from that area (‘‘Spicules en épingle—=150» de long, spic. en 
zigzag—I5-20u”’) is totally at variance with Hancock’s original 
diagnosis, which is supported by good figures, and some mistake 
1 See Topsent, Arch. Zool. expérim. (3) VIII, p. 78 (1900). 
