ZOOLOGY OF THE FAR EAST. 
correlation with life in very shallow water, it must be useful in the circumstances in 
which the sponge was found at Port Weld. In my Malay specimens the processes are 
recumbent or semi-recumbent. 
The other sponge, for which I propose the name Amorphinopsis excavans var. 
robinsonii, is, more strictly speaking, an encrusting sponge. It has a very massive 
structure in spite of its thinness, but contains relatively broad water-channels run- 
ning parallel to the surface a short distance below the external membrane. The two 
other phases of the species already known differ considerably in their mode of life. 
The forma typica was found growing on, or rather in, rotten coral in the Andamans. 
It forms a very thin film on the surface of the coral and sends root-like processes down 
into the burrows of Clionidae, the spicules of which it sometimes incorporates within 
its own substance. The outgrowths on its external surface are very short and com- 
pact. The var. digitifera, on the other hand, was found growing on hard rock and 
had incorporated numerous shells and pebbles, which it had not dissolved or excavated. 
It formed a mass of short, pointed, somewhat compressed upright branches of rather 
irregular outline, joined together by means of a relatively thin crust. The longest 
branches were about lo cm. long by 4 cm. broad. The new variety is almost exactly 
intermediate between these two forms, consisting mainly of a crust about 5 mm. 
deep, but bearing numerous short upright processes not more than 3 mm. long. It 
has no basal root-like outgrowths. Like all the phases of the species it is able to 
close its oscula and pores very tightly. The large holes shown in the photograph of a 
fragment reproduced on PI. II of this volume are due to the burrows of a mollusc 
in the wood below the sponge, and do not open into the interior of the sponge, 
which merely grows round them. 
The two sponges on the landing stage at Port Weld have not, therefore, under- 
gone any special structural evolution in correlation with the particular dangers to 
which they are exposed, viz. those of partial disiccation and of muddy water. They 
possess structural peculiarities that identical or closely related sponges living in totally 
different circumstances also possess, but these peculiarities are of great use to them 
in their peculiar environment. Without peculiarities of some kind, indeed, they would 
hardly have been able to establish themselves in their present habitat. The useful 
structural features are not the same in the two sponges. In the Reniera the principles 
adopted are those of receiving water in large empty spaces and of giving a free passage 
to small particles of inorganic matter through patent channels. In the Amorphinopsis, 
on the other hand, the open spaces are more restricted, the whole structure more 
massive and the orifices capable of complete contraction. The two sponges thus 
afford a parallel to two other cases of a similar nature that I have discussed recently, 
namely that of Nudospongilla asper and Cortispongilla barroisi in the Lake of Tiberias ' 
and that of Tetilla dactyloidea var. lingua and T . limicola in muddy lagoons on the 
coasts of India and Ceylon.'' In both cases we find sponges living in muddy water 
and adopting divergent means of protection, in one species by decreased size of the aper- 
Jouni. As. Snc. Bengal (ii.s.). IX, p. 75 (1913). 
* Mem. Ind. Mus. V. p. 54 (1915)- 
