294 JULIAN 8. HUXLEY 
heterocoelous calcareous sponges than in any other types 
investigated. I further showed that, by certain methods, 
restitution-bodies composed entirely or almost entirely of 
collar-cells could be produced, and that these assumed a form 
quite unlike anything found in normal sponges, but with 
a resemblance to a Choanoflagellate colony. Simple excess 
of collar-cells, or, apparently, larger masses composed 
almost entirely of collar-cells, led to the formation of what 
T called choanocyte blow-outs—a part of the solid mass becom- 
ing blown out to form the segment of a collar-cell sphere. 
Since then I have continued making observations on the 
subject as opportunity offered. Although these cannot pretend 
to completeness, they have brought certain new facts to light, 
which I publish in the hope that other workers may extend 
them by observation on the same favourable material. Some 
of the work was done at Wood’s Hole, Massachusetts, and some 
at the M.B.A. Laboratory, Plymouth. I have to thank the 
authorities at both institutions for their help in getting material 
and in other ways. I have also to acknowledge much efficient 
help at Wood’s Hole from Mr. I. J. Davies, laboratory assistant 
in the Rice Institute, Houston, Texas. 
2. MATERIAL AND METHODS. 
A species of Sycon was used at both places. That used at 
Plymouth was 8S, coronatum, obtained from piles in the 
Millbay docks. Orton (12) has recently drawn attention to 
the fact that this sponge grows actively during the winter 
without reproducing ; but during the summer it reproduces 
so long as the temperature is above a certain level, and scarcely 
crows at all, The same is to be presumed true of other species 
of the same genus. If so, it follows that the best time for 
conducting similar experiments will be during the cooler half 
of the year. 
Experiments were tried on homocoelous sponges such as 
Clathrina and Leucosolenia, but without much success. Restitu- 
tion masses are formed, but are small and do not live well. 
The collars and flagella are withdrawn on very slight provoca- 
