28 GEORGE BIDDER. 
nected with nutrition, as I was formerly (18) in my hypothesis 
as to changes of cell-form accompanying digestion. 
Of my own experiments I printed shortly the main results 
in February, 1888. Omitting the passage (quoted in 19) on 
Sollas’s membrane, I reprint the statement.! 
“Tn Leuconia aspera I find that carmine granules are 
taken in freely by the collared cells, not appearing in the 
mesoderm, and only infinitesimally in the other epithelia. .. . 
“T observed that during four hours a Leuconia plentifully 
supplied with carmine ejected none in its oscular stream, which 
was powerful and continuous. Its flagellate cells proved to be 
heavily charged with carmine grains. Such complete filtration 
would be uneconomical, if not impossible, were the carmine 
arrested merely by the ingestion of cells laterally bounding 
the current. 
“‘T believe, from a consideration of the observations of 
others and the above facts, that the collared cells primitively 
both ingest and digest for the sponge; the function of diges- 
tion being in some sponges, but not in Leuconia, passed to 
cells situated in the mesoderm. I think that probably only 
under exceptional necessities of structure do other cells of a 
sponge ingest food in valuable quantity. 
“My experiments were suggested by a recognition of the 
fact that in the current through a sponge the region of slowest 
motion, and therefore of greatest deposit and easiest arrest, is 
in ‘the flagellate chambers, where the transverse area of the 
total channel for the water is greatest. This fact also explains 
1 Extracted from the ‘ Proceedings of the Cambridge Philosophical Society,” 
vol. vi, pt. iv, ‘Preliminary Note on the Physiology of Sponges.” Fifty 
copies only were printed in full through a mistake owing to change of 
editorship by which an abstract of ten lines was substituted in the ‘Proceedings * 
as issued; for this reason I print a “ Preliminary Note ” of work still, alas! 
unfinished. I hope soon to publish a discussion of the mechanical conditions 
here referred to. The lamellar forms of sponges are naturally independent of 
oscular velocity, since the stream of foul water is 180° from the stream of 
fresh water. It is the increase of this angle which leads to the number of 
stalked forms, from which are usually evolved the flabellar species and 
varieties. 
