302, JULIAN S. HUXLEY 
cells present in a particular mass. That this is not all, however, 
is Shown by the following experiment (Wood’s Hole). 
July 12. Several sponges squeezed through bolting-silk 
into a finger-bowl. ‘Three dilutions of the resulting cell- 
suspension were made: (1) dense, (2) medium, (3) dilute. 
Results :—After one day: (1) large, sometimes irregular, 
masses ; (2) medium-sized spheroids, many with good collars 
and flagella, many with larvae embedded in them; (3) as (2), 
but smaller, and fewer with larvae. 
After two days: (1) no blow-outs ; most seem normal restitu- 
tion masses; (2) most with collar-cell blow-outs ; (8) some 
with collar-cell blow-outs. 
After three days: (1) as before; (2) fewer collar-cells than 
the previous day ; (3) none seen blown out. 
After five days: (1) the smaller masses forming small dermal 
blow-outs ; (2) many with large dermal blow-outs ; (8) solid. 
After nine days: (1) mostly dead ; (2) many attached to 
glass ; (3) as before, none attached. 
A repetition of the experiment gave similar results, except 
that dermal as well as choanocyte blow-outs were formed 
early in the middle dilution. 
It will thus be seen that dermal blow-outs did not begin 
to appear until the fifth (or fourth) day, and that they appeared 
most notably in the same culture which had previously pro- 
duced the best choanocyte blow-outs. Their failure to appear 
in the large masses of (1) may be due to the fact that these 
in this experiment were not very healthy. It would appear, 
since the only difference between (2) and (3) lay in the size of 
the masses formed, that the eventual production of dermal 
blow-outs is determined partly by the total, and not only by 
the relative number of dermal cells present. It appears that 
first of all the collar-cells on the surface protrude collars and 
flagella towards the water ; these are, however, very susceptible 
to noxious influences, and as culture conditions became less 
good they retracted into a spheroidal form. The dermal cells 
then migrated to the surface and covered the collar-cells with 
an epithelium, which they were apparently unable to do when 
