RESTITUTION-BODIES AND FREE TISSUE-CULTURE IN SYCON 308 
the external collar-cells were functional. Since, however, the 
total number of dermal cells in a mass is proportional to its 
volume, while the number required to form a single external 
layer of epithelium is proportional to its surface, there will be 
in large masses an excess of dermal cells above those needed 
to form the epithelium. This excess apparently forms the 
dermal blow-outs. The replacement of choanocytes by dermal 
covering is of interest in view of the greater viability or protec- 
tive capacity of the dermal cells shown by other considerations 
(Huxley, 8). 
Fig. 13 shows another type. A number of very large masses 
had formed in a culture produced by squeezing without gauze. 
The larger masses had first been very irregular in shape, and 
had demonstrably been formed by the coherence of original 
smaller spheroids. (The culture was made in a finger-bowl. 
The flat bottom was covered with small spheroids, while 
a ring of the large irregular masses was found at the foot of 
the sides. This was due to the opportunity given here for 
many masses to come in contact by rolling down the steep 
sides. ) 
These irregular large masses later rounded up, and shortly 
after this produced blow-outs. Some were similar to that 
seen in fig. 12. Others, however, consisted of a much- 
distended sphere surrounded by an epithelium of dermal cells, 
the contaimed gastral cells not forming a well-marked mass, 
but spread in layers of varying thickness over part of the inner 
surface of the sphere (fig. 13). The majority of the larger 
masses in the culture were of this type, while the majority of 
the smallest were not blown out at all, but were normal 
regenerates. This bears out the conclusion drawn above as 
to the réle of size of mass. 
Wilson, in his experiments with Coelenterates, also found 
that size of mass was very important, the larger masses failing 
to metamorphose. A study of restitution-bodies from this 
point of view would probably throw light upon the reasons 
for the sizes of the larvae in many low types. 
An interesting feature of most dermal blow-outs examined 
