No. 396.] HYDROMEDUSA, GONIONEMUS VERTENS. 945 
days they did not assume the typical medusa-form. In another 
case the distal ring was cut into eleven pieces. Four of the 
pieces contained the distal end of a radial canal; the rest 
did not contain any part of the radial canals. About half of 
the pieces showed later little 
resemblance to the medusa- 
form. They became swollen @-.-._ 
and irregular, and although 
kept for twenty days did 
not develop further. Other 
pieces became somewhat 
bell-shaped, but in none of 
them was the medusa-form 
well developed. These 
pieces seemed, therefore, to 
be near the lower limit of 
size necessary for the formation of even the medusa-form, 
although it is possible that even smaller pieces of a different 
shape might produce the characteristic form. 
The following experiment was made to see if, in the absence 
of a definite structure such as 
the radial canal, a piece might 
regenerate a new radial system 
having the typical form. A 
jellyfish was placed oral side 
down on a glass plate, and 
with a knife, or with scissors, 
an inter-radial piece triangular 
in outline was cut out (Fig. 
8, A). A piece of this sort 
does not contain any part of 
the central stomach or of the 
radial canals. It contains a 
part of the endoderm in the form of a plate extending through- 
out the middle of the piece, and a ring canal around the mar- 
gin. In all, forty-eight pieces of this kind were cut out. Most! 
Fic. 7. 
Fig. 8. 
1 Some of them died, especially in one experiment, due to too many being 
kept in the same dish. 
