946 THE AMERICAN NATURALIST.  [Vou. XXXIII. 
of the pieces closed in and the old tentacles formed a ring 
encircling the margin. The piece assumed, in a general way, 
the form of the medusa (Fig. 9), and from the middle of the 
lower side a proboscis-like outgrowth developed (stippled in the 
figure). In the center of the bell two 
cavities are present. In order to inter- 
pret these structures it was necessary to N 
make serial sections. The sections show W NVA 
that the larger cavity (stippled in the Ay 
figure) represents the subumbrella space, 
and is therefore lined by ectoderm. 
This ectoderm continues down into the 
proboscis-like structure of the lower side. 
The outer surface of this structure is covered by the outer 
ectoderm. It represents, therefore, not a new manubrium, but 
a tubular outgrowth of the velum, the latter having completely 
closed over the lower surface. The outgrowth is fringed along 
one side and might easily be taken for a new proboscis, but 
sections show, in the clearest way, that it does not represent 
that organ. The other cavity seen in the figure lies on the 
side where the piece closed in. It represents an enlargement 
of the ring canal and is continu- 
ous with the ring canal around 
the base of the tentacles. The 
results show that the small 
piece, while assuming the form 
of a bell, is completely lacking 
in the essential organs of the 
medusa. 
Fic. 9. 
In this same experiment the 
cross-shaped piece that re- 
mained after the four triangular 
pieces had been removed was it- 
self cut into four pieces through 
the stomach (Fig. 8, B). These small pieces sometimes devel- 
oped into medusa-like forms, each with a manubrium, but did 
not reproduce the other missing parts (Figs. 10 and 11). 
Quite a number of experiments were made at different times 
Fig. 10, Fic. 11. 
