Craspedacusta sowerbyi and Calpasoma dactyloptera — MATTHEWS 
255 
Fig. 11. Atentacular hydranth budding and frustule locomotion. A, Atentacular hydranth; B, frustule, hori- 
2 ontal (and dotted, vertical) at end of mucus tube; C, position of frustule at end of mucus carpet ( B-C ); and 
D, frustule attachment point at end of mucus carpet ( C—D ) . 
position with its smaller end pointing away 
from C. Next morning at 8:10 AM it had ar- 
rived at point D, about 850 /x from C and only 
about 175 g from B. The thicker end was al- 
ready attached, but the thinner end was elevated 
from the bottom of the Petri dish at about a 
45° angle. It was no longer bean-shaped and 
was slowly metamorphosing into a young aten- 
tacular hydranth. It was now only 375 /x long 
and 125 /x wide at the thicker end; the thinner 
end still measured 75 /x wide. Next day, with 
the righting process completed, a capitulum de- 
veloped. Exceedingly small, unidentified proto- 
zoa could be seen in the disintegrating tube 
between A and B. Regions between B-C and 
C-D lacked a lumen and thus were carpet-like. 
Toluidin blue (1:10,000) stained both tube 
and carpet very lightly, but in two days neither 
was visible. Although this copious secretion was 
probably the result of Zn 65 irritation, it is in- 
teresting that the frustule seemed to be more 
affected than the hydranth. The frustule’s de- 
crease in size and its abrupt, almost opposite 
change in direction support this view. Also, its 
change of position from horizontal to vertical 
and vice versa seems more related to mucus 
changes than to changes of buoyancy. 
Frustules of untreated hydranths also remain 
attached for some time, but in no instance is 
their attachment so pronounced. In most in- 
stances, although the remaining connection 
could not be seen, its presence could be demon- 
strated by passing a small dissecting needle be- 
tween hydranth and frustule. 
During Part II of this study many stages were 
observed which at first sight strengthened the 
