Craspedacusta sowerbyi and Calpasoma dactyloptera — Matthews 
253 
mixed cultures would contain fragments of 
atentacular frustules and (as will be shown di- 
rectly) small frustules of tentacular hydranths. 
In isolated cultures of atentacular hydranths, 
where frustules are allowed to divide before 
samples are taken, Figure 9 B would approxi- 
mate A. Since my selected B frustule so nearly 
approximates Buchert’s E, one might assume 
that Figure 9 E frustules were also selected be- 
fore any of them divided (Fig. 13 A-C). 
Measurements in microns of the first frustules 
observed in each of 20 isolated, tentacular cul- 
tures showed these variations: lengths 225-75, 
mean 150; widths, thicker end 75-50, mean 55; 
widths, thinner end 50^37, mean 45. Figure 9 F 
shows the size of this frustule compared with 
Fig. 9- Schematized frustules drawn to the same 
scale. A, Mean length and widths (thicker end, thin- 
ner end) of the first frustule observed in each of 20 
mixed cultures; B, mean length and widths of first 
frustules observed from isolated atentacular hydranths 
before any division of frustules occurred; C, small 
fragment of a frustule; D, large fragment of a frus- 
tule; E, Buchert’s atentacular frustule; F, tentacular 
frustule; and G, Buchert’s tentacular frustule. ( Ques- 
tion marks suggest possible relationships between C—F 
and D-G .) All measurements are in microns. 
sizes of other frustules. Buchert (1960:47) lists 
the following variations for his Type B (ten- 
tacular) frustules (measurements in microns): 
lengths 442-172, mean 303; widths, thicker end 
78-48, mean 62; widths, thinner end 58-38, 
mean 48. Figure 9 G shows the size of this frus- 
tule compared with sizes of other frustules. 
Even without the application of statistical 
analyses the similarity of small fragment Figure 
9 C with frustule Figure 9 F is striking, and sug- 
gests quite convincingly how one might be mis- 
taken for the other. This is undoubtedly what I 
had done. I have no explanation why Figure 
9 G (Buchert’s tentacular frustule) more nearly 
approximates Figure 9 D (except for thicker 
width of D ) than it does Figure 9 F. 
This much is now known: In isolated aten- 
tacular cultures, not only do large frustules (Fig. 
9 B) metamorphose directly into atentacular hy- 
dranths, but, more important, so do their frag- 
ments (Fig. 9 D, C). Furthermore, in isolated 
tentacular cultures frustules metamorphose only 
into tentacular hydranths. 
Change in frustule shape is considered here 
for two reasons: ( 1 ) its relation, if any, to frus- 
tule fragmentation, and (2) its relation to lo- 
comotion. Frustules of both atentacular and 
tentacular hydranths exhibit peculiar waves of 
contractility which slowly pass from one end of 
the frustule to the other. Figure 10 is a diagram- 
matic representation of these waves as observed 
in a 525-/* frustule of an atentacular hydranth. 
(In the figure, numbers correspond to ocular 
micrometer spaces of 25.) When first observed 
(Fig. 10 A), this frustule appeared as if about to 
divide into a fragment 150 /* long, and a frag- 
ment 375 /* long; however, 15 minutes later 
the region of contractility had moved 100 /* 
(Fig. 10B). In another 5 minutes the region of 
contractility had moved an additional 75 /* (C) ; 
5 minutes later another 75 /* brought the area 
of contractility to within 125 /* of the wider 
end (D). Then, 9 minutes later, the wave of 
contractility had completely traversed the frus- 
tule ( E ) — in the observed elapsed time of 34 
minutes. If one adds another 10 minutes for the 
wave to reach the point where first observed 
(A), 45 minutes would be a conservative esti- 
mate of the time required for a wave to traverse 
the entire length of the frustule. Rarely, how- 
ever, is frustule change of shape so simple. 
Figure 10F, G, and H clearly show that more 
than one wave may be operative and, although 
