218
HILARY B. MOORE
variation fist rose with increasing size and then dropped. Dr. T. V.
Borkowski (personal communication) found the same pattern in a
species of Littorim. Dimon (1902) showed that in Nassa obsoleta, a population living under conditions of stress showed an increased coefficient
for one parameter but a decreased coefficient for another parameter.
Several of the aspects of tropical stress which we have discussed
offer the opportunity of examining changes in coefficient of variation
through an array of species over a latitudinal temperature gradient.
By comparing an array of different species rather than Werent
populations of the same species it seems probable, although admittedly
i
0
10
20
30
Temperature -
'
C
.34
FIO. 29. Relation to temperature of the ooeffioient of variation of the peroentage of the
intertidal zone oooupied by speoies.
this cannot be proved, that the effects of changing coefficients with
size may not apply, at least so far as general trends are concerned.
Figures 27-29 show the relation to temperature of the coefficient of
variation of growth after maturity, growth rate and in relation to
intertidal zonation respectively. In all three the lowest coefficient is at
mid-temperatures, with a marked rise towards polar and tropical seas.
In the case of intertidal zonation (Fig. 29) there is a drop again at the
highest temperatures corresponding to the increase in the percentage
of shore occupied at the highest temperatures (Fig. 28).
Determination of the coefficient for mollusc sizes was difficult.
Their size frequency distribution, as discussed earlier, was a strongly
skewed curve. Further, small sizes were inadequately represented.