Lnfluence of Selection 201 
the curve is given by the individuals themselves, but in practice 
this method is seldom employed. The measurements are 
used directly to construct the curve. 
The causes of the differences in individuals are to be found 
partly in the various external conditions under which the indi- 
viduals have developed, and partly in initial differences at the 
beginning of development in the individuals themselves. How 
far these latter may be traced to still earlier effects of external 
conditions, and how far to the possibilities of combinations of 
inherited differences within the organism itself is difficult to say. 
It is clear that a knowledge of the curve of variability does not 
contribute, in any large measure, to an elucidation of the causes 
that bring about the fluctuations. These causes must be studied 
by examining the effects of the external factors that influence the 
organism. 
If the curve of variability is high and does not extend out far 
along the base line, this indicates that the form is little variable. 
If the curve is low and extends far out at the sides, the vari- 
ability is large in amount. One side of the curve is sometimes dif- 
ferent from the other. Such a curve is asymmetrical (Fig. 20, C). 
Thus the curve may rise abruptly on one side and rise very slowly 
on the other side. The meaning of this is obscure. Equally 
important are the cases in which the curves have more than one 
summit. As an example, I may cite the case of the rhinoceros 
beetle (Fig. 20, D). -Its curve hastwosummits. Suchacurve is 
said to be bimodal. This means that there are many long- 
horned beetles and many short-horned ones, but few of the in- 
termediate size. Here we see variations about two means, as 
though the group were about to split into two groups; but 
whether anything of the sort is really occurring is entirely un- 
known. It has been observed in certain plants that the secon- 
dary summits coincide with the points on which in allied species 
the main summit lies, and this appears to give us a suggestion 
concerning the factors producing the change. If, for example, 
we look upon the characters seen in the allied species as dormant 
in the one with two modes, but which may, under certain 
