376 



VARIATION 



(Fig. 220). The larger the number of individuals taken into 

 account, the smoother the outline of the curve. Such variation 

 curves are most commonly symmetrical (Fig. 220, left), but they 

 may be one-sided or ^.symmetrical (Fig. 220, right), as in the case 

 of the meristic variation of the corolla-segments of many flowers, 

 where there are often relatively few examples with less than the 

 normal number of parts. 



The normal symmetrical curve of variation agrees very 

 closely with that representing variation depending on pure 

 chance, a feature which in itself suggests that the manifold 



differences in the conditions of 

 if the environment are involved. 

 The classical example of such 

 chance variation is afforded by 

 the repeated tossing up of two 

 coins, the combination of one 

 head and one tail being most 

 frequent (forming about 50 per 

 cent, of the cases), whilst the 

 combinations two heads or two 

 tails occur with about equal in- 

 frequency. A more illustrative 

 curve of chance variation would 

 be obtained if ten coins were 

 tossed simultaneously for a large 

 number of times in succession. 

 In cases of meristic variation 





mm 



FIG. 219. Seeds of different varie- 

 ties of Castor Oil (Ricinus), to 

 illustrate variation within a 

 species. [Photo. E. J. S.] 



it will be noticed that the differ- 

 ence between the extremes is much greater when the mode 

 corresponds to a large number than when it coincides with a 

 small one. Thus, if variation due to fission of parts (say of the 

 corolla of a flower) be equal in two species, the one having a 

 mode of five (i.e. usually five petals) and the other having a 

 mode of ten (i.e. usually ten petals), there would be just double 

 the chances of fission occurring in the latter as in the former. 

 The extreme condition (i.e. 10 and 20 petals) would be attained 

 where all the petals underwent fission. That is, the range would 

 be greater in the one case than in the other, although the actual 

 frequency of fission of the individual segments was the same 



