EVOLUTION 



355 



showing a gradual transition from one extreme to the other. 

 When, however, the numbers of the individuals with forceps 

 of the varying lengths were plotted, a graph was obtained 

 demonstrating that the whole population was grouped 

 around two modal values of respectively 3 '5 mm. and 

 7 mm. in length of forceps. The insects, except for the 

 exceedingly small number that showed a forceps length 

 of 4*5 mm., were in fact separable into a '' low " and a 

 " high " group. 



The examples that have been given may suffice to show 

 that discrimination between continuous and discontinuous 

 variation must often be difficult. Since the work of Bateson 

 there has been an increasing tendency among students of 

 the subject to lay stress on the importance of discontinuous 

 variations or mutants as the starting-point for further 

 evolutionary progress in the origin of a new species. It 

 will be remembered that many of the observed mutants 

 are distinguished by characters whose factors are segregated 

 in inheritance, and therefore remain stable ; and it is 

 stabiUty, the tendency to breed true, rather than an extensive 

 divergence from the type, that distinguishes a mutant from 

 a mere fluctuation. Hence it is usually necessary, in order 

 to discriminate between types of variation, to carry out 

 observations and experiments in breeding. Even when a 

 continuously graded series of varieties within a species can 

 be demonstrated, it does not follow that such a series re- 

 presents successive steps in the variation. The fruit-flies 

 (Drosophila), so well known now as subjects for the study 

 of heredity, have normally winged individuals and others 

 with the wings more or less reduced so that they may become 

 degenerated to the merest vestiges. These last arise as the 

 offspring of normal winged parents, not as the final result 

 of a series of generations showing a slowly progressive 

 degeneration of the wings. The modern study of variation 

 has at least demonstrated that in a single generation a 

 marked and definite change may in many cases be observed. 



A suggestive example of some of the questions concerned 

 with variation is afforded by the small European butterfly 



