NATURAL SELECTION 213 



hand, it is of the greatest importance that, in all the cases in 

 which selective elimination appears to be established, the 

 distinguishing features of the survivors arc not known to be 

 heritable. 



Lastly, we think it desirable to give in a condensed form 

 some direct observations on the alteration of the composition 

 of natural populations. Sometimes, as in (4), a ' new ' character 

 appears to have spread ; but we do not really know that the 

 character is a novelty in the history of the species. 



(1) Adlerz (1902a). The butterfly Polyommatus vigaureae 

 was very abundant in Sweden in 1896. A peculiar form of the 

 female (with blue spots on the light band of upper side of hind 

 wings) was common. In 1897 the species was not common. 

 The variety was relatively and absolutely rarer. In 1901 the 

 species was again very abundant and the variety made up 

 about half the individuals. Ford and Ford (1930) have found 

 that in Melitaea aurinia there is an increase of variation during 

 local numerical increase. 



(2) Scudder (1889, p. 12 13). Pieris rapae, first introduced 

 at Quebec in i860, appeared in New York in 1868. A variety 

 with yellow wings (var. novangliae) first appeared in Canada in 

 1864. Later it was found also in the United States, where it 

 occurred about once in 500 specimens. It died out again by 

 1878. In Europe the variety is excessively rare, only one or 

 two doubtful specimens being on record. 



(3) Probably the best instance of the appearance and 

 multiplication of a new variant is that of the melanic form 

 (doubledayaria) of Amphidasys betularia, the Peppered Moth. 

 The actual facts are too well known to require repetition here. 

 It is enough to remind the reader that (a) the melanic variety 

 first appeared near Manchester in 1850 and has in many 

 places in England now completely superseded the type form ; 

 (b) a similar course of events occurred on the Continent, though 

 beginning at a later date ; and (c) in the twenty-seven 

 years that have elapsed since the original study (summarised 

 by Doncaster, 1906) was made, the melanic forms (originally 

 largely restricted to the North and Midlands of England) are 

 now far more frequent in the South. An analogous north to 

 south invasion is found in France (Demaison, 1927, p. 295). 

 (d) Similar melanic forms occur in other genera in the same 

 areas, (e) We can find no evidence in contradiction of 



