NATURAL SELECTION 215 



It is a great pity that this problem has not been attacked 

 more resolutely. There has been a tendency to accept some- 

 one's provisional hypothesis and let the matter drop. The 

 fact remains that we have a very clear-cut case of evolutionary 

 change transforming a population rapidly and under our 

 eyes, and the cause has not yet been ascertained. We believe 

 that Bowater's discovery should be followed up. 



(4) Ford (1924, p. 733) states that the varietal constitution 

 of Heodes phlaeas is noticeably different in Madeira from that 

 observed by Wollaston seventy years previously. 



(5) Crampton (1925, p. 17) found that the distribution of 

 the variants of Partula suturalis is very different from what 

 obtained in Garrett's collecting period (1875). The number 

 and range of the sinistral form have increased. So too in 

 P. mooreana {I.e. p. 24) : in 1904 the banded type was 44 per cent, 

 of the population ; in 1919 it was 8 per cent., and in 1923 it 

 was 2 per cent. In mooreana also a ' new ' colour-variety has 

 arisen since 1875. 



(6) Woltereck (1928) summarises the data concerning the 

 appearance and spread of certain new (?) forms of Daphnia 

 longirostris. He (p. 39, supra) attributes their origin to environ- 

 mental causes — a view which is attacked by Wesenberg- 

 Lund (1926), who advances an adaptive explanation. This 

 subject is in its present stage too controversial to discuss 

 in detail. 



(7) Stresemann (1925^. 1 63) states that the melanic variant 

 of Rhipidura flabellifera sixty years ago was known only in 

 S. Island, New Zealand. In 1864 it was taken in N. Island, 

 and has now spread all over N. Island. 



(8) Bateson (1913^. 143) describes the spread of the melanic 

 form of Coereba saccharina, which was originally found on 

 St. Vincent (W.I.) and now is the dominant form, the typical 

 saccharina being ' perhaps actually extinct.' 



Summary. — The value of these observations, in so far as 

 presumptive new characters are concerned, is not very great, 

 because in no instance do we really know that the new characters 

 are, in fact, genetic novelties and had not been previously 

 present in a few individuals which had escaped attention. 

 There is in no instance any evidence as to why the observed 

 increase took place, but there is very definite proof of periodic 

 change in the percentage representation of the classes of 



