274 AN INTRODUCTION TO MODERN GENETICS 



factors involved. Sumner was at first inclined to doubt whether the 

 race differences really did depend on Mendelian factors at all, and 

 drew attention to the apparently adaptive character of some of the 

 colour varieties. However, experiments in which animals were trans- 

 ferred from one region to another failed to show any direct effect of 

 the enviroimient on colouration, and Sumner eventually was able to 

 satisfy himself that the MendeHan explanation was adequate. 



8. Local Races in Plants 



The investigation of local races in plants reveals a situation essen- 

 tially similar to that in animals, but since plants are sessile whereas the 

 most fully analysed animals are mobile, there is an even greater multi- 

 plication of local genetical types. Particularly extended studies have 

 been made by Turesson,^ who has given the name of genecology to the 

 combination of genetical, ecological, and taxonomic investigations. He 

 collected samples of plants from many different locaHties and cultivated 

 them in the uniform environment of an experimental garden. The 

 luxurious conditions of cultivation in some cases allowed stunted forms 

 from unfavourable situations to develop potentialities which had pre- 

 viously been hidden, and apparently uniform populations from such 

 places might reveal considerable hereditary variation. The main result, 

 however, was to show that forms from different localities tended to 

 retain their characteristics, many of which had a clear adaptive signi- 

 ficance. The adaptive modifications of wide-ranging species seem there- 

 fore to be hereditarily fixed, and the progeny of crosses of different 

 races showed segregation and high variability, often paralleling the 

 mixed populations which can be found in nature in intermediate 

 situations. The differences between the races rarely depended on single 

 gene differences, but usually polymeric multiple genes were involved. 

 Within a given ecological situation, however, the population was 

 usually fairly uniform, so that the genes are not distributed at random 

 throughout the whole species, but are collected into complexes, each 

 complex being characteristic of, and adapted to, a certain set of con- 

 ditions. It is only in situations intermediate between two localities with 

 well-marked characteristics that the complexes break up and a free 

 mixture of genes is found. 



Again, these differences between local races can probably be taken as 

 typical of the differences between fully developed species. Harland^ has 

 analysed several species of cotton and has shown that, besides a few 

 1 Turesson 1925, 1930. ^ Harland 1936. 



