26 THE VARIATION OF ANIMALS IN NATURE 



the most obvious cases, such as polymorphic butterflies or snails, 

 the evidence suggests that the various forms differ in a rather 

 small number of genes and the range of variation on crossing 

 is not very great. If we except geographical races and poly- 

 morphic species, crosses within the species rarely give rise to 

 a large series of variants. We are not aware, however, of any 

 serious attempt to discover by prolonged inbreeding how many 

 genes might be present. Duncan (1915) crossed specimens 

 of Drosophila from widely separated localities, but found that 

 no unusual amount of variation resulted. Unfortunately, the 

 flies of this genus are so largely spread by commerce that they 

 are not suitable material for such an investigation. Timofeef- 

 Ressovsky (1927) obtained seventy-eight wild females of 

 Drosophila melanogaster from a house in Berlin. It was supposed 

 that each of these had already mated with more than one male. 

 As a result of interbreeding it was deduced that eighteen of the 

 females and thirty-four of the males were heterozygous for at 

 least one mutant. Ten different genes were identified, some 

 of them already known in cultures. 



Geographical races when crossed often give a consider- 

 able range of variation, usually intermediate between the 

 parents. If the types produced by recombination are few, 

 the chances of a beneficial variant are smaller, while the 

 larger the number of types, the fewer the individuals of each 

 that will appear. As far as the evidence goes, it would seem 

 that most individuals of a species are homozygous for a large 

 common stock of genes, so that little or no recombination 

 would occur on crossing. The geneticists' idea of a ' wild 

 type ' is partly based on this assumption. Of course we cannot 

 say how far this is true of genes producing only very minute 

 external effects, but we must judge by what evidence we have. 

 When forms differ considerably, so that recombination would 

 be expected to produce much variation, sterility in one form 

 or another seems usually to intervene. It is quite possible 

 that the majority of animal species have always been homo- 

 zygous for most of the genes carried at any one time. 



No doubt some crossing between species, subspecies, etc., 

 occurs in nature. How far such unions are fertile is a very 

 debatable point. When we consider the diversity of means by 

 which isolation is brought about (Chapter V) it does not seem 

 likely that successful crossing is very common or that it occurs 



