350 THE EVOLUTION THEORY 



yield fertile descendants should be sought for by crossing Smerintlais 

 ocellata with one of these American species if it is to have any 

 decisive value. 



Experiments of the same kind have been made by Standfuss with 

 different species of indigenous Saturnia, and these have shown not 

 only that crossing is possible, but that the hybrids are fertile in their 

 turn. These results are to be valued the more highly because it is 

 well known that Lepidoptera, and even the usually prolific silk-moths, 

 do not readily reproduce in captivity, even within the same species. 

 We have in Saturnia pyri, spini, and carpini three well-marked 

 distinct species with no intermediate forms in nature, and with quite 

 different colouring in the caterpillars. That these should have been 

 successfully combined in a triple hybrid proves at least that sexual 

 alienation cannot have advanced far in this case. 



We must beware, however, of attributing too much to the con- 

 stant mutual crossing which occurs in a species living on a connected 

 area and of regarding its influence as irresistible. Undoubtedly it 

 must go far towards securing the uniformity of individuals, but not 

 only is it unable to achieve this, but it cannot successfully resist the 

 stronger influences making for variation which may be exerted upon 

 a part of the area of the species. We have already seen that it is 

 quite erroneous to suppose that every new adaptation must be lost 

 sight of again because of the continual crossing with other members 

 of the species upon the same area. Other things being equal, this 

 depends entirely upon the importance of the adaptation in question. 

 Just as climatic influences may be so strong that they entirely over- 

 come the influence of crossing, and give rise to a local race notwith- 

 standing imperfect geographical isolation, so the same may happen 

 in the case of adaptations. It is quite conceivable that the polar 

 hare of Scandinavia may have evolved a whole series of races, each of 

 which is adapted to the duration of the snow in its geographical 

 range, although a crossing of these quick-footed animals must fre- 

 quently occur in the course of time, even as regards forms from widely 

 separated areas, and although the whole region is inhabited without 

 a break by the species, so that a ' mingling ' of the hares of all regions 

 from south to north, and conversely, may take place, and indeed inust 

 be continually taking place, though of course very slowl}^ 



It is precisely this extreme slowness with which the inter- 

 mingling of racial characters take place that seems to me essential 

 for the production of local or, as in this case, regional races. It is not 

 difficult to calculate the rate of ' blood-distribution ' if we assume that 

 the conditions for a rapid dissemination are as favourable as possible. 



