SPONTANEOUS VARIATIONS 41 



He concluded, therefore, that the temperature, acting directly, 

 caused a change not only in the individual but in the species. 

 The horse is said to decrease rapidly in size in the Falkland 

 Islands, and European dogs to deteriorate in India. Many 

 similar examples have been recorded. 1 Medical men are 

 familiar with statements declaring that offspring are affected 

 by various pathological conditions in the parent gout, plumb- 

 ism, alcoholism, phthisis and the like. Bud variations e. g. 

 the appearance on a peach tree of a branch bearing nectarines 

 have also been instanced. 



69. But, though all this evidence appears at first sight very 

 strong, it becomes less convincing after careful examination. 

 Hoffman's double flowers may not have arisen through the 

 direct action of the changed environment. Under the greatly 

 altered conditions the fit that survive were probably not the 

 same as would have survived under normal conditions. Pos- 

 sibly, therefore, the double flowers were correlated to some 

 unperceived but important character, which underwent con- 

 comitant evolution. It is doubtful if Weismann's experiments 

 fulfilled all the necessary conditions. For example, he does 

 appear to have paid due attention to light. 2 Colour is affected 



1 Animals and Plants, vol. ii., p. 260 et seq. 



2 I must admit, however, that this is a pure guess on my part. But 

 in the only account accessible to me (The Germ- plasm, p. 399) Weismann 

 does not mention that he provided for the influence of light. He appears 

 also to have reared only one generation of butterflies, and, therefore, we 

 do not know whether the alterations of colours were transmissible. 

 White men who have spent the major portion of their lives in India, but 

 whose families have resided in England, do not appear to have children 

 darker than those who have always dwelt at home. But, since exposure 

 to strong light tends to produce pigmentation in the skin of human 

 beings, the presumption is that this effect of use is beneficial. Fair skins 

 certainly blister very readily. It follows that people with naturally dark 

 skins must, in the tropics, have a considerable advantage in the struggle 

 for existence, and therefore Natural Selection alone should be able to 

 cause the evolution of a black race. Judging by the colour of the new- 

 born babies, the negro-race was not originally black. Terra-del-Fuegians 

 who live in a cold climate are dark ; Esquimaux and Samoydes who live 

 farthest from the equator are yellow ; American Indians are copper- 

 coloured ; Scandinavians are white. It would seem, then, that exposure 

 to light tends to evolve a dark colour in the naked skin of man, and 

 therefore in the tropics, where this form of selection is most stringent, 

 all races are dark. But nearer the poles sexual selection evolves a variety 

 of colours. In the extreme north no doubt the blistering glare from the 

 snow has an important influence. If, then, light so strongly affects the 

 races of man we may suppose it has some influence on butterflies ; though 

 in view of the facts that so many brilliantly coloured species inhabit the 

 tropics, and that the skin of butterflies is not naked, the influence is 

 doubtless much weaker. 



