198 THE FUNCTION OF SEX 



the sole source of progression. For example, to choose another 

 homely instance, suppose a human race inhabits a region where 

 malaria is prevalent, but where tuberculosis and measles are 

 absent. Then offspring will vary from their parents, favourably 

 and unfavourably, with respect to all three diseases. In other 

 words, they will, as compared to the parents, be more resistant 

 or less resistant to each disease more or less able to survive if 

 exposed to it. It does not follow, however, that an individual 

 who has varied favourably as regards malaria, will also have 

 varied favourably as regards tuberculosis, measles, or any other 

 disease. Since malaria is present, those who have varied un- 

 favourably in relation to it will be eliminated, while those that 

 have varied favourably will be preserved. Like will mate with 

 like, and therefore little or no retrogression caused by blending 

 will occur to check progression caused by Natural Selection. On 

 the other hand, the progressive variations with respect to tuber- 

 culosis and measles will be mere useless redundancies. Since 

 there will be no elimination of those that vary unfavourably, 

 unlike individuals will mate. The result will be retrogression. 1 

 Apply this reasoning to all the characters of the individual, and 

 the function of sex becomes apparent. Natural Selection preserves 

 or causes progression in useful characters by eliminating the 

 individuals that do not possess them in sufficient degree. On the 

 other hand, sex, a product of evolution through Natural Selection, 

 automatically eliminates useless characters, but without eliminating 

 the individuals that possess them. 



326. The fact that races undergo protective evolution only 

 1 When two dwarf breeds of dogs are crossed the offspring tend to be larger 

 than either parent. This would seem to indicate progression as a result of crossing. 

 But such breeds have resulted from the artificial selection of mutations. In them, 

 presumably, the larger size is latent. Again, inbred artificial varieties tend to be 

 delicate ; and Darwin noted that, when two such varieties are crossed, the result 

 is often an increase of health and vigour. It seems not unreasonable to suppose 

 that the apparent progression is really due to retrogression. Nature eliminates 

 unfavourable variations in all directions in evolving varieties. But man, devoting 

 his attention exclusively to one or two points (e.g. speed in race-horses, size, 

 shape, and colour in flowers), neglects or cannot perceive others which, though 

 unfavourable, are not eliminated by selection because the individuals are pro- 

 tected, nor by conjugation because the individuals, being inbred, are much alike. 

 Crossing unites unlike individuals and so eliminates the unfavourable redund- 

 ancies ; for example, it tends to remove the traits that cause ' delicacy ' and so to 

 restore the pristine vigour. It performs that which it is the function of conjugation 

 to perform. If this interpretation be mistaken we have no alternative but the 

 hypothesis of rejuvenescence. It should be noted that the traits that man con- 

 sciously selects are usually mutations. But those which he unconsciously selects 

 (e.g. * delicacy') are usually fluctuations. 



