346 HABIT AND INTELLIGENCE. [cHAP. 



of the fisli ? Onglit it not to take much more than a million 

 fofmf * times as long ? I ought perhaps to repeat, that if accuracy 

 from the of numerical statement here is unattainable, it is also 

 requke, unnecessary ; what we have to compare is not the magni- 

 fy the ^^(jeg themselves, but only the orders of the magnitudes, 

 process ? But the argument, as stated here, is much too favourable 

 to the theory of none but imperceptibly slow variations ; 

 for it compares variation under domestication with varia- 

 tion in a state of nature, as if it were equally rapid in 

 Variation both. And this is not the case : on the contrary, we may 

 aniong" Say, without much risk of exaggeration, that variability 

 wild than jg the rule among domesticated races, but the exception 

 among wild ones. Wild races remain unchanged through- 

 out whole geological ages, though, I have no doubt, they 

 become variable under the influence of changes of habitat, 

 climate, and food; when these changes take place, new 

 varieties and new species, more suitable than the old ones 

 to the new conditions of life, wiU be formed by the conjoint 

 action of self-adaptation and natural selection. But, so 

 long as variation takes place so slowly as not to produce 

 new races at once (which in Darwin's opinion is always 

 How selec- the case in the wild state), the formation of distinct races 

 acTin the ^^ ^^^ same habitat will be prevented by intercrossing, 

 wUd state. ^^^ it will be impossible for the processes of change to 

 do more than to modify the xvhole race together, so as to 

 suit the new conditions of life.^ This will not prevent 

 the action of " natural selection among spontaneous varia- 

 tions," in giving origin to new species, but it must tend 

 greatly to restrict its operation so long as the variations 

 are so small as not to give origin at once to new species, 

 or strongly marked races. The absence of mixtures of 

 race in the wild state will tend to prevent variation, 

 for, as we have seen, pure races are less variable than 

 mixed ones. 



"^ If some individuals of a race were so far to change their habits as not 

 to he brought into contact with their kindred, — as, for instance, in the case 

 of becoming nocturnal instead of diurnal, — they would, no doubt, give 

 origui to a new race. But such a change of habit comes to the same 

 thing as a change of country. 



