288 LIFE AND HABIT. 



plant the descendants of A; but let there be 7001 A's 

 against 7000 B's at first, and the chances are once 

 more equal, while if there be 7002 A's to start, the odds 

 would be laid on the A's. Thus they stand a greater 

 chance of being killed ; but, then, they can better afford 

 to be killed. The grain will only turn the scales when 

 these are very nicely balanced, and an advantage in 

 numbers counts for weight, even as an advantage in 

 structure. As the numbers of the favoured variety 

 diminish, so must its relative advantages increase, if 

 the chance of its existence is to surpass the chance of 

 its extinction, until hardly any conceivable advantage 

 would enable the descendants of a single pair to exter- 

 minate the descendants of many thousands, if they and 

 their descendants are supposed to breed freely with the 

 inferior variety, and so gradually lose their ascen- 

 dancy,'" ("North British Review," June 1867, p. 286 

 " Genesis of Species," p. 64, and onwards). 



Against this it should be remembered that there is 

 always an antecedent probability that several specimens 

 of a given variation would appear at one time and place. 

 This would probably be the case even on Mr. Darwin's 

 hypothesis, that the variations are fortuitous; if they 

 are mainly guided by sense of need and intelligence, it 

 would almost certainly be so, for all would have much 

 the same idea as to their well-being, and the same cause 

 which would lead one to vary in this direction would 

 lead not a few others to do so at the same time, or to 

 follow suit. Thus we see that many human ideas and 

 inventions have been conceived independently but simul- 

 taneously. The chances, moreover, of specimens that have 



