10 DEVELOPMENT AND HEREDITY. 



feet, there are a thousand points where an equally 

 slight change would make it less perfect. When 

 we endeavour to unite the conception of this vast 

 uninterrupted series of useful variations with the 

 idea that the detrimental variations are just as likely 

 to occur as the useful ones, we discover that the 

 two ideas are incompatible. Especially is this the 

 case when we recognise the fact that it has not 

 been a single individual nor a single pair which has 

 survived in each generation, but a species or variety, 

 composed probably of as many individuals as we 

 find to-day in a species or variety. In fact, we 

 would expect it nearly always to have been what 

 Darwin called a "dominant species." When we 

 remember the many cases of wholesale destruction 

 of animals, — for instance, the killing of countless 

 fishes by a sudden change in the temperature of 

 the ocean, the killing of birds and insects by cold 

 and storms of wind and rain, and the killing of 

 myriads of organisms of all kinds by circumstances 

 over which they could have no control, and from 

 which no mere individual variation could save them, 

 — we are led by these facts to doubt that natural 

 selection acts with such mathematical certainty and 

 accuracy in accumulating slight individual variations. 

 It seems more plausible to suppose, since organisms 

 are generally adapted to their environment, and 



