NATURAL SELECTION 131 



leads to its reinstatement. 1 But I must return to the question of 

 chance. It is easy to adduce familiar facts that show that 

 variations become adaptations only by lucky coincidences. A 

 wild poppy in a corn-field is pink, while all the others round 

 about are red. This pink poppy is no better for its singularity. 

 But a gardener happens to notice it. He marks it, keeps all the 

 seed, and next year there are a number of seedlings growing 

 from it. It is clear now that the pink colour has become an 

 adaptation. There was, in fact, a sudden change of environ- 

 ment. The gardener, with his taste or caprice, became an 

 important external condition. But the variation which proved 

 to be an adaptation to this condition, originated before the 

 gardener became part of the environment. It was due to mere 

 coincidence that the variation became an adaptation. 



In the animal world the facts and their teaching are the same. 

 Cattle through a congenital variation may have no horns. Now, 

 if they are wild cattle, it is unlikely, unable as they will be to 

 defend themselves, that they will found a hornless race or species. 

 But if they belong to a domesticated breed, it is very likely that 

 they may be singled out, and that a new breed may be formed. 

 In fact, mere chance will make the variation an adaptation. 



Now that we have decided that adaptations are chance coin- The cause 

 cidences, we must turn our thoughts to the variations themselves. 

 In a former chapter (see p. 33) I have traced them partly to the 

 process of fission, partly to conjugation. When fission takes 

 place, inequality must result. We are bound, if we are Dar- 

 winians, to assume that even the approximation to equality is 

 due to Natural Selection. But absolute equality is unattainable. 

 Moreover, the process of division is often a very complicated 

 one. The elaborate structure of the nucleus goes through a 

 number of transformations before division takes place. There 

 is a thorough shuffling of the cards before they are cut. To 

 adopt another metaphor, how can there be reconstruction on an 



1 As to Hoffmann's experiments see Weismann's Germ Plasm, p. 437, and Essays, 

 vol. i. p. 420. I cannot see that the facts necessitate the concession which he 

 makes, viz., that a new character in the germ-plasm due to the environment was 

 inherited. 



