CH. XX] THE ORIGIN OF SPECIES 207 



must probably have some degree of luck. It may not chance 

 upon a spot where it can grow, the ground being'alreadv fully 

 occupied by a closed association of plants; or it may easily be 

 destroyed by a fire or a flood or other accident (cf. Didtjmocarpus 

 and Christisonia on p. 151). 



The older view, that species arose by gradual accumula- 

 tion of infinitesimal or fluctuating variations (up and down 

 variations, such as always show in any character, as when a leaf 

 varies in length from one to one and a half inches on the same 

 species), is now dying out in favour of mutations, or sudden 

 alterations of form which have their origin in changes that have 

 occurred in the material bearers of heredity. It is conceivable 

 that any changes, however great, might be brought about by 

 the accumulation of fluctuating variation, provided (1) that the 

 variations were fully hereditary, (2) that they were not linear, 

 showing the same character in greater or less degree, but dif- 

 ferentiating, a simple leaf, for example, showing a tendency to 

 compoundness, (3) that the necessary variations appeared, and 

 (4) that natural selection should be able to act, i.e. that the 

 appearance of the variation should give to the plant or plants 

 possessing it such advantages as should ensure their survival in 

 at least the majority of cases. 



In regard to the first supposition, so far as we know, infini- 

 tesimal variation is not fully hereditary, but always regresses or 

 falls back, so that while one may make great impro\-ement.s by 

 selection (as, for example, in the speed of trotting horses, oi the 

 content of sugar in the root of the beet) there always comes a 

 point beyond which one cannot go. It is sometimes stated that 

 the wonderful varieties of our cultivated crops owe their origin 

 to the selection of infinitesimal variations, and that when left 

 to themselves they go back to the wild form, but this is not the 

 case, however; as Hooker long ago pointed out, the cultivated 

 apple goes back, not to the crab, as is popularly supposed, but 

 to crab types of cultivated forms. 



These facts agree with ordinary observation, which gi\es no 

 reason to suppose that continuous change is going on. Hooker 

 (55 a, p. x) has so Avell put the argument in favour of the general 

 permanence of species that it would be presumptuous to try to 

 better it (I have shortened it). 



"(1) The fact that the amount of change produced by ex- 

 ternal causes does not warrant our assuming the contrary as a 

 general law. 



