58 A STUDY IN HEREDITY 



and propagates it by means of cuttings. He repeats 

 the process again and again. He finds that, as a 

 rule, seedlings develop into plants inferior to the 

 parent ; but that now and then a superior plant 

 arises. It is as though, in an attempt to improve 

 the speed of race-horses, we chose the swiftest 

 individual, destroyed the rest of the species, and 

 propagated him by means of slips, thus making the 

 whole race as swift as the swiftest ; and repeated 

 this process generation after generation. It is 

 evident, owing to the facilities for selection afforded 

 by the great number of seeds borne by plants, and 

 by the fact that some plants can be propagated by 

 cuttings,^ that many of our prize plants have been 

 evolved by a process of selection infinitely more 

 severe than that by which our prize animals have 

 been evolved. Indeed, it is probable that only 

 a very few real {i.e. seed) generations intervene 

 between some of our most evolved garden plants 

 and their wild progenitors. Now what happens 

 if we breed indiscriminately from our garden 

 plants, starting with the best? In a very few 

 generations — sometimes in a single generation — 

 the cultivated plants reverts to something very 

 like the wild variety — but not to the wild variety 

 exactly ; for, while under artificial selection the 



^ It is asserted by some botanists that all plants can be so 

 propagated. 



