12 



mate and territory, they more or less vary, and that 

 when they have once departed from their natural state 

 (or commenced varying) they never return to it again, 

 but are removed more and more therefrom by suc- 

 cessive generations — and that finally if their varieties 

 are even carried back to the territory of their ances- 

 tors, they will neither represent the character of their 

 parents or even return to the species from whence 

 they sprung." 



He also estabhshed that so long as plants in a state 

 of nature remain in their native soil they produce 

 seeds which do not degenerate — but that it was dif- 

 ferent with seeds of a tree in state of change — or 

 as we say improvement, whether the variation be 

 produced by change of chmate, territory or other 

 unknown causes, and that the bounds of this change 

 or variation are not known, except that the last seeds 

 from a tree in state of variation will produce a gen- 

 eration nearer a state of nature than those from its 

 first seeds. Hence, the necessity of raising from the 

 first seeds of a new variety if we wish to obtain a 

 tree far removed from a state of nature — as to that 

 state the plant always in age by its seeds, tends, 

 though never able quite to reach it. 



Upon this basis, he established his theory of pro- 

 ducing new varieties of fruits, viz. that when we 

 have produced a variation by removal or cultivation 

 in any tree, let the first seeds be planted, and upon 

 first production of fruit by the new generation, let 

 its first seeds be planted, and so on without interrup- 

 tion as it is expressed from parent to son, and at each 

 remove it is found that the character of the tree be- 



