124 Selection Does Not Lead to Origin of Species. 



ing for the first year only, but in succeeding years become 

 later and later. ^ 



Then there is the well-known case of Buckman who 

 by reversing selection for a few generations converted 

 the cultivated parsnip into the wild form (Pastiiiaca 

 sativa). Watson obtained the same result with the 

 Scottish cabbage in three generations. Darwin's verdict 

 on this point is that a period of selection which need 

 not extend over many generations would be sufficient to 

 convert most of our cultivated plants into wild or nearly 

 wild forms. 



I have already mentioned Schubeler's experiments 

 on the extension of the northern limit of the culture of 

 cereals in Norway. He found that if he took the forms 

 which have been grown at the northern limit — that is. 

 the short lived forms — back to their native place after 

 a few years, that they ripened earlier and bore heavier 

 seeds than those forms of the same sort which had re- 

 mained there all the time; but that after a few genera- 

 tions this distinction vanished. 



Fruit trees grown from seed quickly revert to the 

 original type; the Olive to the Oleaster; apples and pears 

 give smaller and less juicy fruit; and chestnuts become 

 quite unpalatable.^ But our information on these phe- 

 nomena is far too meagre. 



If it was not a so-called species but a subspecies 

 which was subjected to improvement, the new form re- 

 turns, on cessation of selection, not to that of the spe- 

 cies, but to the mean of the subspecies. Double balsams 

 and buttercups tend to become single, Triticitm coin- 

 posittmi turgidimi (Fig. 26) becomes less branched, the 



* Darwin, Das Variiren, II. p. 42. 



^De Candolle, Originc dcs Planfes cultivccs, p. 372. 



