the plants from all these grains, I select the finest one, 

 which 1 accept as a proof that its parent grain was the best 

 of all, under the peculiar circumstances of that season. 

 This process is repeated annually, starting every year with 

 the proved best grain, although the verification of this su- 

 periority is not obtained until the following harvest. 



"In illustration of those principles of selection, I now 

 give the following results, due to their influence alone as 

 the kind of seed, the land, and the system of culture em- 

 ployed were precisely the same for every plant for four 

 consecutive years; neither was any manure used, nor any 

 artificial means of fostering the plants resorted to. 



Table showing the importance of each additional generation 



of selection. 



"Thus, by means of repeated selection alone, the length 

 of the ears has been doubled, their contents nearly trebled, 

 and the "tillering" power of the seed increased five-fold. 



"The following table gives similar increased contents 

 of ear obtained in three other varieties of wheat: 



"It was supposed by the ancient writers that the pow- 

 ers of grains differed in relation to their position in the ear. 

 This I investigated in 1858 by planting the grains of ten 

 ears on a plan showing their several positions in the ear. 

 The only general result, among most conflicting ones, was 



