Methods of Breeding anm) Improving the Potato Crop. 577 



measure of the productivity of the bud, which formed the seed tuber 

 planted, and this is true in a large measure also when single pieces of 

 the tuber are planted in a hill, provided these pieces are of the same size. 



Experiments which have been conducted by several investigators 

 have demonstrated that hills differ greatly in their productivity, and 

 that this tendency is one which is in considerable degree transmitted 

 to the hill or tuber progeny. 



In experiments on the Cornell Farm great variations in the yield of indi- 

 vidual hills have been obtained. In the suinmer of 1908 a fairly well 

 selected strain of Sir Walter Raleigh potato gave an average of two 

 and one-half pounds per hill, which in a perfect stand with rows three 

 feet apart, and the hills one and one-half feet apart in the rows, would 

 give a yield of slightly over 400 bushels per acre. In this same selected 

 strain the minimum-yielding hills gave only eight ounces per hill, which 

 would represent a yield of 80 bushels per acre, planted as above, while 

 the maximum-3'ielding hills weighed slightly over four and one-half 

 pounds per hill, which in a perfect stand would give a yield of 726 bushels 

 per acre. Unselected Early Rose planted in the same soil averaged 

 only one pound per hill, which would give a yield of 161 J bushels per 

 acre. The maximimi-yielding hills of this strain, however, reached as 

 high as three and one-fourth pounds per hill, which represents a yield 

 of 524I bushels per acre. It would thus seem that with our ordinary 

 varieties there is a great difference in the yield of hills, even when pieces 

 of uniform size are planted, and that there is thus great opportunity 

 in selecting high-yielding hills. There have been very few experiments 

 carried out to show what effect could be expected from such selection. 

 The most reliable results of this kind of which the writer has knowledge 

 are those that have been obtained by C. W. Wade of the Ohio Experi- 

 ment Station at Wooster, Ohio. 



In these experiments, which were begun in 1903, ten high-yielding 

 hills and twenty low-yielding hills were selected and the seed preserved 

 separately. In 1904, ten hills each were planted from seed of the ten 

 heavy hills and five hills each from seed of the light-yielding hills, mak- 

 ing 100 hills of each group. To compare wath these as a check, 100 hills 

 were planted from seed which had been selected without reference to 

 individual hills. 



In 1905, 100 hills were again planted with seed from high-yielding 

 hills and 100 hills from low -yielding hills, the seed being selected respec- 

 tively from the high-yielding and low -yielding hills of the 1904 crop. 

 A similar check to that of the preceding year was also planted. In 

 1906 the same policy was pursued, the results reported thus representing 



19 



