Potato Breeding 71 



produce six or more merchantable tubers. This experi- 

 ment has extended over three years. The first year, 

 from 7 per cent to 8 per cent of hills were found that met 

 the requirements ; the second year from 17 per cent to 

 20 per cent. This season the number of hills which have 

 produced six or more merchantable tubers has risen to 

 24 per cent to 27 per cent. 



Whenever pure lines or strains or varieties of potatoes 

 are compared, every effort should be made to eliminate 

 soil differences or other factors of environment, so that a 

 direct test of the heredity only can be made. Potatoes 

 are so easily influenced by enwonment that real heredi- 

 tary differences are often covered up. Absolute imi- 

 formity in time of planting, methods of culture, spraying 

 and so forth are essential. The seed pieces should be as 

 nearly the same weight as possible. Arthur says, " What- 

 ever increases rate of growth at the beginning, increases 

 yield." 



If tuber-units are being compared, they should be 

 planted in duplicate or preferably in triplicate in differ- 

 ent parts of the field and their production averaged. 



If single high-producing liills are saved for the next 

 year's planting, one should make sure that they did not 

 come from some highly fertilized part of the field. Pref- 

 erably they should be chosen from the poorer parts of the 

 field when their production may be high in spite of their 

 environment, the real test of heredity. 



Zavitz, of the Ontario Experiment Station, selected 

 the best hills of seven varieties of potatoes for a period 

 of sixteen years. During the first four-year period 

 the average yield was 120 bushels. During the three 

 succeeding periods of four years each, as a result of 

 selection, the average yields were increased to 216, 218 



