No. 6. DEPARTMENT OF AGRICULTURE. 363 



select a smooth type of the Uncle Sam at the cost of reducing yield. 

 Shape may be affected by the season, character of soil and culture 

 methods. Small tubers may be expected in any hill. Some varieties 

 like the Irish Cobbler normally have a numerous second setting, at 

 least with me. Weight of salable tubers should determine judging, 

 disregarding small ones unless many more than usual. Any mixture 

 of varieties should be thrown out. One is the best for any one set of 

 conditions and only that one should be kept. 



I never saved any poor hills to test till 1912. That year the low 

 yielders row averaged only 70 bu. per acre against 350 for the ad- 

 joining rows treated exactly the same. In 1913 the low yielders 

 averaged 150 against 290 for the normal. This was not a fair test 

 as the low yielders were mature when a frost, September 14, killed 

 the others while in full growth. In 1915 the yields were 74 and 334 

 bu. per acre. In some seasons low yielding strains may be told by a 

 longer and more pointed shape than normal. 



I dig about one acre out of 17 each fall with hooks, either before 

 the rest are ripe enough to dig with a digger or on days too wet to 

 use a machine. Each hill is kept separate. Next I go over the row 

 myself and select 5 to 20% of the best hills. This is the hardest 

 part and requires the best judgment. These selected hills are planted 

 the next year in a seed plot on the best part of the field, if neces- 

 sary extra care should be given. The soil should be one holding 

 water well for times of drought, the time when seed vitality may be 

 injured. Early in the fall and before the tops on the healthy vines 

 die, a hook should be used to remove all weak, early, dying or diseased 

 hills. This year I found only 4 bushels of such on 5 acres yielding 

 1,810 bushels. Small tubers from such a plot are as good as large 

 but there are not very many. 



Owing to the different parts of a field varying greatly in yield it 

 is better to take a percentage of the best hills rather than to use a 

 rigid standard of weight over the whole field. This percentage 

 should be higher in the good parts than where potatoes are poorer 

 because of possible injury to the vigor of seed from poor parts of the 

 field. In case conditions are so poor that degeneration takes place 

 it would be best to save only from the best part. Seed selection can- 

 not prevent loss of vigor, but it can quickly eliminate the damaged 

 strains unless the injury is so severe that all are affected. 



Scientists are not agreed as to whether the yield of a particular 

 strain can be increased by annually selecting the best hills. It is 

 regarded as certain that the average yield of a variety can be in- 

 creasd by finding and multiplying the best strains within the variety. 

 As a farmer I am well satisfied that this can be done with profit 

 by using some of the methods here given. The cost is so small that 

 every hour spent is well paid. I believe that I have received greater 

 returns from the time used in seed selection than in any other thing 

 done on the farm. 



