1899.] PUBLIC DOCUMENT — No. 33. 83 



Individual Variation, Tubers of the Same Variety. 



In view of the frequently reported tests of varieties in 

 which some two or three tubers only of each are used, it was 

 thought best to carry out an experiment to determine if 

 possible the extent to which the product of single tubers 

 will vary when grown under conditions as favorable as pos- 

 sible to uniformity of yield. As a preparation for this test, 

 tiles two feet in diameter and four feet long were set into 

 the ground in a single row, the distance between them being 

 about two and one-half feet. To insure equal drainage con- 

 ditions, a drain tile was laid at about the level of the lower 

 edge of the tile, being given just enough pitch to carry off 

 water. The plot of land in which the tiles were set was sur- 

 rounded with drain tile, to prevent the ingress of soil water 

 from outside. This plot had been uniformly manured for 

 many years, so that the subsoil conditions below the tiles 

 must have been practically uniform. The plots were set so 

 that the surface water from outside was excluded, but the 

 earth outside was brought to within about one inch of the 

 upper edge. 



These tiles so set were filled to within one foot of the 

 top with carefully mixed subsoil, consisting of a very fine 

 sand, this subsoil being settled by the liberal use of water. 

 After this subsoil had thoroughly settled and somewhat 

 dried, equal weights of carefully mixed medium loam were 

 put into each tile, the quantities being sufficient to fill them. 

 The amount used was two hundred and twenty-five pounds 

 for each tile. Conclusive evidence that the work was well 

 and uniformly done is afforded by the fact that the earth 

 in the several tiles remained at practically uniform height 

 throughout the season. 



With the upper four inches in depth of soil in these tiles 

 were most carefully mixed the fertilizers applied, precisely 

 the same weights as determined by chemical balances to each 

 tile. The materials used supplied per tile and at the rate 

 per acre as follows : — 



