JQ08. 



IMPROVEMENT OF THE POTATO. 



439 



SELECTIONS FOR PLANTING, 1902 



To endeavor to determine whether these fluctuations in a chem- 

 ical constituent are transmitted by tuber propagation, two plots 

 were planted, one of high nitrogen selections and one of low nitro- 

 gen. This gives a check upon the seasonal and soil variations. The 

 selections were made upon the fresh substance, because this is their 

 market condition, and because the dry matter determinations were 

 not finished soon enough to make selections on the water free basis 

 practicable. 



As may be seen in the table, potatoes were chosen with as 

 nearly as possible the same average weight. The low protein se- 

 lections, however, averaged about i8g. heavier per tuber. 



The average nitrogenous matter in the high portein plot was 

 2.95 percent of the fresh substance and 14.07 'percent calculated to 

 the water free substance. The following data were taken on each 

 tuber. The long, medium and short diameters give a fair indica- 

 tion of the shapes of the tubers : 



TABLE 16. HIGH PROTEIN SELECTIONS IN 1902 



Each potato was quartered and each quarter planted in a hill. 

 The hills were four by two feet apart. The planting could not be 

 done on account of press of other work until June tenth. 



In exactly the same manner, end to end with these selections, 

 were planted the following low protein selections, which averaged 

 1.78 percent nitrogenous matter in the fresh substance, and 8.75 

 percent calculated to water free basis. 



