34 
Colorado Experiment Station 
shows some inconsistencies, which may be attributed chiefly to the 
effect of the different conditions under 'which the crop was grown 
during the course of the experiment. There appears to be a direct 
relation between the size of the seed piece and the yield where the 
seed was subjected to unfavorable conditions during the period of 
germination, as was the case in 1916 and 1917. On the other hand, 
where conditions were favorabje during this period, this relation- 
ship is not so apparent ; in fact, after 'deducting from the market- 
able tubers the amount of seed used, the larger sizes of seed were 
found to be less economical. It is, however, essential in all cases 
to use a seed piece of sufficient size to furnish the necessary amount 
of food material to give the young plants a good start and especially 
to insure against loss of stand due to unfavorable conditions during 
germination which cannot be foreseen at the time of planting. The 
experimental results indicate that while the most economical size 
of seed to use may vary according to conditions, it should at least 
equal or exceed two ounces in weight. 
