22 
COLOHADO ExPEKIMENT StATION 
could be detected for about six weeks. In 1917 the germination 
appeared to be hastened slightly by greening tho the differences 
in the rate of growth disappeared after a few days. In 1918 the 
effects of greening observed in previous years could not be dis- 
cerned except in the small increase in yield. 
IRRIGATION 
Considerable difference of opinion exists among practical grow- 
ers regarding the stage of growth at which irrigation should be 
started. Some maintain that water should not be applied until after 
the tubers have set even tho the plants have begun to suffer from 
lack of water. Others hold that if the plants show need of water at 
an earlier period it should be applied. 
To determine, if possible, which of these views is correct experi- 
ments were conducted in 1917 and 1918 with two of the late varie- 
ties, the Pearl and Rural New Yorker. Each plat consisted of eight 
rows. In 1917, when only the Pearl variety was used, one-half of 
each plat was planted with dry land seed and the other half with 
Wisconsin seed. In 1918 Pearl seed from Wisconsin and Rural New 
Yorker seed from Colorado were used for the respective halves of 
each plat. To overcome any effects which might arise from lateral 
movement of water in the soil the plats were separated by two dis- 
card rows the yields of which are not included in the experimental 
data. 
The different periods at which irrigation was started were desig- 
nated as ‘'checks,’’ “early,” “medium” and “late.” On the check 
plats irrigation was started in 1917 just as the tubers were be- 
ginning to set. In 1918 tuber development was farther advanced 
at the time of the first irrigation which was delayed because of 
rains. The early irrigated plats received the first irrigation about 
two weeks before tuber formation had started. On the plats desig- 
nated as “medium,” which were employed only in 1917, the first 
application of water was on the same date as that of the check 
plats. The total number of irrigations for the season was, however, 
one less. On the late irrigated plats irrigation was not started until 
after the tubers had made considerable growth. At that time the 
dark color and stunted appearance of the plants plainly indicated 
the need of water. These plats received the same number of irriga- 
tions as the “medium” plats but the first application was mucli 
later. The results are given in Table 10. In addition to the yields 
per acre based on actual weights the corrected yields have been 
