CEREALS ON THE BELLE EOURCHE EXPERIMENT EARM. 45 
the yield data. In 1913 and 1917 the earliest dates of seeding gave 
the highest yields; in 1912, 1915, and 1916 the latest dates gave the 
highest yields, while in 1918 the medium dates were most favorable. 
The flax in the last sowing in 1915 did not emerge and that from 
\ the earliest sowing of 1916 was blown out. Late summer rains in 
1912 favored late sowing of flax, but in general the yields from flax 
sown after May 20 were comparatively low. The best date of seed- 
ing for flax on dry land at Newell, although undetermined, prob- 
ably occurs between April 15 and May 15. The data are shown in 
Table XXXII. 
EXPERIMENTS ON IRRIGATED LAND. 
In 1912 a large part of the Belle Fourche Experiment Farm was 
placed under irrigation, and experiments with cereals on the irrigated 
land were begun at that time. A few varieties of spring wheat, oats, 
barley, and flax were sown on irrigated land in 1912, 1913, and 1914. 
In 1915 the number of plats was increased to include experiments 
with winter wheat and grain mixtures, and in 1916 and thereafter a 
number of additional experiments were in progress. The irrigated 
plats were one twenty-fifth of an acre each in 1912 and 1913, but in 
1914 and thereafter nearly all experiments were conducted in tripli- 
cated fiftieth-acre plats. 
Most of the cereals grown under irrigation were sown on land 
which had produced an intertilled crop, such as corn, potatoes, roots, 
or sunflowers, the previous year. The land was double disked and 
harrowed or " floated " before seeding. The cereal crops received two 
and sometimes three irrigations during the season. The time of irri- 
gation was gauged by the condition of the soil rather than by the 
stage of growth of the grain, because it was practically useless to 
irrigate until the surface soil was sufficiently dry and cracked to take 
up water. The quantity of water applied was not measured, but was 
approximately 4 or 5 acre-inches at each application. Water was ap- 
plied by flooding from field ditches, which usually were constructed 
along the ends of the plats. 
The average yields from the irrigated land are not greatly in excess 
of those from the same crops grown on the dry land. In 1915 the 
yields on dry land were much higher than on irrigated land. 
From 1912 to 1915, inclusive, the irrigation experiments were con- 
ducted on rather unfavorable soil, while in 1916 most of the crops 
were damaged by soil blowing and rust. The irrigated grains were 
sown on previously cropped land following either small grain or an 
intertilled crop, while many of the dry -land experiments were con- 
ducted on summer fallow. In most seasons the grains were sown on 
the dry land earlier than on the irrigated land, chiefly because the 
