WATER UTILIZATION BY SPRING WHEAT. 15 
which the moisture has been utilized varies to some extent in dif- 
ferent years. In the four years of heaviest production the soil-mois- 
ture content of the fifth foot section was not reduced to the minimum 
point. In the two years of low production all of the available mois- 
ture was removed from the fifth foot and use was made of the 
water in the sixth foot. In 1911 the sixth foot of soil was dry at 
harvest. The indications are that the soil at lower depths is most 
completely 'utilized when the crop suffers for moisture. 
At Dickinson there has been a great opportunity to study moisture 
penetration and depth of feeding of crops. The soil at the station is 
open and water penetrates freely, and the precipitation in some years 
has been so high that under all the methods of cultivation the soil has 
been wet to a depth of at least 6 feet. At this station neither plat A 
nor B has shown to advantage over the other in respect to moisture 
stored in the soil. These two plats furnish a striking illustration of 
the fact that roots will not penetrate a dry soil, even though there 
may be available water below. In 1919 plat A used all of the avail- 
able moisture in the first 5 feet of soil but not in the sixth foot. In 
the same year plat B exhausted the available moisture in the first 4 
feet, but did not secure that from the fifth and sixth foot sections. The 
moisture present in the soil in the spring of 1911 and that added by 
precipitation during the growing season moistened the soil to a depth 
of less than 2 feet. In spite of severe drought injury no moisture was 
used below that depth. The dry layer in the third and fourth foot 
sections prevented the roots from extending to moisture in the fifth 
and sixth foot sections. The crops of both plats were destroyed by 
hail in 1912, but the soil did not become wet that year to a depth 
greater than the fourth foot. In 1913, both plats stored moisture 
to a depth of 4 feet. This was enough to wet all of the dry layer in 
plat B, but the fifth foot in plat A still remained dry, as it was left 
by the crop in 1910. The roots of the wheat crop on plat B in 1913 
were able to draw upon the available water in the fifth and sixth foot 
sections, but in plat A the roots were not able to reach the available 
water in the sixth foot on account of the dry layer in the fifth foot. 
The soil in plat C or D shows a deeper storage of moisture in dry 
years than in plat A or B. In all years except 1920 the plat has held 
available water in each foot of soil at all depths. In the two years 
of heaviest production no water was used from below the fourth foot, 
although there was water present in the fifth and sixth foot sections. 
The soil at Mandan is easily penetrated by moisture. In 1915, it 
was wet in all plats to a depth of more than 6 feet. The moisture in 
the fifth and sixth foot sections of plats A and B was not all removed 
until the very dry summer of 1919, in spite of the fact that the wheat 
growing on these plats had suffered from drought injury in the three 
years preceding. The water of the first 4 feet is utilized freely, but 
evidently the stress of a long-continued demand for water is prece- 
dent to its complete utilization at lower depths. 
Edgeley is a striking example of a shallow soil. At this station the 
soil is a heavy clay that changes into undecomposed shale at a depth of 
about 2 feet. Even in the second foot pieces of undecomposed shale 
distributed through the soil are frequent enough to reduce seriously 
the quantity of available moisture that can be held. The depth of 
penetration of plant roots seems to be limited to 2 feet. The quantity 
