8 BULLETIN 546, U. S. DEPARTMENT OF AGRICULTURE. 
IV, one of the two fall-irrigated series. The 3-year average yield of 
the crops on Series IV was 16 per cent lower than the corresponding 
average of Series II, while the 3-year averages of the two check 
series, I and III, were identical. The relatively low productivity of 
Series IV may have been associated with a heavy infestation of 
gumbo weed (Iva axillaris). This weed was more abundant on 
Series IV than on any of the other series throughout the period of 
experiment. At any rate, there is nothing in the data to indicate 
that fall irrigation influenced the crop yields. 
SOIL MOISTURE. 
When the experiments were planned, it was the intention to make 
soil-moisture determinations shortly after the fall irrigation each 
year and also during the early part of the growing season. It was 
found, however, that because of cold weather the soU remained wet 
so long after the fall irrigation that fall sampling was impracticable. 
All the soil-moisture data available, therefore, were secured during 
the spring and early summer of each of the three years of the period 
of the experiment. In each sampling two cores were taken on each 
plat and composited to a single sample for each foot section. 
Soil moisture in 1914- — Three samplings were made in 1914, on 
May 5, June 18, and June 27. The samples secured on May 5 were 
taken from the plats seeded to oats, barley, and wheat, two plats 
of each crop on the fall-irrigated land and the same number of 
check plats, a total of six plats representing each fall treatment. 
These samples were taken 34 days before the first irrigation. The 
second set of samples, taken June 18, was from the same plats as 
the May 5 samples and from the two flax plats, a total of eight 
plats representing each fall treatment. This sampling was done 10 
days after the first irrigation of the oats, wheat, and barley and one 
day after the flax was irrigated. The samples secured on June 27 
were from the same plats as those taken on June 18, and no irrigation 
water was applied between the two sampling dates. These moisture 
determinations were made to a depth of 4 feet. A summary of the 
results is given in Table VI, which also shows the probable errors of 
the averages of moisture content. 
Table VI shows that on May 5 the average moisture content of the 
upper 3 feet of soil on the fall-irrigated plats was higher than that of 
the corresponding depths on the check plats, the greatest difference 
being in the second foot, where it amounted to 8.3 ±0.5 per cent. 
The soil of all the plats, however, contained an abundance of mois- 
ture, the moisture content being well above the wilting coefficient, 
which is approximately 17 per cent. The difference in the moisture 
content of the fourth foot on the same date, 2.2 ±1.2 per cent, was 
insignificant. On June 18, after the plats had been irrigated, the only 
significant differences were found in the second and third feet, where 
the soil of the fall-irrigated plats contained about 5 per cent more 
