6 BULLETIN 1180, U. S. DEPARTMENT OF AGRICULTURE. 
pounds of NH 3 , P 2 5 , and K 2 for the results of 1920 and 1921. 
All mixtures were prepared the first year at the plats and used within 
10 days, but the following two years the preliminary work was done 
at the Washington laboratory and the mixture shipped to Muscle 
Shoals. In the latter case from three to eight weeks elapsed from 
the time of mixing until the time of application. 
The methods used in preparing the soil, application of fertilizers, 
seeding, and cultivation were essentially the same as those followed 
by the farmers of the community. With the exception of one ex- 
periment, where side applications were made, the fertilizers for 
cotton and corn were applied in the row just prior to seeding. The 
materials were in close proximity to the seed but never in direct 
contact. Both corn and cotton were planted in rows 4 feet apart 
with three rows per plat. The corn was thinned to 30 to 36 inches 
in the row and the cotton to 15 to 24 inches. Harvesting was done 
in the usual manner, the cotton being weighed when picked, while 
the corn was allowed to dry in the shock for four to eight weeks 
before husking and weighing. 
The methods of fertilizing, seeding, and harvesting the winter 
grains are considered in connection with the discussion of those 
experiments. 
OBSERVATIONS DURING GROWTH. 
The effects of the different fertilizer mixtures on plant growth 
under the various conditions were intimately related to the weather 
conditions prevailing, particularly the rainfall. A severe drought 
subsequent to planting sometimes resulted in marked injuries due to 
the high concentration of soluble salts. A drought at the time of 
maturity often decreased the grain yields, resulting in a yield of 
stalks considerably out of proportion to the grain production. The 
rainfall records of the United States Weather Bureau at Florence, 
Ala., located at a distance of 3 to 6 miles from the experimental 
fields, are shown in Table 2. 
Table 2. — Rainfall at Florence, Ala., during the 8-year period from 1914 to 1921, inclusive. 
Rainfall records (inches). 
Month. 
Average, 
1914 to 
1918. 
1919 
1920 
1921 
17.37 
3.56 
3.60 
5.90 
3.87 
14.67 
23.70 
6.96 
2.72 
1.54 
5.24 
22.08 
30.41 
5.70 
6.10 
3.32 
9.93 
12.93 
20.73 
1.14 
4.26 
July 
6.74 
3.78 
14.42 
Total 
48.97 
62.24 
68.39 
51.07 
It will be noted from these figures that the total rainfall for the 
three experimental years, 1919 to 1921, was higher in every case 
than the average for the five preceding years. The total of 68.39 
inches for 1920 was higher than for several years previous. How- 
ever, these total figures are of much less significance than the rain- 
fall of the growing season. The year 1919 was quite wet during 
the early spring and dry during the growing period, with adequate 
