18 BULLETIN 1094, U. S. DEPARTMENT OF AGRICULTURE. 
the’ summer until seeding time. The replowing is necessary to 
destroy the scattered deep-rooted weeds that may escape the surface- 
cultivating implements and to correct a fineness and compactness — 
of surface that results from continued surface cultivation. Under 
this system a bare cultivated surface is maintained for about 14 © 
months preceding seeding. 
In 1913 an experiment was started to determine to what extent 
this fallow period might be shortened and the expense of the fallow 
reduced without reducing yields, or, more correctly speaking, to 
determine the effect on yields of reducing the length of the bare- 
cultivated period and the amount of work expended on the fallow. 
Four pairs of plats were set aside to be alternately cropped and fal- 
lowed. On the pair H—I the method above described was employed. 
These plats, therefore, duplicate the pair C-D from which results 
have been given. On the pair J—K the cultivated period is reduced 
by delaying the first plowing, which is the first cultivation, until 
late fall. This goes through the winter rough, as left by the plow, 
unless cultivation is necessary in winter to prevent blowing. The 
summer cultivation and replowing are the same as in the pair H-I. 
The plat to be fallowed in the pair L—M goes through the summer, 
fall, and winter in stubble. Its first cultivation is plowing in spring 
before vegetation makes very heavy growth. This usually has been 
in the last half of April. In 1915 this was replowed July 24; in 1916, 
August 1; and in 1918, June 13. By error it was plowed in the fall 
of 1913. In the summer of 1914 it was replowed August 20. In the 
pair N-O the bare-cultivation period is still further shortened by de- 
laying the first cultivation until in June, when it is plowed. The 
actual dates of plowing have ranged from June 7 to June 24. 
The yields from plats fallowed by these methods and others to be 
outlined are given in Table 7. The 7-year averages for the period 
1914 to 1920 are 21.1 bushels from H-I, 21.6 bushels from J-K, 
21.8 bushels from L—M, and 23.6 bushels from N-O. The first three 
methods run very close together, not only in the average but in each 
of the years. The last method was inferior in the drought of 1917 
and decidedly superior in 1919. In 1919 those methods that had 
the rankest early growth and at one time the highest potential yield 
later suffered from lodging, storm damage, and fungous diseases 
and finally produced the lightest yield. This tendency, however, 
is not an unusual one at the Fort Hays branch station, and the 
performance record of methods that avoid it should not be discounted. 
One year later, the experiment in methods of fallow was extended 
to include two more methods. The fallow plat in the pair P-Q is 
listed instead of plowed in the fall. The time of listing agrees in 
general with the time of plowing in the pair J-K. During the 
spring it is cultivated level, and it is plowed in June at the same time 
| 
