FARM MANAGEMENT STUDY OF COTTON FARMS. 39 
COVER CROPS. 
Scarcely any attention has been given in the region to the utiliza- 
tion of winter cover crops, with the exception of oats and a very little 
wheat. The fact that cotton picking often extends well into the 
winter season accounts in part for this condition. 
It has been clearly demonstrated in other sections that a rotation 
of crops, even where no legumes are used, increases yields. Where 
legumes are included very marked increases are noted. For ex- 
ample, McNair, in Arkansas, found that where cotton followed cot- 
ton the average yield on 50 farms was 168 pounds per acre; follow- 
ing corn the yield was 197 pounds; following corn with cowpeas 
between the rows the yield was 251 pounds; and where cotton fol- 
lowed cowpeas (cut for hay) the yield was 270 pounds per acre. 
At 10 cents per pound the increased yield of cotton due to the pre- 
vious crop of cowpeas was $11.73 per acre. 
It is a well-known fact that the use of any cover crop tends to 
hold plant food in the soil, preventing its washing and leaching 
away. The use of legumes, such as vetch, crimson clover, bur clover, 
and others, actually adds the element nitrogen to the soil by taking 
it from the air. 
These legumes may be planted between the rows of cotton, as win- 
ter wheat is now frequently planted in the black-land region, as soon 
as the cotton is picked. In the case of early maturing cotton, with a 
warm fall, the legumes should make a sufficient growth by the time 
of spring plowing to be of immense value when turned under. It is 
believed to be advisable for farmers on these black lands to experi- 
ment with these winter legumes as a possible means of increasing 
the yield of cotton. 
FARM ORGANIZATION. 
It already has been shown that cotton holds the most, important 
place in the farming of Ellis County. A few farms in the county 
are dairy farms or diversified farms, but only one dairy farm and 
one diversified farm were found in this survey, and these were elimi- 
nated, since no reliable conclusions could be drawn from single in- 
stances. 
It is not always possible to ascertain whether these types of farm- 
ing found on a few farms in a locality could be successfully fol- 
lowed on farms generally in the region, though it is frequently pos- 
sible to say they would not be successful generally. A few cases will 
illustrate this point. In Anderson County, S. C, a single farm was 
found that was doing well with hogs. The owner kept several brood 
sows, raised two litters of pigs a year from each of them, and sold 
the pigs at weaning time to neighboring farmers who wanted them 
