40 BULLETIN 659, U. S. DEPARTMENT OF AGRICULTURE. 
to feed from table scraps and other waste materials. This farmer's 
business was successful because his neighbors ivere not keeping brood 
saws. 
In a survey of 270 farms at Monett, Mo., four dairy farms were 
found. They made good incomes. Three of them did so by supply- 
ing milk for the city of Monett. They occupied this field completely. 
The other made butter, and was successful only because the farm 
family was a very exceptional one. 
A 25-acre farm in South Carolina was devoted exclusively to hay 
growing. It was highly successful because the hay was sold at high 
prices to neighboring farms that did not produce hay. Such cases as 
these tend to make farm-management investigators conservative in 
recommending changes in farm practice. They feel safe in recom- 
mending a new system only where a considerable body of local ex- 
perience has grown up with it. However, such investigators fre- 
quently recommend experiments in a new direction. This is very 
different from recommending a new system of farming. 
There is no well-established system of rotation in the region. 
Cotton occupies so much of the land that extensive rotation is not 
possible without a marked change in the system of farming. It is 
generally the practice to plant corn on the better land of the farm, 
preferably on bottoms, if possible. Legumes are found on only a 
few farms. A few farmers appreciate the value of rotations, and 
these make some effort to have a grain crop appear on all crop land 
once in three years. It is desirable to plant cotton on land previously 
occupied by grain on account of the fact that the fungus disease 
known as " root-rot " is not so prevalent under this condition. This 
disease attacks cotton and legumes (taproot plants), but not grains. 
A study of this character does not give much information as to 
the value of diversified farming in a region like Ellis County. It 
is a study of farm experience, and local experience with diversified 
farming is so meager here that no conclusions can be drawn from it. 
Other types of farming might be better for the region than the one 
prevailing, but this can not be determined definitely in advance of 
experience, or at least extensive experiment, with a new system. 
This study does reveal much of value concerning the details of 
the prevailing system. Some of the lessons bearing on these details 
are given in what follows. 
CROPS. 
Cotton. — In order to arrive at some indication of the proportion 
of the crop area that should be in cotton to make the best returns 
with the system prevailing in this region, a tabulation was made 
using the percentage acreage in cotton as the basis. The results, 
