6 BULLETIN 595, U. S. DEPARTMENT OF AGRICULTURE. 
ceding the seeding of the wheat. In some cases it is necessary to 
plow the land more than once during the period, in order either to 
maintain a surface receptive to water and resistant to blowing or to 
prevent the growthof weeds. As practiced, this is an expensive sys- 
tem of production. Experiments are under way to determine the 
most economical method of summer tillage. The indications are 
that a less intensive method than that practiced in the work here 
reported will give practically as good returns. 
The yields given in these tables begin with the second year of crop 
production at each station. The first year’s crop is produced on land 
uniform in its treatment. 
Where an entire crop has been lost by hail or other agency that 
could not possibly be overcome by cultivation, the year is not counted 
in computing averages. Such failures must, of course, enter into 
the final results of agricultural endeavor. They are, however, of 
such uncertain occurrence that the series of years here considered is 
too short to permit an attempt to establish their normal frequency 
for any locality. This is in effect what would be done by including 
them in averages. It is believed that less error is introduced by 
recognizing their occurrence and excluding them from averages. 
When the’loss of a crop is due to conditions that might possibly have 
been overcome by cultural practices, a zero yield for that year is 
included in the calculations. 
Embodying the basic data given in Tables II, III, and IV, the sec- 
ond part of the table for each station has been compiled. In this 
are brought together in summary form the yields detailed in the first 
part of such table. The value of the average yields thus obtained 
is calculated and given, together with a computation of the cost of 
production. The last line of the table gives the profit or loss result- 
ing from the production of winter wheat by the method stated. 
Loss is indicated by the minus sign. In this second part of each 
table there are two general headings, ‘Tillage treatment”? and 
“Previous crop.” Under the first heading the plats are grouped en- 
tirely by treatment without considering the previous crop. Under 
the second heading treatment is not considered, and the grouping is 
entirely governed by the crop immediately preceding the winter wheat. 
This really makes two tables combined in one, with subdivisions 
common to both. (See Tables V to XVIII.) 
Some of the rotations are calculated to conserve or increase the 
fertility of the soil, while others may perhaps deplete it. In the 
present stage of the work, the effects of rotations as units are greatly 
overshadowed by the effects of the cropping and cultivation of a 
single year. This is due to the fact that the controllable factors are 
the water supply, the physical condition of the seed bed, and a cer- 
tain recognized, if not understood, effect of the crop immediately 
preceding. Uniformity in these factors is largely restored by the 
