4 BULLETIN 1377, U. S. DEPARTMENT OF AGRICULTURE 
crop yields; the average yield without fertilizer and rotation (check- 
plot yield in continuous culture), the yield resulting from the use 
of fertilizer alone (yield from fertilized plot in continuous culture), 
the yield in rotation without fertilizers (check-plot yield in rotation), 
and the yield in rotation with fertilizers (yield from fertilized plot 
in rotation). 
Though this study calls for a consideration of the long-time soil- 
fertility experiments, yet the data of all such experiments can not 
be used, since in some instances, those at State College, Pa., for 
example, none of the crops which make up the rotation are grown 
in continuous culture (3). 
The soil is an important factor in determining the comparability 
of continuous-culture and rotation yields in each experiment. For- 
tunately, in the long-time fertility experiments involving yields in 
rotation and continuous culture, practically the same kind of soil is 
under test in each case. 
The seasonal effects on crop yields are well known and are usually 
given careful consideration when comparisons are made of crop 
yields. In some of the long-continued experiments the rotation plots 
are repeated as many times as it is necessary to give the yield of 
each crop each year; whereas in other experiments there is no repli- 
cation of the rotation plots. In the case of the former experiments, 
an average yield in continuous culture for a series of years is com- 
parable with the average yield of the same crops grown in rotation 
ior the same period of years. Such yields are comparable, since they 
include the same number of years and the same seasonal effects. On 
the other hand, in those experiments where the rotation plots are 
not repeated, an average yield for a series of successive years in 
continuous culture is not comparable with an average yield of the 
same crop grown in rotation for the same period, since the rotation 
average represents a less number of years and reflects a different 
combination of seasonal effects. In such cases comparisons between 
Particular yields in continuous culture and rotation may be made 
y taking the average yield of a crop grown in rotation and com- 
paring it with the average of the yields obtained during the same 
years in continuous culture. At Rothamsted, for example, wheat is 
grown continuously on the same land and also in a four-year rota- 
tion which is represented by only a single series of plots. Experi- 
mental data for 72 years include 18 wheat yields in rotation and 
72 yields in continuous culture. An average yield of the 18 crops 
grown in rotation is comparable with an average yield in contin- 
uous culture only when the latter average is obtained from the 18 
yields which have been obtained in continuous culture during the 
same years that wheat has been grown on the rotation plots. 
Inasmuch as this study involves a consideration of not only the 
effects that rotation and the use of fertilizers have on crop yields 
when acting independently of each other, but also their conjoint 
effects, as compared with their single effects, it is important that the 
crops in rotation be fertilized the same as in continuous culture. This 
raises the question as to the kind of chemical fertilizers that should 
be selected for comparable yields. 
In some of the long-time experiments are included the results from 
chemical fertilizers which carry one, two, or all three of the major, 
nutrient elements; that is, nitrogen, phosphorus, and potassium. In 
