16 BULLETIN 901, U. S. DEPARTMENT OF AGRICULTURE. 
Possible differences in soil condition and natural yielding power 
may be largely eliminated by comparing the relative yielding power 
of the crops in the several rotations in succeeding periods. The 12- 
year period can be divided into three periods of four years each, cor- 
responding to the length of the rotations. When so studied it is 
found that the yields of all crops in rotations Nos. 14, 15, 16, and IT 
have been decreasing instead of increasing, as compared with the 
yields of the same crops in the corresponding fallow rotations, Nos. 
18 and 19. The possibility that the later seasons may have been 
relatively more favorable to bare fallow than the earlier ones might 
be advanced in explanation of the behavior of the first crop follow- 
ing fallow or green manure; but such an explanation could hardly 
account for the behavior of the corn following this crop, and cer- 
tainly not for the crop of wheat or oats which follows the corn and 
has two crops intervening between it and the fallow. 
An exception has been mentioned above. This is noted in the 
sweet-clover rotations, Nos. 31 and 32. In these rotations the total 
yield of corn, which is the second crop after the sweet clover is 
plowed under, has been increasing in comparison with the yield of 
corn in the other rotations of this series. 
Unfortunately, there is no rotation to determine what the effect 
would have been had the sweet clover been harvested for hay or seed 
instead of being plowed under. Rotations to test this have been in- 
corporated in the newer work on section 9, but are not yet advanced 
enough to furnish the desired data. 
As to the relative values of rye and peas for green manure, the 
evidence is somewhat contradictory. Rotation No. 14 with rye has 
yielded heavier than No. 16 with peas. In these rotations wheat fol- 
lows the green manure. The corn in rotation No. 15 with rye has 
outyielded the corn in No. 17 with peas, but the other crops have 
yielded more in No. 17. In these two rotations the green manure is 
followed by oats. The differences are small and probably well within 
the limits of experimental error. 
In view of the fact that in more humid sections increases are 
usually expected from the use of legumes as green manure, it might 
be fair to state that one of the most interesting results of these ex- 
periments is the failure of peas as green manure to increase yields in 
comparison with those obtained on either fallow or nonleguminous 
green manures. 
A result from these experiments more important than the differ- 
ences between green manures or fallow is that on disked corn ground 
the wheat has averaged 1.3 bushels per acre more and the oats 
4 bushels per acre more than the same crops on green manures and 
fallows. The corn following wheat in four rotations has averaged 
6.8 bushels of grain and 3,065 pounds of stover per acre, and follow- 
ing oats in four similar rotations it has averaged 6.9 bushels of grain 
and 3,407 pounds of stover per acre. 
