124 
Journal of Agricultural Research 
Vol. XXI, No. 3 
last-named, in which all of the presumably beneficial treatments were 
combined. 
One other feature of these two-year rotations remains to be considered. 
It is a group in which the land is spring plowed or disked immediately pre¬ 
ceding the cotton crop. The four rotations included in this group lie in 
the south half of series B-6 and have all shown serious injury from rootrot. 
In the first of these the land is plowed only once in two years, and that 
immediately following the cotton crop. After the com crop the land is 
disked and kept fallow by shallow tillage until the cotton crop is planted. 
The mean annual rootrot injury has been 12.4 per cent. In the next two 
rotations the land, though plowed twice in two years, is left undisturbed 
after the corn crop until just before planting time, when it is plowed, and 
one of the rotations is also subsoiled at the time of plowing. The rootrot 
injury on these rotations has been 26.3 per cent for the subsoiled rotation 
and 31.9 per cent for the one that is not subsoiled. In the fourth rotation 
in this group the land is disked after the corn crop and sown to rye for 
green manure which is plowed under in the spring just before the cotton 
is planted. The annual rootrot injury in this rotation is 11.2 per cent. 
While it seems inadvisable to make direct comparisons between these ro¬ 
tations and other cotton-corn rotations such as B-6a, because of the differ¬ 
ence in field location, it is clear that none of these treatments has checked 
rootrot injury. 
As a summary of these results from the two-year rotations, it may be 
said that the extent of rootrot injury appears not to have been influenced 
either by different crop sequences or by different tillage methods or by the 
use of barnyard manure or green manure crops. 
THREE:-YEAR AND FOUR-YEAR ROTATIONS 
On the black lands of Texas where cotton is the leading field crop it is 
hardly practicable to consider the use of a rotation which does not involve 
cotton as frequently as every third year, or at most, every fourth year. 
If a period of two years or three years without cotton is not sufficient to 
secure a rotation effect, then it is clear that rotation effect must be dis¬ 
pensed with. As indicated above, the San Antonio rotation experiment 
includes four three-year rotations and two four-year rotations. The de¬ 
tailed record of rootrot injury in these six rotations is shown in Table IV. 
Two of the three-year rotations and both of the four-year rotations were 
started in 1913. 
While four of the six rotations included in Table IV show very slight 
rootrot injury, the other two—that is, one three-year and one four-year— 
show an extent of injury comparable with the plots cropped to cotton 
continuously. In other words, these results do not justify the hope that a 
relatively long period without cotton leaves the land less likely to show 
rootrot injury when cotton is again planted. It is true that these longer 
rotations differ from each other in certain respects, but these differences 
