408 
Journal of Agricultural Research 
Vol. XI, No. 9 
— 0.394±0.075. This is nearly six times the probable error, and thus 
it is statistically significant. The possibility that height and yield of 
adjacent rows was one and the same thing is dependent upon the question 
as to the relation of height and yield for the various crops. The correla¬ 
tion between height and yield for the 57 varieties of barley is 4- 0.214 ± 
0.085. 
The coefficient as obtained for spring wheat, Farm Crops test, shows 
some competition between adjacent rows due to yield. 
Perhaps the most striking correlation coefficient is that for winter 
wheat, which is +0.315±0.099. As has already been mentioned, 
minus (—) coefficients show competitive effect. Positive (+) coeffi¬ 
cients might be explained as due to soil heterogeneity. 
ACTUAL EFFECT OF HEIGHT OF ADJACENT ROWS ON BORDER ROWS OF 
SAME VARIETY 
The correlation coefficients as calculated show whether there is any 
competition between rod rows of small grains when the rows are spaced 
at a distance of 1 foot apart. The results as computed show that when 
varieties and strains of small grains are grown in rod rows there is 
an effect due to the height of adjacent rows in the case of barley, oats, 
and winter wheat in the Plant Breeding test and no effect for the spring- 
wheat test. The correlation coefficients for the oat and winter-wheat 
results show only a small effect, the coefficients being a little less than 
three and two times the probable errors, respectively. 
In the Plant Breeding barley test the effect is a large one, and it seems 
worth while to demonstrate the actual effect of certain height differ¬ 
ences. In considering these results certain facts will be repeated. The 
height of each plot was measured in centimeters, and there was no appre¬ 
ciable difference in height between the border and central rows in each 
plot. If there is any effect on border rows of the same variety due to 
differences in the height of adjacent rows one might expect, on the aver¬ 
age, that the border row which grew near a short variety might yield 
more than a row of the same sort which happened to be near a tall 
variety. If this is so, the result is considered as a positive, or plus (+) 
difference. The average increase, in bushels per acre, is also given. 
Negative results are preceded by a minus (—) sign. The results are 
given in 3-cm. classes for the difference in height of the neighboring plot 
(Table VI). 
The results in Table VI show that the actual effect on rows of the same 
variety due to the height of adjacent rows is of the utmost importance in 
rod row tests under conditions such as obtained for the barley in the 
Plant Breeding tests. 
