apart. This dif- 
ference is 8.65 
times its probable 
error. The mean 
difference corres- 
ponds to an in- 
crease of 20.1 per 
cent in the spacing 
with three plants 
in hills 12 inches 
apart, whereas the 
difference be- 
tween the mean 
yields is 1.08 
pounds, corre- 
sponding to an in- 
crease of about 
Loss per cent: 
These two spac- 
ings differ widely 
in yield, and 
there appears to 
be considerable 
robbing by the 
hill plantings of 
the outside rows 
of the blocks 
with one plant 
every 12 inches. 
The actual differ- 
ence, however, 
is slight, and a 
comparison of 
mean row yields 
from inside rows 
with the mean 
yield of inside 
and outside rows 
shows that the 
plants in hills 
produced a larger 
crop with no 
serious detrimen- 
tal effect on the 
adjoining rows 
with wider spa@d 
plants. This also 
indicates that 
several stalks in 
a hill are more 
efficient producers 
of cotton under 
very dry condi- 
tions than single 
COTTON-SPACING EXPERIMENTS 
1) 
POLNIDS PER RO#” 
s 
aS 
PONS PLR COA” 
LPP!POUONOS PLR COA 
74 
SHANTS | PLANT FILAMMIS (PLANT FP PLANTS (PLAN 
felled [elWCh fl? NWCA l2 NCR fe 16h f2é 1 
Fig. 16.—Yields (in pounds) of seed cotton per row from compari- 
son of one plant every 12 inches with three plants in hills 12 inches 
apart, Greenville, Tex., 1923 
