COTTON-SPACING EXPERIMENTS 8) 
There is considerable variation in row yields for the different spac- 
ings, the outside rows of adjoining blocks of different spacings showing 
-some wide differences. Comparing the row yields from the different 
spacings shown graphically in Figure 3, the variation of the border 
rows is very evident. In every case where a close-spaced block 
adjoins a wider spaced block the outside row of the closer spaced blocks 
shows an increase in yield. 
Comparing the sum of the yields of the two inside rows of the 
unthinned blocks with the inside rowsof the 12-inchspacing, this spac- 
ing shows higher yields in three out of four comparisons. In three 
comparisons of unthinned rows with 6-inch spacing the 6-inch spaced 
blocks have given slightly higher yields. Comparing the 6-inch and 
12-inch spacing, the 12-inch has exceeded the 6-inch in three out of 
four comparisons. Comparing yields of the two inside rows of 
unthinned blocks with 12-inch spacing, in pounds of seed cotton, there 
is a Mean increase of 0.90+0.102 in favor of the inside rows of the 
12-inch spacing. This difference is 8.8 times its probable error and 
may be regarded as significant. Making a similar comparison of the 
6-inch with the 12-inch spacing, there is a mean difference in the yield 
of 0.475+0.216 in favor of the 12-inch spacing. This difference is 
only 2.2 times its probable error and is not significant. As only three 
comparisons of block yields are available for the 6-inch spacing with 
the unthinned rows, probable error was not figured for them. There 
are indications, however, that the 6-inch spacing was the more desir- 
able. The increase of the 12-inch spacing over the unthinned was 
equal to 10.4 per cent. 
The row yields and block yields in this experiment indicate that 
the 6-inch and 12-inch spacings were a more desirable distance than 
was obtained in the unthinned rows and that the advantage in pro- 
duction appears to have been in favor of the wider spacing. This is 
the only experiment made during the 5-year period covered by this 
report in which the 12-inch spacing produced significantly more cot- 
ton than a spacing of less than 12 inches. 
EXPERIMENT WITH PLANTINGS IN HILLS 
In the experiment with hill plantings, hills with several plants were | 
compared with hills thinned to one plant. The seeds were dropped 
by hand in hills an average distance of 21 inches apart. The test 
consists of three 4-row blocks, of which the two outer ones were 
thinned to one stalk to the hill and the inner block left unthinned. 
The rows which were not thinned averaged 214 plants to the hill. 
(Table 4 and fig. 4.) — . 
This experiment was harvested in one picking. The yields from 
the two inside rows with several stalks to the hill were consistently 
higher than from the inside rows of the blocks thinned to one plant 
to the hill. This test indicates that more than one plant at 21 inches 
is needed if good yields are to be expected. The inside rows from 
the blocks with several stalks to the hill produced about 22 per cent 
more cotton than the inside rows of the blocks thinned to one plant 
to the hill. Each row with several stalks to the hill produced more 
cotton than the thinned rows. 
The development and fruiting habits of plants in rows with several 
stalks to the hill as compared with one stalk to the hill are shown in 
Plate 1, A and B. 
23911—27}-——2 
