564 
Journol of Agricultural Research y 0 i. xxvm, No. 6 
In order to smooth the curve and reduce the probable errors of the means, 
they were computed for each successive three fruiting branches, taken as one 
array. Fruiting branches in Pima cotton are rarely retained at nodes lower 
than No. 9 and the numbers of bolls which matured on fruiting branches higher 
than No. 32 were too small to give reliable data. The mean fiber lengths for 
the 8 groups of fruiting branches between nodes 9 and 32, inclusive, are stated in 
Table II and the data are presented graphically in figure 1. 
E/3ER LENGTH /N MILLIMETERS 
Fig. 1.—Average length in millimeters of Pima cotton fiber from bolls borne at successive heights on the 
plant, the heights being indicated by the numbers of the nodes of the main stalk at which the fruiting 
branches are borne. 
Table II.— Mean length of Pima cotton fiber from bolls borne at successive heights 
on the plant , the heights being indicated by the numbers of the nodes of the main 
stalk at which the fruiting branches are borne 
Node 
numbers 
of the 
fruiting 
branches 
Number 
of deter¬ 
minations 
Mean length 
of fiber 
Node 
numbers 
of the 
fruiting 
branches 
Number 
of deter¬ 
minations 
Mean length 
of fiber 
9-11 
12-14 
15-17 
18-20 
Bolls <* 
10 
19 
11 
20 
Millimeters 
38.75±0. 85 
40. 46± . 36 
41. 27=L . 31 
41. 75± . 40 
21-23 
24-26 
27-29 
30-32 
Bolls & 
10 
19 
12 
17 
Millimeters 
42.20±0. 45 
42.41 ± .25 
42.89± . 40 
42.05± . 40 
o The absence of mature bolls on many of the fruiting branches at the time the collection was made 
accounts for the fact that the number of determinations is in all cases fewer than 30 (10 plants x 3 fruiting 
branches). 
Although the differences in mean fiber length as between any two successive 
groups of fruiting branches, are in no case significant, there is a steady increase 
from the lowest to the next to the highest group. The highest group (nodes 
30 to 32) shows a slight decrease in length of fiber as compared with the group 
just below (nodes 27 to 29). 
The fiber in the bolls borne on fruiting branches at nodes 9 to 14, from which 
a large part of the first picking probably is derived, is decidedly shorter than the 
fiber produced higher on the plant. Comparing the lower half of the fruiting 
zone (nodes 9 to 20) with the upper half (nodes 21 to 32) the means for fiber 
length are 40.75±0.28 and 42.37±0.19, respectively, the difference being 1.62± 
0.34. This difference evidently is significant, being nearly 5 times its probable 
error. 
