FLOWERS AND BOLLS OF PIMA AND ACALA COTTON 7 
from about 5 to 25 days — a period double that of Pima. Very much 
less difference was found between the two varieties in 1925 in respect 
to the period in the age of squares in which shedding was most likely 
to take place, but a slightly longer period seems indicated in the Acala, 
as shown in Table 3. Shedding of younger squares is apparent in the 
Acala, while old squares shed at about the same rate in both varieties. 
DETERMINING PERIODS OF SQUARE AND BOLL DEVELOPMENT 
The earliness of a variety of cotton depends upon two factors — 
(1) the time at which the young plants begin to produce squares 
and (2) the length of time required for these squares to reach the 
flowering stage and for the flowers to develop into open bolls. The 
square and boll " periods," as the intervals between square and flower 
and between flower and open boll are called, have been determined 
for many varieties under different conditions, but generally by con- 
sidering squares and bolls without regard to their location on the 
plants. (See under ''Literature cited," 3, 10, 12, 14, and 17.) But it 
appears that squares or bolls of the same date may differ in the 
length of their period of development, depending on whether they are 
borne on the basal nodes of the fruiting branches or farther out on 
the branches, as shown by the following data. 
DEVELOPMENT OF SQUARES ON SPECIFIC NODES OF FRUITING 
BRANCHES 
The period required for squares to develop into flowers on specific 
nodes of the fruiting branches of Pima and Acala cotton was deter- 
mined in successive 10-day intervals in 1924 and 1925 from 20 plants 
in each variety in an effort to ascertain whether the squares on the 
outer nodes of the fruiting branches required a longer period than 
those that were developing at the same time on basal nodes, closer 
to the main axis of the plant. The periods for squares on specific 
nodes and the number of squares upon which these periods were 
based are shown in Table 4. Because of a fault in the method of 
recording notes in 1924, it was possible to compute the periods for 
only those squares that later developed into open bolls. In 1925, 
however, all squares that reached the flowering stage were used in 
computing the square periods. 
Instances are seen in some 10-day intervals, particularly in 1924, 
where the square period of certain nodes has not been presented. 
This was because these nodes produced too few squares to allow the 
period to be determined. 
In both years the square periods for the nodes in intervals in which 
all four nodes are represented show that squares classed as borne 
on the fourth-to-last nodes in most cases required a longer time to 
flower than squares produced at the same time on nodes 1 or 2, and 
usually longer than those on node 3. The general tendency is 
apparent from the table, and the probable errors for the square 
periods show significant differences toward a lengthening of the 
square periods on the outer nodes in many cases, as with Pima in 
1924; for node 3 compared with node 2 in the interval from July 6 
to 15; for node 4 compared with nodes 1 and 2 in the interval from 
July 16 to 25; for node 4 when compared with nodes 1, 2, and 3 in 
the interval from August 5 to 14. 
