10 BULLETIN 1365, TJ. S. DEPARTMENT OF AGRICULTURE 
for Pima in both seasons, whereas the Acala squares required but 28 
days in 1924 and 29 days in 1925, or 5 days less for the Acala in the 
first year and 4 days less in the following year. These differences 
between the two varieties remained constant throughout the season 
in both years, as reference to Table 4 will show. 
In respect to Pima cotton grown at Sacaton, Ariz., Martin, Ballard, 
and Simpson (17) reported that from their data there was "no signifi- 
cant evidence that the square period is longer because the square is 
produced at the outer nodes, toward the end of the branch." They 
state further that "this is evident when the periods for squares of 
first nodes are compared with periods for squares of fifth nodes that 
appeared on the same dates," but no data were presented to illustrate 
this point. The mean square periods given by them for Pima and 
Acala grown at Sacaton in 1921 were considerably shorter than were 
found in either 1924 or 1925, and squares of Pima were shown to 
require somewhat over 7 days longer to reach the flowering stage than 
did Acala, whereas the difference between these varieties in 1924 and 
1925 amounted to only 5 and 4 days, respectively. 
SQUARE AND BOLL RECORDS FOR SPECIFIC NODES OF FRUITING 
BRANCHES BORNE BY MAIN STALKS 
The numbers of squares produced, the numbers that were shed, and 
the numbers that developed into mature bolls on specific nodes of the 
fruiting branches of the main stalks of 20 Pima plants in 1924 and 1925 
are presented in Table 6. Squares or bolls borne beyond node 6 of the 
fruiting branches were too few in number for separate consideration 
and were included in the table under the heading of "nodes 6 to last." 
A class was included in the table presenting data for the last or outer 
node of all fruiting branches of two or more nodes, and another class 
was included giving the data for fruiting branches of one node only, 
such fruiting branches not being included in the figures given under 
node 1 in the table. 
As shown by the totals in the table, the percentage of square shed- 
ding was very low on node 1 and was at least twice as great on the 
succeeding nodes. In 1924 nodes 4 and 5 had a lower percentage of 
square shedding than did node 3, but exceptionally heavy shedding 
took place on nodes 6 to last. No explanation for the decrease of 
shedding on nodes 4 and 5 was. found. A consistent increase in shed- 
ding of squares occurred from node 1 to nodes 6 to last in 1925, indi- 
cating what appears to be normal behavior for Pima. Squares pro- 
duced on the terminal node of fruiting branches of two or more nodes 
shed at the rate of 69 per cent in 1924 and 95 per cent in 1925; the 
increase was perhaps connected with the fact that the 1925 plants 
were producing a larger crop of bolls. Branches of one node only, 
which are normally produced only at the bottom and top of the 
plants, shed nearly half of their squares, in contrast with the first 
nodes of the longer fruiting branches, which shed less than one-fifth 
of their squares. The percentage of square shedding of these single- 
noded fruiting branches was intermediate between that for node 2 
and node 3 of the longer branches. 
The proportion of bolls matured on specific nodes is seen to have 
been highest on node 1 and to have decreased more or less rapidly to 
