18 
BULLETIN 1018. U. 
DEPARTMENT OE AGRICULTURE. 
in plat 1. Following this week the plants in plats 1 and 2 were 
never so responsive to the influences which affect flowering as those 
in the other plats, and their curves dipped with considerable regu- 
larity after September 3 to a condition of almost complete cessation 
about September 30. The curves of plats 3 and 4 also dipped toward 
a hibernation period, but with greater fluctuation and a greater 
amount of flower production. 
It will be observed that during this period of gradual cessation 
in flowering the different plats held about the same relative position 
in regard to the moisture content of the soil as at the beginning of 
the flowering season, plat 2 having usually the second highest and 
the moisture, content of the other two plats having been almost con- 
stantly smaller than that of plats 1 and 2. 
Data given in Table VII indicate that some economic importance 
may be attached to such an occurrence, especially in seasons when 
the first killing frost of the autumn does not arrive earlier than 
is commonly expected for the vicinity. A summary of the data 
obtained by the Weather Bureau shows that the average date of 
the first killing frost at Phoenix, Ariz., during the years 1895 to 1920 
is December 3, or 33 days later than the date on which cotton was 
killed by the frost in 1919. 
Table VII. — Comparison of flowers produced and stem grotcth prior to August 
15, plants producing vegetative branches, bolls produced on 20 plants during 
September, and September bolls destroyed by frost on four plats of Pima 
cotton in 1919. 
Percentage of- 
Plat. 
Total flowers 
produced 
during first 
45 days be- 
ginning with 
Initial date 
Total stem 
growth pro- 
duced prior 
to August 
15. 
Plants 
producing 
vegetative 
branches. 1 
oi nowenng. 
No.l 
No.2 
No.3 
No. 4 :'. 
- . 
: 60.5±0.97 
58. 6± 1.01 
48. 9± 1.00 
46. 8± 1.08 
85.3 
80.4 
76.2 
76.9 
5 
40 
10 
10 
49±2.2 1 
60±2.3 • 
61±1.8 
59±2.1 
263 
215 
340 
264 
September 
bolls 
destroyed 
by frost. 
Bolls 
produced 
on 20 
plants 
during 
September. 
1 Most of the vegetative branches made only a rudimentary development. The flowers produced were 
so few as to have no material effect on the flowering curves of the final yield. 
From the data in Table VII it is apparent that plats 1 and 2 
would have produced considerably more cotton per plant despite 
the early frost had the later flowering continued at the same rate 
as that on plats 3 and 4. The relative production of both stem growth 
and bolls was considerably greater during the first 45 days of the 
flowering season on plats 1 and 2 than on plats 3 and 4. This may 
be attributed to the fact that plats 1 and 2 received their first irriga- 
tion 16 days earlier than plats 3 and 4. 
