6 BULLETIN 1438, TJ. S. DEPARTMENT OF AGRICULTURE 
a loss of 47.2 per cent and bale No. 5, a loss of 23.8 per cent. Bale 
No. 1 in the warehouse lost 0.4 per cent, or 2 pounds, because of dry- 
ing out in storage. 
Test No. 2 was conducted at Raleigh, N. C. (Table 2). The bales 
were set out on November 20, 1918, and opened for drying on June 
9, 1919. In this test bale No. 4 showed a loss of 43.5 per cent and 
bale No. 5, a loss of 19.2 per cent. 
Test No. 3 was conducted at Dallas, Tex. (Table 3), beginning 
December 23, 1919, and terminating August 3, 1920. The time 
covered in the test was somewhat greater than in former tests and 
the damage was greater. The proportion, however, is about the 
same. 
Test No. 4 was conducted at Raleigh, N. C. (Table 4), extending 
over a period from January 15, 1920, to August 24 of the same year. 
This test was started later than the test in Dallas, Tex., and con- 
tinued later. The only outstanding difference to be noted here was the 
serious damage to bale No. 4, the greatest loss so far noted. The 
Fig. 3.— Bales used in test No. 4 during reconditioning. The damaged cotton has been "picked" 
from all of the bales except No. 4. The damaged cotton or "pickings" have little commercial 
value 
damage to bale No. 7 was comparatively heavy. This is especially 
noticeable, since this bale absorbed comparatively little water 
(fig. 3). 
Test No. 5 was conducted at Jefferson, Ga. (Table 5) from January 
10, 1920, to August 26 of the same year. The outstanding feature of 
this test was the very severe damage to bale No. 4 — 370 pounds, or 
over 73 per cent of its original weight. The losses to bales 6 and 7 
were rather heavy, too, while No. 5 lost considerably less than most 
of the corresponding bales in other tests. 
Test No. 6 was conducted at Dunn, N. C. (Table 6), beginning on 
December 13, 1921, and ending on July 31, 1922. Through an error, 
there was no bale No. 6 in this test. There was an apparent gain in 
bales Nos. 2 and 3. This may be largely explained by the fact that 
the bales used were shipped from a comparatively dry area (Wills 
Point, Tex.), and the gain may represent moisture absorbed in the 
more humid climate of eastern North Carolina. 
The outstanding feature of this test was that there was no loss by 
bales 1,2, and 3, which was as it presumably should be, and the loss 
