- 23 - 
through a tag or marker placed at a fixed point on the bale at the time of 
ginning, to remain in that position throughout the existence of the bale. 
Table 3.- Comparative ability to identify top and bottom sides of 
flat, standard-density, and high-density bales of cotton 
at compresses and warehouses visited, by States, 1933 
State 
Compresses 
and 
warehouses 
visited 
Number 
148 
Proportion of total compresses handling 
specified types of bales at which top and 
bottom of such bales could be identified 
Flat 
Standard- 
density 
High- 
density 
Total 
Percent 
91.9 
Percent 
63.5 
Percent 
60.7 
Alabama 
Arkansas 
Georgia 
13 
12 
23 
14 
2 
12 
11 
6 
46 
2 
100.0 
91.7 
100.0 
92.9 
100.0 
100.0 
66.7 
72.7 
100.0 
95.7 
50.0 
77.8 
90.9 
20.0 
16.7 
100.0 
50.0 
33.3 
66.7 
71.8 
77.8 
87.5 
28.6 
Louisiana 
Mississippi 
Missouri 
21.4 
100.0 
North Carolina 
South Carolina 
Tennessee 
100.0 
33.3 
Texas 
73.2 
Virginia 
Drawing th e Sample .- Although many of the compress and warehouse man- 
agers interviewed seemed to think that the method of sampling employed lo- 
cally was the only method used to any extent in the United States, there 
was a consensus of opinion among buyers that a poorly drawn sample affects 
the classification of the bale, and many buyers stated that classification 
varies with the method of drawing the sample (table 4). 
Only about 3 percent of the buyers were of the opinion that the method 
of sampling affects staple-length classification. There is reason to believe 
that seme buyers who answered in the negative did not take into account the 
fact that variations in classing may result from differences in the degree 
to which different samples pass through the plate. These variations may 
result from hurried stapling of samples inconspicuously plated with cotton 
of the same grade as the remainder of the bale but of different staple length. 
Of all buyers interviewed, 25 percent stated that the method of sam- 
pling affects grade, and nearly 65 percent stated that it affects the appear- 
ance of the sample. The outstanding fact brought out by a comparison of 
the data shown in tables 1 and 4 is that more than 86 percent of the buy- 
ers interviewed in Texas, Louisiana, and Alabama, where the more attrac- 
tive plugged and cut samples are in use, thought that the method of sampling 
