- 35 - 
be incorporated in any plan for standardized rolling of samples. They facil- 
itate rolling and tend to give greatest protection to that portion of the 
sample that is likely to most nearly represent the bale. The classer would 
thus know that the portion of the sample rolled on the outside is the one 
that was taken from the bottom of the bale and may be affected by plating. 
When samples are to be used again, they should be handled carefully 
and in such a way that the two portions of the sample are not changed from 
their original position in the sample. 
The present method of wrapping samples in individual papers for cer- 
tification is the best yet devised for this purpose, but this method is not 
often employed in wrapping other samples. Individual papers for single sam- 
ples are useful for preserving their identity, but the cost of such papers 
is often prohibitive. Wrapping several samples carefully and completely in 
paper, excluding light and air insofar as possible, is another good method 
of preserving their representativeness. Probably the next best way is to 
put the samples in a sack. Using open boxes and baskets is less desirable. 
Although managers of compresses and warehouses were usually familiar 
with only the method of sampling that was employed locally, cotton buyers 
were of the opinion that method of sampling affects cotton classing. Al- 
though only 3 percent of the 161 cotton buyers stated that method of sam- 
pling affects staple-length determination, 25 percent stated that it affects 
grade determination, and 65 percent stated that it has a psychological ef- 
fect on classing through difference in appearance. 
The hand-pulled sample is the smallest and thinnest of the four types, 
and it is drawn with the greatest difficulty, particularly from high-density 
bales. 
The sample pulled by hook is rough in appearance with irregular edges, 
As a rule, some of the trash is missing from this type of sample, and the 
layers of fibers have been disturbed. This sample may be smaller than either 
of the cut samples, as the cotton cannot be pulled loose close to the bands. 
The rumpling effect of the hook is not so evident when the cotton sampled is 
longer than 1-1/8 inches in staple, as such cotton is naturally somewhat 
rough in appearance. 
The sample cut on one edge only can be taken conveniently from high- 
density bales only when the bands are properly spaced. Although there is 
usually plenty of room between bands on flat bales for drawing a sample by 
making the crescent-shaped cut, there seems to be little reason for prefer- 
ring this kind of sample to the plugged sample — it does not roll so easily 
and is more difficult to pack and store. 
The results of this study indicate that plugging is the most desir- 
able method of drawing samples from high-density bales. The plugged sample 
can be wider than other samples, for the edges can be cut close to the bands. 
Such samples are more attractive in appearance and are more compact. They 
