- 34 - 
At 31 percent of the compresses and warehouses visited, after cutting 
the bagging only, samples were pulled from cotton bales by hand. This prac- 
tice was most prevalent in North and South Carolina, Virginia, Georgia, and 
Alabama. No compresses or warehouses employing this method were found in 
Texas, Louisiana, Arkansas, or Missouri. 
Ginning and baling practices are closely associated with sampling. 
The possibility of insufficient mixing of the load of seed cotton brought 
to the gin and of "plating" the bale with cotton from a previous load that 
is of different quality, are matters associated with ginning that sometimes 
affect materially the representativeness of samples. 
Information furnished by buyers interviewed throughout the Cotton Belt 
indicates that about 13 percent of the cotton marketed in the United States 
in 1932-33 was two-sided, and 92 percent of the managers of compresses and 
warehouses visited indicated that the top and bottom sides of flat bales 
can be identified. As a rule, the top and bottom sides of a compressed bale 
can be identified only when it is known that the flat bale was placed in the 
press with the top side up. 
Since buyers, as a rule, are unwilling to pay for cotton of two dif- 
ferent qualities within the bale a price above that which the lower quality 
commands, farmers should benefit from a method of drawing and rolling sam- 
ples that would indicate whether or not the lower quality represents the 
bulk of the ba^e or only a small part of it. Additional benefits would re- 
sult from developments in ginning by which plated or two-sided bales might 
be obviated. If the identity of the two sides of a compressed bale as well 
as of a flat bale could be known from time of ginning throughout its exist- 
ence as a bale, and if the rolling of samples were so standardized as to 
insure the identity of the two portions of the sample, the classer would 
be able to know, in all instances, which portion of the sample is from the 
top side, and therefore most representative of the bale. 
It has been shown that most cotton bales are plated to some extent. 
Under prescut conditions, it is necessary in cutting a sample that the cut 
into the bale be sufficiently deep to insure representativeness in the sam- 
ple; thai, is. that a thin plate will not unduly affect the representative- 
ness of the sample after it is trimmed. If the plate is thick, or if the 
sample is thin, adequate representation of the bale may not be obtained. 
Thick, uneven patches on cotton bales often prevent the drawing of 
a good sample. 
Cotton samples are usually trimmed to a depth that will remove bag- 
ging stain and weather stain. 
At one-half of the compresses and warehouses it was customary to roll 
the samploo from the top and the bottom of the bale with the inside surfaces 
together, separated only by the identifying coupon. At several, care was 
taken to roll the cc .ton from the bottom side of the bale on the outside of 
the combined sample Both of these practices are desirable and could well 
