- 30 - 
Occasionally a sampler will draw a sample from the top side of the 
bale, Irim it, and divide it by layers, mixing the layers several times so 
that the inside and outside surfaces of the sample cannot be distinguished. 
n the sample drawn from the bottom side of the bale has been given the 
same sort of treatment, the two samples are placed together with the coupon 
between them. This practice of mixing layers is decidedly undesirable, for 
there is a possibility that the bale is thickly plated and that the portion 
of the plate drawn will be so placed as to unduly affect classing. In most 
cases it is possible to identify the inside surface of a sample before it is 
trimmed, unless it has been pulled by hand and is not very deep. Other means 
of identification are found in the bagging stain or trash en the outside 
surface and the comparative shortness, narrowness, and roughness of the in- 
side surface. As shown in table 7, employees at compresses are usually able 
to distinguish the inside surfaces cf cotton samples from the outside surfaces, 
but are less able to distinguish the portion of the sample taken from the top 
of the bale from that taken from the bottom of the bale. 
It is well to remember that the sample from the bottom of the bale 
may be composed, wholly or in part, of cotton from the previous bale ginned, 
as is not the case with the sample from the top side of the bale. If, as 
previously suggested, some means of permanently identifying the top and bot- 
tom sides of bales should be adopted, standardized practice with respect 
to rolling would enable the classer to know in every instance, which portion 
of the sample is from the bottom of the bale. The practice of rolling the 
sample from the bottom side of the bale (when the bottom can be identified) 
on the outside of the combined sample would aid the classer in determining 
which portion of the sample is most representative of the bale. This deter- 
mination would be further facilitated by the practice of placing the inside 
surfaces of the two portions of the sample together (with the identifying 
tag between them) . The results of this study indicate that this is the pre- 
dominant method of rolling samples at compresses and warehouses, presumably 
because the outside surface is in most instances larger than the inside sur- 
face. This gives special protection to the inside surface, which is prob- 
ably mest representative of the major portion of the cotton in the bale. 
H andling , Packing, and Care of Samples .- Cotton samples should be 
handled carefully after they are drawn in order to prevent loss of sand or 
other foreign matter therefrom or other changes that may affect their rep- 
resentative character. In some localities samples are left on top of the 
bales fcr varying lengths of time before they are wrapped, whereas in other 
localities it is customary to wrap samples as soon as they are cut. The 
practice of leaving samples on the top of the bales after they are cut is 
not a good one, for the humidity of the sample may be changed, and the color 
also may be changed to some extent by bleaching. 
Samples taken from cotton that is to be certificated are rolled and 
wrapped in individual papers, which helps to keep the humidity and the color 
of the cotton as nearly constant as possible throughout the sample. At some 
of the inland compresses and warehouses in each of the States except Alabama, 
Arkansas, Tennessee, and Virginia, individual papers were used for the sam- 
ples from each bale. In most instances, when a single method is used, the 
