4 BULLETIN 36, U. S. DEPABTMENT OF AGBICULTUEE. 
in the State, with the prices paid, would be prepared in time for exhi- 
bition at the Oklahoma State Fair in 1913. The samples were, un- 
fortunately, destroyed, by a fire which occurred in one of the rented 
buildings of the Department of Agriculture. 
The data resulting from this investigation are far too voluminous 
to be presented in a single publication, and a detailed discussion of 
staple-cotton production in Oklahoma and of the effect of varying 
lengths of staple on the price paid in different sections will not be 
undertaken at this time. It should, however, be distinctly under- 
stood that the facts have all been carefully scrutinized and that none 
of the variations in price which appear in this publication are due 
to variations in the length of staple of the individual bales considered. 
There are some markets in which the prices are rather uniformly 
higher than in certain other markets, the difference apparently being 
due to the reputation of particular compress points for shipping out 
cotton superior in staple or other quality to that shipped from other 
compress points, but every individual bale for which any premium 
was paid on account of its length of staple has been excluded from 
the tables and the discussions presented in this publication. 
METHOD OF SAMPLING. 
Because of the impossibility of starting this survey at the begin- 
ning of the cotton-marketing season and because so few people could 
be assigned exclusively to this work, no attempt was made to take 
the samples in such a way as to show the proportion of the various 
grades of cotton which were produced from this crop. It was 
realized that a very large number of the highest grading bales of the 
year's crop had been marketed before our work was undertaken and 
that a considerable number of bales, nearly all low grade, would be 
marketed after our force had returned to Washington. The plan 
followed was therefore to secure in each market, as nearly as pos- 
sible, samples from bales representing the extreme range of grades 
on the market at the time. Following this plan, if the cotton yard 
contained only a dozen bales of high-grade cotton and a dozen bales 
of low-grade cotton with about a hundred bales grading about mid- 
dling, the 15 or 20 bales which would be selected for sampling would 
probably include more bales above and below middling than of mid- 
dling cotton. For this reason the number of middling bales taken 
at any one place on any one date appears to be small, but in a great 
many cases where a dozen or more bales were sampled early in the 
season every bale would be found to grade above middling, while 
later in the season a similar sampling would show very few bales 
grading as high as middling. 
