COTTON FIBER SCIENCE — PALMER 477 



generations that, by this rather simple — and sometimes mystery- 

 shrouded — process known as "classing," the more skillful manufac- 

 turers have been able generally to select the cottons needed to pro- 

 duce their customary fabrics, some of them of amazingly fine quality. 

 But, classing is a himian hand-and-eye operation and at its best leaves 

 much to be desired. In the normal order of things the process can- 

 not be exact ; cotton quality by its nature is too complex, too intricate 

 and too involved with variations of its individual fibers and of its 

 combinations of individual fibers to lend itself fully to simple methods 

 of measurement or evaluation. Too often, consequently, the accuracy 

 of classing or the reliability of quality descriptions become the sub- 

 ject of dispute. Too often, as manufacturers of the newer industrial 

 textiles have had to meet buyers' rigid specifications, classing has 

 failed to assure them the right cotton for their exacting work. Too 

 often the seed-breeder, seeking to set a goal for the improvement of 

 his strains, has been left groping for guidance. 



Half a century ago studies were begun in this country to establish 

 fixed standards of cotton grades and staples which, under govern- 

 ment authentication, were intended to be universally accepted and 

 applied throughout the industry. In a series of legislative enact- 

 ments between 1914 and 1923, quality standardization and identifica- 

 tion were made a responsibility, first of the Bureau of Markets and 

 then of its successor agency, the Bureau of Agricultural Economics. 

 The first official standards were established in 1914. In 1923, the 

 official United States grade standards were accepted throughout the 

 world as Universal Standards for American cotton. Over the years 

 much has been accomplished by means of these official standards in 

 unifying and stabilizing the concepts of quality employed by cotton 

 classers. But standards created by classers have all the shortcomings 

 of the classing art itself, and even the stamp of the Government of the 

 United States has not always proved a sufficient -fiat to protect them 

 from a challenge of their unifonnity or of their constancy — two of 

 the basic requisites of a standard of any kind. This was a problem 

 that began to vex and embarrass the Department almost from the 

 time the preparation of standards was first undertaken. Some early 

 researches had been started in an effort to find solid ground — some 

 definite relationships to known and accepted constants — on which firm 

 specifications of standard qualities could be rested, but these had been 

 abandoned when they appeared to be bogging down in a morass of 

 confusing and irreconcilable results. Controlled spinning tests had 

 previously been undertaken under direction of such masters as Fred 

 Taylor and Drayton Earle, followed by the late W. G. Blair and H. H. 

 Willis, and with the indispensable cooperation of North Carolina 

 State College and Clemson College had progressed well. But spin- 



