38o 



Journal of Agricultural Research 



Vol. IV, No. s 



than 4 per cent, and between the mean of the sixth and the mean of the 

 seventh hundred there was a difference of 14.63 per cent. He says: ^ 



A result such as this from the first experiment shows that, even if this were an 

 exceptional case, it is at least necessary to test several more samples before deciding 

 to base any conclusions upon means of so few as 100 measurements. 



At the Wyoming Experiment Station as high as 10,000 fibers from one 

 sheep were tested for breaking strain and divided into groups of 5,000 

 each, on which Hill comments as follows : ^ 



Notwithstanding the fact that for most purposes the drawing and testing of 5,000 

 fibers of wool in order to determine the tensile strength of a sample would be highly 

 impracticable. Table III shows that where the samples were not carefully mixed 

 before drawing the sub-samples, the means of this number in two samples differ by 

 more than 5% when only two and three means are compared. 



NECESSITY FOR A METHOD OF TESTING WOOL 



On July I, 1908, a project in wool research was undertaken by Prof. 

 R. W. Clark, of the Montana Experiment Station. The object was to 

 study the effect of various factors — feeding, breeding, care, management, 

 etc. — upon the wool and form of the sheep. 



The first step necessary was to work out a method to test accurately 

 the qualities of the wool fibers. This has proved a long and tedious 

 task, owing to the excessive variation among the fibers. 



In the sample of Rambouillet wool, fibers grown side by side varied 

 in breaking stress from 20 to 140 dgm. Also the stretch or strain varied 

 from I to 15 mm. Hence, it is self-evident that a great many fibers 

 would have to be tested to reduce this variation to a minimum. 



Table I gives the results obtained by testing i ,000 fibers from the same 

 place on the body of the sheep for breaking stress. Five samples were 

 taken from each sheep. 



Table I.— Results of testing fibers of wool for breaking stress 



To explain the notation, let us consider for an illustration the shoulder 

 of sheep No. 324, which is given as 55.994 ±0.404. By this is meant that 

 the average breaking stress of the 1,000 fibers was 55.994 dgm., and the 

 probable variation for the 1,000 was 0.404 dgm. — i. e., the chance is 

 even that if another lot of i ,000 fibers were taken from the same place, the 



* Hill, J. A. Studies on strength and elasticity of the wool fiber, i. The probable error of the mean. 

 Wyo. Agr. Exp. Sta. 21st Ann. Rept., 1910/11, Sup., p. 16. 1911. 



2 Hill, J. A. The value of fiber-testing machines for measuring the strength and elasticity of wool. Wyo. 

 Agr. Exp. Sta. Bui. 92, p. 22. 1912. 



