10 



U. S. BUREAU OF FISHERIES. 



mate, I; petroleum product No. 1, J; a proprietary waterproofing 

 material, K; and the Dutch method, L; untreated white line. A, 

 was used as a control. Seven samples were prepared by each pre- 

 servative, one to hold as an unexposed check; the other six to be 

 exposed. 



These samples were lightly stretched on the roof of the Fishery 

 Products Laboratory, Washington, D. C, on November 25, 1921, 

 At the end of each succeeding month for six months one sample 

 treated by each preservative and one control sample were removed 

 and tested for tensile strength. The experiments made by others in 

 England, Norway, and Holland have usually lasted only about two 

 months. In the present instance it was desired to have a full test of 

 endurance, so the exposures were continued for six months. 



2 3 + 



MONTHS EXPOSED 



Fig. 1. — Tensile strength of cotton lines exposed to weather conditions at 

 Washington, D. C. 



TENSILE STRENGTH. 



This factor was measured on a tensile-strength machine at the 

 U. S. Bureau of Standards.^ The machine was automatic, motor- 

 driven, and recorded the tensile strength by means of a moving pen. 

 Because of variations in the line it is necessary to take the average 

 of a number of tests. The figure given for the control is the average 

 of 60 tests, all others of 15. Table 3 gives the results of the measure- 

 ments of tensUe strength for the lines exposed to weather conditions 

 at Washmgton, D. C. The same results are expressed graphically 

 in Figure 1, wherein the curves are smoothed by the method of moving 

 averages. 



8 See footnote 2, p. 1. 



