ELECTRICAL CONDUCTIVITY OF COTTON 447 



to nearly dryness. These equations, with the respective ranges of 

 relative humidity (and, therefore, of moisture content) over which 

 each is significant, are given on page 432. 



It is concluded that exposure of cotton to high atmospheric humidity 

 causes a change in the gel structure due to absorption of moisture, 

 since the insulation resistance of the material as measured at some 

 comparable condition (75 per cent R.H. at 25° C.) is less after such 

 high humidity exposure than before, even if the cotton is well dried 

 before testing. 



The temperature of such exposure to high atmospheric humidity also 

 afifects the subsequent electrical properties of the cotton. Data to 

 show this temperature effect are given in Table IV. 



Temperature Effects 

 Effect of Temperature at High Humidity on I.R. of Air -dried Cotton 

 Table IV contains the results of a series of tests on the I.R. of 

 samples of raw and washed cotton which were exposed to several 

 cycles of high humidity and dry air, each cycle being as follows: 



(a) Equilibrated and measured at 75% R.H.— 25° C. 

 {h) Equilibrated and measured at 88% R.H .—at t° C. 

 (c) Dried for 16 hours with a stream of dry air at 25° C. 



This cycle was repeated four times, the only difference in each case 

 being the temperature (/° C.) at which the 88 per cent R.H. equilibrium 

 tests were made. These temperatures were successively — 22°, 30.2°, 

 38°, and 22° C. In all, eight samples of washed cotton and four 

 samples of raw cotton were used in the test. Two of the raw cotton 

 samples (1 and 4) were not exposed to the 88 per cent humidity con- 

 ditions, but were used as control samples to check the reproducibility 

 of the 75 per cent humidity conditions in each cycle.'" 



Table V is a condensation of Table IV. The decreases in insulation 



-•* Five measurements each were made on these two control samples during the 

 course of the test, giving a mean value of 4.52 kilomegohms, with a standard deviation 

 of but 0.13 kilomegohms. 



The differences in the initial values of I.R. for the eight washed samples are not 

 due to lack of control, either in the method of washing or in the method of testmg, 

 but to actual differences in the equilibrium moisture contents. For example — saniple 

 1 gave 73 kilomegohms initially, and sample 6 gave 102 kilomegohms. Their 

 respective moisture contents, under the test conditions, were 8.17% and 8.00 ,o- 



Using Equation II, and with the constant A = 10, the values of B were calculated 

 in this equation as 13.99 and 14.05 respectively for samples 1 and 6. Assuming these 

 samples to be of equal purity, since they were washed in an efficient manner,^' it 

 is reasonable to take B = 14.03 for both samples. From this value of B, the 

 I.R. of sample 1 was calculated at a moisture content of 8.00 per cent, giving 98 

 kilomegohms, a satisfactory check with sample 6 at the same moisture content. 



-1 Walker & Quell, Jour. Text. Inst. 24, T141, 1933. 



