252 



BELL SYSTEM TECHNICAL JOURNAL 



evidence to show that several hours are required for them to come 

 to complete equilibrium. A temperature of 100 degrees Fahrenheit 

 and relative humidity of 75 per cent has been found suitable for 

 cotton testing and 100 degrees Fahrenheit and 87 per cent relative 

 humidity for silk. 



The number of turns of yarn or thread wound around the electrodes 

 will vary with the size of the thread since the winding space is fixed 

 and a single layer of thread is applied. For No. 30/2 cotton approxi- 

 mately 90 turns, 180 parallel threads, have been found to give satis- 

 factory readings. This same space accommodates about 256 turns, 

 512 parallel threads, of No. 62/1 spun silk. 



Fig. 7 — Humidity cabinet disassembled. 



The distance between the electrodes is not particularly critical. 

 That is, it is not important, for example, whether the distance is 

 Ys in. or % in. It is important, however, that having decided upon 

 a certain separation, say % in., this separation be accurately main- 

 tained for all electrodes if the readings are to be comparative. Of 

 course if the separation is too great, an unreasonable number of turns 

 of textile is required to bring the resistance of the sample within the 

 range of the galvanometer. On the other hand, if the separation is 

 too small, the error due to variation in separation for different sets of 

 electrodes increases in magnitude. 



In actual practise, ^/i in. separation with a winding space of 2 in. 

 accommodating, as mentioned above, about 90 turns of No. 30/2 



