620 PROCEEDINGS OF THE AMERICAN ACADEMY. 



It is this second condition which has been specially studied by 

 previous experimenters, and it has been due to the measurements made 

 while the contacts between the electrodes and the specimen were in 

 the condition above described that the negative coefficient of resistance 

 for heat has been attributed to silver sulphide. Increasing temperature 

 of the specimen or increasing difference of potential at its terminals 

 simply promotes the ionization at the imperfect contact or contacts and 

 permits a greater flow of current. 



Consideration of Condition III. The condition which I have called 

 the third is that in which fairly good connections have been established 

 between both electrodes and the specimen, and is characterized by an 

 absence of noise in the telephone, with the current passing through 

 the specimen in either direction. The figures given in Table IV. are 

 those obtained from continued tests of the sample, the apparent resis- 

 tances of which are given in Table III. after it had passed into the 

 third condition. 



TABLE IV. 



