G18 PROCEEDINGS OF THE AMERICAN ACADEMY. 



The existence of these burning noises shows the presence of an im- 

 perfect connection between the specimen and one or both of its elec- 

 trodes and that arcing is taking place at one or both of these places. 

 An examination of all specimens tested has shown that violent action 

 of some kind ha,s taken place at the point of connection between the 

 sulphide pellet and the electrode connected with the positive pole of 

 the battery ; the surface of the specimen has been blackened and some 

 of the fine wires of the electrode have been burned away. On the other 

 hand, the negative ends of the specimens and the electrodes by means 

 of which they have been connected with the negative pole of the battery 

 have been invariably bright and clean. 



The probable explanation of the physical action which takes place 

 during the first condition is that despite the close mechanical associa- 

 tion of the electrodes and the specimen there is no electrical connection, 

 and an arc must take place between one or both electrodes and the 

 specimen before a current can pass. To produce such an arc requires 

 ionization either by heat or by a sufficient difference of potential at the 

 point or points of discontinuity. It appears from all tests that this 

 action takes place most readily at the negative electrode of the speci- 

 men, and that gradually a more or less perfect connection is established 

 at this point. When this action takes place the excessive apparent 

 resistance of the specimen is broken down, and the specimen may be 

 considered as passing from the first into the second condition. In 

 support of this explanation of this phenomenon, the following tests may 

 be described. 



The direction of the current through a specimen, the resistance of 

 which had been broken down, was reversed. The noises in the tele- 

 phone which previously had been present were no longer apparent, and 

 after a short time the changes in resistance characteristic of the second 

 condition ceased to exist and the specimen passed into the third con- 

 dition. Again, a specimen which was in the third condition had one 

 of its electrodes remured and, after the surface of the specimen had 

 been carefully cleaned where the electrode had previously been, it was 

 clamped under great pressure in the jaws of a steel vise with soft copper 

 surfaces. The figures obtained in this test are given below. 



From these figures it will be seen that a perfect connection had 

 been established between the original terminal and the sulphide, but 

 that a good connection had not been made at the terminal which had 

 been clamped in the vise. During the test, while the current was pass- 

 ing from the vise to the sulphide, the noise in the telephone was very 

 great, but when the current was passing through the imperfect connec- 

 tion in the opposite direction no noises could be detected. After the 



