116 Mr. T. 0. Mendenhall on the Resistance of 



model of that described iu the report referred to above, the 

 writer undertook a short time ago to investigate the quantita- 

 tive relation between pressure and resistance for the carbon 

 disk which belonged to it. In a series of preliminary experi- 

 ments, the use of the toothed wheel and screw, by means of 

 which the pressure is communicated to the disk, was found to 

 be extremely objectionable on account of the impossibility of 

 exactly reproducing a given pressure. This portion of the 

 instrument was therefore entirely removed, and an arrange- 

 ment made by means of which any definite pressure might be 

 quickly brought to bear upon the disk or removed from it. A 

 slender brass rod was placed in a vertical position upon the 

 centre of the upper contact piece, the upper end of which 

 rested lightly in a small conical cavity made on the underside 

 of the scale-pan of a balance. The weight was suspended 

 above by a fine thread passing over a pulley; so that by raising 

 or lowering it, the pressure was applied or removed us was 

 desired. The carbon disk was made one of the branches of a 

 Wheatstone's bridge, as described by Mr. Edison. In lowering 

 the weight, care was taken to make the movement slow enough 

 to avoid any shock to the disk. When the apparatus stood 

 with the weight lifted, the adjustment of the galvanometer to 

 the zero was made without any difficulty, the resistance of the 

 disk appearing to be quite constant. When the pressure was 

 applied, however, the adjustment became very troublesome ; 

 and after a few trials it was discovered that time was a very 

 important element in the problem. The addition of a pressure 

 of fifty grams reduced the resistance to nearly one fourth of 

 what it was in its normal condition'instantly; but it was found 

 that the minimum was not reached at once. The resistance 

 continued to fall during the first two or three minutes with 

 considerable rapidity, and after that more slowly. A series of 

 experiments was accordingly undertaken for the investigation 

 of this phenomenon. After a number of trials, the bridge was 

 adjusted so that when the key was closed simultaneously with 

 the application of the pressure the needle of the galvanometer 

 would remain momentarily at zero; for the instantaneous effect 

 of this pressure seemed to be quite constant. In a few seconds, 

 however, the needle began to move, showing that the resist- 

 ance was diminishing. With this constantly decreasing resist- 

 ance it was, of course, difficult to obtain balances which were 

 very accurate; but generally one could be obtained within a 

 minute after the application of the pressure, and another a 

 minute or two later, and so on. The operation was repeated 

 many times, and a number of points for the curve shown 



