222 



O. U. VONWILLER. 



contact resistances are as important as with other materials 

 possessing the rectifying property, and whether the actual 

 conductivity of the substance does vary with the potential 

 gradient. 



Accordingly experiments with molybdenite were carried 

 out on the lines employed with other materials, special 

 attention being given to results obtained with very small 

 currents. This material is of a nature which allows speci- 

 mens suitable for these tests to be easily prepared. A thin 

 strip of molybdenite was cut, the dimensions in most of the 

 trials being within the following limits: length 2 to 4 cms., 

 width 2 to 5 rams., thickness 0*3 to 1 mm.; contact was 

 made at four regions along the strip, A, B, C and'D, spaced 

 at approximately equal intervals, by means of narrow strips 

 of thin copper sheet bent over the molybdenite, as shown 

 in figure 1, and clamped so as to touch it above and below. 



Fig. 1. 

 Copper wires soldered or otherwise connected with each 

 copper strip served as means of conveying the current into 

 aud out of the molybdenite and allowed any two of the four 

 points to be connected with a potentiometer so that the 

 potential difference between them could be determined. 

 The current entered and left by the end terminals A and D 

 and was measured by determining the potential difference 

 between the terminals of a coil of known resistance, usually 

 10 ohms, in series with the molybdenite. 



The tests consisted in sending currents of various mag- 

 nitudes through the preparation first in one direction and 



