624 BRIDGMAN. 



All of the electrical properties of tellurium are unusual, and are 

 very variable; it is seldom that two observers obtain the same numeri- 

 cal value for any of its properties. The explanation for this seems to 

 be that tellurium under ordinary conditions contains two modifica- 

 tions in unstable equilibrium, and the proportions of the two varieties 

 changes greatly with the manner of treatment. The temperature 

 coefficient of the same piece may be positive or negative according to 

 the thermal treatment. The pressure coefficient of tellurium has been 

 very recently measured by Beckman ^ who used small rods of tellu- 

 rium cast in glass. His specimens were subjected to various thermal 

 treatments, and it was possible to find a connection between the pres- 

 sure coefficient and the manner of treatment. The specific resistance 

 of his samples varied from 0.0493 to 0.617 ohms per cm. cube. The 

 pressure coefficient varied linearly with the resistance from —9.1 X 10~^. 

 to-26.6X10-^ 



The specific resistance of the specimen used above was much lower 

 than any found by Beckman, being only 0.00645 ohms per cm. cube. 

 The initial pressure coefficient was —0.00012, which is included in 

 Beckmann's range. The average temperature coefficient between 24° 

 and 0° was —0.0063. This is larger numerically than any of the nega- 

 tive coefficients found by Beckman. It is evident therefore, that the 

 properties of the extruded metal are quite different from those of the 

 cast metal. 



Bismuth. Runs were made on three dift'erent samples of varying 

 purity. It is known that the specific resistance and temperature 

 coefficient of bismuth are very sensitive to small impurities; a frac- 

 tion of a per cent of lead or tin may change the temperature coefficient 

 from positive to negative and increase the specific resistance several 

 fold. Minute impurities also introduce complicated hysteresis effects. 

 The samples used in this work were Kahlbaum's ordinary grade, 

 Kahlbaum's "K" grade, and electrolytic bismuth of my own prepara- 

 tion. The purification of bismuth is a matter of some difficulty; it is 

 known that no chemical method will suffice. I used the method of 

 electrolysis from the solution of the fluo-silicate described by Foerster 

 and Schwabe.^° It is essential to keep the current density low and to 

 continually stir the bath. 



The wire, which was 0.013 inch in diameter, was formed in each case 

 by extrusion when warm. It was wound loosely on a bone core and 

 connected to the leads with "fine" soft solder. At 100° the purest 



20 F. Foerster und E. Schwabe, ZS. Elektrochem. 16, 279 (1910). 



