130 BRIDGMAN. 



resistance itself was measured at every point with enough accuracy 

 so that it would have been permissible to keep throughout the table 

 a constant number of decimal places, but the pressure itself at the 

 lower pressures is not known with a high enough accuracy to justify 

 keeping more significant figures than shown. 



When the logarithm of resistance is plotted against pressure a 

 nearly linear relation is found at all three temperatures. This means 



1 dR . . , 



that ^ ^ IS approxmiately constant at all pressures at constant 



temperature, where R is the instantaneous value of the resistance at 

 the pressure in question. The instantaneous pressure coefficient is a 

 function of temperature, however. The average value of the instan- 

 taneous coefficient between and 12000 kg. is — 0.000293 at 0°, 



- 0.000277 at 50°, and - 0.000250 at 100°. The deviation of the 

 logarithm from exact linearity changes sign with rising temperature. 

 At 0° and 50° the instantaneous coefficient becomes greater with 

 increasing pressure, which is not what one might expect, whereas 

 at 100° it becomes less. At the two lower temperatures the deviations 

 from linearity run smoothly with the pressure, but at 100° the varia- 

 tions, although much less numerically, show one or two points of 

 inflection with rising pressure. At 0° the initial value of the instan- 

 taneous coefficient is — 0.000200 and at 12000 kg. it has risen to 



— 0.000320; the corresponding values for 50° are — 0.000231 and 

 0.000290, and for 100° - 0.000262 and - 0.000249. 



The specific resistance was also determined. At 0° this was found 

 to be exactly 1.000 ohms per cm. cube. This is higher than the 

 value previously published for the other specimen, which was 0.711. 

 The effect of temperature on the new specimen is also greater than 

 on the previous one. The Aalues for this specimen are shown in the 

 table. The resistance decreases with increasing temperature, and 

 the effect is not linear, as of course it cannot be, for otherwise the 

 resistance would become zero at some finite temperature. The coeffi- 

 cient found for the other specimen was also negative, but smaller 

 numerically and within the temperature range, the relation was linear. 

 Previously the relative resistance was found to drop from 1.000 to 

 0.711 between 0° and 50°, whereas here the drop is found to be from 

 1.000 to 0.622 for the same temperature interval. In view of the 

 greater precautions in the preparation of this sample, there can be no 

 doubt but that the present values are to be preferred. 



It was considered of sufficient interest to measure the thermal 

 e.m.f. of this specimen of black phosphorus. The details of the 



