ELECTRICAL RESISTANCE UNDER PRESSURE. 143 



The pressure coefficient of strontium is abnormal. At 0° the in- 

 stantaneous coefficient varies only little with increasing pressure, but 

 what change there is is a decrease, which is abnormal. The range 

 of values is from 0.04503 to 0.04492. At 50° the pressure coefficient 

 at first increases with rising pressure, which is what we have come to 

 regard as normal for this type of substances, but between 2000 and 

 3000 kg. passes through a flat maximum, and from there on decreases. 

 The range is from 0.04469 to 0.04451. At 100° the behavior is like 

 that at 50° except that the maximum with pressure is very much more 

 pronounced, and the maximum occurs at 7000 kg. At 100° the initial 

 value of the instantaneous coefhcient is 0.04351, the maximum at 7000 

 is O.O4452, and at 12000 kg. it has dropped to 0.04432. There is, 

 however, nothing abnormal in the temperature coefficient. The 

 average temperature coefficients between 0° and 100° are 0.00383, 

 0.00311, and 0.00275 at 0, 6000, and 12000 kg. respectively. In 

 respect therefore to the variation of pressure coefficient with tempera- 

 ture and temperature coefficient with pressure strontium is like the 

 other metals with positive coefficient, but the variation of pressure 

 coefficient with pressure is like that of the others over only a part of 

 the range. It is to be remarked that the absolute value of the pres- 

 sure coefficient of strontium is much higher than that of any other 

 metal. 



Summarizing, the behavior of the five abnormal metals, with the 

 exception of the pressure variation of the pressure coefficient of stron- 

 tium, is alike in that the instantaneous pressure coefficient increases 

 with rising pressure and decreases with rising temperature, and the 

 temperature coefficient falls with rising pressure. 



Carbon, in the form of graphite, is the only other element at present 

 known with a positive pressure coefficient of resistance. Since it is 

 not metallic, comparisons are unprofitable. Furthermore, it was 

 not possible to obtain results that were numerically reproducible. 

 It may be worth mentioning, howcAcr, that graphite is like the metals 

 above in that the pressure coefficient decreases with increasing tem- 

 perature, but that it is different in that the pressure coefficient is very 

 much less at the higher pressures. ' 



Normal Liquids. The only liquid metal previously measured was 

 mercury. It was found for it that the instantaneous pressure coeffi- 

 cient decreases with rising pressure and increases with rising tempera- 

 ture, and that the temperature coefficient decreases with rising 

 pressure and rising temperature. The beha\^o^ is in all respects 

 that which appeals to us as normal. It is worth while to give the 



