THERMO-ELECTRIC QUALITY UNDER PRESSURE. 301 



According to Colien,^ lead also is in unstable internal equilibrium, and 

 has allotropic modifications. No evidence whatever of anything of 

 this kind was found above. If there had been, it should certainly 

 have manifested itself as a change of zero. The maximum departure 

 of any single observed point from linearity was 1.7% of the maximum 

 effect, and the average numerical departure of all the observed points 

 was 0.25%. Since no readjustment whatever was necessary in passing 

 from the curves at constant temperature to those at constant pressure, 

 the results may be accepted with considerable confidence. 



The numerical results are shown in Tables VII and VIII and Figures 

 8 and 9. The thermal e.m.f . rises regularly with pressure and tempera- 

 ture, as do also the Peltier heat and the Thomson heat. All three of 

 these are positive. The fact that there is an appreciable Thomson 

 heat between the compressed and the uncompressed metal has an 

 interesting bearing on the question whether there is a Thomson heat 

 at atmospheric pressure. It is a matter of rather general experience 

 that zero, or atmospheric, pressure, is no preferred pressure, but the 

 properties of a substance change without discontinuity on going from 

 negative to positive pressure. This is in accord with a point of view, 

 fruitful in many situations, that a substance at atmospheric pressure 

 is actually under a high internal pressure, and that changes in external 

 pressure produce the same effect as proportionally small changes in 

 total pressure (internal plus external). The fact, therefore, that 

 compressed lead has a Thomson heat against uncompressed lead 

 makes it probable that the Thomson heat of uncompressed lead is 

 not rigorously and exactly zero but is merely too small to be measured 

 conveniently. 



Wagner is the only previous observer on lead; he gives + 5.6 X 

 10"i2 volts per degree per kg. for 0° to 100° and to 300 kg. The 

 results above for the same range would be interpolated as + 5.2 X 

 10"^-; the agreement is satisfactory. 



Zinc. This was Kahlbaum's "K" grade, from the same rod as 

 the resistance specimen, but it was not the identical piece of wire. 

 It was extruded to 0.020 inch diameter at a temperature of 330°, and 

 was annealed in an electric oven for several hours at 120°. 



At atmospheric pressure its thermo-electric behavior against lead 

 is given by the formulas: 



E = (3.047 ^-0.00495 <2) X lO"^*^ volts, 



P = (3.047 -0.0099 (t + 273) X lO"" volts, 



a = -0.0099(/ -h 273) X lO'^ volts/°C. 



9 E. Cohen and W. D. Helderman, Proc. Amst. Acad 17, 822-828, 1914. 



