286 



BRIDGMAN. 



TABLE I. 



Tin. 



Thermo-electromotive Force, volts X 10" 



is the way in which they were obtained experimentally) they are 

 equally curious. At 25°, the effect is initially positive, passing through 

 a very flat maximum of about 0.005 X 10~^ volts at 2500 kg., and from 

 here on becomes rapidly negative, reaching — 0.185 X 10~® at 12000 

 kg. This curve is concave downwards. The approximate effect of 

 higher temperatures is merely to rotate this curve bodily, without 

 distortion, anticlockwise about the origin. At 50° the effect is through- 

 out positive, the curvature displaying itself as a maximum near 7000 

 kg., and at 75° and 100° the rotation has become great enough to 

 obliterate the maximum. 



The Peltier and Thomson heats deduced from these measurements 

 are shown in Figure 3 and Table II. Of course when the fundamental 

 data are so irregular there must be much uncertainty in the deriva- 

 tives, but I believe that the essential features of the situation, with 

 respect to changes of sign etc., are correctly given, and that the nu- 

 merical values of the Thomson heat are correct to within perhaps 50%. 

 Both Peltier and Thomson heats change sign in the range. At low 

 temperatures the effect of all pressures is to give a negative Peltier 

 heat; that is, heat is absorbed by the positive current in flowing 

 across the junction from compressed to uncompressed metal, but at 



