380 BRIDGMAN. 



General Survey of Results. 



The first impression given by the measurements under pressure 

 is of bewildering and perhaps hopeless complexity, but there are 

 nevertheless certain uniformities and normal types of behavior. Let 

 us for the first discussion disregard detail and take as a measure of the 

 effects the maximum e.m.f., Peltier heat, and Thomson heat listed in 

 the tables above. The normal effect of pressure on e.m.f. is positive; 

 there are only three out of 20 cases, manganin, Mg, and Co, in which 

 the effect is persistently negative to the highest pressures and tem- 

 peratures; and for only three other metals, Fe, Al, and Sn is the effect 

 negative over any part of the range. The range of values of maximum 

 e.m.f. is from +710 X lO-^ volts for Bi to -20.6 X lO'^ for Co. 

 Similarly, the normal pressure effect on Peltier heat is positive, that 

 is, heat is absorbed by the positive current in flowing from uncom- 

 pressed to compressed metal. There are only four cases of negative 

 Peltier heat under pressure, manganin, Sn, Mg, and Co. The range 

 of values is from 2960 X lO^^ for Bi to -90.5 X lO'^ for Co. The 

 Thomson heat under pressure is also normally positive, Co, Fe, and 

 Bi being the only negative cases. The range of values is from 

 220 X 10-8 for Zn to -280 X 10-* for Bi. 



When, however, we try to correlate the thermo-electric behavior 

 with other electrical properties of the metals we find complete lack of 

 connection. In table XLV the metals are arranged in a number of 

 columns, in each column the order being the order of relative magni- 

 tude of the property standing at the head of the column. In the first 

 place, the pressure effect on e.m.f. is not greatest for the most com- 

 pressible metals, as we might expect, but the effect is apparently 

 quite independent of the compressibility. Compare, for example, 

 the positions of Pb and Mg in the compressibility column with their 

 positions in the pressure e.m.f. column. It was pointed out in a 

 previous paper that the effect of pressure on resistance is approxi- 

 mately proportional to its effect on volume; this is shown in the 

 table by the similarity of the columns of compressibility and pressure 

 effect on resistance. There is accordingly very little connection 

 between the pressure effect on resistance and that on e.m.f. The 

 conclusion suggests itself that there is little connection between the 

 mechanism which determines the thermo-electric behavior of a metal 

 and its electrical conductivity. This is further borne out by the lack 

 of parallelism between the columns of pressure effect on resistance 

 with those of pressure effect on Thomson and Peltier heat. 



