108 Dr. 0. V. Burton on a 



thought, but evidently a body of large bulk would be best. 

 Such a body (perhaps hollow) o£ suitable conductivity, well 

 jacketed with a material of the highest possible thermal and 

 electrical resistance,, and further enveloped in a double-walled 

 vessel containing ice, might be examined for a gradual rise 

 of temperature. Even if all laboratory experiments on these 

 lines should give negative results, it might still be held 

 possible that some part of the earth's heat was due to the 

 electromotive effect of compressional setherial waves. This 

 is only mentioned as a possibility suggested by the theory 

 of this paper, and not as tending to remove any existing 

 difficulty in the domain of geophysics. 



67. Some attempt having already been made in Appendix 

 A to trace the possible electromotive effects of a gravitational 

 field, the modifications which must be made in the equations 

 of § 17 above, when we wish to discriminate between positive 

 and negative electrons, will now be very briefly indicated. 

 With the notation explained in § 58, the source whose strength 

 is expressed by (10) must now be represented by the more 

 complete expression 



— (Hi/^ + Hj^) -=~vdxdydz, . . (85) 



which is to be equated to y*$dxdydz ; hence (11) becomes 



while (16) is replaced by 



^X 4-7T ~&t 



I x , . . . . (87) 



2 ^z' 



where 1^, as before, has the meaning indicated by (17). 



In place of (18) we have now two equations, expressing the 

 forces acting on positive and negative electrons respectively. 

 The force on the positive electrons in unit mass of matter 



the force on the negative electrons being similarly 



~ ^ 2 ^~ ^\ 4tt W Ix '' (89) 



