Modified Theory of Gravitation. 107 



the value of the conductivity being then given by c = s/47rV 2 ; 

 so that the specific resistance would be 47rV 2 /s=l , 4 x 10 11 in 

 absolute measure, which is equivalent to 140 ohms as the 

 resistance of a centimetre cube. If the slab were of much 

 higher or much lower conductivity, the dissipation of energy 

 (corresponding to the same set of assumptions) would of 

 course be much less. 



65. Thus with the special assumptions of § 59 the rate of 

 development of heat, through the action of the primary waves 

 on matter, though excessively minute compared with the 

 rate at which we have supposed energy to be propagated in 

 those waves, is nevertheless, at its maximum, much greater 

 than could be reconciled with experience. But in obtaining 

 the estimates of the previous paragraph, it was assumed 

 that the " extrusion " of the aether per unit mass of matter * 

 was entirely due to the negative electrons, and that the 

 inertia of matter was almost entirely due to the positive 

 electrons. Unless we were to suppose the extrusions due to 

 positive and negative electrons to be of opposite signs, this 

 would be the assumption corresponding with the greatest 

 electromotive effects from the action of the primary waves on 

 atomic matter, and with the greatest consequent dissipation 

 of energy ; the inference being that, if we are to retain our 

 tabulated estimates, or any others not wholly different from 

 these in order of magnitude, we must adopt some more 

 favourable assumption as to the relative " extrusions " and 

 masses of positive and negative electrons. Now it is evident 

 from (71) that if the "extrusions" of positive and nega- 

 tive electrons were proportional to their masses (that is 

 F 1 = F = F 2 ), the electromotive effects of the primary waves 

 would vanish. In that case there would be no heating of 

 matter through dissipation of the energy of the primary 

 disturbance, nor would there be any corresponding production 

 of electromagnetic waves. Experience suggests, then, that 

 the relation 



F 1 = F = F 2 (84) 



is exactly or very nearly fulfilled. 



66. If indeed the aether is traversed by waves of other 

 than electromagnetic type, to the penetration of which no 

 sort of matter forms a barrier, it would be interesting to test 

 for any residual heating effect due to the interaction of such 

 waves with atomic matter. To devise the most sensitive 

 arrangement within given limits of size wound need some 



* Cf. § 7. 



