January 20, 1905.] 



SCIENCE. 



89 



The production of absorption of heat 

 within a single unequally heated metal, the 

 calorimetrie aspect of the Thomson effect, 

 is, apparently, easily accounted for without 

 additional assumptions. Thus, according 

 to the theory already stated, the line CC 

 in Fig. 4 represents a case in which the 

 attractive pressure of the positive elec- 

 tricity is greater at the cold end than at 

 the warm end, while the attractive pres- 

 sure of the negative electricity is greater 

 at the warm end than at the cold end, of 

 metal 1. According, positive electricity 

 moving from the cold end to the warm end 

 of this metal will expand more, and there- 

 fore absorb more heat, than the mere rise 

 of temperature requires, while the negative 

 electricity in moving from hot to cold with- 

 in the same metal will contract less, and 

 therefore give out less heat, than the mere 

 fall of temperature requires. That is, to 

 use the conventional mode of expression, 

 the current absorbs heat where it flows 

 from cold to hot in metal 1. For the line 

 ir and the metal 2 the case is vice versa. 



The conception of electricity, each kind 

 of electricity, as acting within a metal like 

 a perfect gas seems very revolutionary to 

 one who has been strongly impressed by 

 Maxwell's discussion of the analogy which 

 the behavior of electricity in Faraday's 

 'ice-pail' experiment presents to the be- 

 havior of an 'incompressible fluid,' though 

 Maxwell in pointing out this analogy warns 

 us against being too much influenced by it. 



The ice-pail experiment, however, as I 

 understand it, proves merely the difficulty 

 in p\itting an appreciable excess of either 

 kind of electricity into a given space, a 

 difficulty which still exists after all the 

 assumptions of this paper are made. Con- 

 sider, for example, the difficulty of putting 

 any considerable excess of positive or of 

 negative ions into an electrolyte. Indeed, 

 the idea of the electric current within a 

 solid as consisting of two oppositely mov- 



ing perfect gases is so like the familiar and 

 commonly accepted idea of the current in 

 an electrolyte, where we apparently have 

 two oppositely moving bodies of ions, each 

 body obeying the gas law in its osmotic 

 pressure, that, instead of being troubled 

 by the heretical character of this view of 

 the current in a solid, I am somewhat con- 

 cerned lest I am failing to give due credit 

 to some one who has already proposed it. 

 Of course, Drude in his electron theory 

 does apply the gas laws in some particulars 

 to the electrons within metals, and I can 

 not be sure that he has not anticipated me 

 in much that is given in this paper, though 

 I did not, so far as I am aware, get from 

 him any of the main features of the theory 

 here proposed. 



The question naturally arises. Why not 

 determine the direction and magnitude of 

 the local electromotive forces of the thermo- 

 electric circuit, and so get a decisive trial 

 of the case between the ordinary and the 

 proposed view of thermo-electric action? 

 The reply is that physicists have been try- 

 ing for more than a hundred years to get 

 a satisfactory determination of a single 

 one of these local forces, the one measured 

 by the true contact difference of potential 

 between any two metals, and have, appa- 

 rently, not yet succeeded in the attempt. 

 It is the old question of the Volta effect. 

 Some months ago I was of the opinion that 

 Mr. John Brown, F.R.S., of Belfast, had 

 found a way of getting rid of the dis- 

 turbing effect of the medium surrounding 

 the two metals, zinc and copper in his case, 

 by heating them for several hours in a cer- 

 tain kind of oil. Considerable recent ex- 

 perience with various kinds of oil at the 

 Harvard Physical Laboratory has led me 

 quite unwillingly to the conclusion that the 

 kind of treatment to which Mr. Brown 

 subjects his metal plates may substitute 

 for the disturbing surface condition ac- 

 quired in air an equally baffling surface 



