PRESIBENT's address — SECTION A. 49 



liquid-liquid junction is due to the change of thermal capacity of the 

 substances there present, in other words, that the Peltier electro- 

 motive force itself is, at such junctions, the indirect result of chemical 

 action, "Whether this be true of the Peltier effect generally is quite 

 another question. For my part, 1 have inclined strongly to this view 

 ever since the publication of Roberts- Austen's discovery (q) of the 

 inter-diffusibility of solid metals in contact, though I admit that much 

 experimental work remains to be done before a definite pronouncement, 

 one way or the other, can be made. Should the magnitude of the 

 Peltier effect at a metallic junction be found to run a parallel course 

 with the change of thermal capacity during the alloying of the metals, 

 a good prima facie case would be made out for this view ; but at present, 

 so far as I am aware, the necessary information respecting the thermal 

 capacities of alloys is lacking. All the same, the smallness of the Peltier 

 effect at a zinc-copper junction, coupled with the approximate equality 

 of the specific heats of zinc, copper, and brass, looks rather like a hint 

 in this direction, (r) 



Richards, as I have said, does not make this comparison, but 

 utilises the new facts for an altogether different purpose. He suggests 

 that the difference between the observed values of the bound energy 

 and those of the changes in thermal capacity indicates a real difference 

 between the free energy and the energy of chemical affinity — a difference 

 which he ascribes to energy evolved during change of atomic magnitude. 

 As far as the thermodynamic theory is concerned, this is possible enough ; 

 it would simply necessitate the inclusion of an additional parameter in 

 the fundamental equations. But in view of Jahn and Gill's experiments 

 the matter takes on a different complexion, inasmuch as they show 

 that the bound energy is fully accounted for by reversible changes of a 

 purely thermal character. I find it difficult to associate the idea of such 

 a reversible thermal change with that of energy liberated by atomic 

 compression, just as I can attach no intelligible meaning to the tem- 

 perature of an atom or a molecule. It seems, moreover, fair to remark 

 that the only instances we have of the liberation of intra-atomic energy 

 — if indeed they be such — are those supplied to us by radio-activity, 

 and here we find the energy taking a form which is far removed from 

 that involved in reversible thermal change. It seems to me, therefore, 

 more reasonable to suppose that intermoleciilar changes, other than those 

 involved in the change of thermal capacity, go to make up the balance 

 of the Peltier effects than to cut the knot by ascribing them to intra- 

 atomic changes. Nor need we overlook a series of considerations which, 

 if correct, show that the foregoing portion of this paragraph is merely 

 a tilt at a visionary windmill. The outstanding difference between the 

 bound energy and the change of thermal capacity, on which Richards 

 rests his argument, may, after all, not be real, but merely due to the im- 



(q) Proc. Roy. Soc, 1900. 



(r) Lodge (Proc. Phys. Soc, Lond., XVII.) says that if you pass a current through 

 a zinc-copper junction you get no brass ; but this statement was made before 

 the interdiffusibility of solid metals was discovered. Bar oxidation films I 

 suspect that we shotdd find brass at the junction, whether a cui-rent pass 

 through it or not. 

 D 



