304 SECTIONAL TRANSACTIONS.— A. 



region does not exceed o • 0005°. We investigated the influence of the crystal 

 lattice on grey and white tin, which diff"er only in this respect : grey tin does 

 not show supra-conductivity, white tin does. Gold-bismuth alloys show 

 the same influence — the alloy becomes supra-conductive though neither of 

 the components do ; but X-ray experiments showed that it has a crystal 

 lattice of its own. 



Investigations of the thermal conductivity of supra-conductors showed 

 an influence of the supra-conductive state. At the transition point indium 

 shows a small sudden increase of thermal conductivity. When the supra- 

 conductivity is disturbed by a magnetic field the thermal conductivity is 

 increased for pure metals. The results for PbTlg are very complicated, 

 probably as a result of the lack of homogeneity of the alloy. The specific 

 heat of tin increases when the metal becomes supra-conductive. In a 

 magnetic field, high enough to disturb supra-conductivity, this increase 

 disappears. 



Prof. O. W. Richardson, F.R.S. 



There is one point on which I should like to hear the opinions of 

 Prof. McLennan and Prof, de Haas before this meeting closes. It concerns 

 the views of Dorfman to which Prof. McLennan referred. These go 

 further than the relations between the frequency and the magnetic field 

 necessary to destroy the superconductivity which have been mentioned. 



There is some resemblance, even though it may be only superficial or 

 accidental, between superconductivity and ferromagnetism. Following this 

 idea, Keesom and his associates at Leiden measured the specific heat of 

 superconductors in the neighbourhood of the critical point, where one might 

 expect an abnormality similar to the abnormality in the specific heats of 

 ferromagnetic substances in the neighbourhood of the Curie point ; but 

 no such effect could be detected. This, however, is not entirely conclusive. 

 The number of electrons concerned in the superconductive phenomenon 

 might be too small a fraction of the total number, or of the number of atoms 

 present, to exert any appreciable influence on the specific heat, or, alterna- 

 tively, there might be some compensating effect on the atoms which might 

 counterbalance any changes in the specific heat of the whole substance 

 arising from changes in the energy of the electrons. 



Dorfman has pointed out that what is in some respects a more direct test 

 of this particular issue can be made if the specific heat of electricity (Thomson 

 effect) in the superconductive region of temperature is considered. The 

 magnitude of this can be deduced from the thermoelectric measurements 

 of Keesom and his associates which refer to lead and tin. These show that 

 there is such an abnormality in the Thomson effect. It is true that it does 

 not occur exactly at the superconductive critical temperature. For example, 

 in the case of lead this critical temperature is 7-2° K. ; whereas the anomaly 

 in the Thomson effect sets in at about 5° K. and rises to a maximum at a 

 little over 10° K., after which it falls. This anomaly is quite similar to the 

 corresponding anomaly in the case of ferromagnetic substances near the 

 Curie point. 



If it is admitted that this anomaly in the Thomson effect is associated with 

 the establishment of superconductivity, it is a natural inference that it is a 

 result of the change in the energy of an electron connected with this pheno- 

 menon. On this basis the thermoelectric data enable the difference AWo 

 between the energy of a superconducting and a non-superconducting electron 

 to be estimated. The interesting fact then emerges that, approximately, 



AWo = !^Ho = Avo 



