604 NATURE NOVEMBER 21, 1942, VOL. 150
Superconductivity and Liquid Helium II phenomena . Indeed the shape of these curves is similar; they both start with a finite slope at the WHEN discussing our observations on the transfer temperature at which the first anomalous behaviour of liquid helium II along solid surfaces above the is evident, and they both become temperature in• liquid level', we pointed out that this mechanism of dependent near absolute zero. surface flow might also be responsible for the 'trans• (6) When developing the thermodynamics of a port phenomena' in the bulk liquid (high heat con• superconductor with the current density rather than duction, low viscosity and the fountain effect)•. the magnetic field as the variable of state, one arrives Subsequent experiments by ourselves 3, Allen and at a formalism which also fits well the observations Rookie' and by Kapitza• strongly support this in liquid helimn II. While the rigorous derivation assumption. It appears that the transfer film of this cannot be given here, it appears that R• observed above the liquid level also extends below it, assumes the same significance in determining the covering all solid surfaces in contact with liquid free energy of liquid helium II as the square of the helium II, and it is evidently in this film that the threshold field has for superconductors•. anomalous transport phenomena take place. (7) One would expect that in such a model the It has been suggested by F. London6 that the specific heat of the system of superconductive elec• )..phenomenon in helimn may be caused by the trons as well as that of the 'unexcited' helimn atoms existence of helium atoms of very low (or zero) would be zero. This seems actually to be the case. thermal energy, and the mechano-caloric effect The facts that a super-current in a temperature found by us• strengthens this view. This effect shows gradient produces no Thomson heat8 and that no that by flow along a solid surface, such thermally heat transport exists in a helium film in a direction 'unexcited' atoms can be separated to a certain ex• opposite to that of the fiow 1 strongly suggest that tent from the bulk liquid. no heat is transferred to the particles while they In elaborating our hypothesis we have now come remain in the anomalous state. We ascribe the ob• across a striking analogy between the phenomenology served high specific heat of superconductive metals of liquid helium II and superconductivity, which and liquid helimn II to the energy taken up in lifting seems to go much farther than a superficial similarity. particles from the lowest state into thermal equili• While we have a good deal of information on the brimn with the rest of the substance. superconductive state, our knowledge concerning The question naturally arises what physical liquid helium II is very limited. A certain amount of significance has to be given to this striking analogy. generalization has thus been necessary in the formula• It seems to us that the fundamental phenomenon is tion of our thesis, which therefore must be considered the passing at finite temperatures of a number of as of an approximate nature only. The main points particles into the lowest quantum state, where no of the analogy are as follows : energy exchange can take place between them and (I) As it is impossible to establish a difference of the remainder of the substance. This new form of electrical potential at the ends of a superconductor, aggregation of matter evidently follows quite general so it is impossible to establish a temperature difference ru1es, so that the fact that the particles are electrons along the transfer film of helium II. in one case and atoms in the other is only of secondary (2) In both cases the frictionless transport of importance. The phenomenon of frictionless trans• particles (electrons or atoms) breaks down as soon port confined, as it is in both cases, to the geometrical as a certain value of flow is exceeded. This limiting surface, might be explained by the fact that, as the value is given by the current threshold in super• particles are free from interaction with the bulk of conductors, J, and by the rate of transfer, R, in the substance, they can only be accelerated by ex• helimn1• ternal forces. The range to which these external (3) In both cases this limiting value of flow is forces can p enetrate the substance limits the layer solely dependent on temperature and is independent in which the transport takes place. Thus, in a super• of the length of the path the particles have to travel. conductor, electrons can only be moved within the It is directly proportional to the width of the con• depth of penetration of an external magnetic field•, necting surface (this holds rigidly for plane surfaces, while the flow in a helium film is limited by the whereas for curved surfaces the limiting rate of flow extent of the van der Waals' forces of the container. will be reduced in both cases). It appears from these considerations that the (4) In liquid helium as well as in superconductors, theoretical interpretations of the superconductive the frictionless transport seems to be confined to the state and of the ).-phenomenon in liquid helium II geometrical surface of the substance. Kapitza6 has should be similar in their fundamental aspect, and demonstrated that an anomalously high heat con• it is with regard to such a generalized theoretical duction does not take place in the bulk of helimn II ; treatment that we hope the analogy pointed out in that is, there is no frictionless transport of atoms this note will be of value. unless this can take place along a solid - liquid J. G. DAUNT. boundary surface. An analogy t o this exists in a Clarendon Laboratory, K. MENDELSSOHN. superconductor, since it seems theoretically im• Oxford. possible to remove an electric charge from inside a Nov. 5. completely closed superconducting shell except by 1 Daunt and Mendelssohn, Proc. Roy. Soc., A, 170, 423 (1939). the process of normal conduction. 1 See, for example, Darrow, Rev. Mod. Phya.J.. 12, 257 (1940), and (5) It appears that in both cases the maximum Jones, "Reports on Progress In Physics", o, 280 (1940). cw·rent density of flow on the surface (J, R) indicates 1 Daunt and NATURE, 143, 719 (1939). the number of particles which can at each tempera• • Allen and Iteekle, NATURE, 144, 475 (1939). ture be transported frictionlessly (provided one 'Kapitza, J. Phya. U.S.S.R.,
© 1942 Nature Publishing Group