388 SECTIONAL TRANSACTIONS.— A. 



establish the temperature equilibrium between spin and the surroundings ; 

 the mechanism of this process is not completely understood. 



The case of Hell shows that in problems of the first group there are also 

 interesting difficulties. 



Dr. J. F. Allen. — The properties of liquid Helium II (10.30). 



Liquid helium is a substance which differs most remarkably from any 

 other liquid. The phase diagram of helium possesses no triple point for 

 equilibrium between gas, liquid, and solid. Instead, as far as one can 

 ascertain, the liquid phase persists down to the Absolute Zero, and the 

 liquidus and solidus curves become parallel at that temperature. The 

 liquid phase consists of two modifications, and the transformation between 

 them occurs at 2-19° K (the X-point). The modifications, called Hel and 

 Hell, are totally different phenomenologically. Hel is a normal Liquid, 

 while Hell possesses properties completely different from those of any 

 other known substance. The most striking phenomena exhibited by Hell 

 are as follows : — A negative temperature coefficient of expansion ; a very 

 high specific heat which suffers a discontinuity at the X-point ; a thermal 

 conductivity which is approximately five hundred times as great as that of 

 copper at room temperature ; and a mode of heat transport which appears 

 to involve a transfer of momentum. When one measures the viscosity of 

 Hell by means of a rotating disc, one obtains a value of lo"* CGS units, 

 i.e. comparable to a gas. On the other hand, when measured by the flow 

 method the viscosity becomes immeasurably small and is certainly less than 

 lo"^" CGS units. So far no comprehensive theory has been developed to 

 explain all of the properties of Hell. 



Prof. J. H. Van Vleck. — The molecular field and the determination 

 of very low temperatures (11. 15). 



In experiments on magnetic cooling, it is customary to determine the 

 temperature by assuming that the susceptibility obeys Curie's law x = C/T. 

 Actually, this law carmot hold because of (a) the Stark splitting of energy 

 levels caused by the crystalline fields from the non-magnetic atoms surround- 

 ing the paramagnetic ion and (b) the dipole-dipole and perhaps exchange 

 forces coupling together paramagnetic ions. The effect (a) is wanting in 

 CsTi(S04)2i2H20, while (b) disappears at infinite magnetic dilution. 

 Before reliable determinations of the temperature can be made from sus- 

 ceptibility (in distinction from thermodynamic) measurements, it is necessary 

 to devise an adequate theory of (a) and (b). Difficulties in the way of doing 

 this, as well as the progress so far made, are summarised. In particular, 

 the usual Lorentz field H + 47tM/3 is only an approximate representation 

 of (b) valid only if the temperature is not too low. It is not even clear 

 whether dipole-dipole forces can ever make a body become ferromagnetic. 

 A discussion is included of the analogous electrical case, where possibly 

 the hypothesis of hindered rotation may not be necessary to prevent spon- 

 taneous polarisation in isotropic dielectrics. Ultimately, magnetic cooling 

 experiments should throw considerable light on inter-molecular forces in the 

 solid state. 



Dr. F. Simon. — Experiments below 1° abs. (11.45). 



The experiments carried out by Dr. Kurti, Dr. Lain^, Dr. Squire and 

 the author with the magnet at Bellevue (Paris) are described. 



