208 
MAJOR A. E. OXLEY ON THE INFLUENCE OF i\IOLECULAR 
exjDlanatioii, I think, is to be found in the finite though small angular oscillations 
which constitute a portion of the thermal energy of the molecules. The molecules 
are fixed relative to one another and form a definite space lattice but they are 
oscillating with small amplitude under the molecular field. This allows them to 
retain a finite susceptibility. Suppose we could double the molecular field, the 
limiting susceptibility would become one-half its former value and the saturation 
intensity of magnetization would be slightly increased. If we could increase the 
molecular field indefinitely, the susceptibility would get indefinitely small, the product 
of the two however tending to a finite limit equal to the true saturation intensity of 
magnetization. The amplitude of the molecular oscillations would, under the 
infiuence of this indefinitely large forcive, be indefinitely small. This state might be 
attained in a practical manner by cooling the substance, say in liquid hydrogen, when 
the limiting susceptibility would become vanishingly small. 
As N is the reciprocal of the limiting susceptibility the constant of the molecular 
field will become indefinitely large. Weiss, however, supposes N to be constant."^ 
The tendency of xl Io approach a small limiting value as the temperature is lowered 
is confirmed experimentally for ferro-magnetic substancesf and is particularly noticeable 
in the case of weak magnetic fields. The reduction of the amplitude of vibration of 
the molecules as the absolute zero is approached merely implies a higher frequency of 
angular oscillation under the increasing molecular field and does not necessarily imply 
that the rotational energy becomes vanishingly small. In this case it should be noted 
that the saturation intensity of magnetization we are considering is smaller than that 
which would be given by the simple summation of all the magnetic moments of the 
molecules in unit volume. In other words, the difficulty of producing this latter 
saturation by an external field becomes increasingly difficult on account of the larger 
molecular forcive at low temperatures, in agreement with the vanishingly small 
susceptibility referred to above. At higher temperatures the susceptibility to an 
external field is far greater ; the molecules are, as it were, helped over their difficulties 
with respect to the molecular field, when the external field is applied, by the increased 
energy of the rotational oscillations, and having passed this critical point they are held 
in new comhinations. Beyond the critical point the molecular state is chaotic, the 
molecules being interlocked {cf. p. 265), and the external field has sufficient control to 
produce a paramagnetic eftect only. 
* Following \\''eiss, we have taken the molecular field proportional to I. Weiss writes the molecular 
field NI and assumes N to be constant. This applies with sufficient accuracy in a temperature region just 
below the critical temperature, but cannot be true over the whole region down to absolute zero, because, 
as the molecular translational vibrations die down, the molecules approach one another more closely and 
the molecular field must necessarily increase considerably although I remains practically constant. This 
increase is accounted for by the increase of the coefficient N, which is the reciprocal of the limiting 
susceptibility. 
t Ewing, ‘ Magnetic Induction in Iron and other Metals,’ p. 172 et seq., where curves are given for iron, 
hard steel, nickel and various nickel steels. See also p. 269 infra and Ewing, Ioc. cit., ji. 3-54. 
