Abstract
Titanium salts should behave more simply than most materials under magnetic cooling, since the ground level of the paramagnetic ion is here inherently doubly degenerate and so unaffected by crystalline fields. De Haas and Wiersma show that as long as the magnetic moment remains constant during adiabatic demagnetization, the temperature is a linear function of a field strength, a most valuable fact since it permits the establishment of a thermodynamic temperature scale. The present paper computes when deviation of the moment from constancy should commence because of the perturbing effect of second‐order terms (the first nonvanishing order) in the dipole‐dipole coupling energy. The calculated moment does not remain constant to as low field strengths as reported experimentally. The discrepancy cannot be ascribed to a counter‐balancing effect of exchange coupling but may be due to the neglect of third and higher order dipole terms. Because of these large dipole perturbations, the status of the experimental temperature scale in titanium alum becomes rather uncertain. The present stage of the theory may be tested more accurately by using much lower initial fields than in the existing Leiden experiments.