THE thermodynamics of the orientation of nuclei by strong magnetic fields at very low temperatures was worked out by Simon1 in 1939. He found that with fields of 100,000 gauss and temperatures of 0·01° K., that is, under conditions which can be realized experimentally, the entropy of nuclear spin systems should be decreased by about 20–30 per cent, and would remain, of course, at this reduced value after adiabatic demagnetization. This would be sufficient to influence properties which depend on nuclear orientation, and the first experiments he contemplated were on the angular distribution of radioactive disintegration products originating from a partially oriented nuclear spin system. Preparations for these experiments, which were interrupted by the War, are now going ahead at the Clarendon Laboratory. It should be added that in the meantime experiments on nuclear induction have shown that the ‘reaction velocity'—at least in some nuclear spin systems—is still very high even at temperatures as low as 1° K. 2.