Definition of the Second of Time

Abstract

SIR EDWARD BULLARD¹ solicits views on the desirability of abandoning the astronomical second for the most refined physical measurements, in view of the high precision² of atomic frequency standards. He suggests instead a ‘physical second’, to be defined in terms of the natural period of the cæsium atom, the advantage being that the physical second would be more immediately available than; the astronomical second in cases where the highest precision is sought. The purpose of this communication is to direct attention to two consequences that should be well understood before the astronomical second is abandoned by physicists. I leave aside the question of desirability, on which views are certain to differ, because the change is desirable from some points of view and undesirable from others.