IN 1939 Professor W. F. Giauque, of the University of California, proposed that we change the manner of defining the number scale of the absolute, thermodynamic (Kelvin) scale of temperatures by assigning a number to asinglereproducible thermal state (e.g., triple point ofH2O). At present the number scale is defined by assigning a number (100) to the interval betweentwofixed points (“ice” and normal boiling points ofH2O). J. P. Joule and Lord Kelvin made the same proposal one hundred years ago but it was forgotten until it was made anew by Giauque. Joule and Thomson in 1854 said the proposed change in definition “must be adopted ultimately” because the definition based on asinglefixed point is “preferable”.