Abstract: | It is shown that it is technically possible to use the boiling point of nitrogen when calibrating platinum resistance thermometers. An analysis of the origins and an estimate of the components of the error are given, and the limits of the error of the results of measuring the boiling point are calculated. |