The normal method used to measure the temperature of the windings of transformers, reactors and other heavy electrical plant is by change of resistance. By means of a simple d.c. test, resistances are measured before and after the heat run. The temperature at shutdown is calculated using the known temperature coefficients of resistance for copper.It is well recognised that inductance may cause errors in these measurements, especially in large transformers.A simple test procedure is presented which enables the presence of inductive errors to be detected. Should they be present, a numerical correction to the measured resistances is readily applied.The principle involved is to place a d.c. millivoltmeter across a winding other than the one under test and note its readings during the period of the cooling curve. By transformer action, the induced e.m.f. may be related to the winding under test and a quantitative correction in ohms derived.The method may be easily applied at any test bench, irrespective as to whether the volt-ammeter, Wheatstone bridge or Kelvin bridge circuits are employed. The only additional equipment required is a vacuum-tube millivoltmeter of common type.