A feature of the method described is that it involves neither temperature nor heat‐flow measurements. The procedure consists merely of taking a stop‐watch reading (over a few minutes) and finding a value on a calibration chart. The two flat ends of a cylindrical sample, about 18 mm in diameter and 1–30 mm in length, are put in contact with two silver plates which are kept at constant temperatures by means of two boiling liquids of suitable boiling points,viz.differing by about 10–20°C. The time is measured in which a quantity of heat flows, in the steady state, through the sample. This quantity is determined very simply by the quantity of liquid evaporated at the ``cold'' end and which is collected as a condensate. In practice, two calibrated samples are measured first and a calibration line is plotted from which the thermal resistance of subsequent samples can be found at a glance. It is then not necessary to know the exact boiling points of the two liquids, nor the absolute amount of evaporated liquid, nor its heat of vaporization. A measurement takes about 15 min and the error is not more than ±3%. The construction of the apparatus and the performance of measurements in the temperature range of from 20 to 200°C are described in detail. Moreover this article shows a new way of measuring the conductivity of very small samples (e.g., small single crystals).