The use of spontaneous voltage fluctuations for the measurement of low temperatures

Abstract

A discussion is given of the method proposed by Lawson and Long to use the noise across a resistance as a measure of its absolute temperature. The statistical fluctuations in the measured noise are calculated, taking into account the bandwidth of the amplifier. This calculation shows that is not necessary to use a first tube with lowered voltages on plate, screengrid and heater and it shows that a quartz crystal is essentially worse than a pure resistance or a L—C circuit as a thermal noise generator for thermometrical purposes.