Specific Heat Measurements in 1–10°K Range Using Continuous Warming Method

Abstract

We describe a system for measuring specific heats by a continuous warming method with accuracy at least as high as has been attained with the usual heat burst method. Using this system data can be obtained very rapidly, even by unskilled operators, and reduced to final form with straightforward analysis and small amounts of computer time. We analyze the method in terms of simple models which take into account uncertainties in the stray heat, thermal gradients in the specimen assembly, the effects of electrical noise and finite bandwidth in the thermometer circuits, and their relation to the measurement of small heat capacities. Data on lead, mercury, and gallium are presented which indicate that the models lead to fair estimates of the anticipated scatter once certain characteristics of the apparatus are known. The observed scatter varied from 0.3% for gallium, where electrical noise in the amplifiers was a dominant factor, to less than 0.1% for lead and mercury, where no single factor was conspicuous. In addition, completely independent measurements were reproducible to about 0.1%. In these experiments the measuring intervals were never greater than 1% of the absolute temperature.