Specific Heats of Solids at Very Low Temperatures

1 General.—The equations (7) which constitute the substance of my Heat Theorem, at once suggested a new experimental problem, namely, the determination of specific heats down to temperatures as low as possible. It is only if the specific heats are known that we can quite safely calculate the quantities A and U at low temperatures by means of the two already known laws, and so test the new Heat Theorem for these quantities are, for many reasons, practically always inaccessible to direct measurement at low temperatures. 

I therefore set myself to perfect methods for the determination of the specific heats of solids at very low temperatures, and I think I may say that I have solved the problem. 

We shall see in the next chapter that the results obtained have given a decisive bent to our conceptions of the energy content of solids, and hence also of the theory of the solid state in general. 

Thanks to the work of Behn, Tilden, Dewar, and others, something was already known of the behaviour of specific heats at low temperatures but there was no method which would give, not merely the mean energy content over a considerable temperature interval, but the true specific heat down to the lowest possible temperatures. 

From what was known, it was by no means certain even that the problem was soluble. It was conceivable that at low temperatures phenomena of retardation caused, say, by a large decrease in conductivity might have made measurements impossible. Such difficulties did not arise. As the investigations of Eucken (82) in particular show, the conductivity has, on the contrary, a tendency to increase considerably at low temperatures. Since also the thermal capacity falls rapidly at the same time, temperature differences are very quickly equalized at low temperatures, a circumstance which is extremely favourable to accuracy in calorimetry.  

---