On the effect of temperature on the viscosity of air
It appeared that further experiments on the viscosity of air were desirable in order to discriminate between the results of F. A. Williams* and those of most previous observers, and to test his conclusion respecting the validity of Sutherland’s law of the variation of viscosity with temperature. It happened that this could be done easily and expeditiously in the laboratories of the Imperial College. Mr. R. S. Edwards, the author of the following paper, was in the midst of preparations for determining the viscosity of neon at a number of temperatures ranging from atmospheric temperature to the normal boiling point of sulphur, and at my suggestion diverted his attention to the behaviour of air at the same temperatures. It is true that this range (about 430 centigrade degrees) is not so extensive as the thousand degrees covered by Williams’ experiments, but it includes all that region in which, according to Williams, the value of Sutherland’s constant displays the large increase upon which I have cast doubt. Edwards’ method of temperature control and estimation involves heating by the saturated vapour of selected substances of well-established boiling points, and would appear to be more reliable than the electrical heating, and particularly the temperature measurement by a single thermocouple, as employed by Williams. In the present experiments also, considerable variations of the pressure conditions have been made, with consistent results, thus proving the validity of the transpiration formula assumed. No such internal evidence of accuracy was provided in Williams’ experiments.