The effect of temperature on the viscosity of air

The recent measurements of F. A. Williams of the viscosity of air from 15°C. to 1002°C. have produced unexpected results. The greater part of this extensive temperature range has hitherto been quite inadequately investigated, and the new data, if correct, will obviously be of great value in relation to the theories of gaseous viscosity as a means of comparing their validities. It is proposed to extend the measurements by the same experimental method to argon, which, being monatomic, is of special importance theoretically. For these reasons it is essential that the accuracy of Williams’s data should be satisfactorily established before acceptance. An examination of these data in comparison with data from other sources in respect of the lower part of the temperature range reveals differences which cast doubt upon the reliability of Williams’s values, and there are, besides, some features of his experimental work which are open to serious criticism. Of previous investigations, that of Breitenbach covers the largest temperature range—namely, from — 21.4°C. to 302°C. For purposes of comparison with Williams’s results given in Table V of his paper, Breitenbach’s values, interpolated where necessary, are included in Table I, the last column of which gives the differences, expressed as percentages of Breitenbach’s values.