The thermal accommodation coefficients of gases. I. An investigation of the effect of flashing

Abstract
Modifications to the hot-wire method of Knudsen (1911) for determining accommodation coefficients have been introduced by various workers in recent years. These methods have in common the flashing of the wire in the gas (Roberts 1930; Mann 1934) or in vacuo (Mann & Newell 1937) immediately before heat-loss measurements are begun. This treatment was originally suggested by Roberts (1930) and was assumed to give a ‘clean’ surface by removing adsorbed gas films. The heat loss from the wire was reduced by flashing, and this was ascribed to a reduction in the accommodation coefficient. The fall in heat loss after flashing has been confirmed, but it is accounted for, not by a reduction in the accommodation coefficient, but by the persistence of the thermal effects of flashing. Thus the effect of flashing is accentuated by working with silvered and platinized vessels; it is reproduced by flashing an adjacent wire and is particularly evident when working in vacuo . Abnormal temperature rises can also be observed when pipette methods are used to introduce the gas to an evacuated vessel (Mann & Newell 1937; Raines 1939; Rolf 1944). The results obtained in similar experiments were shown to be due mainly to the slowness of the diffusion process.