In a recent paper Professor Tyndall and the writer described experiments on the variation with temperature of the mobility of positive ions of helium moving in helium gas. Owing to lack of knowledge of the precise effects of the phenomenon of electron transfer, it was not possible to deduce from the results the forces between an ion and a neutral atom of this gas. But it was pointed out that if alkali ions were used instead of helium ions, electron transfer between the atoms and ions would not occur. In that case the main objection to a treatment based on a classical which had already been proposed. Further it was suggested that the most marked variation of mobility with temperature might be expected for a relatively large ion moving in a weakly polarizable gas. In the work described below, it was this condition that determined the choice of caesium ions in helium. Sodium ions also have been studied mainly for purposes of comparison. Experimental Fig. 1 is a scale diagram of the apparatus. In principle, the apparatus is similar to experimental tubes employed previously with the alkali ions in this laboratory; but it was necessary to modify the old design so that the enclosed gas could be maintained at various temperatures ranging between that of liquid nitrogen, and about 250° C. A Kunsman sources S, run at a dull red heat, was used as emitter of ions, and this had to be placed as far as conveniently possible from the section M in which the ionic speeds were measured, so that it would not affect the temperature of the gas in that region. According the apparatus was arranged as in the figure, with the source S at the upper end the measuring section M at the lower end, the ions being carried across the intervening space by means of a uniform electric field maintained by six guard rings. The measurement of the speed of an ion in the gas at any chosen temperature was made in the manner previously described.
Read full abstract