Chemistry and Physics. 477 



6. Diffusion of Gases through Mercury Vapor. The "Diffu- 

 sion Air Pump." — When a vessel, such as an X-ray bulb, is evac- 

 uated by means of a mercury pump while the pressure is read on 

 a MacLeod gauge it becomes a matter of no little importance to 

 know whether the total pressure in the vessel is really as low as the 

 gauge indicates or whether the pressure can never fall appreciably 

 below 0*0013 mm , that is, below the pressure of saturated mercury 

 vapor at ordinary room temperature. Since decided differences 

 of opinion exist with regard to this question, the more general 

 problem of the diffusion at low pressures of various gases through 

 mercury vapor has been attacked both experimentally and theo- 

 retically by W. Gaede. A striking and important consequence 

 of this investigation is the invention of a new and remarkable kind 

 of air pump. Since the original paper is comparatively long and 

 involves a number of fairly complicated mathematical formulae, it 

 will only be possible to give a fragmentary account of this very 

 interesting research in these columns. 



The first experiments were made upon the diffusion of air and 

 of hydrogen through saturated mercury vapor. A bulb, A, was 

 connected by a diffusion tube to a cooling bulb, C, which in 

 turn communicated with another bulb, M. The vessel, A, and 

 the diffusion tube were wrapped with a coil of wire through 

 which a heating current could be sent and the temperature regu- 

 lated at will. The lower end of A opened into a vertical glass 

 tube, which was connected to a mercury reservoir in such a man- 

 ner as to enable the experimenter to introduce mercury into A or 

 to lower the top of the column below the region of relatively 

 high temperature, as occasion required. The lower end of A 

 was also immersed in an oil bath to facilitate the heating of 

 the upper portion of the mercury column. Each experiment 

 involved three sets of observations, (i) The pressure of the gas 

 in A and M was observed with the whole apparatus at room tem- 

 perature, (ii) The pressures of the gas in A and M were ascer- 

 tained when A was heated by the electric current, but when the 

 top of the mercury column was well below the level of the bottom 

 of A. In cases (i) and (ii), therefore, the pressure of the mercury 

 vapor was negligible, (iii) The pressure in M was measured 

 under the same thermal conditions as in case (ii), the only change 

 consisting in the elevation of the mercury meniscus to the level of 

 the hot oil bath. From the observed pressures together with the 

 volumes of the vessels on both sides of the cooling bulb it was 



easy to calculate the fraction — - , where c (which is nearly unity) 



denotes an empirical constant involved in case (ii), and p\, ji? 2 sym- 

 bolize the partial pressures of the gas in A and M respectively for 

 case (iii). The formula for this fraction, in terms of known quan- 

 tities involves two legitimate assumptions, (a) the adequacy of 

 the " Boyle-Gay Lussac " gas law, and (b) the constancy of the 

 total mass of gas in the apparatus for cases (i), (ii), (iii). The 



P' 



ratio k — - is a measure of the disparity of pressure in A and 



cp 2 r " 



