PRODUCED BY HOT PLATINUM IN DIFFERKNT GASI-s 63 



conditions. It was, therefore, never steady, so the minimum values were taken. 

 This kind of effect was much less marked, if it occurred at all, in the other gases. 



The minimum value of the positive ionisation was found to remain practically constant 

 with a wire heated during three months at various times, for 1 50 hours altogether, in 

 oxygen at 900 C. to 1000 C. Moreover, four different wires of different dimensions, 

 after continued heating in oxygen, gave nearly the same value for the ionisations 

 at the same temperatures and pressures. 



The positive ionisation in air at constant temperature is smaller than that which 

 would be obtained if the nitrogen were withdrawn, so as to leave only oxygen at a 

 low pressure. The nitrogen, therefore, exerts an inhibiting effect on the oxygen. 



The minimum value of the positive ionisation at a definite pressure in all gases 

 appears to be connected with the temperature by the relation first deduced by the 

 author for the negative ionisation. This relation may be written t = A0*e~ <J/a *, where 

 t is the ionisation, 6 is the absolute temperature, and A and Q are constants. The 

 value of the constant Q, which is a measure of the energy associated with the 

 liberation of an ion, is in most cases smaller for the positive than for the negative 

 ionisation. 



These results refer to wires which have been heated in a vacuum and sul>sequently 

 in the gas in question for a long time. New wires exhibit peculiar properties, 

 especially in regard to their behaviour under different electromotive forces. Old 

 wires also exhibit hysteretic effects with change of pressure and temperature. 



The view is developed that the positive ionisation is caused by the gas adsorbed by 

 the metal and the consequence examined of supposing the ionisation to be 

 proportional to the amount of the adsorbed gas present. In the case of oxygen, by 

 making the assumption that the rate of increase of the adsorbed gas is proportional 

 jointly to the concentration of the external dissociated oxygen and to the area of 

 unoccupied platinum surface, whilst the rate of breaking up is proportional to the 

 amount present, a formula is obtained which agrees with the experimental results. 



This formula is that the ionisation t = -. * , where p = (kP+k 3 )* %k, P being the 



external pressure and k the dissociation constant of oxygen. A, B and k are 

 constants depending on the temperature and are of the general form a0*e~*'. Thus 

 this view accounts for both the temperature and pressure variation. 



The positive iouisation from the outer surface of a hot platinum tube in air is 

 increased when hydrogen is allowed to diffuse through from inside the apparatus. 

 The increase in the ionisation is proportional at constant temperature to the quantity 

 of hydrogen escaping from the surface in unit time. For different temperatures the 

 effect produced by a given quantity of hydrogen is greater the higher the 

 temperature. 



The negative ionisation from hot platinum in air is unaltered when hydrogen is 

 allowed to diffuse out through the platinum. 



