RY HOT PLATINPM IN PIKFKKKNT HASRS. 



:i hi'\\ \\irr was llU' to the presence of Sonic ^.is \vl)ich was |x>sit ivdv 

 \\licn emit led liv (In- \\iiv. Sucli .-i gas might lc fxpri-trd In dill'iisc avvav from the 

 wire more readily at a !<>\v than a high pressure and under a high than a low field. 



10. Tlit-ori/ of tin' Sti'inlif I'nx'it /')/ /,</// iii n.ri/tjen. 



The view which the author has l>een led to as the result of the foregoing experi- 

 ments is that the positive, as well as the negative, ions are produced in the first 

 instance inside the metal, or, at any rate, so far within the surface that their rate of 

 production is in general uninfluenced by the external field. The experiments, in 

 addition, afford strong support to the view that the positive ions are furnished not hy 

 the platinum itself, but by the oxygen absorbed or dissolved in the transition layer 

 l>etween the metal and the surrounding gas. That the positive ions do not arise in 

 the space outside the free metal surface appears to be distinctly indicated by the 

 following considerations. 



We have seen that the positive and negative ionisations from hot platinum vary 

 with the temperature according to the same kind of law, but quite independently of 

 one another. This is exemplified by the fact that at a pressure of T5 millims. of 

 oxygen the two leaks are equal to one another at about 1230" C., whereas at 700 C. 

 the saturation current with the wire negative is about 10~ 7 of its value with the wire 

 positive. The disparity between the two leaks points strongly to the view that they 

 are the result of separate actions ; it proves, in addition, that there is no considerable 

 amount of volume ionisation, but it does not prove that the ions are not formed 

 outside the wire. For instance, an oxygen molecule might attack a positively- 

 charged platinum atom, with the result that of the two oxygen atoms present in the 

 molecule the one carrying the negative charge combined with the platinum, whilst 

 the other one was set free, forming a positive ion. Of course, the negative ion would 

 in a sense be free while the interaction was taking place, but it would always be 

 within a distance from the attracting positive ion comparable with the radius of the 

 sphere of molecular action, and the interionic attraction would probably be enormous 

 compared with the greatest external field of force (about 10* volts per centimetre) 

 which could l>e applied. 



There are two main objections to the foregoing view, which otherwise has the 

 advantage of simplicity to recommend it. One of these is the way in which the leak 

 lap* U0iind change of pressure. It will be noticed that the above theory would 

 make the leak proportional to the rate of reaction between the O, and the free platinum. 

 In order to xplain the variation of the leak with pressure under steady pressures, it 

 is necessary to assume, what is otherwise probable, that the proportion of uncombined 



* This conclusion receives strong support from previous experiments by the writer, in which he showed 

 that the ionising power of a new wire could be transferred to a second wire, previously rendered inactive 

 by heating, it" this \\riv negatively charged. 'Phil. Mag.,' [6], vol. 6, p. 86. 



