PRODUCED BY HOT PLATINUM IN DIFFKl'KNT CASKS. 



liv tin- copious negative ionisation. It was impossible to test tin- 

 question with the apparatus u-nl. 



A second wire, 0*2 milliin. in diameter, was tested and found to give results 

 almost identical with the ;il>ove when it was heated in hydrogen for the first time at 

 860 0. This wire was afterwards heated in hydrogen and later in oxygen for several 

 days, mostly at a temperature of atxmt 1100C. A long time after the wire had 

 again got into a steady condition in regard to the ionisation in oxygen the oxygen 

 was pumped out and hydrogen re-admitted The wire was now heated in hydrogen 

 at 26 millims. pressure at a temperature of 900 C., when it was found that the above 

 slow time changes had almost disappeared. The negative leak when first measured 

 about 10 minutes after first heating the wire was 672xlO~* ampere, whilst the 

 positive was 4*5 x 10~ la ampere. They subsequently rose and fell to 8x 10~* ampere 

 and 67 x 10~ 13 ampere respectively. The difference between the two cases seems to 

 indicate that heating in hydrogen produces a permanent change in the constitution 

 of the platinum. There are two other facts which support this contention. One 

 is the permanent reduction of the value of the steady positive leak in oxygen 

 produced by continued heating in hydrogen, which was mentioned on p. 27. The 

 other is that the surface of a wire which has been heated for a long time in hydrogen 

 becomes visibly pitted and cracked. This change does not appear to be produced by 

 heating, to moderate temperatures at any rate, in oxygen. 



Observations were also recorded of the variation in the ionisation when the gas in 

 which the wire was heated was changed from hydrogen to oxygen. The change from 

 one gas to the other was carried out in the same way as in the previous case. Oxygen 

 was admitted to a pressure of T067 millims. and the wire maintained at 900 C. 

 Under these circumstances the negative leak was found to fall at once to the 

 small value previously obtained in oxygen. The negative leak when first measured 

 registered 2xlO" 13 ampere per square centimetre, and it was found to possess the 

 same value 20 minutes later. The positive leak, on the contrary, fell gradually 

 during 3 hours from 4'8 x 10~" to 8*9 x 10"" ampere, more than half the fall occurring 

 in the first half hour. This decrease in the iouisation was accompanied by a slight 

 decrease in the pressure of the oxygen, which fell to 1*026 millims., indicating that 

 hydrogen had been evolved by the wire, had combined with the oxygen, and that 

 the water formed had been absorbed by the phosphorus pentoxide. This experiment 

 indicates that hydrogen diffusing out of a hot platinum wire increases the positive 

 leak in oxygen, but is without effect on the negative leak. This conclusion will be 

 more fully established by experiments to be described later. 



The last experiment shows that although a considerable amount of hydrogen may 

 remain in the wire, the addition of a small quantity of oxygen at once reduces the 

 negative ionisation to a small value. This indicates that the great negative ionisation 

 produced by hydrogen in platinum is due to some change of a very superficial 

 character. On the other hand, a wire which has previously only been heated in 



