by Hot Platinum in Aiv at Low Pressures. 83 



under the influence of heat into at least one positive ionjand 

 possibly other products. The reaction is supposed to be 

 monomolecular, and mav be represented by the equation 



where X represents the unknown substance, n is the number 

 o£ positive ions formed from each molecule of X, and Y 

 represents the other products if they exist. 



The laws of chemical dynamics then give at once 



where /l is the velocity of the reaction ; so that 



But if the induced leak be denoted by ?/, 



y = n^ = - '^^ =kX = kX,e-^-' = Ae-'-', 



so that logi/ = B — Id. 



(A2, A, and B are different constants.) Evidently on this 

 view, if we plot the logarithm of the difference between the 

 actual current and the steady leak with the time, a straight 

 line ought to be obtained. 



The observations have been reduced in this way in fig. 3, 

 and it will be s^een that all the points fall very nearly on the 

 straioht line drawn. The irregularities are due to the fact 

 that the wire could not be treated in exactly the same way 

 between each two observations, and it is probable that the 

 rate of decay depends not only on the time which has elapsed, 

 but also on some other circumstances. 



A series of experiments was now made with a new tube in 

 which all the metal parts were of platinum, so that the whole 

 could be carefully cleaned out by boiling in nitric acid and 

 washing with distilled water. After this treatment, it was 

 found that the wire gave quite as considerable a leak as one 

 which had not been cleaned, so that the positive ionization 

 is not due to superficial impurities soluble in nitric acid. 



With this wire an attempt was made to see if the rate of 

 decay of the variable leak depended on the temperature to 

 which the wire was heated. The method used was to maintain 

 the wire at a constant temperature, and take readings of the 

 current every eight minutes. The wire was then heated to a 

 somewhat higher temperature and the process repeated. 

 The following table gives the results in the order in which 

 the experiments were made. Owing to the magnitude of the 



G2 



