146 Miss E. W. Hobbs on the Change in the 



supposing that the material originally deposited in an amor- 

 phous state undergoes a slow process of crystallization ; in 

 other words, that this is another example of the metastability 

 of metals observed by Cohen and his co-workers. Unfortu- 

 nately, no light is thrown upon the true nature of the change 

 by an analysis of the ageing curves, as they are not expressible 

 in terms of any simple formula. 



On both views, the effect of heat, unless it caused a break 

 in the film, would be to bring about a marked fall in resistance,, 

 in the former case owing to the decrease in viscosity and the 

 consequent acceleration of the process of coalescence, and in 

 the latter because of the increased velocity of the allotropic- 

 change at higher temperatures. Fawcett (loc. cit.) showed 

 that the decrease in resistance which occurred in his expe- 

 riments when a film was heated was accompanied by a 

 marked hardening of the surface of the film. This effect was 

 observed in the present work. 



It is important to note that the ageing must be proceeding 

 during deposition, and consequently it is another factor influ- 

 encing the nature of the film obtained at the cessation of 

 discharge. Again, while the discharge is taking place, films- 

 being deposited near the cathode will be affected to some 

 extent by its temperature, which Hodgson and Mainstone*" 

 have shown is a function of the current through the 

 discharge-vessel and the cathode-fall of potential. All 

 attempts, therefore, to obtain quantitative results of the 

 rate of laying down of the film by measurements of resist- 

 ance during deposition are doomed to failure. 



Deposition in Hydrogen. 



An investigation of the change in resistance with time was 

 made for films deposited in hydrogen, and the ageing curves 

 in this case were found to be similar to those obtained with 

 air as the residual gas during discharge. In fig. 1 the dotted 

 curves G and H represent two ageing curves for platinum 

 films (R = 8400 ohms and 9300 ohms respectively) deposited 

 in hydrogen, and the early portions of the dotted curves in 

 fig. 3 give similar results for palladium. It will be noted 

 in the case of palladium that rates of ageing are shown which 

 are far greater than those obtained under any other conditions 

 (recurves W and Y, fig. 3). There is strong evidence that 

 in many instances this change is a continuation of a change 

 which has been proceeding at a very rapid rate during the 

 discharge itself, and this probably applies to platinum films as 

 well, though their case has not received such full investigation. 

 * Phil. Maj. vo 1 . xvi. p. 411 (1913). 



