174 DR CARGILL G. KNOTT ON THE 



sets in is of course a function of the time as well as of the temperature. I 

 believe that, if the wire were kept at a steady temperature of 260°, or even 

 lower, it would ultimately lose all its hydrogen. Observations on the 

 change of resistance give, indeed, the most delicate means of studying the 

 manner in which the hydrogen escapes, and would well repay a careful investi- 

 gation. 



These earlier experiments were made with a view to establish the broad 

 features of the case. They suggested, however, various lines of further and 

 more careful inquiry, of which one has just been mentioned. Another is 

 obviously a following up of the remark made at the end of the description of 

 the experiment of 27th January, which was indeed the very first of the series. 

 The problem, expressed in its generality, is, What relation, if any, exists 

 between the temperature-coefficient of resistance and the charge of hydrogen 

 present ? At first sight there is a tendency for the total change for a given rise 

 of temperature to remain the same whatever charge of hydrogen is present. 

 That is if we use throughout the whole series of experiments the same wire at 

 different saturations, and draw curves of the resistance (as measured) in terms 

 of the temperature, we shall obtain a family of curves which in their initial 

 portions run parallel to each other, the lowest curve being that of pure 

 palladium, the highest that of saturated hydrogenium. A closer study of the 

 various experiments showed that this relation did not strictly hold ; but before 

 anything definite could be obtained, it was necessary to make a series of careful 

 experiments with this special object in view. The results of these later 

 experiments I shall now give, comparing them when possible with the results 

 of the earlier series. 



A palladium wire was taken of resistance - 927 ohms at 18° C. Its 

 resistances at different temperatures up to 110° or so were carefully measured 

 in an ordinary Wheatstone bridge. It was then charged with a small charge of 

 hydrogen, and the same process of measurement of resistance gone through, 

 and so on, with necessary additions of hydrogen, till the wire became saturated 

 with the gas. 



From the observations, the values of the resistance for each wire were 

 interpolated so as to correspond to the temperatures 18°, 28°, 38°, &c. The 

 subjoined table gives these interpolated values for six different hydrogeniums 

 besides the pure palladium itself. In one experiment four terms only appear. 

 This resulted from the breaking of the large glass beaker in which the wire 

 was being heated : — 



