324 PROCEEDINGS OF THE AMERICAN ACADEMY. 



volumes absorbed), but this line does riot point toward the origin. The 

 resistance was first measured with no hydrogen in the wire, then with 

 about 175 volumes, then with still greater amounts. No observations 

 were taken to obtain points on the curve between the origin and that for 

 175 volumes, but, if the results as plotted are correct, it is evident that 

 the curve, as it approaches the origin, must either bend very sharply 

 toward the left or else it must drop below the horizontal axis. In other 

 words, it would appear from the curve that a small quantity of occluded 

 hydrogen, say 50 volumes, would change the resistance of the wire very 

 slightly or, possibly, make it actually smaller than when there is no 

 hydrogen in the wire. Knott comments on the peculiar shape of the 

 curve, but offers no explanation, and there appears to be no record of 

 any later attempt to repeat the experiment or to seek for a more accurate 

 method of studying the phenomenon. The work outlined in this paper 

 was undertaken in the hope of making a more careful determination of 

 the curve given by Knott and, more particularly, with a view to studying 

 its form near the origin. 



The method, employed by Knott and others, of determining the 

 amount of hydrogen occluded — by measuring the increase in weight 

 of the wire — appears to be incapable of giving very exact results. 

 Accordingly, in the work described below, the wire is charged with 

 hydrogen by being made the negative electrode of a gas voltameter and 

 the amount absorbed is determined as follows. The electrolysing cur- 

 rent is accurately measured at regular intervals during the charging 

 process and the total quantity of hydrogen liberated from the electrolyte 

 is thus easily calculated ; the hydrogen which escapes from the wire 

 is carefully collected, and its volume measured, and this volume being 

 subtracted from the total volume liberated by the current it is assumed 

 that the difference represents the volume of hydrogen occluded by the 

 wire. 



In the earlier experiments the voltameter was simply a glass jar of 

 about three liters' capacity filled with a two per cent solution of sulphuric 

 acid ; a light hard-rubber frame held in place the two electrodes, the 

 palladium wire being coiled in a slender spiral in the middle of the jar, 

 while a platinum wire, coiled around the inner wall of the jar, formed 

 the positive electrode ; a burette with an inverted funnel fused to its 

 lower end served to collect and measure the escaping hydrogen. Two 

 objections to this form of apparatus at once suggested themselves. In 

 the first place it was found that even a large funnel would not catch all 

 of the rising hydrogen bubbles, especially toward the end of the experi- 



