500 



Mr. T. Graham on Hydrogenium. 



[June 17, 



VI. " Additional Observations on Hydrogenium." By Thomas 

 Graham, F.B.S., Master of the Mint. Received June 10/1869. 



From the elongation of a palladium wire, caused by the occlusion of 

 hydrogen, the density of hydrogenium was inferred to be a little under 2. 

 But it is now to be remarked that another number of half that amount 

 may be deduced with equal probability from the same experimental data. 

 This double result is a consequence of the singular permanent shorten- 

 ing of the palladium wire observed after the expulsion of hydrogen. In 

 a particular observation formerly described, for instance, a wire of 609*14 

 millims. increased in length to 618*92 millims. when charged with hy- 

 drogen, and fell to 599*44 millims. when the hydrogen was extracted. 

 The elongation was 9*78 millims., and the absolute shortening or retrac- 

 tion 9*7 millims., making the extreme difference in length 19*48 millims. 

 The elongation and retraction would appear, indeed, to be equal in 

 amount. Now it is by no means impossible that the volume added to 

 the wire by the hydrogenium is represented by the elongation and re- 

 traction taken together, and not by the elongation alone, as hitherto 

 assumed. It is only necessary to suppose that the retraction of the 

 palladium molecules takes place the moment the hydrogen is first ab- 

 sorbed, instead of being deferred till the latter is expelled; for the 

 righting of the particles of the palladium wire (which are in a state of 

 excessive tension in the direction of the length of the wire) may as well 

 take place in the act of the absorption of the hydrogen as in the ex- 

 pulsion of that element. It may indeed appear most probable in the 

 abstract that the mobility of the palladium particle is determined by the 

 first entrance of the hydrogen. The hydrogenium will then be assumed 

 to occupy double the space previously allotted to it, and the density 

 of the metal will be reduced to one half of the former estimate. In 

 the experiment referred to the volume of hydrogenium in the alloy will 

 rise from 4*68 per cent, to 9*36 per cent., and the density of hydro- 

 genium will fall from 1*708 to 0*854, according to the new calculation. 

 In a series of four observations upon the same wire, previously recorded, 

 the whole retractions rather exceeded the whole elongations, the first 

 amounting to 23*99 millims., and the last to 21*38 millims. Their 

 united amount would justify a still greater reduction in the density of 

 hydrogenium, namely to 0*8051. 



The first experiment, however, in hydrogenating any palladium wire 

 appears to be the most uniform in its results. The expulsion of the hy- 

 drogen afterwards by heat always injures the structure of the wire more 

 or less, and probably affects the regularity of the expansion afterwards in 

 different directions. The equality of the expansion and the retraction in 

 a first experiment appears also to be a matter of certainty. This is a 

 curious molecular fact of which we are unable as yet to see the full 

 import. In illustration, another experiment upon a pure palladium wire 



