258 EXPERIMENTAL INVESTIGATIONS. 



hydrogen, and having been for some time connected with a 

 voltameter (fig. 7), the tubes of which are of the same size as 

 those of the battery. 



In the form last described (figs. 4 and 6) the tubes were 

 all as nearly of the same size as could be procured ; they con- 

 tained each about one and a half cubic inch ; in the first 

 form (figs, i and 3) the portion o y r of the narrow tube con- 

 tained one and a quarter cubic inch, and the portion h, / of 

 the wide tube contained two and a half cubic inches. A por- 

 tion of the apparatus with which I wrought was constructed 

 by my order for the London Institution, and another portion 

 belonged to Mr. Gassiot, and was by him very kindly placed 

 at my disposal for the purpose of these experiments. Had it 

 not been for this valuable addition, I should have been obliged 

 to make all my experiments on a much smaller scale ; they 

 would have taken more time and been by no means so 

 satisfactory. 



As I have already stated, a third form has occurred to me 

 while writing this paper, which I think in many respects more 

 advantageous than either of the two preceding, and which, as 

 it may be some time before I can experiment with it myself, 

 I will here describe for the benefit of those who are differently 

 situated. One cell of it is shown in fig. 8 : a, a, is a Woulfe's 

 Fig. 8. Fig. 9. bottle with three necks ; in the centre neck is 

 fitted a glass stopper, b ; in the other two 

 tubes o, h fit accurately by means of glass 

 collars (c c, fig. 9.) welded to them, and 

 ground on the outside ; the platinum is her- 

 metically sealed into the tops of the tubes, 

 which may be charged in a similar manner 

 to fig. i. By immersing this apparatus in 

 the water-trough each tube with the gas it 

 contains may be detached and examined 

 separately ; but its principal advantage is, that by slightly 

 greasing the stopper and collars it may be made perfectly 

 air-tight, which, for reasons that will be apparent in the 

 course of this paper, is a most material point. This apparatus, 

 moreover, being entirely composed of glass and platinum, 

 concentrated acid, alkaline or other corrosive solutions may 



