122 Improved Sliding-Rod Eudiometer and the Volumescope. 



ter. The pressure of the screw G upon the spring F being 

 restored, and an explosion effected, agreeably to the directions 

 already given, by returning the rod into its tube, more or less, 

 and appealing to the gauge, the deficit may be ascertained. 

 If no error shall have taken place, expelling the residual gas 

 will just return the rod to the situation which it occupied when 

 the experiment commenced. Of the deficit, of course one-third 

 is dwe to oxygen. It may be proper to mention that some de- 

 lay is necessary, in order to permit the residual gas to part 

 with the heat acquired from the combustion of the hydrogen 

 and oxygen. 



As for the analysis just described, the eudiometer may, as 

 represented in the preceding figure, be seated in a cup of mer- 

 cury, instead of being placed over a mercurial reservoir ; and 

 since the apparatus, when once put into operation, enables us 

 to multiply experiments with great facility, it will be found 

 peculiarly well calculated for a series of observations inider 

 circumstances in which access to a pneumatic cistern cannot 

 be had. 



Eiidiometrkal Ap2)aratus analogous to the preceding, excepting 

 that it is constructed of Brass, used xmth Water, and that Ex~ 

 plosions are caused in it hy an Electric Spark. 



In the analysis of atmospheric air, agreeably to the process 

 last described, no gaseous product being generated which is 

 absorbable by water, it is not necessary to employ mercury, 

 and, consequently, to have the metallic part of the eudiometer 

 of iron and steel. It is in fact preferable to have it of brass, 

 as in that case it will not rust, and may be kept in operation 

 for many months without requiring much adjustment. I have 

 an apparatus thus made, and so contrived as to be ignited by 

 an electric spark. Excepting the substitution of brass for iron, 

 there is no material difference between that apparatus and the 

 one represented by the figure, excepting that the receiver E 

 is exchanged for one of which there is a representation ni 

 Phil. Mag. vol. Ixvii. p. 22, fig. B. 



In the brass eudiometer last described, the cock C is omit- 

 ted ; while, at right angles to the receiver, a small cock is in- 

 serted, which supports a glass vessel holding water. By these 

 means, any excess or deficiency of this liquid is easily remedied, 

 and the employment of the cup beneath the eudiometer ren- 

 dered unnecessary. 



[To be continued.] 



XX. On 



