Drs. Williamson and Russell on the Measurement of Gases. 525 



and trouble are expended in observing the exact temperature and 

 pressure to which the gas is subjected at the time of measurement ; 

 and also in calculating from these data the volume which the gas 

 would occupy at the normal temperature and pressure. Frankland's 

 excellent apparatus, on the other hand, protects the gas from the 

 influence of variations of atmospheric" pressure, and, under favour- 

 able conditions, even from the influence of- change of temperature ; 

 but the complication of this apparatus; and its liabihty to derange- 

 ment, seem likely somewhat to limit its use. 



If, when a fall of temperature takes place, we could diminish the 

 pressure on the gas exactly in proportion to the diminution of elas- 

 ticity which it undergoes, such fall -of temperature would evidently 

 not alter the volume of gas in the eudiometer. In Uke manner a 

 rise of temperature might, if known, be counteracted by lowering the 

 eudiometer-tube. The same remarks apply to variations of barome- 

 tric pressure ; as an increase of this influence might be counter- 

 balanced by raising the eudiometer, and a diminution by depress- 

 ing it. 



It is therefore a question of some interest, to find, for any atmo- 

 spheric temperature and pressure, at what height of the eudiometer 

 the enclosed gas will occupy the same volume as at the normal tem- 

 perature and pressure. This is easily found by introducing a stand- 

 ard quantity of air into a tube over mercury, marking off" the height 

 of the mercury in the tube at the normal temperature and pressure, 

 then, at any other temperature or pressure, raising or lowering the 

 tube in the mercurial trough so as exactly to bring the enclosed air 

 to its nprmal volume. The mercurial pressure needed for this pur- 

 pose is evidently the same as that needed under the same circum- 

 stances for the reduction of any quantity of gas to the volume which 

 it would occupy at the normal temperature and pressure. 



The apparatus we use in applying this principle to gas-analysis 

 (fig. 1) consists essentially of the ordinary Bunsen's eudiometer, 

 and a "pressure-tube," which is simply a tube of some 6 or 7 inches 

 in length, and about the diameter of an ordinary eudiometer. It is 

 closed at one end ; and to the other is fixed a smaller tube of about 

 the same length. Such a quantity of air is introduced into this 

 pressure-tube, that when it is inverted in the trough the mercury 

 stands at a convenient height in the narrow tube. At this point a 

 mark is made, which indicates the height of mercury needed at any 

 temperature or pressure to reduce the enclosed air to its original 

 volume. The mercurial trough which we have used diifers only 

 from the ordinary one in being provided with a well at one end, thus 

 enabling the operator to raise or depress the eudiometer at pleasure, 

 so as always to bring the gas which it contains to the same pressure 

 as the air in the pressure-tube. IJoth the eudiometer and the 

 pressure-tube are held in a perj)endicu]ar position by means of 

 clamps which slide on uj)right rods. Each clamp is provided with 

 a simple kind of slow movement, by which the tube can be raised or 

 lowered by the operator whilst he is looking through a horizontal 

 telescope at a suitable distance. We place the j)ressure-tube in front 



