292 LECTURES TO SCIENCE TEACHERS. 



such a case it is necessary to introduce some explosive gas 

 obtained by the electrolysis of water to aid the explosion. 

 The gases are generally measured moist ; for this purpose 

 a little water is introduced into the end of the eudiometer 

 by placing a drop on the end of a long stick ; this keeps 

 the gas saturated, but great precaution must be observed 

 with regard to the presence of water, for in some cases the 

 gas is dry after your measurement, although it was wet 

 before you began, so that you must make a correction for 

 the tension of aqueous vapour in one case and not in the 

 other. In this case the carbonic anhydride was quite wet, 

 but after the potash was put in it will have dried it, and 

 therefore after the removal of the bullet of potash it must 

 be looked upon as a dry gas. 



When the oxygen has to be determined, it is usually 

 done by the diminution of volume by the disappearance 

 of a certain quantity of gas after hydrogen has been intro- 

 duced and the mixture exploded ; but in some cases it is 

 necessary to actually measure the steam. For this purpose 

 Bunsen devised this instrument, consisting of a tube 

 through which a quantity of steam from boiling water can 

 be passed, in which the eudiometer is suspended by means 

 of a holder after it has been removed from the trough : A 

 powerful current of steam is sent through, and a thermo- 

 meter is placed within so that the temperature may be 

 determined ; or you may assume it to be that of boiling 

 water. The process is very rarely adopted, and perhaps I 

 am not committing any breach of confidence when I say 

 that a celebrated chemist who has done a great deal of 

 work with gas analysis when he saw this instrument did 

 not recognise it ; which shows that it is not very much 

 used. 



Many improvements have been made in this process of 

 analysis. You see the great slowness of it. The absorp- 

 tions take a long time, and the time which has to be lost 

 in allowing the gas to acquire the temperature of the air 

 is very considerable, and the calculations also are by no 

 means short, although they are not difficult. An improve- 

 ment was made by Regnault and Reiset, who devised an 

 apparatus in which the eudiometer is connected at the 

 lower end by means of an iron three-way cock to another 

 glass tube longer than but of the same diameter as the 



