Apjmratvs for Rapid Gas-Analasis. 373 



gas. In treutiug a mixture of gases with several reagents, it is 

 necessary to remove one reagent before adding another. This 

 is accomplished by washing out with water in such a manner 

 that the gas is not lost. Eaoult performed this treatment of the 

 gases and washing out of reagents, in a graduated tube with two 

 stoj^-cocks, one at each end ; one of the stop-cocks was sur- 

 mounted with a funnel to introduce the fluids. But the whole 

 affair was not easily managed, and the gases were submitted to 

 an unnecessary amount of washing while removing the excess of 

 reagents used. 



Wilkinson modified this method, and devised a very simple 

 and useful apparatus, in which the clumsy manipulations of 

 Eaoult were overcome by using a tube with one stop-cock above, 

 the lower end of the tube dipping into water in another tube of 

 much larger diameter. By this means the gases could be treated, 

 with liquid reagents, introduced through a funnel attached to 

 the stop-cock above ; and by introducing or removing water from 

 the outer tube, the gas could be measured at atmospheric press- 

 ure. To facilitate the removal of liquids from the outer tube, 

 the latter has a stop-cock attached below. But, as in the appa- 

 ratus of Raoult, the gases are submitted to an unnecessary 

 amount of washing when water is introduced to remove the re- 

 agents. This Avashing becomes very important in many cases. 

 For example, take the case of illuminating gas. We introduce 

 potassic hydrate solution to remove the carbonic acid, then 

 potassic ])yrogallate to remove oxygen ; and now we must wash 

 out the alkali before adding bromine to absorb the illuminants. 

 To do this, much water is needed, and this large quantity of , 



water will wash out some of the illuminates, often as much as 7^/ 

 two per cent. ^ 



To overcome this difficulty of excessive washing, I have de- 

 vised the apparatus which is the subject of this paper. In this 

 process, the gas is removed from the absorbent liquid and 

 measured in another vessel, without washing. 



The cipparatus is shown in Plate XXII, The tube A is of 

 about 125 c.c. capacity, whilst B, although the same length, 

 holds only 100 c.c. from the point D, or zero, to the mark on 

 the capillary tube at C, and is carefully graduated in ,1, c. c. 



