182 



Dr. W. Marcet. On the 



allowing the bell-jars to be moved up and down alternately in the 

 tanks. The inside of the bell-jars communicates with the outside air 

 by means of a U- sna P e d. iron pipe, one limb of which opens inside the 

 bell-jar above the fluid it contains, and the other outside on a some- 

 what lower level for sake of convenience. The tanks or troughs 

 contain glycerine, over which floats a layer of almond oil an inch or 

 two in thickness, and filling approximately the trough. The opening 

 of the pipe above the surface of the oil enters loosely, so as to leave 

 free passage of air, a neck in the bell-jar admitting a thermometer 

 fitted air-tight through it. By this arrangement when the bell-jar is 

 lowered, the thermometer descends into the inner limb of the [J -pipe 

 or stand-pipe without dipping into the oil or glycerine. 



A measured volume of air to be analysed is drawn into one of the 

 air-holders (selected for that purpose) by raising it to a certain 

 height, and on lowering it slowly, after cutting off communication 

 with the outside by a stop-cock, the air is driven first of all through 

 an absorption apparatus and then into the second bell-jar, which has 

 at that time an ascending motion. The whole apparatus is connected 

 by brass tubes and stop-cocks, on one hand with a pressure-gauge, 

 and on the other hand with a graduated glass receiver, immersed in 

 mercury, and fixed to a bracket and rack movement by which it 

 can be raised and lowered at will. The various parts of the instru- 

 ment are all in air-tight communication with each other while the 

 carbonic acid is being absorbed, so that the absorption will tell on 

 the pressure-gauge, and by depressing the measured glass receiver 

 into the mercury trough, and driving the air it contains into the 

 air-holders, while keeping the eyes fixed on the pressure-gauge, it will 

 be easy to restore the atmospheric pressure in the whole apparatus, 

 and thus determine the volume of gas absorbed. 



I shall now beg to describe in succession the various parts of this 

 instrument, and next state how the analysis is made. 



The Tanks. — One of the great difficulties I met with was the 

 selection of the fluid to be used in the tanks. Water would not 

 answer, first of all because it absorbed carbonic acid, and next on 

 account of the tension of aqueous vapour. I tried a nearly saturated 

 solution of common salt in water, so as to avoid any loss of carbonic 

 acid by absorption, but after experimenting for a considerable time I 

 had to reject the plan because of the vapour-tension, which I found 

 impossible to take into account. I came to the conclusion that it was 

 necessary to dry the air and submit it to analysis in the dry state, 

 keeping it entirely out of reach of aqueous or other vapours. Bearing 

 this in mind, glycerine appeared to me to be possessed of the required 

 properties as a medium for confining the air in the bell-jars, and 

 therefore for filling the tanks. It had the advantage of holding no 

 water, but its thickness appeared at first sight a serious objection, as 



