94 



CIIEMISTEY. 



to unite with the nitrogen, this passes on. The weight of this tube com- 

 pared with its weight before the experiment shows the amount of oxygen 

 in the air. Then subtracting the sum of the weights of the oxygen and 

 carbonic anhydride from the weight of the air examined, we have the 

 weight of the nitrogen. 



Another mode which ascertains the proportionate volumes of the gases 

 as well as their weights is represented in Fig. 28. We 

 have here a tube, a b c, with a very minute opening at 

 a, a bulbous enlargement at 6, and its larger orifice, c, 

 made to fit air-tight in the top of the vessel, d. This 

 vessel is filled with mercury, and is graduated, as you 

 see. There is a cock at e, by which you can let the 

 mercury run out in as small a stream as you please. 

 Before fitting the tube, a ft c, to the vessel, it is filled 

 with loose cotton having bits of phosphorus scattered 

 in it, which, by warming, is spread over the fibres of 

 the cotton, and then the tube is accurately weighed. 

 Fitting the tube to the vessel, the cock, e, is now slight- 

 Fig. 23. ]y opened. AS the mercury flows slowly out, air pass- 

 es in at a to take its place, and in passing in it loses all of its oxygen, for 

 the phosphorus which it finds every where in the cotton takes it, forming 

 with it phosphorous anhydride, which remains in the cotton. We have 

 therefore nitrogen alone in the vessel, </, to take the place of the mercury 

 that runs out. When the volume of air employed is used up we close the 

 stop-cock at e. We can ascertain the volume of the nitrogen by the grad- 

 uation on the vessel, or more accurately by measuring the mercury which 

 has run out, for this, of course, exactly equals in bulk the nitrogen that 

 has taken its place. And from its volume we know its weight, because the 

 specific gravity of the gas has been ascertained by chemists. Then we find 

 the weight of the oxygen, by weighing the tube, and comparing its weight 

 with that which it had before the experiment ; and its volume is found from 

 its specific gravity. In this process you see that no account is taken of the 

 carbonic anhydride that is in the air. It is indeed so small in amount 

 ( 115) that it would make but little difference in the result. There is al- 

 ways moisture in air, and this must be got rid of in order to make the an- 

 alysis accurate. This can be done by letting the air to be analyzed first 

 pass through a tube containing some substance which has a great affinity 

 for water, as the fused calcium chloride. 



119. The Gases of the Air Obedient to Gravitation. You 



