ANALYSIS OF GASEOUS MIXTURES. 



301 



rect. I have made a great number of experiments to deter-* 

 mine this point. A quotation or two will serve to show 

 how fallacious the method is. 



1. Passed 2*9 cubic inches of nearly pure chlorine, col- 

 lected over water that had been boiled to expel atmospheric 

 air, into a graduated eudiometer standing over water at 50* 

 Fah. The gas measured 1*9 cubic inch. After standing 

 15 minutes, the gas measured 1'68 cubic inch. This indi- 

 cated an absorption amounting to 22 parts. 



2. Passed equal measures of chlorine and atmospheric 

 air into the same eudiometer. The mixed gases measured 

 1*9 cubic inches. After the lapse of 15 minutes, the mixed 

 gases measured 1*8 cubic inch. The absorption, there- 

 fore, in the same time and under the same circumstances 

 as in No. 1 explanation, was only 10 parts. 



The rate of variation is not constant. It varies with the 

 purity of the chlorine itself, and with the quality and pro- 

 portion of the gas added to it. Besides, as I shall prove 

 further on, the assumption by Dr. Fyfe, that the gases con- 

 densed by chlorine consist entirely of defiant gas, is not 

 correct; and therefore, even if the objections to the method 

 just stated did not exist, it could only afford a crude 

 approximation to the composition of the gas, and to its 

 relative illuminating power. 



A better method of determining the amount of olefiant 

 gas, and which was employed by Dr. Henry, and is recom- 

 mended in some of the best analytical works, is to allow 

 the chlorine and gas to re-act on each other, and then to 

 remove the whole of the chlorine by an absorptive solution, 

 determining the amount of olefiant gas by the diminution 

 in bulk of the original quantity employed. This method 

 I have subjected to a very rigid examination. Pure chlo- 

 rine is recommended to be employed; but it has been over- 

 looked that it is almost impossible to obtain pure chlorine 

 over water, and mercury absorbs it so rapidly that experi- 



