Dr. Dalton oji the Constitution of the Atmosphere. 159 



no portion of it can be abstracted from any one of them with- 

 out some chemical operation; whereas nitrous gas will im- 

 mediately seize the oxygen from any of the afore- mentioned 

 mixtures, the same as if it was alone, whatever may be the 

 proportions. Atmospheric air itself, or any artificial mixture 

 of the two gases in the same proportion as common air, is 

 equally affected by nitrous gas and by every other agent. 



Waving at present any consideration as to the nature and 

 properties of the above chemical compounds, I shall now pro- 

 ceed to state the means by which the proportions of oxygen 

 and azote in mixtures of these two gases may best be deter- 

 mined. Having been engaged in this investigation occasion- 

 ally for more than forty years, I may be entitled to give my 

 opinion on this important subject in practical chemistry. 



Various methods of analysing common air have been dis- 

 covered in the last fifty years. I have principally directed my 

 attention to three, namely, (1.) by the use of Volta's eudio- 

 meter and hydrogen, or (2.) by nitrous gas, or (3.) by quadri- 

 sulphuret of lime, to abstract the oxygen from the azote. 



First Method, by Voltas Eudiometer. 



Mr. Cavendish was one of the first to investigate the changes 

 produced by firing mixtures of hydrogen and common airs in 

 various proportions. (Vid. Philos. Trans. 1784.) The fol- 

 lowing table will exhibit a lasting monument of his skill in 

 effecting such an investigation. Many have attempted since 

 to improve the methods of analysis, and have brought out re- 

 sults widely differing from those to be derived from his table ; 

 but it is now universally allowed that his results are nearer 

 approximations to the truth than most of those we have seen 

 since. 



His method was to take 100 measures of common air and 

 mix them with various proportions of hydrogen, beginning 

 with upwards of 100, and gradually descending till about 20 ^ 

 then, firing each mixture by an electric spark, he marked the 

 diminution of the mixture each time as under. 



The following results are extracted from Mr. Cavendish's 

 Table, except the last column, " Amendment," which I have 

 attached, for reasons assigned below. 



Exp. Common 

 Air. 



1. ... 100 measures mixed with 124-1 



2. ... 100 



3. ... 100 



4. ... 100 



5. ... 100 



6. ... 100 



