16^ Dr. Dalton o?i the Constitution of the Atmosphere^ 



spark in its passage will re-alight the flame, to suffer a quicker 

 extinction, and so on till at length the combustion is com- 

 plete. This reason will also explain the excessively slow 

 combustion of azote by the electric spark, as ascertained by 

 Mr. Cavendish, and as I have found by repeated experience. 

 Query, might not this experiment succeed better by heating 

 the eudiometer? 



From what we have stated it must be obvious that in order 

 to secure the complete abstraction of either oxygen or hydro- 

 gen from mixtures by Volta's eudiometer, we should avoid 

 too near an approach to the limitations we have pointed out ; 

 or if that cannot be, we should carefully examine the residue 

 for both gases. The best test for very small portions of oxy- 

 gen is undoubtedly nitrous gas ; for somewhat larger portions 

 of oxygen or hydrogen, additions of those gases might be 

 made so as to bring the mixtures into proportions capable of 

 being exploded. 



Second Method, hy Nitrous Gas, 



The nitrous gas eudiometer is of singular utility on many 

 occasions. No other can exceed it in accuracy when mixtures 

 contain very little, as one or two per cent, of oxygen ; or on 

 the other hand when nearly the whole of the gas is oxygen. 

 But when the mixture of gases contains from twenty to eighty 

 per cent, of oxygen, as in the case of common air, it is not the 

 best when great exactness is required. The reason is well 

 known ; when oxygen and nitrous gas combine, the combina- 

 tion is not like that of oxygen and hydrogen, in uniform pro- 

 portion. We may take one third of the diminution for oxy- 

 gen, when mixed over water; but this can be considered only 

 as a first approximation. One hundred parts of oxygen may 

 combine with 1 30 or 360 parts, or any intermediate quantity 

 of nitrous gas, according to circumstances. When only 1 or 

 2 per cent, of oxygen are expected I put in 5 or 10 per cent, 

 of nitrous gas, and take one third of the diminution for oxy- 

 gen. When the oxygen (freed from carbonic acid) is judged 

 to be 90 or more per cent, pure, I put 100 parts of nitrous 

 gas of known purity (say 98 + ) to 100 of the oxygen, and 

 mark the diminution ; I next put in 40 nitrous and mark the 

 diminution, and so on, till there is manifestly a slight portion 

 of nitrous left; then this is to be removed by a small portion 

 of oxygen; finally, knowing the quantity of azote which was 

 in the nitrous gas, the rest must have been introduced by the 

 oxygen. 



In this way I find a perfect agreement, whether the nitrous 

 test or the hydrogen is used ; but with common air the residue 



