264 
Book I. 
Eudiometer 
of Berthol¬ 
THE ATMOSPHERE. 
is reduced very neaaly to o.22. The trifling variations 
perceptible in his experiments were no doubt owing to 
the quantities of the mixture of sulphur and iron, by 
which he abstracted the oxygen, not being exactly the 
same at different times ; the consequence of which 
would be an unequal absorption of azotic gas. 
4. In the fourth kind of eudiometer, the abstraction 
of the oxygen of air is accomplished by means of phos 
phorus. This eudiometer was first proposed by A 
chard *. It was considerably improved by Reboult, 
and by Seguin and Lavoisier t; but Berthollet § has 
lately brought it to a state of perfection, as it is equal 
ly simple with the eudiometer of De Marti, and scarce 
ly inferior to it in precision. 
Instead of the rapid combustion of phosphorus, this 
last philosopher has substituted its spontaneous com 
bustion, which absorbs the oxygen of air completely; 
and when the quantity of air operated on is small, the 
process is over in a short time. The whole apparatus 
consists in a narrow graduated tube of glass containing 
the air to be examined, into which is introduced a cy 
linder of phosphorus fixed upon a glass rod, while the 
tube stands inverted over water. The phosphorus 
should be so long as to traverse nearly the whole 
of the air. Immediately white vapours rise from 
the phosphorus and fill the tube. These continue till 
the whole of the oxygen combines with phosphorus. 
They consist of phosphorous acid, which falls by its 
weight to the bottom of the vessel, and is absorbed by 
F Four. de Phye. 1784, vol. i. 
t Ann. de Cbim. xii. 38. 
Ibid. ix. 293. 
5 Ibid. xxiv. 73. and Jour. de l'Eecdle Pohteba. I. ii. 274.