VOL. LXIX.] PHILOSOPHICAL TRANSACTIONS. 565 



II — 17, III — 30, when common and nitrous air gave ii — 1, ii + 14, 

 III + 14. The 3d air, before it was shaken in water, crystallized with the oil 

 of tartar like fixed air. An equal bulk of it was absorbed by water, which by 

 this mixture became acidulous; it precipitated the lime in lime water, extin- 

 guished a light several times, and killed an animal instantly. It is therefore 

 partly fixed air, and partly air which is not only better than common air, but 

 likewise than that extracted from Seine water, even when this last has been 

 boiled in glass vessels. The experiments being repeated with the same water of 

 Arqeuil, but in glass vessels, the air obtained, after being shaken in water, was 

 much better than that obtained from the same matter, when boiled in vessels of tin. 



He also extracted the air from distilled water in glass vessels, and having 

 shaken 1 measure — 32 parts of it in water, it was reduced to 1 measure — 35 

 parts. With the test of nitrous air it gave i — 6, when equal parts of 

 common and nitrous air gave i — 2, which showed that it was better 

 than common air. He extracted the air again in the manner above de- 

 scribed; but it was not sensibly diminished when shaken in water. Two mea- 

 sures — 49 parts of it, with the test of nitrous air, gave i — 2, 1+8, when 

 common air, &c. gave i — 1, i -f- «8: It is therefore better than common air. 



He extracted the air from a great quantity of distilled water in the usual 

 manner, and found that it did not sensibly diminish in water. With nitrous 

 air it gave 11 — 14, 11 — 25, 11 + 25, when common and the same nitrous air 

 gave II — 14, II -\- 10, III -|- 10; consequently it was dephlogisticated air, viz. 

 purer than the air of the Seine and Arqeuil, which are much better than 

 common air. 



He tried whether any difference would arise from boiling distilled water in a 

 matrass of tin, instead of glass vessels ; and found that the 1st air was diminished 

 -jL by being shaken in water, and afterwards with the nitrous air gave 11 — 13, 

 II — 16, III — 18, when common air gave 11 — 12, 11 + 8: which shows that 

 it was dephlogisticated air, but not so good as that extracted from the same 

 water when boiled in glass vessels. The 2d quantity of air was not sensibly 

 diminished in water, and with nitrous air gave 11 — 13, 11 — 20, in — 30; that 

 is, it was more dephlogisticated than the first. 



The air extracted from distilled water, is to that extracted fi-om the water of 

 the river Seine, as 13 to 32 nearly; whence distilled water does not give more 

 air than ^-o o( its bulk: but as the air extracted from the water of the Seine is -|- 

 fixed air, it may be concluded, that the quantity of respirable air produced by 

 both kinds of waters is nearly the same, and that they only differ a little in 

 purity. It is however true, that other experiments had shown him that water in 

 general absorbs about twice as much of dephlogisticated as of common air; for 

 which reason, he thinks that the respirable air of Seine water is rather less than 



