564 PHILOSOPHICAL TRANSACTIONS. [aNNO 1779. 



nished in the proportion of 10 to 7; and by the test of nitrous air gave ii — 14, 



II + I, III + Ij when the common air gave ii — 12, ii -|- 6, in -j- 6. The 2d 

 quantity of air was diminished in the proportion of 3 to 1 ; and when examined 

 by the test of nitrous air gave ii + 0, in + O. Whence it may be concluded 

 that the 1st air was better than the atmospheric air; whereas the 2d was worse 

 and mixed with much fixed air. 



Being in doubt whether the tin vessel employed in the experiment above- 

 mentioned might not alter the nature of the air, &c. he made use of glass 

 vessels. Having therefore filled 1 of these vessels, having a long neck bent in 

 2 directions, with the Seine water, he obtained some air which seemed not 

 sensibly diminished when shaken in water. Having introduced 1 measure and 

 37 parts of this air into the tube used to try the diminutions, it gave with the 

 nitrous air 1 + IQ, 1 -f 48, when the same quantity of nitrous and atmos- 

 pheric air gave ii +2(5, ii + 6: it is therefore certain, that the air extracted 

 from Seine water is purer than common air. Another time he extracted, in the 

 same manner, and from the same water, the air; 1 measure and 24 parts of 

 which being introduced into the tube, &c. and shaken, was reduced to 1 mea- 

 sure — 31 parts, that is, -i- of it was absorbed. Treated with the nitrous air it 

 gave I — 4, when equal measures of common and nitrous air gave i + O: it was 

 therefore better than common air. A 3d time he extracted the air from the 

 water of the river Seine, contained in 3 matrasses; this air was about Vt of the 

 bulk of the water, and it gave with the test of nitrous air, ii — 14, ii — 9, 



III — 9; when the common air mixed with nitrous air, as usual, gave 11 — 14, 

 II -|- 8, III -f- 8. It is therefore clear, that the air extracted from the Seine 

 water, by the action of fire in glass vessels, is much better than common air, or 

 than the air which is extracted from the same water when boiled in tin vessels. 



Another time he filled a glass retort, which had a long and doubly bent neck 

 with Seine water. The water weighed about 3 lb. The air that came out of it 

 lost 4- of its bulk by being shaken in water; and afterwards being tried with 

 nitrous air it gave 11 — 16, 11 — 16, in — 16, when common and nitrous air 

 gave II — 12, II -|- 12. This experiment being repeated, the air was diminished 

 of 4- by being shaken in water. One measure — 16 parts of this airintroduced 

 into the measuring tube gave 11 — 32, 11 — 2, when common air ai d the nitrous 

 gave II — 28, 11 -f 4. 



The water d'Arqeuil at Paris is considered as very pure. Mr. F. filled the 

 tin vessel with it, and received the air that came out of it into 3 vessels. Being 

 shaken in water, the 1st of them was diminished -L; the 2d, 4; and the 3d -{-f, 

 by the operation in water. A light burned with a flame, more luminous than in 

 common air, in the 1st air after it had been shaken in water. This air being 

 tried with the nitrous air gave 11 — 10, 11 — 10, in — 10. The 2d gave n — lO, 



