20 A N- A L Y S 1 S, &c. 



gravity of 682.5, ^ expofed it to the atmofphere, after 

 which it yielded as much as before, viz. about one third 

 or one fourth of its bulk, all ftrongly inflammable. I had 

 the fame rcfult in the fubfcjuent procefs. After another, 

 the air was exploded like a mixture of inflammable and 

 atmofpherical air, and the next produce burned with 

 a lambent flame. Being then examined, its fpecific 

 gravity was 692.4 ; fo that it had acquired fome weight 

 by imbibing atmofpherical air. 



Having, in like manner, expelled air which I found 

 to be inflammable from a quantity oifpirit of turpentine^ 

 I made it imbibe atmofpherical air, and expelling it again, 

 found it to be not inflammable, but phlogifticated. This 

 fluid had alfo gained fomething in its fpecific gravity by 

 the procefs. 



The only objection that, after giving much atten- 

 tion to the fubjeft, I think can be made to the con- 

 clufion that I flrft drew from thefe experiments, viz. 

 that air is actually produced from water, is the very 

 fmaU quantity that is produced in proportion to the bulk 

 of the water, after the air naturally contained in it is 

 wholly expelled. But if it fliall appear, after a long 

 courfe of time, the^t this fmaU production of air from the 

 fame water is conftant, and equable, I do not fee how 

 the conclufion, extraordinary as it may be thought, can 

 be difputed. This air being wholly phhg'iflicated is a 

 fufflcient proof that the air fo produced is not abfoibed 

 from the atmofphere in the courfe of the procefs. For 

 then it would have been dephlogifticated, or at leafl" 

 purer than that of the atmofphere, which water always 

 feizes upon in preference to that which is impure. 



To 



