i6 ANALYSIS of 



Then expofing it again to the lieat, I never failed to ge 

 more air ; and having done this, in fome cafes, not lefs 

 than twent)'- or thirty times, I was fatistied that even the 

 imalleft quantity of water will never ceafe to yield air, 

 and in feveral caies I have by this means procured more 

 air than the bulk of the waiter. 



Asfomev/ater would necelfarily infmuate itfelf between 

 the mercury and the glafs, I expofed almoft the whole 

 of the tube containing the mercury to the heat ; by this 

 means converting that water into vapour, and making it 

 afcend to the top of the mercury ; then throwing out 

 both the w^ater and the air produced from it again and 

 again, I at length found nothing but air above the mer- 

 cury. Still, however, the whole of the water was not 

 converted into air. For when, by means of heat, the 

 mercury was made to defcend, the water which had been 

 confined between the mercury and the glafs made its ap- 

 pearance, though by the afcent of the mercury it would 

 again difappear. 



I have alio found that when there was any fenfible quan- 

 tity of water above the mercury, and have expofed it to heat 

 day alter day, the quantity of air, in this cafe as well as 

 the preceding without heat, came to a 7naximum, and no 

 repetitions of the procefs would increafe it. This induces 

 me to conclude that the longeft continuance of any quan- 

 tity of water in the ftate of vapour would not convert it 

 into air. It may, however, be worth while, if there 

 fhould be an opportunity of doing it without much ex- 

 pence, to make the experiment. 



The pureft diftilled water fhould be ufed in thefe ex- 

 periments. Inftead of this, 1 once ufed pump water; 

 but found that, after the produdlion of air was advanced 

 to its maximum, it began to yield a confiderable quantity, 

 at leaft ten times more than it had done before, at the 

 fame time becoming ji little turbid. But when it was 

 4 clear, 



