relating to Phlogijlon , &c. Part III. i > 
water) it muft be fuppofed, that the elements, or compo- 
nent parts of fixed air are in it ; but one of thefe elements is 
pure air, and the mixture of nitrous air fhews, that it contains 
no fuch thing, though, according to M. Lavoisier, fixed 
air contains 72 parts in 100 of pure air. 
However, being apprized of this objection to inflammable air 
from iron, I made ufe of inflammable air from tin, and I had the 
fame refult as with that from iron. I alfo calculated the weight 
o 
of the fixed air which 1 got in the procefs, and comparing 
it with the plumbago which the iron neceflary to make the 
inflammable could have contained, I found, that, in all the 
cafes, it far exceeded the weight of the plumbago ; fo that it 
was abfolutely impoflible, that the fixed air which I found 
fhould have had this origin. For the greater fatis faction, I (hall 
recite the particulars of a few experiments of this kind. 
In ten ounce meafures of inflammable air from malleable 
iron I revived red precipitate till there remained only 1.1 oz. 
meafure of air, and of this 0.07 oz. m. was fixed air,, beings 
completely abforbed by water. The weight of this air would 
be 0.063 gr. But, ^ lice 96° grains of iron will yield 1054 
oz. meafures of inflammable air, the iron employed in procuring 
all the inflammable air that was ufed in this experiment, viz* 
8.9 oz. meafures (without allowing for any that went to the 
revivification of the mercury) would be 8.1 grains; and fince 
M. Bergman fuppofes, that 100 grains of iron contains 0.12 
gr. of plumbago, the quantity of it in this iron would 
only be 0.01008 gr. which is not quite a fixth part of the 
weight of the fixed air. 
With the precipitate perfe, fent me by M. Bertholletv 
I revived mercury till 8f oz. m. of inflammable air was reduced 
to zi oz. m.* and of this 0,04 oz, at leaft was fixed air. 
TThis 
