[ 2 49 ] 
I once imagined that, by mere dagnation, air 
might become unfit for refpiration, or at lead for 
the burning of candles; but if this be the cafe, 
and the change be produced gradually, it mud 
require a long time for the purpofe. For on the 
22d of September 1772, I examined a quantity of 
common air, which had been kept in a phial, 
without agitation, from May 1771, and found it 
to be in no re(pe<d worfe than frefh air, even by 
the ted of the nitrous air. 
The crydallization of nitre makes no fenfible 
alteration in the air in which the procefs is made. 
For this purpofe I didolved as much nitre as a 
quantity of hot water would contain, and let it 
cool under a receiver, danding in water. 
November 6, 1772, a quantity of inflammable 
air, which, by long keeping, had come to ex- 
tinguifh flame, I obferved to fmell very much like 
common air in which a mixture of iron filings 
and brimdone had dood. It was not, however, 
quite fo drong, but it was equally noxious. 
Bifmuth and nickel are diffoived in the marine 
acid with the application of a confiderable degree 
of heat; but little or no air is got from either of 
them; but, what I thought a little remarkable, 
both of them duelled very much like Harrowgate 
water. This fmell I have met with feveral times 
in the courfe of my experiments, and in procefles 
very diderent from one another. 
As I generally made ufe of mice in the ex- 
periments which relate to refpiration, and fome 
perfons may chufe to repeat them after me, and 
purfue them farther than I have done ; it may be 
Vol. LXIL K k of 
