1861.] attending Combustion in Bare/ied Air. 



SECOND SERIES. 



333 



An inspection of these results indicates that even the natural oscil- 

 lations of atmospheric pressure cause a considerable variation in the 

 amount of light emitted by gas flames. In order to determine these 

 variations, the following special series of experiments was made, the 

 pressures being very accurately ascertained by means of a water-gauge. 



THIRD SERIES. 



Pressure of 

 Air in inches 

 of Mercury. 



30-2 

 29-2 



28-2 

 27-2 



Illuminating Power of 

 Experimental Flame. 



100 

 95-0 

 89-7 

 84-4 



100 

 94-9 

 89-8 

 84-7 



It is thus evident that the combustion of an amount of gas which 

 would give a light equal to 100 candles, when the barometer stands at 

 31 inches, would afford a light equal to only 84*4 candles if the baro- 

 meter fell to 28 inches. 



The results of these three series of observations taken together, 

 show, that beginning at atmospheric pressure, and with 100 units of 

 light, a decrease of almost exactly 5 * 1 units of light is the result of 

 each diminution of mercurial pressure to the extent of one inch, until 

 the barometer stands at 14 inches, below which the diminution of 



