550 



Royal Society : — 



this was fixed at one end of a Bunsen's photometer, whilst the flame 

 to be submitted to various pressures, and which I will call the expe- 

 rimental flame, was placed at the other. The experimental flame 

 was made to burn a uniform amount of gas, viz. 0'65 cubic foot 

 per hour in all the experiments. 



The products of combustion were completely removed, so that 

 the experimental flame, which burnt with perfect steadiness, was 

 always surrounded with pure air, the supply of which was, however, 

 so regulated as to secure a maximum of illuminating effect in each 

 observation. 



In all the following series of experiments, the illuminating power 

 given under each pressure is the average of twenty observations, 

 which accord with each other very closely. In each series, the 

 maximum illuminating effect, that is the light given by the expe- 

 rimental flame when burning under the full atmospheric pressure, 

 is assumed to be 100. The following is a summary of the results : — 



1st Series. 



Pressure of air in inches of 

 mercury. 



Illuminating power of experi- 

 mental flame. 



29-9 

 24-9 

 19'9 

 14-6 

 9-6 

 6-6 



100* 



75-0 



52-9 



20-2 



5-4 



•9 



2nd Series. 



30-2 

 28-2 

 26-2 

 24-2 

 22-2 

 20-2 

 18-2 

 16-2 

 14-2 

 12-2 

 10-2 



100- 

 91-4 

 80'G 

 73-0 

 61-4 

 47-8 

 37-4 

 29-4 

 19-8 

 12-5 

 3-6 



These numbers indicate that even the natural oscillations of atmo- 

 spheric pressure must produce a considerable variation in the amount 

 of light emitted by gas-flames, and it was therefore important to 

 determine, by a special series of observations, this variation in lu- 

 minosity within, or nearly within, the usual fluctuations of the 

 barometrical column. In order to attain greater delicacy in the 

 pressure readings in these experiments, a water-gauge was used, 

 but its indications are translated into inches of mercury in the 

 following tabulated results, each of which represents, as before, the 

 average of twenty observations. 



