312 F. J. Rogers — Magnesium as a Source of Light. 



seen, is very high and what is still more surprising, the ratio of 

 radiant energy to total energy is not far from "75. The small 

 loss by convection which renders this ratio possible is partially 

 explained by the fact that the product of combustion is a solid. 

 It may be also that this unusually high percentage of radiation 

 is in some way related to the fact that so large a proportion of 

 the radiant energy itself is luminous. 



The iinal experiments in my study of magnesium as a source 

 of light consisted in the determination of the amount of light 

 produced by burning one gram of the metal. Measurements 

 were made by means of the "horizontal slit" photometer. 

 The standard for four determinations was an Argand lamp 

 with a Methven slit, which allowed two candle power to pass 

 through. In these measurements the comparison was made at 

 the D line. In four other determinations the standard was an 

 incandescent lamp ; in which case the measurements were made 

 at two points on opposite sides of the D line and their mean 

 taken. In Table VII are given the final results of these eight 

 determinations. 



Table VII. 



Standard = Argand lamp. Standard = In candescent lamp. 



C. p. minutes C. p. minutes 



No. per g. of Mg. No. per g. of Mg. 



1 248 J 233 



2 252 2 236 



3 228 3 288 



4 250 4 272 



A mean of the above eight measurements gives 251 candle- 

 power-minutes produced by the combustion of one gram of 

 magnesium. 



This measurement of the candle-power-minutes produced by 

 one gram of magnesium may serve as a check on the determi- 

 nation of the total efficiency. Assuming that the efficiency is 

 •10, one gram produces 601 calories of luminous energy ; there- 

 fore the thermal equivalent of one candle-power- minute of 



magnesium light is — = 2 "39 calories. Even in the 



most unfavorable circumstances, namely, by using the maximum 

 heat of combustion from Table VI and the minimum candle 

 power minutes from Table VII we have, thermal equivalent 



= = 2 "74 lesser calories. In 1865 Julius Thomsen* 



228 



published some determinations of the thermal equivalent of 

 light, which he found to vary from 4/4 calories in the case of 



* Annalen der Physik und Chemie, vol. cxxv, page 348. 



