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



509 



needle due to the face of the thermopile being suddenly ex- 

 posed to a source of radiation is proportional to the final per- 

 manent deflection. Use was made of this fact in obtaining the 

 results given in Table V. In this case there was no difficulty 

 in determining the limit of the " first swing." The difficulty 

 still remained, however, that the magnesium might be giving 

 out a maximum of radiation at the instant of exposing the 

 thermopile to it while at the next exposure it might be far 

 from a maximum. A mean of the results by the method of 

 permanent deflections and ''first swings" gives "135 as the 

 radiant efficiency of magnesium light. The only artificial light 

 whose efficiency is greater than this is the light from a Geiss- 

 ler tube* of which the radiant efficiency is about '34. L. B. 

 Marksf finds the radiant efficiency of the arc light to vary 

 from *08 to "127 depending upon the character of the carbons. 









Table V. 







No. of 



Experiment. 



Galvanometer "swings." 

 Two alum 



cells. Direct radia- 

 R. of circuit tionK. ofcir 

 7-46 cuit 7-46 



Light transmit- 

 ted through 



alum cells divi- 

 ded by total 

 radiation. 



Preceding 

 ratio correc- 

 ted for heat 

 transmitted. 



Radiant effi- 

 ciency (pre- 

 ceding' ratio 

 corrected for 

 light reflected 

 and absorbed.) 



No. 



1 



3-8 



47-0 



•081 



•078 



•122 



a 



2 



3-5 



40-0 



•088 



•085 



•133 



u 



3 



2-9 



31-8 



•092 



•089 



•139 







Sensitiveness of galvanometer changed. 





No. 



4 



4-9 



55-1 



•089 



•086 



•134 



a 



5 



5-2 



57'0 



•091 



•088 



•137 



a 



6 



5-0 



48-3 



•103 



•100 



•156 



a 



7 



5-4 



56-3 



•096 



•093 



•145 



a 



8 



4-6 



54-9 



•084 



•081 



•126 



a 



9 



4-3 



45-3 



•095 



•092 



•144 



Average 



oy the method of " 



first swings 



it 



•137 



H. Nakano,:}: in a series of experiments to determine the 

 relation of the size of carbons to efficiency, found that in the 

 case of carbons *25 in. in diameter and a potential difference of 

 38 volts the " spherical efficiency " was '166 but with carbons 

 of the usual size the efficiency was about '10. 



The radiant efficiency of incandescent lamps is much less 

 than that of the arc light. Ernest Merrit§ found the radiant 

 efficiency of incandescent lamps under the best conditions com- 

 patible with their continued existence to be about - 06. 



In all the experiments above referred to, the efficiency was 

 measured in the same way, viz : by taking the ratio of the 



*Beiblatter zu den Annalen cler Physik, xiv, page 538. 



\ Am. Inst, of E. E., 1890, vol. vii. % Am. Ins. of E. E., vol. vi, p. 30S. 



§This Journal, vol. xxxvii, p. 167. 



