Comparison of Sources of Artificial Illumination. 1 11 



These values have been brought to. a common scale, as in all 

 former comparisons, the region of the D line in both spectra 

 being taken as unity. 



It would not seem unreasonable to ascribe the difference in 

 question, to the atmospheric conditions under which the obser- 

 vations were made. 



The Incandescent Lamp. 



Our study of the incan- 

 descent lamp included the 

 comparison of lamps of 22 

 high and low resistance, 

 heated to normal candle- 20 

 power, and determinations 

 of the color of the former is 

 in different stages of incan- 

 descence. The lamps in 16 

 question were the Edison 

 lamp which had been used u 

 as a standard throughout 

 our investigation and a 12 

 Bernstein lamp giving 16 

 candles at 20 volts. 10 



The results, which are 

 given in table YI (see also 8 

 figure 4), show that the 

 temperature of the lamp 6 

 of low voltage was not so 

 high as that of the Edison 4 

 lamp at the same candle- 

 power, and that the spec- 

 trum of the former was 

 correspondingly of greater 

 relative brightness in the 

 red than in the violet. abcd Eb f g 



Previous to the detailed Curves of relative intensities in the spec- 

 ci-nA-rr rv£ +"U T?A' 1 trum of day-light. Curve VI is from mea- 



Siuay 01 me -UiCllSOn lamp, sur ementof day-light by unclouded sky, curve 

 the lamp was placed in the VII of day-light by heavily clouded sky. 

 Bunsen photometer and the Curve VIII > introduced for comparison is same 



relation between pWrnmn as curye ni ' figure 2 ' and re P resents the s P ec " 

 leiduon uetween eieciiomo- trum of the e i ectr i c .arc. The horizontal line 



tive force and candle-power represents the spectrum of the standard in- 

 determined for a range of candescent lamp. 



brightness between 1 candle and 30 candles. It was then main- 

 tained successively at potentials corresponding to 4 candles, 10 

 candles, 16 candles, 22 candles and 28 candles, and its spectrum 



/VI 



/ 



/ 1! 

 / j 1 



1 1 



1 1 



; 1 { 



,-vnj ! 



// i 



1 1 



1 



1 



,;'''' ^J™ 



J----^^ 



---T--^'^ 



1 1 1 II 1 1 



