10S Nichols and Franllin—A Spectro-photometric 



photometry, as applied to so difficult an object as the arc-light, 

 concerning which Yogel, whose measurements of this source 

 of light, and of sunlight, daylight, and the light of the moon 

 and of various fixed stars, are of uncommon accuracy, makes 

 note that the limit of accuracy of his determinations of the 

 electric light is about 16 per cent. 



Table III. 



A comparison of various measurements of the spectrum of the Foucault 



regulator. 





From measure- 



From measurements 



From measurements 





ments by A. Crova. 



by W. H.Pickering. 



by H. C. Yogel. 



Wave-lengths. 



Arc-light 



Arc-light 



Arc-light 



Colza oil light. 



Gas light. 



Petroleum. 



6760 



0-59 







6562 







0-82 







6333 







- _ - . 



0-71 



6050 



0-94 











6000 









0-88 



5890 



1-00 



1-00 



1-00 



5600 



1-19 











5550 



.... 





1-33 



5230 



1-76 









5180 







1-17 





5170 











2-08 



4860 



2 '27 







311 



4640 











4-16 



4590 



4-97 









4550 







5-88 







4440 











5-33 



4260 











6-66 



These values, taken in comparison with those given in Table 

 II, afford us abundant evidence that the light of arc-lamps 

 with small carbons is very much whiter than that emanating from 

 the large commercial lamps of the present day, with carbons of 

 half inch diameter. This conclusion is quite in accordance 

 with the results of recent investigations of the economy of the 

 arc-light, in which it has been shown that the candle-power in- 

 creases with the current-density.* Certain experiments upon 

 the ratio between total radiation and luminous radiation from 

 the electric arc, now in progress in the Physical Laboratory, 

 point unmistakably to the same conclusion. 



Daylight. 



The method pursued in the determination of the color of 

 daylight, was in every respect similar to that employed in the 

 study of the lime and arc-lights. The rays falling upon 



* Schreihage (La Lumiere Electrique, xxix, p. 585), has shown that the result 

 of diminishing the cross section of the carbons of an arc-lamp from 254 dmm to 

 40 dmm was to increase the mean spherical candle-power from 161 candles to 471 

 candles, with corresponding increase in the efficiency of the lamp. 



