18 proceedings: philosophical society 



ized vacuum tungsten lamp. A cylindrical acetylene flame was used 

 as a source of spectral light. The distribution of energy in the spec- 

 trum of the acetylene flame was determined with great care, in view 

 of the fact that the disagreements in previous work seemed to be due, 

 in part, to uncertainties in radiometrically evaluating the light stimu- 

 lus. Sensibility curves were obtained for 130 persons, of which num- 

 ber 5 were color-blind. The visibility curve of the average normal 

 eye, using 125 observers, was found to be wider than previously ob- 

 served. These data, obtained by the use of a flicker photometer, were 

 given. Only a few observers were able to make accurate settings with 

 the equality-of-brightness photometer. 



As was to be expected, the visibility curves of no two persons appear 

 to be exactly alike. When a visibility curve does not coincide with 

 the average there is usually a marked departure from the average visi- 

 bility in a given spectral region. This gives rise to (1) wide visibility 

 curves with the maximum shifted toward the red, i.e., "red sensitive," 

 (2) narrow curves with a sharp maximum in the green, and (3) curves 

 with the maximum shifted toward the violet. The data available 

 indicate that among a group of persons having normal color vision 

 about 20 per cent are (1) red sensitive, (2) blue sensitive, or (3) 

 average, while in 10 per cent of the cases examined the color sensi- 

 bility falls below the average (1) in the red or (2) in the blue or (3) falls 

 below the average in both the red and the blue, thus giving rise to an 

 apparently high sensitivity in the green. One person in 20 has a very 

 wide visibility curve, while 4 per cent are color-blind, i.e., they confuse 

 colors. The point of maximum sensibility was found to be very dif- 

 ferent for different observers; for the 125 persons the mean maximum 

 is at \ m = 0.5576 /x. The curve of average visibility, when corrected 

 for the selective transmission of the ocular media, including the yellow 

 spot, is very symmetrical. 



An empirical equation of the visibility curve determined was given. 

 Using this visibility equation and Planck's equation of black-body ra- 

 diations, calculations were given of the luminous energy emitted by a 

 black body at various temperatures; also the luminous efficiency, the 

 Crova wave-length, and the mechanical equivalent of light. Using 

 the recent measurements of the brightness of a black body, as deter- 

 mined by Hyde, Cady, and Forsythe, and the most probable values 

 of the radiation constants (C2 = 14,350, a = 5.7) the value is 1 lumen 

 = 0.00161 watt of luminous flux; or 1 watt (of radiation of maximum 

 visibility) = 621 lumens = 49.5 candles. The direct determination of 

 613.5 lumens of green mercury radiation, X = 0.5461 /x, (made by 

 Ives, Coblentz, and Kingsbury, using 61 observers) when corrected 

 for visibility V (at X = 0.5461 M ) = 0.985 V m gives 622.8 lumens per 

 watt, in good agreement with the present determination. 



Discussion. Mr. Emerson emphasized the importance of taking 

 the last observations; if only the first 40 had been used the curves would 

 have been shifted. Mr. White asked whether the two eyes of an ob- 



