450 Table 159 (continued) 



RELATIVE SPECTRAL LUMINOSITY FOR THE HUMAN EYE AT VARIOUS 



LUMINANCE LEVELS 

 Basis of Table: 



The relative spectral luminosity data in Table 159 A are given for wave-length inter- 

 vals of 10 millimicrons. Those data presented in column 2 are applicable for normal 

 vision under good lighting conditions ; that is, with high luminance, viz, about 0.55 

 footlambert or more 8 (logarithm of the luminance in f ootlamberts = 0.25 or more). 

 The values of y\ in column 2 are those adopted by the International Commission on 

 Illumination (I.C.I.) in 1924," and apply only in the case of photopic vision (approaching 

 cone vision; observer adapted to high luminance as specified above). The maximum 

 ordinate (unity) of the relative spectral luminosity curve for photopic vision occurs at 

 a wave length of 555 millimicrons. 



At extremely low levels of luminance, the human observer uses scotopic vision (rod 

 vision), while in the intermediate levels of luminance he uses mesopic vision (mixed 

 rod and cone vision). 



According to the well-known Purkinje effect, the abscissa at which the maximum of 

 the relative spectral luminosity curve occurs is displaced towards shorter wave lengths 

 as the level of luminance to which the observer is adapted falls below about 0.55 

 footlambert (see columns 3-10, compared with column 2). 



Weaver 1 states : "Because of the shift of the luminosity function with luminance, 

 it is necessary to choose arbitrarily a light source of specified spectral composition 

 as a standard for the specification of the luminance scale through the mesopic and 

 scotopic regions. The illuminance or luminance obtained from sources of different 

 spectral composition may be evaluated by comparison with the standard source, by 

 means of a suitable photometer, or by computation, using a set of tables, such as are 

 described here." Weaver has adopted and employed a tungsten lamp at a color tempera- 

 ture of 2370° K. for G = 1.438 cm. deg. (2360 °K. for C» = 1.432 cm. deg.) as the 

 standard source for low level photometry and this standard underlies the data in columns 

 3-10. For reduction of the given data to radiant sources at other color temperatures the 

 reader is referred to Weaver's original article 1 (especially Figure 2). 



The values of the luminosity data in columns 3-10 must be regarded as provisional, 

 since Weaver derived them by interpolation and extrapolation. He had on the one hand 

 the I.C.I, luminosity curve with its maximum at a wave length of 555 millimicrons for 

 log luminance —0.25 (where luminance is in footlamberts). And, on the other hand, 

 he had a luminosity curve for scotopic vision based on the average of observational 

 data obtained by Hecht and Williams, 10 and by Weaver u for a luminance of 6.5 X 10"* 

 footlambert (log luminance —4.187). In addition Weaver 1 had some unpublished data 

 for the upper luminance range. Weaver * by means of interpolation and extrapolation 

 based on the above-mentioned observational data obtained the values presented in 

 columns 3-10. 



8 See reference 4, pp. 249-250. 



9 Proceedings of the Sixth Session, International Commission on Illumination (Geneva, 1924), p. 67. 

 See also Judd, D. B., Journ. Opt. Soc. Amer., vol. 23, p. 359, 1933. 



10 Hecht, S., and Williams, R. E., Visibility of monochromatic radiation and the absorption spectrum 

 of the visual purple, Journ. Gen. Physiol., vol. 5, p. 1, 1922-23. 



n Weaver, K. S., The visibility of radiation at low intensities, Journ. Opt. Soc. Amer., vol. 27, 

 p. 36, 1937. 



(continued) 



SMITHSONIAN METEOROLOGICAL TABLES 



