50 



HOW WE SEE 



during one publication year, 1944, there appeared 

 in the Journal of the Optical Society of America 

 12 articles in which photometric measurements 

 and concepts were used in connection with visual 

 data. (This count does not include the reports 

 of the OSA Committee.) Seven of these articles 

 used the old terminology and concepts, two the 

 new OSA nomenclature, and three the Moon (67) 

 nomenclature . This situation is confusing enough 

 to the visual e.xpert when he tries to convert 

 concepts and measurements of one kind to those 

 of another. The Naval engineer or scientist, 

 who may be unaware of this situation, however, is 

 likely to conclude that an article which talks 

 about "pharosage," "helios," and "heliosent," 

 and e.xpresses measures in "blondels," for ex- 

 ample, has nothing in common with another 



The Techniques of Photometry. If the spectral 

 energy distribution of a stimulus is known, it is 

 possible to compute the corresponding photo- 

 metric quantity without making any additional 

 measurements. It is possible to do this because 

 the spectral sensitivity of the "average" eye is 

 known. Actually, of course, the "average" eye 

 is an abstraction based on the data obtained 

 from a fairly large sample of normal eyes. The 

 average normal luminosity values adopted by the 

 International Commission on Illumination and 

 the American Standards Association (100) are 

 shown in Fig. 39. This curve— the photopic 

 relative luminosity curve discussed in the chapter 

 above — shows that one watt of radiant flux of 

 555 mix falling on an average eye will result in 

 660 lumens of light. One watt of radiant flux of 



400 500 600 700 



WAVELENGTH INlnM 



Fig. 39. Absolute and relative photopic lumi- 

 nosity values for radiant flux of various wave- 

 lengths computed for the average eye. (Data 

 from I.e. I., 100) 



article which talks about "illuminance" and 

 "luminance," and expresses its measures in 

 "millilamberts." The development of a common 

 language is essential if the visual scientist wants 

 to be understood not only by his colleagues, but 

 by other scientists and engineers as well. 



Standardization should be on an international 

 scale and should be sponsored by the Bureau of 

 Standards of the major countries of the world. 

 Other government agencies, e.g., the Navy 

 Department, can probably do much, however, to 

 expedite the completion of such a program. In 

 the meantime, individual investigators can also 

 further this goal by using a consistent set of 

 nomenclature and units in their own work. The 

 proposals of the OSA Committee (16, 17, 18, 19) 

 represent the considered opinions of a group of 

 leading visual scientists in this country. They 

 are consistent, simple and rational, and, as such, 

 are recommended for use in all visual work. 



5 60 



500 600 



WAVELENGTH IN m^ 



Fig. 40. The range of individual differences in 

 the relative luminosity function among 52 ob- 

 servers studied by Gibson and Tyndall (28). 



627 m^i will result in 200 lumens of light, etc. 

 It is apparent, then, that if the radiant flux per 

 wavelength interval is known for a stimulus, the 

 corresponding amount of light seen by an average, 

 light-adapted eye can be computed from the data 

 in Fig. 39. Moon's textbook on illuminating 

 engineering (68) actually defines photometric 

 measures in this mathematical way. Thus, for 

 example, he defines luminous flux as "radiant 

 power evaluated with respect to the standard 

 visibility [luminosity in the new OSA terminology] 

 function." 



Actually, this method of photometry is very 

 difficult and is practically never used by the 

 ordinary visual scientist. Most commonly, as 

 was described in the text above, the light to be 

 evaluated is compared with a standard light of 

 known luminosit}^ using an observer's eye as a 

 null indicator. A summary of these techniques 



