82 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No. 5 
to all three mentioned, with the added one that for many would-be 
observers there is great difficulty in deciding on any values at all. 
Moreover the magnitude of the effects concerned is not, in general, 
negligible in comparison with the accuracy expected in practical 
measurements. 
I think I have enlarged on this point sufficiently to make clear 
what I wish to emphasize, namely, that the ratio of the subjective 
brightness of two lights of different color is not a definite and single- 
valued quantity, but depends upon the conditions of observation as 
well as upon the characteristics of the individual observer’s eye. In 
particular, in order to make values definite it is necessary to specify 
more or less precisely the size and brightness of the photometric field, 
which determine the parts of the retina used and its level of adapta- 
tion. It is also necessary to include observations by a sufficient 
number of individuals to approximate the results which would be 
obtained by an imaginary average, or normal, observer. It is obvious 
also that relative values determined under these limited conditions 
must not be applied indiscriminately. For example, a comparison 
of two lights under the standard conditions may not show at all 
accurately their relative effectiveness at the low illuminations often 
used on highways, nor does a measurement of brightness necessarily 
show accurately the relative merits for work on fine details. 
Nevertheless it is useful and almost necessary to have some means 
of describing in a quantitative and comparable way the different 
kinds of light we have occasion to use. We actually are measuring 
them continually and I shall try to outline briefly the present status 
of methods for such measurements. 
Direct Equality-of-Brightness Observations.—The earlier illuminants 
were nearly alike in color. Consequently the complications just 
described did not become practically important until recent years. 
By. the time these problems arose, excellent instruments had been 
developed and photometric laboratory apparatus was more or less 
standardized throughout the world. The natural course, therefore, 
was to use these instruments and methods and to do the best one 
could with them in measuring the newer illuminants. The rapid 
development in succession of metallized carbon, tantalum, vacuum 
tungsten and gas-filled tungsten lamps has accentuated the diffi- 
culties and has made them affect the most common types of com- 
mercial photometric work. Other developments, such as mercury 
vapor lamps, nitrogen discharge tubes, and neon lamps, as well as 
special filters and glasses to give color effects, have created still more 
serious difficulties for the photometric laboratory. 
ly ee 
