MAR. 4, 1923 CRITTENDEN: MEASUREMENT OF LIGHT 89 
flicker data and the results of the other determinations mentioned. 
All of these were carried out with such care and thoroughness that 
they should serve to establish a standard curve which would approxi- 
mate as closely as can be expected to the characteristics of the hypo- 
thetical average observer. 
When such a standard visibility curve is established, the total light 
value of radiant energy of any quality for which spectral distribution 
curves can be obtained will become merely a matter of calculation 
or of graphic determination. It is evident, however, that these 
visibility curves are subject to all the limitations of brightness meas- 
urements. Using them, consequently, does not solve all our diffi- 
culties. Moreover, the experimental difficulties in obtaining reliable 
spectral distribution curves are considerable, and the aggregate errors 
of calculated values may often exceed the uncertainty in the direct 
comparison by simpler methods. Values calculated from spectral 
measurements and visibility curves are in fact hardly capable of an 
accuracy such as is most commonly desired in photometric results. 
In this connection mention should be made of two interesting possi- 
bilities in the way of more reliable comparisons by means of visibility 
curves, both of which have been described in recent years before the 
Society. One is the combination of Nicol prisms and quartz plates, 
devised by Priest, which constitutes a color filter whose spectral 
transmission curve is adjustable over considerable limits and can be 
calculated with a high degree of certainty. These calculations as 
expressed in terms of light must, of course, depend on the visibility 
curve, but having a standard visibility curve, the results obtained by 
this device should ke reliable. | 
The other possibility is that of a physical photometer consisting of 
an instrument for measuring radiant energy, combined with a filter 
which in effect adjusts the spectral sensibility of the radiometer to 
make it correspond to that of the standard curve adopted as repre- 
senting the sensibility of the eye. The refinement of laboratory 
technique required to make accurate measurements of radiant energy 
is such that this method of measurement does not appear likely to 
find very extensive application, but its possibilities may be worth 
further investigation. 
CONCLUSION 
While applications of the visibility curve will undoubtedly con- 
stitute a valuable supplement to more direct measurements, the situ- 
ation will never be entirely satisfactory unless the values thus calcu- 
