This photometer was exhibited at the Washington meeting of the Optical Society when this paper was presented, Oct. 28, 1922. (J. O. S. A. and R. S. I.7, p. 75–76.)
To save space and needless duplication, explicit definitions are given only for terms which might not be understood generally by photometricians in the sense in which they are here intended by the author. All technical photometric terms used but not explicitly defined in the paper are to be understood in the sense of the definitions given by the I. E. S. Committee on Nomenclature and Standards.
Sensation is here understood to mean a discrete element of consciousness, normally dependent upon the normal functioning of a sense organ. Colors are thus the discrete elements of which the visual consciousness is composed. All that is immediately contributed to consciousness by monocular vision is a pattern of colors composing the field of view. The visual recognition of shapes and forms is possible only as a result of sensibility to color differences. Therefore, form cannot logically be placed in the same category with color as is sometimes done; it is not elemental in the same sense.
Luminosity and “brightness” have been generally used in this sense. Their use as sensation terms is objectionable because of their other better recognized connotations which are distinctly physical, referring to the stimulus.
All of the foregoing definitions, while not verbatim copies, are entirely consistent with the definitions given in the “Report of the Colorimetry Committee,” Jour. Opt. Soc. Am., August, 1922. Reference should be made to this report for a more detailed discussion. They are not entirely consistent with the definitions given by the I. E. S. Committee on Nomenclature and Standards (e.g. definition 18, in the Report for 1918). A full statement of the arguments pro and con will be found in footnote 2, p. 531, Jour. Opt. Soc. Am., August, 1922.
Cf. “Report of the Colorimetry Committee …” Jour. Opt. Soc. Am., 6, p. 538; 1922.
In such extreme cases, where the temperature is such a purely hypothetical thing, I have heretofore sometimes spoken of “apparent color temperature” but this expression is essentially redundant, since all color temperatures are merely apparent temperatures. H. E. Ives has suggested that, in case our primary concern is the specification of color rather than temperature, we might well speak of “temperature color.”
Cf. , Fig. 1.
That is, a distribution which may be represented by the Planckian formula for some value of T which need not be the true temperature.
It is not possible to take space to even outline here the elementary knowledge of the polarization of light, the rotation of the plane of polarization by crystalline quartz and the rotatory dispersion associated with this rotation. This information may be found in the general text books on physical optics; and, on the basis of it, the formulas here given for the relative transmission of the system of nicol prisms and quartz plates may be readily derived. The reader who does not care to go into this may take it as an empiric fact that a quartz plate placed between two nicol prisms so that the path of light (line of sight) is coincident with the crystalline optic axis does constitute a selective filter of radiant energy; and he must take, on faith, the formulas (2 and 3) which give the relative transmission of such filters as a function of wave-length.
The accuracy with which Planckian spectral distribution curves can be duplicated by the use of the rotatory dispersion filter, has been illustrated heretofore (, Fig. 7, p. 242, and , Fig. 6, p. 230); and will be illustrated and discussed more fully in the B. S. Sci. Pap. edition of the present paper.
The technic of constructing such curves will be explained more fully below.
In graphic terms λc is the λ-coordinate of the “center of gravity” of the area included between the “luminosity curve” and the λ-axis.
These values apply only to certain experimental conditions. Absolutely smaller probable errors can probably be obtained under improved conditions of observation.
It is worthy of note that “color blind” observers, that is those whose discrimination of quality of color is not keen, make intensity measurements by way of brilliance match with a precision better than normal.
. Preliminary announcement of results, Jo. Opt. Soc. Am.7, pp. 68–69; 1923.
These intermediate results which cannot be considered in this paper will be dealt with fully in a forthcoming B. S. Sci. Pap. under the same title.
It is assumed that the reader is already familiar with the ordinary use of a pair of nicol prisms in photometry.