OSA's Digital Library

Journal of the Optical Society of America

Journal of the Optical Society of America

  • Vol. 23, Iss. 2 — Feb. 1, 1933
  • pp: 41–45

Optics InfoBase > JOSA > Volume 23 > Issue 2 > A Proposed Scale for Use in Specifying the Chromaticity of Incandescent Illuminants and Various Phases of Daylight

A Proposed Scale for Use in Specifying the Chromaticity of Incandescent Illuminants and Various Phases of Daylight

IRWIN G. PRIEST  »View Author Affiliations

JOSA, Vol. 23, Issue 2, pp. 41-45 (1933)

View Full Text Article

Acrobat PDF (635 KB)

Browse Journals / Lookup Meetings

Browse by Journal and Year


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools


No abstract available.

IRWIN G. PRIEST, "A Proposed Scale for Use in Specifying the Chromaticity of Incandescent Illuminants and Various Phases of Daylight," J. Opt. Soc. Am. 23, 41-45 (1933)

Sort:  Author  |  Journal  |  Reset


  1. For meaning of the term "chromaticity" see O. S. A. Chrometicity Report, 1920–21, J. O. S. A. 6, 535 (1922), footnote 8.
  2. These considerations apply to strictly pyrometric as well as colorimetric computations.
  3. Consider, for example, the formulas of Wien and Planck for spectral distribution of energy; and see especially the following papers for illustration of the convenience of reciprocal temperature in computation and specification: Langmuir and Orange, Trans. A. I. E. E. 32, 1946 (1913); Ch. Fabry, Trans. I. E. S. 8, 317–318 (1913); Foote, Mohler and Fairchild, J. Wash. Acad. Sci. 7, 545–549 (1917); H. P. Gage, Trans. I. E. S. 16, 428–429 (1921). Note especially the convenience and simplicity of reciprocal temperature in expressing the additive value of a blue filter in changing the apparent reciprocal color temperature of a source viewed through this filter.
  4. The practical use of the color of an incandescent solid as an index of other properties dependent fundamentally upon temperature, even though the observer may have had no concept of temperature, is an exceedingly ancient practice, the origin of which is lost in the mists of antiquity. It was a technique doubtless used by ancient prehistoric smiths. In recent times, within the past century, this practice of estimating temperature by means of observations on the color of a furnace or an incandescent body has been developed and used industrially to a considerable extent but largely by tradition and "rule of thumb" without definite scientific formulation. In connection with the development of the incandescent filament lamps, the concept of color temperature finally came to be formulated much more definitely. The formal definition of the term color temperature, the relation of color temperature to the several "brightness temperatures" and the relation of color temperature and brightness temperature to "true temperature" were discussed comprehensively by Hyde, Cady, and Forsythe in their paper Color Temperature Scales for Tungsten and Carbon, Phys. Rev. [2] 10, 395– 411 (1917). Prior to this, Patterson and Dudding in their paper on The Estimation of High Temperatures by the Method of Color Identity (Proc. Phys. Soc. (London) 27, 230–262 (1915)) had made extensive use of the concept of color temperature which they called "the color identity temperature." Hyde and his several associates had used the concept of color temperature in a quite definite manner in several papers earlier than this, although the term color temperature was apparently not used in these earlier papers. See particularly the following: Hyde, New Photometric Methods of Studying the Radiating Properties of Various Substances, Phys. Rev. 27, 521–522 (1908); Hyde, Cady, and Middlekauff, The Selective Emission of Incandescent Lamps as Determined by New Photometric Methods, Trans. I. E. S. 4, 334–354 (1909); Hyde, The Physical Production of Light, J. Frank. Inst. 169, 439–466 and 170, 26–45 (1910); Hyde, The Synthetic Development of Radiation Laws for Metals, Astrophys. J. 36, 89–132 (1912); Langmuir, Tungsten Lamps of High Efficiency—I, Trans. A. I. E. E. 32, 1913–1933 (1913); Langmuir and Orange, Tungsten Lamps of High Efficiency—II, Trans. A. I. E. E. 32, 1935–1946 (1913).The idea of grading illuminants in terms of color temperature was further developed and illustrated by Hyde and Forsythe in their paper Color Temperature and Brightness of Various Illuminants, Trans. I. E. S. 16, 419–430 (1921). The definition of the term color temperature was given further consideration in the discussion which followed this paper (pp. 429–430). The following are subsequent papers bearing on this same subject by Hyde’s associates at the Nela Research Laboratory: Forsythe, Accuracy in Color Matching of Incandescent Light Sources, J. O. S. A. 6, 476–482 (1922); Forsythe, The Temperature and Brightness of Tungsten Lamps, Gen. Elec. Rev. 26, 830–834 (1923); Forsythe, Color Match and Spectral Distribution, J. O. S. A. 7, 1115–1121 (1923); Forsythe and Worthing, The Properties of Tungsten and the Characteristics of Tungsten Lamps, Astrophys. J. 61, 146–185 (1925); Forsythe, Temperature Radiation, J. O. S. A. 16, 307–328 (1928).
  5. It should be noted that in giving color temperatures for this purpose we should adhere strictly to the following definition: "The color temperature of a radiator (source, lamp) is the (absolute) temperature at which a complete radiator must be operated in order to emit energy competent to evoke a color of the same chromaticity as the color evoked by the radiant energy from the source in question." Cf. Priest, J. O. S. A. and R. S. I. 7, 1180 (1923). For certain other purposes a slightly different definition of color temperature is sometimes adopted either explicitly or tacitly. Such a definition is: "that temperature of a black body at which the relative emission intensities in some chosen two wave-lengths are the same as those of the radiating metal under investigation." (Phys. Rev. [2] 10, 396 (1917). The adoption of this definition would not be appropriate to our purpose of color-grading sources and in some cases the choice of the definition might make a significant practical difference in the statement of color temperatures characteristic of certain sources. It is for this reason that certain so-called physical methods for the determination of color temperature (Campbell and Gardiner, J. Sci. Inst. 2, 177–187 (1925); Winch, J. Sci. Inst. 6, 374–379 (1929); Sharp, J. O. S. A. 20, 62–70 (1930)) are not entirely adequate for the color-grading of sources. They would be fully adequate only in case it were known that the chromaticity of the source being investigated would match the chromaticity of the standard (the complete radiator) when the source in question had the same ratio of intensities as the standard for two specified wave-lengths (or specified spectral regions). The term "color temperature" is not an entirely happy one from the point of view of colorimetrics. It was once remarked by Dr. H. E. Ives that, where our interest is in specifying the quality of color (chromaticity) of the source rather than the temperature of the source, it would be logical to speak of temperature color rather than color temperature. Perhaps an improvement in terminology might be made in connection with the present proposal by giving the name "thermal chromaticity" or the name "radiant chromaticity" to the reciprocal of color temperature as defined here.
  6. Priest, The Complete Scale of Color Temperature . . ., Phys. Rev. [2] 20, 93–94 (1922); Priest, Colorimetry and Photometry of Daylight and Incandescent Illuminants . . ., J. O. S. A. 7, 1175–1209 (1923), particularly pp. 1190–1192, and Fig. 2 on p. 1184. See also Priest, A Method of Obtaining Radiant Energy Having the Visible Spectral Distribution of a Complete Radiator at Very High Temperatures, J. O. S. A. 5, 178–183 (1921).
  7. See, for example, Smithsonian Physical Tables, 7th Ed., 3rd reprint, p. 411 (1927).
  8. J. O. S. A. 7, 1190–1192 (1923), and Fig. 2 on p. 1184.
  9. Judd, J. O. S. A. 23, 7 (1933).
  10. Note carefully, however, that the conditions of these experiments were not such as to facilitate the finest discrimination, the purpose of these experiments being merely to obtain relative values of equally perceptible differences for different values of 1/θ, not to determine the very least values of this difference under the most favorable conditions for making the difference small. Under conditions of observation highly favorable to precise observation, the doubtfully perceptible difference is found to be approximately one-fifth of the values found in these experiments. (See Judd, Bur. Standards J. Research 5, 1161–1177 (1930).)
  11. Letter, H. P. Gage, to K. S. Gibson, December 9, 1925. Dr. Gage had noticed that a linear relation existed between the thickness of a plate of "Daylite" glass and the difference between the reciprocal of the color temperature of a source and the reciprocal of the apparent color temperature of the same source viewed through the plate of Daylite glass. Dr. Gibson had pointed out that a linear relation existed between the thickness of a plate of Daylite glass and the spectral centroid of a source as seen through the plate of Daylite glass. (J. O. S. A. 11, 473–476 (1925).) From these two statements Dr. Gage drew the conclusion that a linear relation must exist between the spectral centroid and the reciprocal of the color temperature.
  12. R. Davis, B. S. Res. Pap. No. 365, 672 (1931). There is a typographical error in the ninth line from the bottom of this page, . . . θ should be small δ which is the temperature difference (measured in degrees) which is equally perceptible at all values of temperature over a wide range. The essence of the relation which Davis points out is contained in his equation, δ = Kθ2 where K is a constant.
  13. Note that the rate of change of temperature with respect to reciprocal temperature is minus the square of temperature. In symbols dθ/d(1/θ)= - θ2. With sign reversed, the proportionality constant in the equation stating Davis' rule (see reference 12) is equal to the increment in reciprocal temperature corresponding to the doubtfully perceptible difference in temperature.
  14. It should be noted, however, that the relation between λc and 1/θ is not exactly linear and there are circumstances in which one should use a more accurate empiric relation for obtaining values of 1/θ corresponding to given values of λc, and conversely.
  15. B. S. Res. Pap. No. 365, 677–678 (1931).
  16. Forsythe, Trans. I. E. S. 16, 423–424 (1921).
  17. Computed from data given by Judd in terms of spectral centroid (λc), B. S. Res. Pap. No. 252, Fig. 5, 1175 (1930).
  18. B. S. Res. Pap. No. 252, 1170–1171 (1930).
  19. Priest, Measurement of the Color Temperature of the More Efficient Artificial Light Sources by the Method of Rotatory Dispersion, B. S. Sci. Pap. No. 443, July (1922).
  20. B. S. Sci. Pap. No. 443, Table I, p. 223. Considering all of these data, it is also found that the ratio of the maximum deviation of a single observation from the mean of ten to the probable error of a single observation varies, for the different observers, from 4.1 to 5.0.
  21. B. S. Sci, Pap. 443, Table II, p. 224.

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.

« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited