Page:Journal of the Optical Society of America, volume 33, number 7.pdf/5

 II. SAMPLES MEASURED

Prior to his death, Walter T. Spry, then manager of the Munsell Color Company, deposited one or more samples of all of the original paintings of the standards in the Munsell Book of Color with the Colorimetry Section of the Bureau. He also deposited repaints of all colors the original paintings of which had become depleted, together with new colors prepared up to 1935. In selecting the samples of each color to be measured, that painting was chosen which matched the color chip of the same designation in the Munsell Book of Color. In most instances the color differences between the originals and their repaints were negligible but in several it was important to specify which painting was used. Therefore, for the purpose of accuracy and as a matter of record, the painting number of each sample measured is given.

The 2-value 2-chroma samples for the intermediate hues (10R, 10YR, 10Y, etc.) were painted independently of the other 2-value 2chroma samples and the colors and the data are not as congruent with the other samples as they are with each other. These samples, as well as several 8-value 2-chroma samples for the intermediate hues, are not included in the Munsell Book of Color, but they were measured and the data are included in the present paper for the sake of completeness. One new sample, 10YR 8/8, recently received, is included. The complete list of samples measured is given in Table II.

The samples in the Munsell Book of Color were inspected under a strong source of ultraviolet radiant energy and also under a strong yellowish green light for fluorescence which might vitiate the spectrophotometric measurements (15). No fluorescence was observed under either illuminant.

IlI. METHODS OF MEASUREMENT AND COMPUTATION

Spectral reflection curves of all of the samples noted and listed in Table II were run on the General Electric recording spectrophotometer at the National Bureau of Standards. The samples were run relative to magnesium oxide (16), with approximately 4-mμ slits and over a wave-length range from 400 to 750 mμ. The samples were backed with black paper for these measurements. Calibration curves were run on each sheet enabling corrections to be applied to the data for wave-length errors, for 100 percent and zero curve deviations, and for aging of the magnesium oxide comparison surface, in accordance with methods regularly used at the National Bureau of Standards (17), (18).

As already noted, the colorimetric computations were made for four different illuminants. ICI Illuminants A and C have become well

established in colorimetric work. Illuminant A is the Plankian radiator or blackbody at 2842°K (C$2$=14,320 micron-degrees, or 2848°K with C$2$= 14,350); the color temperatures of common incandescent illuminants vary from about 2600°K to about 3100°K. Illuminant C is that produced by a source at 2842°K combined with a certain Davis-Gibson daylight liquid filter (79). On the “OSA excitations” basis (used in the design of the Davis-Gibson filters) the resulting color matched that of a Plankian radiator at 6500°K. On the basis of the ICI data the approximate color temperature of this lamp-and-filter combination is 6800°K. The color and spectral energy distribution of ICI Illuminant C satisfactorily match those of overcast sky or average daylight for colorimetric use. Illuminant D is that produced by an illuminant at 3000°K combined with a Macbeth (Corning) daylight glass filter giving a color temperature of approximately