Page:Journal of the Optical Society of America, volume 33, number 7.pdf/32

 W.

TABLE III.—Continued.

3 All but seven of the following 26 papers are made from different pigments; seven are mixtures in order to fill up wide hue gaps. This series is particularly useful in disk colorimetry.

4 These papers have semi-gloss surfaces as evidenced by the difference in the Y tristimulus values for the two conditions of viewing, and as a result, the corresponding values of P. for the R and YR papers differ considerably. If these papers were very glossy, a difference in the Y values as great as 0.04 would be obtained for the two conditions of viewing, with a correspondingly greater difference in Pe. Thus, viewing and illuminating geometry me increasingly critical as surfaces depart from non-specularity, and if the geometry is not known, measurements on glossy chromatic samples are subject to misinterpretation.

5 Specular component included in spectrophotometric measurement.

6 Specular component excluded in spectrophotometric measurement.

ments have indicated that the viewing geometries of the spectrophotometer as originally manufactured, as well as the new geometry with specular component both included and excluded, give similar values of reflectance for samples of matt surfaces. Since the usual Munsell color chips have nearly matt surfaces, the two methods of viewing geometry are believed to give values of apparent reflectance that differ by less than 0.002. A few “special” papers possessed glossy surfaces, and are so noted in the tables. They were measured with the specular component both included and excluded.

Some of the samples also exhibit a slight iridescence which often has been termed as “bronze.” As bronze increases, viewing and illuminating geometry become increasingly critical.

Tristimulus values and trilinear coordinates have been determined for I.C.I. Illuminant C, using the 30 selected ordinate method. The graph paper on which the spectrophotometric curve was recorded had the selected ordinates printed thereon. Dominant wave-length and purity were read from large-scale sections of the I.C.I. mixture diagram in the Handbook of Colorimetry (8). Trilinear coordinates are reported to three decimal places, instead of the usual four, in order to call attention to the fact that the fourth place is accurate only when corrections for all instrumental and recording errors are applied.

The spectrophotometric measurements were made in the Interchemical Corporation Research Laboratories, and a complete set of calculations were compiled in the laboratories of the Food Distribution Administration. Each of the authors has had a part in checking the data.

This work was started in order to supply hue sensibility data for surface colors (9), also in