named surface colors adequate for the purpose, and any arrangement or organization of the colors serves only the secondary purpose of assisting the user to find the one which most nearly matches.
The Maerz and Paul Dictionary of Color [97] is the foremost authority on color names. It contains about 7000 different samples of color printed on semiglossy paper, and there are listed about 4000 color names which are keyed to one or another of the color samples. These names are drawn from usage in many fields: paint, textile, ceramic, scientific, technical, and artistic. The samples are also identified by plate, column, and row, and because of their large number and fairly uniform color distribution it is usually possible to find among them a sample approaching what is called a "commercial color match" for any given uniform opaque surface. On this account the Maerz and Paul Dictionary finds a considerable application as a collection of color standards quite separate from its primary function of defining color names. There are noticeable color differences between corresponding samples in different copies of the Dictionary, but the differences have been held to a reasonably small amount by discarding the less satisfactory printed sheets.
The accepted authority for color names in the textile and allied industries is the Color Association of the United States. This association has issued nine editions of a standard color card since 1915, the current edition [149] containing 216 color samples of pure dye silk. Furthermore, the association issues to its members several seasonal cards each year. All colors of these standard and seasonal cards are identified by name and cable number. The standard colors have been measured by spectrophotometric and colorimetric procedures, and luminous reflectance, , relative to magnesium oxide, and chromaticity coordinates, , for source C have been published [137].
A color dictionary much used for the specification of the colors of flowers, insects, and birds was prepared in 1921 by Eidgway [139]. This outstanding pioneer work contains about 1000 named color samples of paper painted by hand. Each chart shows columns of colors of the same dominant wavelength progressing from each chromatic color at the middle of the column toward white and the top, and toward black at the bottom; and there are five series of such columns, each one encompassing the entire hue circuit, but at different purities. Many of the names were coined at the time of publication to fill in gaps in popular color nomenclature and so have not much descriptive value. Each sample is arbitrarily identified by column, row, and series, however. In addition, there is an alphabetical list of the color names giving this identification.
The notation of the Ostwald system is based on the properties of idealized pigment surfaces having spectral reflectance constant at a certain value between two complementary wavelengths and reflectance constant at a certain other value at other parts of the spectrum [30, 128]. The full colors are those that have the low values of spectral reflectance equal to zero and the high ones equal to 100 percent. The difference between these two reflectances for other idealized pigment surfaces is the full color content, the value of the low reflectance is the white content, and the difference between the high reflectance and 100 percent is the black content. The complete Ostwald notation consists of a number and two letters. The number indicates dominant (or complementary) wavelength on an arbitrary but approximately uniform perceptual scale, and is called Ostwald hue number. The first letter indicates white content, a being a white content of 89.13 percent, which is as near to 100 percent as is practicable for usual pigment-vehicle combinations, and other letters in alphabetical sequence indicating decreasing white content on a logarithmic scale. The second letter indicates black content, a being a black content of 10.87 percent, which is as near to zero as is practicable, and other letters in alphabetical sequence indicating increasing black content on a logarithmic scale. The logarithmic scales were thought by Ostwald to insure uniform color scales, but this is true only to a rough approximation. Since the percent white content, black content, and full-color content must necessarily add up to 100, no explicit indication of the latter is required.
The Ostwald ideas have been a considerable aid in thinking about color relationships on the part of those who duplicate colors by mixtures of chromatic pigments with white and black pigments, and they have served as a guide in the selection of combinations of such colors to produce pleasing effects. However, the use of these idealized pigment surfaces as a basis for a system of colorimetry has been hampered by the fact that actual pigment surfaces approximate them rather poorly, and by the fact that not all actual pigment surfaces can be color matched by one of these ideal surfaces. Still, color charts made up more or less in accord with the Ostwald principles have been widely used for color standards and for the selection of harmonizing colors [63, 127, 129, 148]. Of these, the Jacobson Color Harmony Manual [63] is pre-eminent not only because of its technical excellence, but also because Foss [30] has given a clear statement of which of the somewhat contradictory Ostwald principles were followed in its construction, and Granville and Jacobson [47] have made a spectrophotometric study of the color chips and have published luminous reflectance, , relative to magnesium oxide, and chromaticity coordinates, for every chip. These chips are therefore valuable for use in colorimetry by difference from a working standard (see sec. 3), and the fact that the chip is a member of an orderly arrangement of; colors facilitates the selection of a working standard for any particular purpose.
26
