tive reproduction should be so selected that mixtures of the transmitted radiations are capable of matching a maximal number of colors. This means that the radiations in question should evoke hues approximating as nearly as possible to spectral saturation, and so distributed in the color triangle that the lines joining them lie maximally close to the spectral locus. Two-color reproducing filters in practice usually resemble closely the red and green members of a three-color set but may have a somewhat reduced saturation, and the dominant hue of the green is usually shifted somewhat more towards the blue than is that of the red number.
B. Standard Solutions.—The spectral transmissions of dyes and inorganic salt solutions of known purity and concentration are in course of determination at the Bureau of Standards, and some of the most important of these will be presented in later Reports by the present Committee.[1]
C. Lovibond and Other Colored Glasses.—Similar statements apply to the Lovibond glasses, which are widely used as technical standards in the ranking of oils and other materials as regards color. There are a very large number of these glasses, and the accuracy with which they are reproduced in different sets is often relatively low (77). The spectral transmissions of a considerable number of precision-made glass plates of various colors have been determined at the Bureau of Standards, and these plates can be borrowed under proper restrictions by those desiring to check their spectrophotometric equipment.
4. Standards of Spectral Reflection
