Transactions of the Optical Society
Introduction, T w o main methods of measuring to a high degree of accuracy the refractive properties of a specimen of glass are in generalsuse. I n the first method the specimen is cut in the form of a prism, usually of angle 60". It is placed on the table of a spectrometer and the angle of minimim deviation for any given spectrum line is measured. The refractive index of the glass for light of the given wave-length is then calculated from the values of the angle of the prism and of the angle
... and of the angle of minimum deviation. A modification of this method has been employed by Abbe and others*. The glass is cut in the form of a 30' prism and the incident beam of light, after refraction at the first face and normal reflection at the second face, is made to retrace its path, the observations being made by means of an auto-collimating arrangement. The second method depends on the determination of the path of a ray which enters a block of glass of known refractive index at grazing incidence at the surface of contact with the specimen, the block having a higher refractive index than the specimen to be measured. In this case the specimen of glass must be provided with one or two optically plane surfaces. The Abbe and Pulfrich refractometers are familiar examples of instruments which depend on this method of measurement. I n addition to the methods mentioned there are of course others, such as those which depend on interference measurements ; here again the provision of one or more optically plane surfaces is necessary. None of these methods, however, can be applied directly when one wishes to determine the refractive properties of a specimen of glass which is unpolished, or is in the form of a lens, or is very small. It is frequently either inadvisable or impossible to grind and polish optically flat surfaces on a specimen on account of its being a component of some optical system which it is essential to keep intact, or because of its small size. In such cases the usual method adopted for measuring the refractive index for any given wave-length is to immerse the specimen in a transparent liquid having approximately the same index, and to vary the concentration of the liquid until the refractive indices of liquid and specimen for that wavelength are the same. T h e refractive index of the liquid is then measured by means of one of the ordinary methods. If the concentration of the liquid is so adjusted as to equalise the refractive indices for one wave-length, it will be found that the 45 (1894, 609. * E. Abbe, A'eue Apparate ziw Bestawmting des Bueclrzingstoid Zerstreuungs-uevmr3gens festev ttndflz2ssi:er Kdvfier, Jena (1874); S. Czapski, 2 . f . Instvk. IO (1890), 361; C. Pulfrich. W i d . Ann.