44 Merwin and Larson — Mixtures of Amorphous 



incandescent electric, for example — is preferable. The wave- 

 lengths to which artificial light transmitted by thin films — 

 •05 to "10 thick — of several of the mixtures is approximately 

 equivalent are given in the table. , 



The material under examination should be finely powdered 

 so that the mixture may be pressed out into a thin film. If the 

 material is colorless and no screen is used the excess of white 

 light passing through it may obscure the light-aud-shadow 

 effects at the boundaries of the larger grains. Smaller grains 

 that are more deeply covered by the mixture should be 

 observed. 



If it is desired to obtain results with errors as small as 

 ± "005 it is essential to use monochromatic light or suitable 

 screens, owing to the extremely high dispersion of Se. For 

 Se the two Na lines have a difference of refraction of "004. 

 Two lines having the same difference of wave-length near the 

 Li line have a difference of refraction of '0013. Lithium light 

 can be used with all the mixtures. If it is contaminated by 

 sodium light the sodium light may be filtered out by using the 

 screen of Se above described ; or a film of celluloid or gelatin 

 stained with methyl violet may be used. The Na-flame can 

 not be used with mixtures containing more than about 70 per 

 cent Se. 



A spectroscope monochromatic illuminator may be used to 

 advantage, particularly as a time-saver. With it the necessity 

 for making up a mixture which exactly matches the refractive 

 index of the substance studied for a particular wave-length is 

 obviated. An approximate match may be made in white light, 

 then with the same mixture the wave-length at which the 

 refractive indices are alike can be found by means of the 

 illuminator. As most transparent solids have very much less 

 dispersion than the mixtures, another mixture can now be 

 selected which will match the refractive index of the substance 

 for a very different wave-length. Then the refractive index 

 of the substance for one or more of the commonly used 

 standard wave-lengths {Li, JVa, Tl) can be obtained without 

 great error by interpolating along a straight line drawn on 

 the chart between the twO points representing the . observed 

 refractive indices* 



* All the available high refractive indices of minerals of known dispersion 

 have been tested by plotting on the chart. Those for which interpolation 

 gives the greatest errors are sphalerite and sulphur (fi). The refractive 

 indices of these minerals for certain wave-lengths are represented on the 

 chart. The slopes of lines through any given point indicate the relative 

 dispersion of the substances which the lines represent. For example, the 

 slope of the line for sphalerite is about 45° ; if it were 60°, the dispersion 

 would be greater ; if 90°, the dispersion would be still greater and equal to 

 that of the mixture. For most minerals the slope is about 45°. It is obvious 

 that after having obtained and plotted a first value of the refractive index 

 for a chance wave-length, a line through the point with a slope of about 45° 

 will indicate approximately the mixture to use in finding the refractive index 

 for any desired wave-length. 



