ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 183 



to severe strain and tires quickly. Fortunately, however, it is possible, 

 by modifying the conditions of observation slightly, to render the 

 phenomena more easily visible, and thus to relieve the eye-strain to a 

 large extent, and at the same time to increase the accuracy of the 

 determinations. These modifications involve both the sources of light 

 and a new method of two-fold oblique illumination. 



Sources of Light.— In place of the sodium-flame ordinarily used as 

 source of monochromatic light, the following light sources have been 

 substituted : Mercury light, helium light, and either a calcium -flame or 

 a molybdenum- or tin-spark. With this array of lights set up side by 

 side on the dark-room table, and in conjunction with a monochromatic 

 illuminator, or a dispersion prism or suitable ray filters (Wratten 

 mercury-line filters), the following spectral line sources are available : 

 A = 546*1, 558 to 561 (average about 560), 577 and 579 (average 578), 

 and 588 /x/x. With these lights it is not difficult to determine between 

 which two of the four available lines (546, 560, 578, 588 /xfx) the 

 refractive indices of mineral and liquid coincide, the liquid having the 

 higher refractive index for the shorter wave-length, and the mineral 

 the higher index for the longer wave-length. Now, the refractive index 

 of solids increases about O'OOl for a decrease in wave-length of 

 10-20 nfji, while for liquids the change is approximately twice as great. 

 If, therefore, the refractive index of a mineral be accurately measured 

 for any wave-length between 546 and 58.S, its index for the wave-length 

 589 fjifji (D line) can be estimated with an error not exceeding ± 0"001, 

 and a liquid then prepared of exactly this index, whereupon the esti- 

 mated refractive index of the mineral grain can lie checked by im- 

 mersion in the new liquid. By use of this arrangement a considerable 

 amount of time has been saved in the routine measurement of the re- 

 fractive indices of fine crystal grains. Occasionally the monochromatic 

 illuminator (Hilger type, with Nernst light filament and ground-glass 

 diffusing screen) has been found useful for ascertaining approximately 

 the wave-length for which the refractive index of the grain coincides 

 with that of the enveloping liquid. 



Xew Method involving Two-fold Oblique Illumination. — Oblique 

 illumination is obtained ordinarily by means of a sliding stop below the 

 condenser of the Microscope. This stop is purposely not sharply 

 imaged in the object field, but appears as a shadow with a hazy edge, 

 which passes gradually into the brightly illuminated part of the field. 

 The mineral grains are placed in this transition shadow-edge between 

 light and dark, and the illumination of their edges both in white and in 

 monochromatic light is observed. Because of the prismatic refraction 

 of the inclined edges of such grains, the intensity of illumination of 

 edges adjacent to the shadow is different from that of the opposite 

 edges, when the refractive index of the grains is different from the re- 

 fractive index of the liquid in which they are immersed. These dif- 

 ferences become less distinct lis the refractive index of the liquid 

 approaches that of the mineral ; and, if the refractive indices differ by 

 only ± O'OOl, the intensity differences in illumination are difficult to 

 see, because of the relatively large amount of light in the field. To 

 reduce the field illumination, and thus to increase the differences in 



