REFRACTIVE INDICES. 93 
On combining equations (i) and (2) we find 
N-n t = N-m 
N'-n't N'-m 
from which equation the refractive index n\ of the substance can be deter- 
mined with a high degree of accuracy. (Probable error in the fourth 
decimal place.) 
Although not strictly germane to the present subject, these observations 
and conclusions of Christiansen have been outlined above because of their 
bearing on the apparent refractive index of fine microscopic mineral aggre- 
gates embedded in glass, in which case the observed refractive index of the 
whole as determined by the immersion method is neither that of the mineral 
nor of the glass. Although such conditions may never arise in the study of 
natural rocks, in artificial silicate melts they have been observed and have 
been so interpreted. In case the mineral particles are birefracting and 
distributed irregularly and in overlapping aggregates, the scattering of the 
light by refraction and reflection is pronounced because of the differences 
in the refractive indices of differently oriented particles and the slide 
appears dusty and not properly transparent. 
In 1 892 J. L. C. Schroeder van der Kolk* described in detail practically 
the same method outlined by Maschke except that he obtained oblique 
illumination by placing a metal strip directly above the condenser and 
immediately below instead of just above the grain under examination. He 
added greatly to the number of liquids suitable for such work and published 
a listfof over 300 minerals arranged according to their refractive index, with 
notes on special individual features useful for their determination. The 
immersion method is frequently called the Schroeder van der Kolk method. 
In 1896 H. AmbronnJ described an immersion method for determining, 
under the microscope, the relative dispersion of two adjacent bodies and 
suggested its use in practical microscopic diagnosis. He observed both the 
spectral colors produced by dispersion in the rear focal plane of the objec- 
tive and also the colored fringes formed along the margin of an immersed 
glass plate. With large grains the spectrum in the rear focal plane of the 
objective can be measured by means of the Bertrand lens and ocular with 
scale as used for interference figure measurements, and the refractive in- 
dices for the different colors may be thus determined, provided the focal 
length of the objective be known. His method has not been used, how- 
ever, by microscopists to any extent. 
In 1900 several changes were suggested by the present writer to facilitate 
the application of this immersion method to petrographic microscopical 
work. In place of the opaque strips immediately above or below the prepa- 
ration, a sliding shutter with different-sized apertures was attached below 
the polarizer and could be inserted or withdrawn at will. Still simpler and 
equally efficient is the finger, which is placed just below the polarizer and 
thus casts a shadow over any part of the field. Practically the same effect 
*Zeitschr. f. wiss. Mikroskopie, 8, 456-458, 1892; Kurze Anleitung zur mikroskopischen ICrystallbestim- 
mung. Wiesbaden. 1898. 
tTabellcn zur mikroskopischen Bestimmung der Mineral ien. Wiesbaden. 1900. 
JFarhenerscheinungen an d. Grenzen farbloser Objekte im Mikroskop. Her. Sachs. Gesel. d. wissen. math, 
phys. KJ.. 1896. i-8. 
Tschermak's miner, petrog. Mitteil.. 30, 339, 1901; also Amer. Jour. ScL (4). 17, 385-387. 1904. 
