108 METHODS OF PETROGRAPHIC-MICROSCOPIC RESEARCH. 
of 5 to 10 with the true direction (optic normal or bisectrix) will be in 
error about 2 per cent of the true value for (7 a), (7 /3) or (0 a) ; an incli- 
nation of 10 to 15, about 5 per cent, while for 15 to 20 inclination the 
error may be as much as 10 per cent of the correct value desired. By means 
of the coordinate scale ocular described in Chapter V, the angular inclination 
of the section can be ascertained and the probable error due to this cause 
thus eliminated. 
In actual practice, therefore, the method of procedure in the determina- 
tion of the birefringence of a mineral plate in the thin section or a mineral 
grain is to measure first the thickness by one of the methods noted above 
and then to insert the graduated combination wedge and determine, under 
crossed nicols and in monochromatic light, the path-difference between 
the interfering light-waves. For less accurate work the direct determina- 
tion of the interference color and equivalent path-difference, as indicated 
on the standard Michel-Le'vy color chart, is sufficient. The actual error 
of such a determination should not exceed 10 per cent of the correct value 
of the birefringence of the section. The probability of finding a section 
making an angle within 10 of a particular direction is about i in 66; and a 
section within 20 about i in 16. 
THE NEWTON COLOR SCALE. 
The succession of colors produced by the interference of light-waves 
reflected from the surf aces of two plates of glass or other substance including 
a thin film of air resembles closely the series of interference colors observed 
on the insertion of a quartz wedge between crossed nicols and is usually 
reproduced in text-books on petrology as the Michel- L^vy standard color 
chart.* In the formation of the Newton color scale the following factors are 
of importance. 
If monochromatic light be used the series of brilliant interference colors 
is replaced by a set of dark bands or lines with intervening light strips. 
For any given thickness or path-difference the intensity of the mono- 
chromatic light emitted is found from the standard equation 
where /= intensity of illumination; A = path-difference, and X = wave- 
length of light used. For white light the expression for total intensity, 7\is 
= f/' (A) </X = f 
In order to solve this expression it is essential that we know tlu intensity 
of each color of light in the- illuminating source. A standard white light 
has not yet been adopted. Nuttingf has recently suggested as a standard 
for white light that which is reflected by "a mat non-selective surface of 
high-reflecting power illuminated at the earth's surface by the midday 
sun." Even this standard varies considerably on different days, but it 
Bull. Soc. Min. Pr. 6. i 4 s I'.t. i8t; 7, 43-44. 1884: Minlraux des Roches. 64. 1888. 
fP. G. Nutting. Circular. Bureau of Standards. No. 28, 6. 191 1. 
