BIREFRINGENCE. Ill 
300 to 500 MM the total intensity decreases rapidly to approximately 5 . 85 
percent of the total possible intensity at 555 MM- At this point the total 
amount of light in the blue end of the spectrum is about equal to that in the 
red as indicated in Fig. 69. From Fig. 69 and Fig. 70 it is evident that in 
this region a relatively slight change (5 MM) in the path-difference produces 
a great change in the relative intensities of the red and blue end of the 
spectrum, while the intensity for the central part of the spectrum changes 
only slightly and is very low. A slight change in the path-difference at 
this point will produce, therefore, a decided shift in the interference color 
from red to blue. Although the color can not be determined directly be- 
cause of the fact that parts of the spectrum will combine to produce the 
sensation of white light, and the resulting hue will be simply the residual 
light combined with this white light, in the present instance the ends of the 
visible spectrum play the important role and the color does shift from 
dominant red to dominant blue. This path-difference of about 555 MM 
produces a tint more sensitive to slight change in path-difference than any 
other and is the violet of the second order. As the total intensity is low 
for this tint, both rod and cone vision are active in the eye and themaximun 
intensity is no longer at 545 MM. but lower down the scale; care should be 
taken, in making the observations, to have a strong source of light, other- 
wise the actual tint will change color and be no longer the most sensitive 
tint. The effect of the variation in chromatic sensibility of the eye on the 
Newton color scale has still to be determined. 
In the text-books on petrography the Newton color scale is given with 
path-differences as determined by observers, especially by Wertheim and 
Quincke* in the middle of the last century, at a time when practically noth- 
ing was known of standard white light. From the above descriptions the 
dependence of the interference tint obtained on the source of illumination 
is obvious and must be considered in practical work. The path-difference 
for the sensitive violet in the Newton color scale is usually stated as 575 MM- 
This is evidently too high for most conditions of illumination and should 
be 10 or 20 MM lower. The exact position of the sensitive tint varies with 
the conditions of the weather, a deep blue sky causing it to shift toward the 
blue end of the spectrum, while an overcast sky gives more nearly white 
light. This shift in the sensitive tint can be readily seen in the microscope 
by observing the interference tints on a graduated quartz wedge, such as 
that described above. On tilting the mirror so that the light is received 
from the horizon or a white cloud, and then from the blue sky, a shift of 
10 MM in the position of the sensitive tint is often observable. Ordinarily 
the sensitive tint is at about 550 MM for quartz. In view of these conditions 
all determinations of path-difference based on interference colors alone 
are inaccurate and methods like that of the Michel-Le*vy color interference 
chart or of the Michel-Le*vy comparator are less accurate than those re- 
quiring monochromatic light. 
More accurate determinations of the Newton color scale have been made 
by A. Rolletf and C. Kraft.t The values obtained by Kraft for sunlight 
reflected from freshly fallen snow agree closely with those recorded by 
*G. Wertheim. Ann. de Chim. et. Phys. (3). 40, 180, 1854; G. Quincke. Fogg. Aon.. 129, 177, 1866. 
tSitzb. Akad. d. Wissen. zu Wien. Abteil. 3. 77, 177-761. 1878. 
{Bull, intern, acad. sci. Cracovie. Classe des Sci. Math, et Nat.. 310-353. 1902. 
