112 
METHODS OF PETROGRAPHIC-MICROSCOPIC RESEARCH. 
Rollet and demonstrate the inaccuracy of the data on which the Michel- 
LeVy color chart is based. Kraft found that not only does the position of 
the sensitive tint vary with different sources of light (gray cloudy sky, 
clear sky, sunlight reflected from white surface, Welsbach burner, arc light, 
incandescent bulb), but its relative width on the color scale also varies, and 
even the character of the tint changes with the different sources of illumina- 
tion. In his paper Kraft discusses the different factors which enter the 
problem and their bearing on the accuracy of the results. Until a definite 
standard for white light has been set and the Newton color scale expressed 
in terms of this standard, the values of Rollet and Kraft for sunlight from a 
mat white surface should be adopted in preference to the old Wertheim- 
Quincke color scale. 
In using quartz or selenite wedges still another factor enters which tends 
to reduce the accuracy of all white-light determinations. The minerals 
show dispersion such that the difference between the principal refractive 
indices for the different wave-lengths differ somewhat.* In table 5 the 
TABLE 5- 
X 
Wavelength 
inpji. 
Quartz 
a. 
Selenite 
7-0. 
Calcite 
a'. 
Quartz 
P 
G 
4)1 
C.OOQ4) 
F 
:> 
4)4 
(1 (MK)jS 
o. 18122 
42 .O 
r 
486 
527 
O.OO930 
0.009183 
0.00932 
0.17709 
32.7 
Cd 
fa* 
O 174)4 
Tl 
Mi 
0.0002} 
26.6 
D 
C 
Li 
589 
656 
0.00910 
0.00904 
0.00916 
O.OOOX>5 
O.OO9O2 
0.17194 
o. 16978 
21.7 
7-3 
16.5 
ft 
686 
O.OO9OO 
o. 169)8 
birefringence of quartz, selenite, and calcite is listed, also the specific rota- 
tion, p, of quartz, the values of different wave-lengths having been obtained 
from the standard books of references, Landolt-Bornstein and Dufet.f 
Using white light as a standard and neglecting in the case of selenite the 
dispersion of the bisectrices, the values for the intensity curves throughout 
the spectrum for a path-difference of 555 nn for light of this wave-length 
are indicated in Fig. 71 (calculated from the standard intensity formula 
for crossed nicols, see page 108). In the case of quartz, parallel to the axis 
(Fig. 71) the dispersion shifts the path-difference for the sensitive-tint to 
about 550 nn; for a quartz plate normal to c, the sensitive-tint is at about 
547 nn; for selenite the path-difference for the sensitive-tint is about 554 MM. 
and for calcite it is about 551 /i/x-t 
*A. Cornu appears to have been the first to direct attention to this factor. Bull. Soc. Min. Pr.. 6, 1.15. 
1881; compare also C. Hlawatsch. T. M. P. M. 21, 107. 1902. 23,415-430. 1004: P. Becke, Denkschr. 
I K K. Akad. d. Wissen., zu Wien. 75, 38. 1004- The present writer ha* unfortunately been unable to 
obtain a copy of this important paper by Professor Becke and the above reference was taken from tin- 
Traitc of Duparc-Pearce. 
tin a recent paprr ! A Wnlfing. Sitz. Bcr d. Hcitlelberger Akad. d. Wissen. 1-16. 1910. has also listed 
these values for quartz and selenite and his values agree practically with those of table 5 . See also M ichcl- 
Levy. Les Mincraiu des Roches. 57. 1888; P. Becke. T. M. P. M. 22. 178. 1001; V. de Sousa Brandao, 
Centralblatt f. Miner . z.v 29. 1003; H. Dufet. Jour. Phys. 303-306, 1888; D. B. Brace. Phil. Mag. (3). 48, 
343. 1899: K. J. Rrndtorff. Phil. Mag. (6). I, 339. 1901. 
{These value* are only tentative, pending the establishment of the white-light standard at which Dr. P. 
G. Nutting, of the Bureau of Standards, is at present engaged. As soon as this standard has been defi- 
nitely fixed, the Newton color scale is to be studied in detail by Dr. Nutting and the writer. 
