1895.) NEW YORK ACADEMY OF SCIENCES. 177 
in which wv and ¢ are variables determined by the relations 
sin @ 
cot (2 u+e)=cot e cos (2 arctan —% ) 
Sin 2e (3) 
2—= ——______ 
< sin (2 u+2 e) 
in which a is the angle of incidence, and ¢« and # are given by the 
final formule 
sin 2e — tan”? Asin (4 H— 2e) 
sin 47 (4) 
ea Ai a 2 
sin AY aa 
in which A is the angle of maximum polarization called the 
‘principle angle of incidence,” and # is the azimuth of re-estab- 
lished plane polarization when the incident beam is polarized in 
azimuth 45°. # is called the “prime azimuth.” The forms 
here given for « and sare due to Jochman (see Pogg. Ann. 
CXXXVI., p. 8561). These formule, first published by Cauchy 
in 1839, were shown by Jamin, by an elaborate series of measure- 
ments, to very closely represent the facts of reflection from me- 
tallic surfaces. The prime angles of incidence and the prime 
azimuths for all the common metals and for the different Fraun- 
hofer lines were determined by Quincke in ’74 (see Phil Mag. 
MEVIT., p 221). 
Now, in order to apply these formule to calculations similar 
to those which have already been made with Fresnel’s formulz 
upon uranium glass, it was necessary to assume, as before, 
that the whole of the light emitted had undergone refraction, 
and it was also necessary to know the two optical constants 
for platinum at the temperature of incandescence. These 
constants could not be determined. However, in a number 
of experiments made by W. R. Grove (see Phil. Mag. (4) 17, 
p. 177) upon the reflection of light from incandescent plati- 
num, he was unable to detect any change in the reflecting 
properties of the platinum due to the fact of incandescence. 
Plane polarized light being reflected from the cold surface, the 
plane of polarization of the refiected beam was not affected by 
heating the platinum to the incandescent temperature. These 
experiments were not performed with delicate apparatus,’ yet 
they give reason to assume that the optical constants of plati- 
num are not greatly altered by temperature. 
Assuming, then, the values of A and H given by Quincke for 
the sodium line, the cajculations of the amount of polarization 
in the emitted beam were made for all the angles of emergence 
TRANSACTIONS N. Y. ACAD. ScI., Vol. XIV., Sig. 12, May 28, 1895. 
