1895. ] NEW YORK ACADEMY OF SCIENCES. 161 
also takes the place of the Wollaston prism of the first polari- 
scope. The glass, being in a state of strain, is doubly refract- 
ing and exhibits with polarized light the familiar dark or light 
cross which is characteristic of doubly refracting crystals, when 
cut perpendicularly to the optic axis and viewed by convergent 
light. With this instrument a polarization of two or three per 
cent. could be easily detected, and it had the further advantage of 
indicating immediately the azimuth of the plane of polarization. 
Also, by careful observation of the distinctness of different parts 
of the figure, it was possible, after a small amount of practice, 
to estimate with considerable correctness the degree of polariza- 
tion. 
‘1. In all experiments which have been previously performed 
upon this subject, the white-hot body has been in immediate 
contact with the air. The emitted light was therefore obliged 
to pass through layers of air of varying density before it reached 
the eye of the observer. That the light might not thus suffer 
a large number of refractions between the incandescent body 
and the eye, and so be endowed with the property in question, 
seemed entirely possible. It was therefore necessary to make 
some experiment in order to determine whether or not this was 
the entire or partial cause of the phenomenon. For this pur- 
pose the contrivance shown in the diagram was employed. A 
strip of platinum foil A about four cm. in length and five mm. 
in width was attached to the platinum and copper wires B and 
C. The former was sealed into the glass tube G, and the latter 
was passed through the cork F which closed the other end of 
the tube. 
The instrument was first sealed with wax and then connected 
with the air pump by means of the small tube D,and with a 
strong electric current by means of the wires B and C. Care 
was taken to place the platinum strip as near the axis of the 
tube as possible, in order that light emitted by it might pass 
normally through the sides of the tube. Otherwise polarization 
TRANSACTIONS N. Y. ACAD. ScI., Vol. XIV., Sig. 11, May 28, 1895. 
