1895. ] NEW YORK ACADEMY OF SCIENCES. ~ 159 
research upon platinum seemed to be inconsistent with such 
an explanation; for, when a well polished platinum strip was 
heated to incandescence by means of an electric current and the 
glowing surface examined by means of a double Wollaston 
prism, the polarization was found to be so nearly complete for 
angles in the neighborhood of grazing emergence that one of the 
images almost disappeared. But, since platinum is known to be 
altogether opaque, except in the case of exceedingly thin lamine, 
it would seem as though the surface molecules must play a con- 
siderable part in the luminosity of the glowing metal; so that, 
even if the assumption were made that the laws of reflection and 
refraction would require complete extinction of the ray polarized 
parallel to the plane of emergence, there still ought to be a con- 
siderable amount of light emitted in this plane from the surface 
molecules; at least, a sufficient quantity to prevent so nearly 
complete extinction as experiment showed to exist for angles of 
88 or 89 degrees. 
The only apparent method of reconciling the facts with Ara- 
go’s explanation was to assume that the opacity of the platinum 
was greatly diminished by an increase in its temperature. And 
yet, such experiments as were made to determine whether or not 
this was the case, gave only negative results. The thinnest sheet 
of platinum which was capable of being heated to incandescence 
without melting, was placed in the focus of a powerful beam of 
light from an are lantern, the beam having been first polarized 
by transmission through a Nicol. The plane of the glowing 
platinum being perpendicular to the beam, the light emerging 
normally on the other side of the platinum was examined by 
means of a delicate polariscope. No trace of polarization was 
detected. Neither could the outlines of the focus be distin- 
guished on the side of the platinum away from the lantern. 
The sheet of platinum employed was evidently just as opaque 
-as at a lower temperature. 
This difficulty of accounting for the extreme polarization no- 
ticed at large angles of emergence appeared to be considerably 
diminished if another cause for the phenomenon were assumed 
than that given by Arago. 
According to the conclusions of Fresnel, Cauchy, Stokes, 
Mascart and most of the advocates of the elastic solid theory 
of light, the direction of vibration of the ether particles in 
plane polarized light is perpendicular to the plane of polariza- 
tion. It would follow that the light emitted at large angles 
by platinum vibrates mainly in the direction of the normal to 
the surface. It is not unnatural to suppose that at the bound- 
ary between very dense and very rare media, like platinum and 
