1895. ] NEW YORK ACADEMY OF SCIENCES. 157 
periments he drew the conclusion “that a considerable portion 
of the light which enables us to see incandescent bodies is pro- 
duced in the interior and at depths which are not yet completely 
determined.” ‘‘ Even when the surface of a solid or liquid was 
not well polished,” Arago still found that he ‘ was able to detect 
evident traces of polarization.”” The substances upon which he 
experimented and from the observation of which he drew his 
conclusion were only four in number, viz.—solids, wrought iron 
and platinum; liquids, molten iron and glass (See Astronomie 
Populaire IT., p. 103.). He made no quantitative measurements, . 
nor even used an instrument which was capable of indicating 
roughly amounts of polarization. His polariscope consisted of 
a single quartz crystal cut perpendicularly to the optical axis 
and a crystal of Iceland spar. The latter produced a double 
image of an opening in a diaphragm placed just beyond the 
erystal of quartz. The two images were of course colored when 
the light was polarized and uncolored when it was natural. 
Arago applied the results of his experiment to the determina- 
tion of the character of the sun’s surface. Being unable to detect 
any trace of polarization in the light emitted by the outer edge 
of the sun’s disk, he drew the well-known conclusion that the 
surface of the sun can be neither liquid nor solid, but must be 
gaseous. 
After the discovery of the polarization of heat, and the con- 
struction of an instrument by Melloni for its detection and 
measurement, Provostaye and Desain examined the heat rays 
emitted by luminous platinum and found that they, like the light 
rays, were polarized in the plane perpendicular to the plane of 
emergence. Their experiments were few in number and con- 
fined entirely to platinum. In 1866 Magnus extended this 
method of experiment to obscure heat rays, making quantitative 
measurements upon the heat emitted at the temperature of 100° 
C,and at an angle of 35°. His experiments embraced the fol- 
lowing list of substances: Parattine, glycerine, white wax, 
melted calophony, rubol, black glass, transparent glass, quick- 
silver, aluminium, copper and tin. For these substances he 
found a polarization at 35° ranging from 5% to 27%. He drew 
the conslusion that obscure heat, like light, must undergo re- 
fraction in emerging from the surface of the radiating body.. 
Verdet, in the paragraph upon polarization by emission pre- 
viously referred to, while stating that little has been done in the 
investigation of the subject, gives the same explanation of the 
phenomenon as that first offered by Arago. He says that “it is 
due to the fact that it is not alone the surface molecules which 
radiate light; those of the interior layers also radiate, at least. 
