TEMPERATURE AND STRUCTURE OF THE SUN. 
93 
therefore a sharp limit between a liquid and a gaseous mass 
on our sun is physically not possible. This difficulty, over- 
looked in earlier times in our physics of the sun, was removed 
by Schmidt, a high school teacher in Germany. According 
to Schmidt’s theory, the sharp limitation between the white 
incandescent photosphere and the gaseous chromosphere is 
an optical illusion, produced by the sun’s being a radiating 
gaseous sphere, the density of which decreases gradually from 
the center to the outside. 
It is evident that gases at a high density, such as must 
exist in the central mass of our sun, emit a continuous spec- 
trum, as observed by our dispersion apparatus, like solid or 
liquid substances of a similar density. Corresponding to the 
Jftg. 4 - 
decrease of density from the interior to the exterior parts of 
the sun, the refractive index also decreases gradually, so 
that light rays emitted by the central parts follow not 
straight, but curved lines. 
When a parallel beam falls on a plane surface, AD, of a 
medium, ABCD, the refractive index of which decreases 
gradually from below upward, the rays are deflected as 
shown by the accompanying figure. The deviation is great- 
est at normal incidence. This deviation accounts for the 
fact that we see the sun before its rays can reach our eyes 
directly. The bending of light rays is also the cause of the 
phenomenon of “Fata Morgana,” observed in the desert. 
In 1860 our mathematical authority, Rummer, concluded 
from his theory of curved light rays that on Neptune we 
could see our own backs without using any mirror, since 
