154 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
out yet, extending a distance of several solar radii, is found the 
corona which we have not yet succeeded in observing except at total 
eclipses of the sun (pl. 1). 
PHYSICAL CONDITION OF THE SUN. 
Formerly the continuous background of the solar spectrum was 
attributed to an incandescent solid or liquid nucleus. The prevalent 
theory at present is that the sun is entirely gaseous. For this belief 
there are several good reasons: Apart from the mean low density of 
this star (1.4 relative to water), its temperature is, as we shall see, 
higher than that at which all known bedies volatilize. Further, a 
sphere not gaseous but solid or liquid would near its edge emit 
polarized light, of which we observe not a trace. Finally, the way 
in which the sun rotates is in complete contradiction to a rotating 
rigid solid body. The objection based upon the continuous back- 
ground of the spectrum alone remains. This was overcome when it 
was found that gases can give such a spectrum. The bright lines in 
the spectrum of a gas are narrow and separate when the gas is under 
a weak pressure, but they broaden as soon as the pressure is increased, 
and, finally, a continuous background appears which may indeed 
become very bright. 
We find, therefore, that in order to account for the observed facts 
it is sufficient to assume that the heavy vapors gather at the center, 
where they are under great pressure, while, according to their densi- 
ties, the lighter gases in successive layers make up the outer portions. 
This explanation, however, has not satisfied all, and physicists ask 
whether the great diffusive tendencies in a gas would not finally tend 
to transform the whole mass into a perfectly homogeneous mixture. 
Further, the form of the sun, so perfectly round and, moreover, so 
sharp, requires an explanation which the laws of refraction for a 
moment seemed to give and about which we will say a word. 
We know that in our own atmosphere the path of a beam of light 
is curved by refraction, especially when the beam is near the horizon. 
(Fig. 1.) Now we may conceive (theory of Schmidt) that the law of 
densities in the sun’s atmosphere is such that, when the beam of light 
is sufficiently inclined to the vertical, the path is so curved that it 
never leaves the sun. (Fig. 2.) On such a star, in the upper very 
rare layers, everything would appear as on the earth; in the lower 
layers, however, only those rays near the vertical would succeed in 
escaping. These two regions would evidently be separated by 
another where the luminous trajectories would encircle the star 
many times before emerging. (Fig. 3.) It is this very thin layer 
which, according to Schmidt, constitutes the photosphere. Further, 
the dark lines of the spectrum may be explained by the optical 
