and Terrestrial Elements in the Sun. 403 
of the photosphere, because all the radiation more refrangible 
than the solar line U is cut off before it reaches us by some 
absorbent medium. There is, however, one argument which 
makes strongly for the supposition that the spot is a region of 
a temperature generally not lower than that of the rest of the 
photosphere, which is this : If we suppose the photosphere to 
be of the nature of clouds formed by condensation of metallic 
and other vapours in the sun's atmosphere, then the gaseous 
part of that atmosphere will be saturated with the vapour of 
the matters forming the cloud, and any depression of its tem- 
perature must cause a fresh formation of cloud, which would 
tend to destroy the character of a spot. Indeed an apparent 
dissolution of photospheric matter carried into spots has fre- 
quently been observed, according to Secchi, and seems to 
suggest an evaporation of the cloudy matter in the spot. 
Dismissing, then, the a priori argument for the coolness of 
the spot in comparison with the photosphere as insufficient, 
and likely to encumber us in the interpretation of the observed 
facts, let us come to the recorded phenomena. 
There is, first, the widening of the Fraunhofer lines. This is 
observed in all spots, but the lines are by no means all widened or 
all equally widened . The diagram (PI. VI.) represents the lines 
in two spots (taken as samples) observed at Greenwich (Obs. 
1881) to be widened. The range observed is from F to b, and 
the length of the lines drawn under the corresponding lines in 
Angstrom's map indicates the amount of widening, except 
where the widening is excessive, when the actual breadth is 
given as it appeared. As to the precise cause of the widening 
of spectral lines, physicists are not agreed ; but as a fact we 
know that increased density of the emitting vapour is directly 
or indirectly a cause of the widening of the spectral lines of 
terrestrial elements ; and we may reasonably attribute the 
widening in the spots to the increased thickness of the solar 
atmosphere, denser of course in the lower part, through which 
the light comes to us. But, then, why are not all the Fraun- 
hofer lines widened ? and why are those that are widened so 
unequally widened? Whatever the cause be, similar phe- 
nomena are observed in the emission-spectra of terrestrial 
elements. In the first place, the lines of some metals are 
much more readily expanded than those of others. The ready 
expansibility of the lines of hydrogen and of the D lines of 
sodium is well known; while the lines of iron and titanium 
are much less easily expanded, but are nevertheless very 
sensibly enlarged when the electric arc is freely supplied 
with the metal. But even so some lines expand more than 
others. This is well seen in the case of magnesium, which 
2F2 
