SPECTRUM ANALYSIS OF THE HEAVENLY BODIES. 145 
A few stars, however, stand out from the rest, and appear to be charac- 
terized by a peculiarity of great significance. These stars are represented by 
Betelgeux and § Pegasi. The general grouping of the lines of absorption in 
these stars is peculiar, but the remarkable and exceptional feature of their 
spectra is the absence of the two lines which indicate hydrogen, one line in 
the red, and the other in the green. ‘These lines correspond to Fraunhofer’s 
Cand F. The absence of these lines in some stars shows that the lines 
C and F are not due to the aqueous vapour of our atmosphere. 
It is worthy of consideration that the terrestrial elements which appear 
most widely diffused through the host of stars are precisely some of those 
which are essential to life, such as it exists upon the earth, namely, hydro- 
gen, sodium, magnesium, and iron. Besides, hydrogen, sodium, and magne- 
sium represent the ocean, which is an essential part of a world constituted 
like the earth. 
We learn from these observations that in plan of structure the stars, or at 
least the brightest of them, resemble the sun. ‘Their light, like that of the 
sun, emanates from intensely white-hot matter, and passes through an atmo- 
sphere of absorbent vapours. With this unity of general plan of structure, 
there exists a great diversity amongst the individual stars. Star differs 
from star in chemical constitution. May we not believe that the individual 
peculiarities of each star are essentially connected with the special purpose 
which it subserves, and with the living beings which may inhabit the 
planetary worlds by which it may possibly be surrounded. 
When they had obtained this new information respecting the true nature 
of the stars, their attention was directed to the phenomena which specially 
distinguish some of the stars. 
CoLours oF THE SraRs. 
The colour of the light of the stars which are bright to the naked eye is 
always some tint of red, orange, or yellow. When, however, a telescope is 
employed, in close companionship with many of these ruddy and orange 
stars, other fainter stars become visible, the colour of which may be blue, or 
green, or purple. 
Now it appeared to be probable that the origin of these differences of 
colour among the stars may be indicated by their spectra. It was obvious 
that if the dark lines of absorption were more numerous or stronger in some 
part of the spectrum, then those colours would be subdued in power, rela- 
tively to the colour in which few lines only occur. These latter colours 
remaining strong would predominate, and give to the light, originally white, 
their own tints. 
This supposition was confirmed by observations of the spectra of several 
white and coloured stars. The grouping of the dark lines in the stars Sirius, 
a Lyre, a Hereulis, (3 Cygni, and some others which were examined for this 
purpose, was such as to account for the difference of colour exhibited by their 
light. The spectra of the two stars forming 3 Cygni are represented in 
fig. 4, Plate IIL. 
It appears, therefore, that the colours of the stars are in genera] produced 
by the vapours existing in their atmosphere. The chemical constitution of a 
star’s atmosphere will depend upon the elements existing in the star and upon 
its temperature. 
VARIABLE Srars. 
The brightness of many of the stars is found to be variable. From night 
to ae from month to month, or from season to season, their light may be 
; M 
