366 
POPULAR SCIENCE REVIEW. 
described would happen ; and if our perception of colour were 
but sufficiently delicate, we could tell whether a star were 
moving from or towards us by the colour of its light. But 
this supposition implies that a star’s colour should be mono- 
chromatic ; and we know that the light from the stars consists 
of a combination of all the prismatic colours. But again, if the 
spectrum had definite extremities, and if no action of any sort 
took place beyond those extremities, then something like what 
Doppler conceived would take place. For then the light-waves 
of all lengths would be affected by a star’s motion ; so that if a 
star were approaching us, all the waves would be shortened, and 
a part of the red end of the spectrum would suffer extinction, 
while in the reverse case the blue end of the spectrum would be 
shortened. We know, however, that beyond the visible ends of 
the spectrum, waves too long and too short to affect the eye as 
light-waves are really in existence. Thus instead of the red end 
or the blue end suffering, in the cases imagined above, all that 
would happen would be that the heat-waves beyond the red end 
or the chemical rays be^mnd the blue end would become light- 
waves, replacing the red or blue end of the spectrum, as the 
case might be. 
But these considerations, while showing that nothing can be 
hoped for from Doppler’s suggested consideration of star-colours, 
show that a much more delicate and satisfactory test can be 
applied. We see that the whole spectrum is shifted bodily. 
ThereforG all its lines, whether dark or bright, must he shifted 
U'iih it. This is a motion we may ho'pe to estimate, because we 
can bring into comparison with any line in the shifted spectrum 
the corresponding line belonging to some terrestrial element. 
We have, in fact, a test of the most extreme delicacy ; and were 
it not that the most rapid stellar motions can produce but the 
minutest change in the position of the star’s spectrum, we might 
read off the stellar motions of recess or approach as readily as 
we can determine the general character of the star’s light by the 
same mode of analysis. 
But when it is remembered that the velocity of light is about 
1 85,000 miles per second, we see that a star must be moving 
with enormous velocity that its spectrum may exhibit any 
appreciable change of position. Our sun is supposed to be tra- 
velling at the rate of about 5 miles per second, and we have 
reason to believe, from some researches of Mr. Stone’s, that the 
average motions of the stars may be about one-third greater, — 
say about 7 miles per second. Now, it has been estimated by ' 
Mr. J. Clark Maxwell that a velocity equal to that of the earth 
in her orbit, that is, rather more than 18 miles per second, 
would shift the sodium line i), through a space equal to about 
the tenth part of that which separates i), from Ug, these lines 
