496 The Lunar Atmosphere (October, 
the moon’s face will not fali either much below zero, for 
there is more heat to be lost than onthe earth. It is, more- 
over, to be remembered, that the proportion of carbonic 
anhydride in the moon’s atmosphere would be probably much 
greater than in the earth’s, and this powerfully retards the 
radiation of heat. From these considerations, then, we have 
at sunrise the moon’s surface at any position near the 
equator, the temperature about —20°, which will increase 
until a little after full moon, when it will have risen to 
about 200°, and will then slowly decrease, until at sunset it 
will be about 60°. During the lunar night the temperature 
will fall slowly until a few days before sunrise, when it may 
become as low as —30, rising again to about —20° at sun- 
rise. But it is evident that as the temperature is solely due 
to the solar rays, these must be the extreme results; and 
that for any place away from the lunar equator the climate 
will be much milder, never rising to so high a temperature, 
as the sun will not be in its zenith, nor ever falling so 
low.* This circumstance has usually been overlooked. 
The temperature of the lunar surface having been there- 
fore approximately determined, the equations can be at once 
applied to find the conditions of the greatest lunar atmo- 
sphere possible under known circumstances. 
An examination of all the methods that have been applied 
to detect a lunar atmosphere shows that all the most delicate 
—and, in fact, only applicable ones, to such an atmosphere 
as we have seen would probably exist—are based upon the 
refraction of light by the atmosphere. And the condition 
under which this refraction is most marked is that which 
requires attention, namely, the horizontal refraction. And 
even of these selected methods, the most delicate and only 
thoroughly reliable one is that based on the retardation of 
the occultation of a star, at the dark limb of the moon, by 
the refraction caused by the light traversing the atmosphere. 
For when the diameter of the moon is known, the exact 
* The values in the text are arbitrary, but have been arrived at from theo- 
retical consideration in conjunction with Lord Rosse’ determination, and can 
be expressed by the following equation, which gives the theoretical results for 
the maximum and minimum taken. Assume © is the solar altitude at any 
spot, and increases from o to its maximum, and then falls to 0, at which it 
remains while the sun is beneath the horizon. Put ¢ equal to 300 degrees at 
sunrise, and rising uniformly by 27 during each lunation; then if X be the 
latitude of the spot we have roughly— 
to = 170°(1-Acos + Bsin € —C sin X). 
It has been taken that A=o'g. B, which varies as the cosine of the latitude, 
is 0°28 at the equator, and C = 0°08, which values but represent the theoretical 
probabilities. 
