THE SUN’S COKONA. 
383 
five miles on all sides of him the earth is in shadow. Now we 
may regard the moon’s shadow as rising from this circular base, 
growing somewhat wider upwards (because at the moon’s dis- 
tance it must have the same diameter as the moon), hut not 
increasing much in width for each hundred miles or so of 
height. Where does this shadow meet the limits of the earth’s 
atmosphere? We may regard those limits as forming a hori- 
zontal plane above the observer ; but the question is, “ At what 
height does this plane lie ? ” It has been thought by some 
that the upper limits of the atmosphere lie but forty or fifty 
miles from the sea-level. Others assign a much greater height. 
Bravais, from a discussion of Lambert’s observations of the 
twilight curve, deduced a height of nearly 100 miles. 
From observations on the aurora, the height of the atmospheric 
limits have been set at more than 120 miles. Observations 
of meteors have resulted in the deduction of a yet more con- 
siderable elevation. And lastly, from polariscopic observations, 
the elevation of the atmospheric limits have been set at more 
than 200 miles. But let us take a value far beyond any of them, 
and assign to the atmosphere a height of 1,000 miles ; and 
let us add to this supposition, which surely will be regarded as 
giving to the atmosphere a sufficient extension,* the suppo- 
* Recently Mr. M. Williams, in his “Fuel of the Sun” — a work of con- 
siderable interest, and, in many parts, of considerable scientific value — has 
endeavoured to prove that the atmosphere has no limits ; founding his proof 
on the probability that perfect gases admit of indefinite expansion. It 
results, according to his views, that every celestial body has an atmosphere 
proportioned to its own mass, according to a sufficiently simple relation. 
We know so little of the extent of the atmospheres of planets that I do not 
care to oppose Mr. Williams’s conclusion , though I must say that there is 
no reason, even if his premises were admitted, for believing in this special 
law of distribution, since it is unquestionable that, if a planet’s atmosphere 
were doubled or halved, there would be no effective forces to take away the 
excess or supply the deficiency, even on the supposition of that unlimited 
atmospheric extension and consequent intercommunion imagined by Mr. 
Williams. But I think it very necessary to point out one important flaw 
in the reasoning by which he disposes, as he considers, of the ordinarily 
accepted theory that the atmosphere has limits. It is usually argued that 
there must be a height where atmospheric expansion is so great that the 
gravity of the particles of the atmosphere is exactly balanced by the repulsion 
they exert inter se ; and it is concluded that above this level there can be no 
atmosphere. But Mr. Williams remarks that an objection to this reasoning 
has hitherto escaped notice. If at the limits of the atmosphere the forces 
were so exactly balanced, the ether would have power to brush off the outer 
layer of particles, then the next, and so on, until the earth is stripped of its 
atmosphere altogether. It is always well, when one supposes one has 
detected some point which nobody else has noticed, to make sure that one is 
