458 The Atmospheres of the Planets. (October, 
There is at present no reason for believing that the atmo- 
sphere of Jupiter possesses a greater depth than from 25 to 
50 miles; the only observation that can be considered as 
indicating any very much greater depth being those of the 
form of the shadow of the satellite, which though highly 
interesting, cannot legitimately be regarded as pointing in 
this direction. Whatever, therefore, may be held to be the 
cause of the slight mean density of Jupiter and its com- 
panions in this respet, Saturn, Uranus, and Neptune, it 
cannot arise simply from their possessing an immense depth 
of atmosphere. 
There is one other point in connection with the atmo- 
sphere of Jupiter requiring examination, namely, the influence 
of the presence of an atmosphere on the occultations of its 
satellites; for if a fringe of atmosphere of any recognisable 
breadth were present, the satellites, instead of disappearing 
sharply behind, and re-appearing as distin¢tly from beyond 
the planet’s limb, would fade slowly away. Thus from the 
effects of its refracting the rays of light, it would appear 
that if any considerable layer of atmosphere to Jupiter 
extended beyond his disc formed by the cloud strata, it 
could not fail to be recognised by retarding occultation and 
eclipses, and accelerating the re-appearances. With suffi- 
cient approximation, the horizontal refraction for any 
layer of Jupiter’s atmosphere assumed similar to our own 
and whose density is 6,’ would be— 
Te— 3102 On 
Accordingly, for a strata of air only one-thousandth the 
density of our own the horizontal refraction would be still 
strongly marked. 
In all probability the real density of the strata of Jupiter’s 
atmosphere immediately beyond the uppermost clouds is in 
density over one-tenth of our own, and would, therefore, 
exert a horizontal refraction equal to half of our own, and 
be therefore most marked. ‘The true reason of its being un- 
detectable consists of the very slight breadth of the fringe of 
atmosphere from its very quick decrease in density, so that 
only twenty-four miles beyond the upper cloud layers of 
Jupiter’s, the density of the atmosphere would be utterly 
insensible, and with it the horizontal refraction. As this 
distance of twenty-four miles would at Jupiter’s distance 
subtend an angle of barely one-hundredth of a second of arc, 
whatever changes might be produced by it would be entirely 
unrecognisable ; so that whatever may be the density of its 
atmosphere, the effects of the refraction of light through it 
