ry respectable masses, as the nuclei are
frequently not more than 5,000 miles in diameter, and consequently it
becomes important to examine the principle. From all experiments the
density of an elastic fluid is directly as the compressing force; and if
a cylinder reached to the top of our atmosphere, compressed by the
gravitation of the earth, considered equal at each end of the cylinder,
it would represent the actual compressing force to which it owes its
density. If the gravitation of the earth were diminished one thousand
times this atmospheric column would expand one thousand times,[44]
(taking no account of the decrease of gravitation by increase of
distance;) so that the diameter of the aerial globe would be increased
to 108,000 miles, taking the atmosphere at 50 miles. But the mere
increasing the _bulk_ of the atmosphere 1000 times would increase the
diameter to little more than double. Even giving the correct expansion,
a comet's mass must be much greater than is generally supposed, or the
diameters of the nuclei would be greater if composed of any gas lighter
than atmospheric air.

It is very improbable that a comet is composed of only one elementary
gas, and if of many, their specific gravities will vary; the lighter, of
course, occupying the exterior layers. With such a small mass,
therefore, the upper portion of its atmosphere must be very attenuated.
Now let us remember that the density of the ether at a comet's aphelion,
is greater than at the perihelion, in the direct ratio of the square
roots of the distances from the sun nearly. At the aphelion the comet
lingers through half his period, giving ample time for the nucleus to be
permeated by ether proportionally dense with the surrounding ether of
the vortex at that distance. Thus situated, the comet descends to its
perihelion, getting faster and faster into a medium far less dense, and
there must consequently be an escape from the nucleus, or in common
parlance, the comet is positively electric. This escaping ether, in
passing through the attenuated layers composing the surface of the
nucleus, impels the lighter atoms of cometic dust further from the
centre, and as for as this _doubly_ attenuated atmosphere of isolated
particles extends, so far will the escaping ether be rendered luminous.
It may be objected here, that a contrary effect ought to be produced
when the comet is forsaking, its perihelion; but the objection is
premature, as the heat received f