hs of space. If we may
employ the analogy of terrestrial vapours to guide us in our reasoning,
then it would seem that, as the comet retreats, its tail would condense
into myriads of small particles. Over these small particles the law of
gravitation would resume its undivided sway, no longer obscured by the
superior efficiency of the repulsion. The mass of the comet is, however,
so extremely small that it would not be able to recall these particles
by the mere force of attraction. It follows that, as the comet at each
perihelion passage makes a tail, it must on each occasion expend a
corresponding quantity of tail-making material. Let us suppose that the
comet was endowed in the beginning with a certain capital of those
particular materials which are adapted for the production of tails. Each
perihelion passage witnesses the formation of a tail, and the
expenditure of a corresponding amount of the capital. It is obvious that
this operation cannot go on indefinitely. In the case of the great
majority of comets the visits to perihelion are so extremely rare that
the consequences of the extravagance are not very apparent; but to those
periodic comets which have short periods and make frequent returns, the
consequences are precisely what might have been anticipated: the
tail-making capital has been gradually squandered, and thus at length we
have the spectacle of a comet without any tail at all. We can even
conceive that a comet may in this manner be completely dissipated, and
we shall see in the next chapter how this fate seems to have overtaken
Biela's periodic comet.

But as it sweeps through the solar system the comet may chance to pass
very near one of the larger planets, and, in passing, its motion may be
seriously disturbed by the attraction of the planet. If the velocity of
the comet is accelerated by this disturbing influence, the orbit will be
changed from a parabola into another curve known as a hyperbola, and the
comet will swing round the sun and pass away never to return. But if the
planet is so situated as to retard the velocity of the comet, the
parabolic orbit will be changed into an ellipse, and the comet will
become a periodic one. We can hardly doubt that some periodic comets
have been "captured" in this manner and thereby made permanent members
of our solar system, if we remark that the comets of short periods (from
three to eight years) come very near the orbit of Jupiter at some point
or other of their