luric or selenic origin yet advanced, can stand
against the weight of evidence against it. Their fragmentary character
rather favors the views of Sir David Brewster, and when we consider that
they have been revolving for thousands of years with planetary velocity,
and in very eccentric orbits, through the ether of space, continually
scathed by the electric blast of the radial stream, their rounded
angles, and black glossy crust of an apparently fused envelope, may be
accounted for, without difficulty, from the non-vitrified appearance of
the interior. The composition of aerolites as far as known, embrace
nearly one-third of all known simple substances according to Humboldt,
and are as follows: iron, nickel, cobalt, manganese, chromium, copper,
arsenic, zinc, potash, soda, sulphur, phosphorus, and carbon.

The theory we have thus given of the common occurrence of shooting
stars, will render a satisfactory general account of their sporadic
appearance; but there are other phenomena of greater interest, viz.: the
occasional recurrence of swarms of such meteors, which defy all
numerical estimates, being more like a fiery rain than anything they can
be compared to. The most interesting feature of this phenomena, is the
_apparent_ periodicity of their return. In the following table we have
set down the most remarkable epochs mentioned by Humboldt, (and no man
has devoted more attention to the subject,) as worthy of notice:

About April 22 to 25
" July 17 to 26
" August 9 to 11
" November 12 to 14
" November 27 to 29
" December 6 to 12

Besides these, he mentions two showers, from Arabian authority, in
October; one in October, observed in Bohemia; one observed by himself,
in the Pacific, on March 15; one February 4, just preceding the terrible
earthquake of Riobamba, in 1797. The Chinese annals also contain many
showers of stars, before the present era commenced. Some were in March,
more in July, and others in different months. How, then, in view of
these numerous dates, can we attach so much importance to the
periodicity of these showers? The great shower of 1833, in the United
States, on the 12th and 13th of November, brought to mind the great
shower at Cumana, observed by Humboldt and Bonpland just thirty-three
years before, to a day; and it must be confessed that more than ordinary
displays have been seen on this date. Yet, on the strength of this,
every meteoric s