must be allowed that the first appearance
of shooting stars occurs at a very
considerable altitude.

The velocity with which shooting stars
move, is more difficult to determine than
their distance from us. It is certain that
their speed is great, compared with the
velocities which we have occasion to
observe on the surface of the earth; but the
numerical value of that speed still remains
so indeterminate that it is absolutely
impossible to make it the base of any
conclusions. It cannot be employed for
determining the orbit described in space by
the moving body to whose presence the
phenomenon is due. Nevertheless, the
determination of that orbit is very important,
and it will be easily understood that
observers have turned their efforts in that
direction.

When it is proposed to determine the
orbit of a new star, planet, or comet, the
first thing is to observe it as accurately as
possible, in three different positions. The
data furnished by these three observations
suffice to deduce from them the orbit of
the star; and the more distant from each
other the three positions are in which the
moving body has been observed, the more
correct is the result. A like mode of
proceeding is evidently impracticable for
determining the orbit described by a shooting
star. The short duration of its visibility
does not allow it to be observed in
three distinct positions with the requisite
precision; which precision ought to be all
the greater, because the three successive
positions can only extend over a very small
arc of the trajectory of the moving body.
It is only by combining the knowledge of
the position of the shooting star, at a given
instant, with the amount and direction of
its velocity at that instant, that we can
hope to succeed in determining the orbit
which it describes The great difficulty of
the question lies in discovering the rate
and the direction of the velocity. It has
just been stated that it is almost impossible
to make this much-needed discovery by
direct observations. Astronomers have
succeeded in overcoming the difficulty by
considering the phenomenon of shooting stars
as a whole, instead of persisting in the
observation and study of these luminous
bodies one by one.

The most striking feature of the curious
phenomenon we are examining, is the
occurrence of extraordinary displays of
shooting stars. Brandos relates that, on the
6th of December, 1798, while travelling to

Bréme in a public conveyance, ho counted
four hundred and eighty through one of
the diligence windows; from which he
reckons that at least two thousand must
have appeared in the heavens during the
course of the night.

In the night from the 11th to the 12th
of November, 1799 the above dates are
important to note Humboldt and Boupland
witnessed, at Cumana, in South America, a
perfect shower of shooting stars. The
phenomenon, already remarked in the evening,
acquired great intensity in the middle of the
night, and continued to increase until four in
the morning, when it gradually diminished
until daylight. Boupland says that there
was not a portion of the sky equal in extent
to three times the moon's diameter, which
was not every instant full of shooting stars.
The inhabitants of Cumana were frightened
at this unusual sight. The oldest amongst
them remembered that the great earthquakes
of 1766 had been preceded by a
similar phenomenon.

These extraordinary facts were in some
measure forgotten, when a fresh shower of
shooting stars was observed in America on
the 13th of November, 1833. Professor
Olmsted, of Newhaven, published a very
important memoir on the subject.
Calculating from the data sent to him, he
estimated the number of shooting stars, which
were seen at certain spots during the night
of the 12th and 13th of November, at more
than two hundred thousand. The numerous
accounts recorded of this event, and
the publicity given to it by the journals,
recalled the general attention in this
direction, and everybody began to watch the
case more carefully than hitherto. Regular
observations of shooting stars were organised,
and little by little there resulted from
them a clearer idea of the general course
and march of the phenomenon.

In Olmsted's opinion, the grand November
display was periodical, and ought to
recur every year at the same epoch. It
was ascertained, in fact, that every year,
about the 12th and 13th of November, there
was a very marked increase in the number
of shooting stars appearing in the sky; but
that was far from reproducing the extraordinary
spectacle beheld in America in 1833.
In 1837, the astronomer Olbers wrote:
"Perhaps we shall have to wait till 1867
before we witness a repetition of the magnificent
phenomenon presented to our view
in 1799 and 1833:" a bold prediction which
we saw completely realised a year sooner,
namely, in 1866. The remembrance, by