two hundred and fifty feet; but subsequent
experiments convinced Mr. Eddy that it was by no means to be relied
upon as an indication of the height of kites. Not that the law is
false; but it holds good only when the meteorological conditions above
are the same as at the earth's surface, which is very far from being
the case always.

Out of these experiments Mr. Eddy evolved an important theory which
has since been abundantly verified. Seeing the frequent variations in
the thermometric readings from what the law had led him to expect, he
concluded that these were due to meteorological variations overhead;
and that changes in the weather, say the approach of warm waves or
cold waves, make themselves felt in the air strata above the earth's
surface several hours before they can be detected at the surface.
Observations extending over months at the Blue Hills Observatory, near
Boston, and elsewhere, have abundantly confirmed this theory.

With this fact established, it followed, in Mr. Eddy's opinion,
that it was perfectly possible to use kites in making weather
prognostications; and, indeed, he has been doing this himself for
several years with the best results. Whenever his kite-thermometers,
sent to a fixed height which he determines independently by a
specially devised kite-quadrant, show actual readings which are either
warmer or cooler than the theoretical readings, he prophesies that
the weather will, within a few hours, become warmer or colder at
the earth's surface, and these prophecies are fulfilled in a
large majority of cases. If the kite-thermometers show exactly the
temperature which the law would call for, he prophesies that there
will be no change in the weather.

[Illustration: CITY HALL PARK AND BROADWAY FROM A KITE.

From a photograph taken from a kite by Mr. W.A. Eddy. City Hall Park,
New York City, appears in the foreground, with Broadway back of it.]

It has also been demonstrated that kites may be used by meteorologists
to indicate the approach of storms, which they foretell by a sudden
and continuous veering over a considerable arc, usually about sixty
degrees. This veering begins usually six or seven hours before a
storm, and often as much as twelve hours. And another sure sign of a
storm is the continuous and sudden dropping of the kites followed by
a quick recovery, which shows that the wind is blowing in gusts
interspersed with periods of calm.

In making a series of meteorological experim