r classes start at the same point, one of them
going as far south as Cincinnati, and the other as far south as
Montgomery, Ala., and both turning at these points northeasterly to the
Gulf of St. Lawrence. Two other classes originate in Colorado, one
moving in a northeasterly direction slightly curved, and the other
directly east. Still others have their origin farther south in the Gulf
of Mexico, and move in a northeasterly direction. Very rarely they
originate in the Atlantic east of Savannah, moving first in a
northwesterly direction, but finally bending to the northeast.

Every day there is a weather map made up showing the locations of the
high and low barometers, direction of wind, lines of equal pressure, as
well as those of temperature. By study from year to year all of these
phenomena have become systematized, so that by tracing an area of low
barometer from its origin in its progress easterly it is soon seen to
fall under one of these classes and we are able to predict about what
its course will be. Knowing the speed of its movement as well as the
velocity of wind and all the conditions attending it, taken in
connection with the weather conditions in the region for which the
prediction is made, an expert can ordinarily forecast with some degree
of accuracy. After all that can be said, however, weather predictions
based upon maps are and have been far from satisfactory. One who has
been a close student of local conditions for a number of years will
often predict with as great accuracy as the weather bureau. Areas of low
pressure are followed sooner or later by a fall of temperature; this is
especially true in the winter months. Sometimes this fall is very
marked, and then it is called a cold wave. These sudden changes of
temperature are not thoroughly understood, but are supposed to be due
partly at least to rapid radiation of heat into the upper regions, as
the clear atmosphere which usually attends areas of high pressure is
favorable to such a condition. Undoubtedly, too, there are dynamic
causes, forcing the colder air from the upper regions to the earth, when
it immediately flows off toward an area of low barometer.

Long-time predictions are purely guesses. They sometimes guess on the
right side, and this gives them courage to make another. It is an old
saying that "all signs fail in dry weather." In time of a drought it is
true that the indications which at ordinary times would be surely
followed by a rai