d, as all of us have reason to
know, weather predictions made by those who have the matter in charge
and are supposed to know all about it often fail to come to pass. The
real trouble is that they do not know all about it. There are so many
conditions existing that are outside of the range of barometers,
thermometers, anemometers, and telegraphs that no one can tell just when
some of these unknown factors will step in to spoil our predictions.

In very many cases, perhaps in a large majority of them, the predictions
made by the weather bureau substantially come to pass. It has been
stated in former chapters that the changes of weather accompany the
movements of what are called cyclones and anti-cyclones, the cyclone
being accompanied by low barometric pressure and the anti-cyclone by a
higher one. The winds of the cyclone move spirally around the center of
lowest depression with an upward trend, the motions being in a direction
reversed to that of the hands of a clock. In the centers of high
pressure the current is downward instead of upward and the direction of
the wind around it is opposite to that around the low-pressure area. The
fundamental factor in predicting the weather is the direction of
movement of these areas of low pressure. In almost all cases the
direction of movement is from the west to the east, but not always in a
straight line. These movements, however, are classified so that after
the direction has become established one can predict with considerable
accuracy as to whether it will move in a curved or a straight line. By
movement we do not refer to the direction of the wind at any particular
point, but the onward movement of the whole cyclonic system, which is
usually from twenty-five to thirty miles an hour, but in some cases the
speed is much greater.

Not only does the upward movement of the whole system vary, but the
velocity of the wind around any given cyclonic center varies. There are
about eleven classes of cyclones that appear in the United States, each
class having its own path of movement and origin. A large number of
these appear to originate north of the Dakotas, and move directly east
to the Gulf of St. Lawrence. Three other classes originate on about the
same line, a little west,--say, north of Montana,--moving first in a
southeasterly direction, passing over the center of Lake Michigan and
bending northerly through Lake Ontario and finally landing in the Gulf
of St. Lawrence. Two othe