which the number of spots visible on the sun first increases to a
maximum, then diminishes to a minimum, and finally increases again to
a maximum. For unknown reasons the period is sometimes two or three
years longer than the average and sometimes as much shorter.
Nevertheless, the phenomena always recur in the same order. Starting,
for instance, with a time when the observer can find few or no spots,
they gradually increase in number and size until a maximum, in both
senses, is reached, during which the spots are often of enormous size
and exceedingly active. After two or three years they begin to
diminish in number, magnitude, and activity until they almost or quite
disappear. A strange fact is that when a new period opens, the spots
appear first in high northern and southern latitudes, far from the
solar equator, and as the period advances they not only increase in
number and size, but break out nearer and nearer to the equator, the
last spots of a vanishing period sometimes lingering in the equatorial
region after the advance-guard of its successor has made its
appearance in the high latitudes. Spots are never seen on the equator
nor near the poles. It was not very long after the discovery of the
sun-spot cycle that the curious observation was made that a striking
coincidence existed between the period of the sun-spots and another
period affecting the general magnetic condition of the earth. When a
curved line representing the varying number of sun-spots was compared
with another curve showing the variations in the magnetic state of the
earth the two were seen to be in almost exact accord, a rise in one
curve corresponding to a rise in the other, and a fall to a fall.
Continued observation has proved that this is a real coincidence and
not an accidental one, so that the connection, although as yet
unexplained, is accepted as established. But does the influence extend
further, and directly affect the weather and the seasons as well as
the magnetic elements of the earth? A final answer to this question