JANUARY 11, 1884.] 
storms that are raised in temperate latitudes, 
although the latter outnumber the former on 
account of the more effective aid of the earth’s 
rotative deflection at a distance from the equa- 
tor. 
We must next examine the cause that deter- 
mines the season of cyclones, throws them near 
the western shores of their oceans, and requires 
them to move toward or parallel to the east- 
ern coast of the adjoining continents. This 
will be found to depend on the general circula- 
tion of the winds, as may be seen on examin- 
ing the air-currents of the North Atlantic at 
the seasons of the most frequent hurricanes. 
Poey has compiled a list of hurricanes observed 
in the West Indies since 1493, amounting to 
three hundred and sixty-five in all; and of 
these, two hundred and eighty-seven, or nearly 
eighty per cent, occurred in July, August, Sep- 
tember, and October. Now, these are the very 
months when the equatorial calms or doldrums 
are farthest north of the equator, and hence in 
a position to allow the embryonic storms to 
develop by the aid of the earth’s deflective 
force. At other seasons the trade-winds ex- 
tend nearer to the equator; and then, in a lati- 
tude where storms might grow if once started, 
the steady blowing trades prevent even the for- 
mation of an embryo. The few storms that 
oceur at these other seasons have less evident 
causes: they may arise in conflicting winds, 
and may be fairly thrown among those unex- 
plained effects that we call accidental. Once 
formed, the storm is carried along, by the gen- 
eral circulation and by the strong winds, to- 
ward the West Indies. On nearing them, it 
moves to the north-west and north, mostly be- 
cause branches of the trade-winds here turn to 
that direction in the cyclone season, so as to | 
avoid the mountains farther west, and to run 
up over the warm land of our country ; partly 
because of the continual polar tendency, or 
excess of deflection on the northern side of the 
storm. Even if the general surface-winds do 
not blow along the storm-tracks, it is very prob- 
able that the upper current, returning from the 
equatorial calms toward the prevailing westerly 
winds of the temperate latitudes, follows a 
course closely parallel to the average of the 
cyclone paths ; and there is good reason to be- 
lieve that the upper winds have a great control 
over the storm’s progression. If the storm 
should begin on the eastern side of the Atlan- 
tic, it would probably be held so near the equa- 
tor by the indraught of the trade-winds that it 
could not reach a destructive size. The greater 
Atlantic hurricanes are therefore those that 
begin in the western part of the calms or dol- 
SCIENCE. 
_gseribed for the North Atlantic. 
4] 
drums when they are farthest from the equa- 
tor, and then, passing along their carved paths, 
take the West Indies and our south-eastern 
coast or their way up into the North Atlantic. 
As they go, their diameter greatly increases ; 
because they draw their wind-supply from 
longer distances, and because in the temperate 
latitudes the earth’s deflective force is greater 
than it was in the tropics. But with this in- 
crease in diameter there comes a diminution 
of intensity, because the winds are cooler and 
contain less vapor; and finally the storm dies 
away when the weakened updraught at the cen- 
tre fails to throw its overflow outside of the 
limits of the whirl. The storm is then not work- 
ing its way: friction will soon cause the winds 
to cease, and the disturbance will come to an 
end. 
As for the South Atlantic, it possesses no 
cyclone region, because the doldrums never 
extend south of the equator. In spite of the 
sun’s passing to the south in winter, the heat- 
equator, which determines the position of the 
doldrums, hardly passes the geographic equa- 
tor in the Atlantic; the excess of land in the 
northern hemisphere, and the strong general 
winds of the southern hemisphere, keep it 
back: and so the South Atlantic has no ecy- 
clones such as occur in all the other oceans. 
The cyclones of the Pacific and Indian oceans 
depend on conditions such as have been de- 
They are 
commonest in the southern hemisphere in Feb- 
ruary for the same reason that they are most 
frequent in the northern in the months about 
September. 
We have now considered the origin and mo- 
tions of the cyclones and hurricanes, and the 
regions of their occurrence. This study has 
its highest aim in giving timely warning of 
their approach and in devising rules for avoid- 
ing them. If their tracks lay over the land, 
the telegraph could in all cases give sufficient 
notice of their coming, for their motion is slow ; 
but they are at sea during much of their life, 
and the questions now arise, How can the cap- 
tain of a vessel gain the first intimation of their 
coming? and, What should he best do to avoid 
their dangerous centre? 
The storm’s earliest effect on the atmosphere 
is shown by the barometer. It is ordinarily 
stated that the first effect is seen in a diminu- 
tion of pressure; but it is very probable, both 
from theory and from careful observation, that 
a slight abnormal increase of pressure precedes 
this diminution. ‘The tropical seas, where cy- 
clones are most violent, have, as a rule, very 
small and very rare irregular changes in at- 
