that from the S. to the tropic of Cancer. But while travelling in these lofty regions they become cold and heavy, and, sinking to the surface at the tropics, each proceeds to the opposite pole from which it set out. Now, however, they have a greater rotaplaces they successively arrive at, so the bias is to the W., and they become the N.W. and S.W. extratropical winds.

tory motion than the

If on arriving at the poles the air were to accumulate there, the circulation of the winds would cease; but currents rise into the upper regions, and flow back again to the tropics, where they sink down to fill the vacuum caused by the great precipitation of vapour in these regions, and then flow to the equator as trade-winds (N. 177). So the currents of air cross again at the tropics and produce two belts of calms which surround the globe, named by Lieutenant Maury the Calms of Cancer and the Calms of Capricorn, but generally know to sailors as the Doldrums. Thus the winds go from pole to pole and back again, alternately as under and upper currents. In their circuits the winds cross each other five times, producing regions of calms at the poles, the tropics, and equator. The trade-winds generally extend for about 28° on each side of the equator, but, on account of the greater quantity of land in the northern hemisphere, the N.E. trade-wind is narrower than the S.E.

The sun is perpetually raising enormous quantities of vapour from the ocean which the trade-winds carry to the equator: it is condensed when it rises with the air into the higher strata, and forms a ring of clouds along the southern side of the belt of equatorial calms that surrounds the earth, which, during the day, is perpetually pouring down torrents of rain, while the sun continually beating upon its upper surface dissolves the clouds into invisible vapour which is carried by the winds and condensed into rain on the extra-tropical regions. The whole system of trade-winds, equatorial and tropical calms, with the cloud ring, follow the sun in declination; consequently in its journeys back and forwards it annually travels over 1000 miles of lätitude, and regulates the dry and rainy season in the tropical parts of the earth.

The monsoons, which are periodic winds in the Indian Ocean, in part depend upon this movement. For when the sun is in the northern hemisphere the trade-winds come northward with him ;

and when his intense heat expands the air over the Great Gobi and other arid Asiatic deserts, it ascends; the N.E. trade-wind is drawn in to fill the vacuum and ascends with it; then the S.E. trade-wind, being no longer met and balanced by the N.E. trade, passes into the northern hemisphere, and as it proceeds northward from the equator it is deflected to the west by the rotation of the earth, combined with the indraught over the heated deserts, and becomes the S.W. monsoon, which blows while the sun is north of the equator, but as soon as he goes south, and no longer rarefies the air over the Indian deserts, the S.E. tradewind resumes its usual course, and is then known as the S.E. monsoon. The influence of the heated deserts is perceptible to the distance of 1000 miles from the shore; the monsoons prevail with great steadiness over the Arabian Gulf, the Indian Ocean, and part of the China Sea. At the change, torrents of rain and violent thunderstorms accompany the conflict between the contending winds.

The Sahara desert in North Africa, and those of Utah, Texas, and New Mexico, occasion the monsoons which prevail in the North Atlantic and on both sides of Central America, and the monsoons which blow to the north of Australia show the sterility of the interior, even if other proofs were wanting. From the powerful effect of the land in drawing off the winds from their course, it may be seen why the N.E. trade-winds are narrower than the S.E. trades.

In the extra-tropical winds in the North Atlantic, which blow from the 40th parallel to the pole, the north-westerly are to the easterly as 2 to 1: hence there would be an accumulation of air at the pole at the expense of the equator, did not a current rise at the pole and return to the equator through the high regions of the atmosphere, which confirms the theory of the rotation of the wind.

There are many proofs of the existence of the counter-currents above the trade-winds. On the Peak of Teneriffe the prevailing winds are from the west. Light clouds have frequently been seen moving rapidly from west to east at a very great height above the trade-winds, which were sweeping along the surface of the ocean in a contrary direction. Rains, clouds, and nearly all the other atmospheric phenomena, occur below the height of 18,000 feet, and generally much nearer to the surface of the earth.

They are owing to currents of air running upon each other in horizontal strata, differing in their electric state, in temperature and moisture, as well as in velocity and direction.

When north and south winds blow alternately, the wind at any place will veer in one uniform direction through every point of the compass, provided the one begins before the other has ceased. In the northern hemisphere a north wind sets out with a smaller degree of rotatory motion than the places have at which it successively arrives, consequently it passes through all the points of the compass from N. to N.E. and E. A current from the south, on the contrary, sets out with a greater rotatory velocity than the places have at which it successively arrives, so by the rotation of the earth it is deflected from S. to S.W. and W. Now, if the vane at any place should have veered from the N. through N.E. to E., and a south wind should spring up, it would combine its motion with the former and cause the vane to turn successively from the E. to S.E. and S. But by the earth's rotation this south wind will veer to the S.W. and W., and, if a north wind should now arise, it would combine its motion with that of the west, and cause it to veer to the N.W. and N. Thus two alternations of north and south wind will cause the vane at any place to go completely round the compass, from N. to E., S., W., and N. again. At the Royal Observatory at Greenwich the wind accomplishes five circuits in that direction in the course of a year. When circumstances combine to produce alternate north and south winds in the southern hemisphere, the gyration is in the contrary direction. Although the general tendency of the wind may be rotatory, and is so in many instances, at least for part of the year, yet it is so often counteracted by local circumstances, that the winds are in general very irregular, every disturbance in atmospheric equilibrium from heat or any other cause producing a corresponding wind. The most prevalent winds in Europe are the N.E. and S.W.; the former arises from the north polar current, and the latter from causes already mentioned. The law of the wind's rotation was first described by Dr. Dalton, but has been developed by Professor Dove, of Berlin.

Hurricanes are those storms of wind in which the portion of the atmosphere that forms them revolves in a horizontal circuit round a vertical or somewhat inclined axis of rotation, while the axis itself, and consequently the whole storm, is carried forward

along the surface of the globe, so that the direction in which the storm is advancing is quite different from the direction in which the rotatory current may be blowing at any point. In the West Indies, where hurricanes are frequent and destructive, they generally originate in the tropical regions near the inner boundary of the trade-winds, and are caused by the vertical ascent of a column of rarefied air, whose place is supplied by a rush of wind from the surrounding regions, set into gyration by the rotation of the earth. By far the greater number of Atlantic hurricanes have begun eastward of the lesser Antilles or Caribbean Islands.

In every case the axis of the storm moves in an elliptical or parabolic curve, having its vertex in or near the tropic of Cancer, which marks the external limit of the trade-winds north of the equator. As the motion before it reaches the tropic is in a straight line from S.E. to N.W., and after it has passed it from S.W. to N.E., the bend of the curve is turned towards Florida and the Carolinas. In the southern hemisphere the body of the storm moves in exactly the opposite direction. The hurricanes which originate south of the equator, and whose initial path is from N.E. to S.W., bend round at the tropic of Capricorn, and then move from N.W. to S.E.

The extent and velocity of these storms are great; for instance, the hurricane that took place on the 12th of August, 1830, was traced from eastward of the Caribbee Islands, along the Gulf Stream, to the bank of Newfoundland, a distance of more than 3000 miles, which it passed over in six days. Although the hurricane of the 1st of September, 1821, was not so extensive, its velocity was greater, as it moved at the rate of 30 miles an hour: small storms are generally more rapid than those of greater dimensions.

The action of these storms seems to be at first confined to the stratum of air nearest the earth, and then they seldom appear to be more than a mile high, though sometimes they are raised higher; or even divided by a mountain into two separate storms, each of which continues its new path and gyrations with increased violence. This occurred in the gale of the 25th of December, 1821, in the Mediterranean, when the Spanish mountains and the Maritime Alps became new centres of motion.

By the friction of the earth the axis of the storm bends a little forward. This causes a continual intermixture of the lower and

warmer strata of air with those that are higher and colder, producing torrents of rain and violent electric explosions.

The breadth of the whirlwind is greatly augmented when the path of the storm changes on crossing the tropic. The vortex of a storm has covered an extent of the surface of the globe 500 miles in diameter.

The revolving motion accounts for the sudden and violent changes observed during hurricanes. In consequence of the rotation of the air, the wind blows in opposite directions on each side of the axis of the storm, and the violence of the blast increases from the circumference towards the centre of gyration, but in the centre itself the air is in repose: hence, when the body of the storm passes over a place, the wind begins to blow moderately, and increases to a hurricane as the centre of the whirlwind approaches; then, in a moment, a dead and awful calm succeeds, suddenly followed by a renewal of the storm in all its violence, but now blowing in a direction diametrically opposite to its former course. This happened at the Island of St. Thomas on the 2nd of August, 1837, where the hurricane increased in violence till half-past seven in the morning, when perfect stillness took place for forty minutes, after which the storm recommenced in a contrary direction.

The sudden fall of the mercury in the barometer in the regions habitually visited by hurricanes is a certain indication of a coming tempest. In consequence of the centrifugal force of these rotatory storms the air becomes rarefied, and, as the atmosphere is disturbed to some distance beyond the actual circle of gyration or limits of the storm, the barometer often sinks some hours before its arrival, from the original cause of the rotatory disturbance. It continues sinking under the first half of the hurricane, is at a maximum sometimes of two inches in the centre of gyration, and again rises during the passage of the latter half, though it does not attain its greatest height till the storm is over. The diminution of atmospheric pressure is greater and extends over a wider area in the temperate zones than in the torrid, on account of the sudden expansion of the circle of rotation when the gale crosses the tropic.

As the fall of the barometer gives warning of the approach of a hurricane, so the laws of the storm's motion afford the seaman knowledge to guide him in avoiding it. In the northern