Page images



we must, however, just glance at those winds which are produced by whirls or rotatory movements in the atmosphere of greater or less extent, the winds which have hitherto engaged our attention being general or periodical winds of the nature of currents.

Rotatory disturbances in the air display a variety of phenomena which have been distinguished by various names, such as whirlwind, water-spout, sand-spout, sand-pillar, tornado, white squall, pampero, &c. These terms have been applied to rotatory storms of small extent. The hurricanes of the West Indies and the typhoons of the Indian seas are whirls of greatly increased extent, often extending from one to four or five hundred miles in diameter, and consisting of a revolving movement propagated from place to place, not by bodily transfer of the whole mass of air which at any moment constitutes the hurricane from one geographical point to another, but by every part of the atmosphere in its track receiving from that before it and transmitting to that after it this revolving movement.

It was formerly supposed, when accounts of hurricanes were received, as occurring at different islands on various dates, with marked differences also in the direction of the wind, that those violent storms were rectilinear in their course, and that such accounts related in most cases to different storms. Their true nature was first clearly established by Mr. Redfield of New York, and afterwards by Lieutenant-Colonel Reid, Governor of Barbadoes, by constructing charts on a

of the winds brought together in a popular form, in a little book by the author of this work, entitled "The Tempest; or, an Account of the Nature, Properties, Dangers, and Uses of Wind in various Parts of the World." Published under the direction of the Committee of General Literature and Education, appointed by the Society for Promoting Christian Knowledge.

large scale of some of the most remarkable hurricanes on record. In this way a single storm was traced successively from one island or locality to another, and the direction of the wind at any one point or place was found to have no connection with the general progress or direction of the storm. For example, one of the tracks thus laid down is that of the memorable gale of August, 1830, which passing close by the Windward Islands, visited St. Thomas's on the 12th; was near Turk's Island on the 13th; at the Bahamas on the 14th; on the Gulf and coast of Florida on the 15th; along the coast of Georgia and the Carolinas on the 16th; off Virginia, Maryland, New Jersey, and New York on the 17th; off George's Bank and Cape Sable on the 18th; and over the Porpoise and Newfoundland Banks on the 19th of the same month, having occupied about seven days in its ascertained course from near the Windward Islands, a distance of more than 3000 miles, the rate of its progress being equal to 18 miles an hour. Now, the actual velocity of the wind in its rotatory movement is probably five times greater than this rate of progress, which would be equal in a rectilinear course to about 15,000 miles; but the whole length of the track does not exceed 3000 miles: we thus have strong evidence of the rotatory nature of these storms; for, if the wind move 90 miles an hour, and the whirls or eddies which constitute the storm move in a body at the rate of only 18 miles an hour, it is clear that the motion of the wind must be in circles. It is curious also that these eddies turn in a direction contrary to the sun's apparent daily motion, and, therefore, in our hemisphere, contrary to the motion of the hands of a watch, but similar to that motion in the southern hemisphere, a result which would naturally be expected from their common origin, just as two wheels set in



motion by anything passing between them must necessarily turn in contrary directions. The general progress of these storms is always away from the equator, and therefore reversed in the two hemispheres; but in both they have first a westward motion, until they escape the influence of the easterly trade winds, when they turn round in a hyperbolic curve and are drifted eastward by the prevailing westerly winds of each temperate zone. Throughout their course they increase in size, but diminish in intensity, until they are lost in the winds of high latitudes, the variable and fluctuating nature of which they greatly increase.

Sir John Herschel explains the origin of one of these rotatory storms by supposing a column of air, intensely heated at a particular point of the intertropical plains of America, to rise bodily from the lower stratum of the atmosphere with sufficient ascensional force to carry it into the upper current, but retaining the full westerly energy which it has derived from the earth's rotation. Now nothing is more likely than that a ripple in its course should thus be produced, and that the portion thus driven upwards should, on its return, strike down far below, into the lower current. All the conditions necessary for a rotatory storm would then arise. A mass of air, animated with immense velocity, has to force its way through an atmosphere either at rest or moving in a contrary direction: a state of things which, in the movements of fluids, is invariably accompanied with vortices on one or both sides of the moving mass, which continue to subsist and to wander over great tracks long after the original impulse is withdrawn. In such vortices the motions of translation and rotation may have any proportion; but the former is usually slow compared with the latter. An illustration of this kind may be witnessed at a mill-dam when the sluice is


closed so as to allow the water to escape at some small hole. A funnel-shaped depression will appear on the surface, in which air descends, often to the actual hole of escape, though many feet below the surface, but often also as an interrupted column. All the movements of a rotatory storm are here represented by the revolving fluid. So long as the hole is kept open it retains a fixed position, at least at the lower extremity, fluctuating only by a bend of the column. But if the hole be closed, it immediately begins to wander, continuing often a long time, but gradually retreating upwards.


46. It must be obvious that all these atmospheric currents, depending as they do entirely on the variations of elasticity brought about in different parts of the atmosphere by the solar heat, must be most intimately connected with the indications of the barometer. the tendency of all these currents is to establish equality of pressure over the earth's surface, it was formerly supposed that the mean height of the barometric column would be found equal in every part of the world. Numerous careful and long-continued registers of its height, however, kept at various spots distant from each other, have shown a decided though very small local variation of this element, and the dependence of this variation on the latitude we will now endeavour to explain.

The atmospheric pressure is greatest at, or at a short distance beyond, the tropics, or about the outer limit of the trade winds, where the mean height of the barometer is about 30 inches. From this point it declines both towards the equator, where it is about 299, and also towards high latitudes. In England, for example, it is reduced to 29'8, and it continues to fall at about the same rate to the highest latitudes that have been



reached. But towards the South Pole, as appears from the late expedition under Ross, the fall appears to increase rapidly, so much so, that in the Antarctic regions the mean pressure is reduced even below 29 inches; but it must be remembered that this is only the mean of three summers, and may possibly be compensated by increased pressure in winter.

The maximum of mean pressure occurring just outside the tropics, is referred by Professor Daniell, with great probability, to the obstruction caused by the meeting and crossing of the two main currents as already described (37), which must produce an accumulation of air upon that parallel where it occurs.

47. But the mean local pressure appears to depend less on latitude than on the proximity of great masses of land or water, it being highest in the interior of continents and lowest on the Pacific. Captain King, in a voyage on that hemisphere of waters, having observed the barometer five times a day for five months, obtained a mean of only 29 462 inches. Connected with this fact is the lower mean pressure in the Antarctic regions than in the Arctic, and also the well-known fact of the barometer on the borders of continents (as in this country) standing lowest during moist winds, or those which come from the ocean, and highest during the dry winds from the continent.

48. This brings us to the fact, that the local differences between the mean pressure observed at the most distant parts of the globe are very trifling compared with the variations constantly occurring at most places, indeed all places, outside the tropics. Thus in this country we have at the sea-level the mercury not unfrequently rising to 30.5 inches and sometimes falling to 28.5. This difference of two inches indicates a difference of about a pound per square inch in the

« PreviousContinue »