Physical Geography of the Sea and its Meteorology/Chapter 19

by Matthew Fontaine Maury

CHAPTER XIX.

§ 781-808.—STORMS, HURRICANES, AND TYPHOONS.

781. Plate V.—Plate V. is constructed from data furnished by the Pilot Charts, as far as they go, that are in process of construction at the National Observatory. For the Pilot Charts the whole ocean is divided off into "fields " or districts of five degrees square, i. e. five degrees of latitude by five degrees of longitude, as already stated in the "Explanation of the Plates." Now, in getting out from the log-books materials for showing, in. every district of the ocean, and for every month, how navigators have found the winds to blow, it has been assumed that, in whatever part of one of these districts a navigator may be when he records the direction of the wind in his log, from that direction the wind was blowing at that time all over that district; and this is the only assumption that is permitted in the whole course of the investigation. Now if the navigator will draw, or imagine to be drawn in any such district, twelve vertical columns for the twelve months, and then sixteen horizontal lines through the same for the sixteen points of the compass, i. e. for N. N.N.E. N.E., E.N.E., and so on, omitting the by-points, he will have before him a picture of the "Investigating Chart" out of which the "Pilot Charts" are constructed. In this case the alternate points of the compass only are used, because, when sailing free, the direction of the wind is seldom given for such points as N. by E., W. by S., etc. Moreover, any attempt, for the present, at greater nicety would be over-refinement, for navigators do not always make allowance for the aberration of the wind; in other words, they do not allow for the apparent change in the direction of the wind caused by the rate at which the vessel may be moving through the water, and the angle which her course makes with the true direction of the wind. Bearing this explanation in mind, the intelligent navigator will have no difficulty in understanding the wind diagram (Plate V.) and in forming a correct opinion as to the degree of credit due to the fidelity with which the prevailing winds of the year are represented on. Plate VIII. As the compiler wades through log-book after log-book, and scores down in column after column, and upon line after line, mark upon mark, he at last finds that, under the month and from the course upon which he is about to make an entry, he has already made four marks or scores, thus (////). The one that he has now to enter will make the fifth, and he "scores and tallies," and so on until all the abstracts relating to that part of the ocean upon which he is at work have been gone over, and his materials exhausted. These "fives and tallies" are exhibited on Plate V. Now, with this explanation, it will be seen that in the district marked A (Plate V.) there have been examined the logs of vessels that, giving the direction of the wind for every eight hours, have altogether spent days enough to enable me to record the calms and the prevailing direction of the winds for eight hours, 2144 times : of these, 285 were for the month of September; and of these 285 observations for September, the wind is reported as prevailing for as much as eight hours at a time: from N. 3 times; from N.N.E., 1; N.E., 2; E.N.E., 1; E., 0; E.S.E., 1; S.E., 4; S.S.E., 2;S., 25; .S.S.W., 45; S.W., 93; W.S.W., 24; W., 47; W.N.AV., 17; N.W., 15; N.N.W., 1; Calms (the little O's), 5; total 285 for the month in this district. The number expressed in figures denotes the whole number of observations of calms and winds together that are recorded for each month and district. In C, the wind in May sets one third of the time from west. But in A, which is between the same parallels, the favourite quarter for the same month is from S. to S.W., the wind setting one third of the time from that quarter, and only 10 out of 221 times from the west; or, on the average, it blows from the west only 1⅓ day during the month of May. In B, notice the great "sun swing " of the winds in September, indicating that the change from summer to winter, in that region, is sudden and violent; from winter to summer, gentle and gradual. In some districts of the ocean, more than a thousand observations have been discussed for a single month, whereas, with regard to others, not a single record is to be found in any of the numerous log-books at the National Observatory.

782. Typhoons.—The China seas are celebrated for their furious gales of wind, known among seamen as typhoons and white squalls. The seas are included on the plate (VIII.) as within the region of the monsoons of the Indian Ocean. But the monsoons of the China Sea are not five month monsoons (§ 681); they do not prevail from the west of south more than two or three months. Plate V. exhibits the monsoons very clearly in a part of this sea. In the square between 15° and 20° north, 110° and 115° east, there appears to be a system of three monsoons; that is, one from the north-east in October, November, December, and January; one from east in March and April, changing in May; and another from the southward in June, July, and August, changing in September. The great disturber of the atmospheric equilibrium appears to be situated among the plains and steppes of Asia ; their influence reaches up to the clouds, and extends to the China Seas; it is about the changing of the monsoons that these awful gales, called typhoons and white squalls, are most dreaded.

783. The Mauritius hurricanes.—In like manner, the Mauritius hurricanes, or the cyclones of the Indian Ocean, occur during the unsettled state of the atmospheric equilibrium which takes place at that debatable period during the contest between the trade-wind force and the monsoon force (§ 699), and which debatable period occurs at the changing of the monsoon, and before either force has completely gained or lost the ascendency. At this period of the year, the winds, breaking loose from their controlling forces, seem to rage with a fury that would break up the very fountains of the deep.

784. The West India hurricanes.—So, too, with the West India hurricanes of the Atlantic ; these winds are most apt to occur during the months of August and September. There is, therefore, this remarkable difference between these gales and those of the East Indies: the latter occur about the changing of the monsoons, the former during their height. In August and September, the south-west monsoons of Africa and the south-east monsoons of the West Indies are at their height; the agents of one drawing the north-east trade-winds from the Atlantic into the interior of New Mexico and Texas, the agents of the other drawing them into the interior of Africa. These two forces, pulling in opposite directions, assist now and then to disturb the atmospheric equilibrium to such an extent that the most powerful revulsions in the air are required to restore it. "The hurricane season in the North Atlantic Ocean," says Jansen, "occurs simultaneously with the African monsoons; and in the same season of the year in which the monsoons prevail in the North Indian Ocean and the China Sea, and upon the Western coast of Central America, all the seas of the northern hemisphere have the hurricane season. On the contrary, the South Indian Ocean has its hurricane season in the opposite season of the year, and when the north-west monsoon prevails in the East Indian Archipelago."

785. The cyclone theory.—Under the head of hurricanes, typhoons and tornadoes, I include all those gales of wind which are known as cyclones. These have been treated of by Redfield in America, Reid in England, Thom of Mauritius, and Piddington of Calcutta, with marked ability, and in special works. I refer the reader to them. The theory of this school is, that these are rotary storms; that they revolve against the hands of a watch in the northern, and with the hands of a watch in the southern hemisphere; that nearer the centre or vortex the more violent the storm, while the centre itself is a calm, which travels sometimes a mile or two an hour, and sometimes forty or fifty; that in the centre the barometer is low, rising as you approach the periphery of the whirl; that the diameter of these storms-is sometimes a thousand miles, and sometimes not more than a few leagues; that they have their origin somewhere between the parallels of 10° and 20° north and south, travelling to the westward in either hemisphere, but increasing their distance from the equator, until they reach the parallel of 25° or 30°, when they turn towards the east, or "recurvate," but continue to increase their distance from the equator—i. e., they first travel westwardly, inclining towards the nearest pole; they then recurve and travel eastwardly, still inclining towards the pole; and that such is their path in both hemispheres, etc.

786. Puzzling questions.—The questions why these storms should recurve, and why they should travel as they do, and why they should turn with the hands of a watch in the southern, and against them in the northern hemisphere, are still considered by many as puzzles, though it is thought that their course to the westward in the trade-wind region, and to the eastward in the counter-trades, is caused by the general movement of the atmosphere, like the whirls in an angry flood, which, though they revolve, yet they are borne down stream with the currents as they do revolve. The motion polarward is caused, the conjecture is, by the fact that the equatorial edge of the storm has, in consequence of diurnal rotation, a greater velocity than its polar edge. There seems, however, to be less difficulty with regard to their turning than with regard to their course; the former is now regarded as the resultant of diurnal rotation and of those forces of translation which propel the winds along the surface of our planet. This composition of the forces of the revolving storm, and the resolution of them, are precisely such (§ 215) as to produce opposite rotation on opposite sides of the equator.

787. Espys theory.—Many of the phenomena connected with these storms still remain to be explained; even the facts with regard to them are disputed by some. The late Professor Espy, after having discussed for many years numerous observations that have been made chiefly on shore, maintained that the wind does not blow around the vortex or place of low barometer, but directly towards it. He held that the place of low barometer, instead of being a disc, is generally an oblong, in the shape of a long trough, between two atmospherical waves; that it is curved with its convex side towards the east; that it is sometimes nearly straight, and generally of great length from north to south, reaching in America, from the Gulf of Mexico to the great lakes and beyond, and having but little breadth in proportion to its length; that it travels east, moving side foremost, requiring about two days to go from the Mississippi to St. John's, Newfoundland; that on either side of it, but many miles distant, there is a ridge of high barometer; that the wind on either side of the line of low barometer, in which there is little or no wind, blows toward it, etc., and, in support of these positions, he advanced this theory: "When the air in any locality acquires a higher temperature or a higher dew-point than that of the surrounding regions, it is specifically lighter, and will ascend; in ascending, it comes under less pressure, and expands; in expanding from diminished pressure, it grows colder about a degree and a quarter for every hundred yards of ascent; in cooling as low as the dew-point (which it will do when it rises as many hundred yards as the dew-point at the time is below the temperature of the air in degrees of Fahrenheit), it will begin to condense its vapour into cloud; in condensing its vapour into water or cloud, it will evolve its latent caloric; this evolution of latent caloric will prevent the air from cooling so fast in its farther ascent as it did in ascending below the base of the cloud now forming; the current of the air, however, will continue to ascend, and grow colder about half as much as it would do if it had no vapour in it to condense; and when it has risen high enough to have condensed, by the cold of expansion from diminished pressure, one hundredth of its weight of vapour, it will be about forty-eight degrees less cold than it would have been if it had no vapour to condense nor latent caloric to give out—that is, it will be about forty-eight degrees warmer than the surrounding air at the same height; it will, therefore (without making any allowance for the higher dew-point of the ascending current), be about one tenth lighter than the surrounding colder air, and, of course, it will continue to ascend to the top of the atmosphere, spreading out in all directions above as it ascends, overlapping the air in all the surrounding regions in the vicinity of the storm, and thus, by increasing the weight of the air around, cause the barometer to rise on the outside of the storm, and fall still more under the storm-cloud by the outspreading of air above, thus leaving less ponderable matter near the centre of the upmoving column to press on the barometer below. The barometer thus standing below the mean under cloud in the central regions, and above the mean on the outside of the cloud, the air will blow on all sides from without, inward, under the cloud. The air, on coming under the cloud, being subjected to less pressure, will ascend and carry up the vapour it contains with it, and as it ascends will become colder by expansion from constantly diminishing pressure, and will begin to condense its vapour in cloud at the height indicated before, and thus the process of cloud-forming will go on. Now it is known that the upper current of air in the United States moves constantly, from a known cause, towards the eastward, probably a little to the south of east; and as the upmoving column containing the cloud is chiefly in this upper current of air, it follows that the storm-cloud must move in the same direction. And over whatever region the storm-cloud appears, to that region will the wind blow below; thus the wind must set in with a storm from some eastern direction, and, as the storm-cloud passes on towards the eastward, the wind must change to some western direction, and blow from that quarter till the end of the storm."^[1]

788. Doves law.—According to Dové's "Law of Rotation," which is said to hold good in the northern hemisphere, and is supposed to obtain in the southern also, the wind being N.W. and veering, it ought to veer by W. to S.W., and so on, against the hands of a watch. This "law" is explained thus: Suppose a ship be in S. lat. off Cape Horn as at a, with a low barometer to the north of her, as at C, where the air ascends as fast as it comes pouring in from all sides. The ship, let it be supposed, is just on the verge of, but exterior to the vortex, or that place where the wind commences to revolve. The first rush of the air at a will be directly for the centre C; consequently, a ship so placed would report the storm as commencing with the wind at south.

For the sake of illustration, we will suppose this place of low barometer to be stationary, and the air, as it rushes in, to ascend at the disc C. Thus the area of inrushing; air will gradually enlarge itself by broad spreading, like a circle on the water, until it be compassed by a circle with a radius C S, of indefinite length. The air then, on the meridian S C N, but to the south of a, will not blow along this meridian and pass over the ship; in consequence of the diurnal rotation of the earth, it will take a direction, S a', to the westward; and the arrow d a, representing a S.S.E. wind, will now show the direction of the wind at a. Thus the ship will report that the wind commenced at south, and gradually hauled to S.S.E., i.e. against the hands of a watch; and so the arrows b' a' will represent the direction of the wind at each station, a' a' a' when the storm commenced, and the arrows d' a' the directions afterwards, thus showing it to have veered against the hands of a watch. And this is the direction in which the forces of diurnal rotation, when not mastered by opposing forces, always require the wind, when not blowing round in spirals and a whirl, to haul in the southern hemisphere. Now, paradoxical as it may at first seem, it is also the forces of diurnal rotation that give that same wind, when it is blowing round in spirals, its first impulse to march round in the contrary direction, or (§ 786) with the hands of a watch; but this is as it should be—it hauls one way, and marches the other. After passing a, and each of the other stations, a' a', the rush of wind is sufficient, let us suppose, to create a whirl. The wind at a' a' a', continuing on with a circular motion, is represented thenceforward in its course by the curved arrows a e, a' e'. The wind coming from the east and the west has no direct impulse from diurnal rotation, but the wind on either side of it has, and hence the prime vertical wind is carried around with the rest. If, now, we imagine the disc C to be put in motion, and the storm to become a travelling one, we shall have to consider the composition and resolution of other forces also, such as those of traction, aberration, and the like, before we can resolve the whirlwind.

789. Bernouilli's formula.—But the cyclonologists do not locate their storms in such high latitudes as the parallels of Cape Horn. Hence we might safely infer, one would suppose, that in high southern latitudes a north wind has a tendency to incline to the ^westward and a south wind to the eastward; and the cause of this tendency is in operation, whether the place of low barometer be a disc or an oblong, for it is in obedience to the trade-wind law, as expounded by Halley, that it so operates; and it will also be the case whether the wind be caused by an influx into the place of low, or the efflux from the place of high barometer; or, as is generally the case, by both together. If the distance between the place of high and low barometer were always the same, then a given difference of barometric pressure would always be followed by a wind of the same force of velocity. By expanding Bernouilli's formula for the velocity of gas jets under given pressures, Sir John F. W. Herschel has computed^[2] the velocity and the force with which currents of air or winds would issue under certain differences of barometric pressure. Under the most favourable conditions, i. e., when the places of high and of low barometer are in immediate juxtaposition, as on the inside and outside of an air-pump, an effective difference of 0.006 inch in the barometric pressure would create a breeze with a velocity of seven miles the hour. Such a wind is capable of exerting a horizontal pressure of 0.2 lb. the square foot, thus :

Diff. barometric Pressure.	Velocity of Wind.	Horizontal Pressure.	Strength of Wind.
.006 inch	7 miles per hour.	0.2 lbs. per sq. ft.	Gentle air.
0.010,,	1-4,,,,	0.9,,,,	Light breeze.
0.016,,	21,,,,	1.9,,,,	Good sailing breeze.
0.06,,	41,,,,	7.5,,,,	A gale.
0.14,,	61,,,,	16.7,,,,	Great storm.
0.25,,	82,,,,	30.7,,,,	Tempest.
0.41,,	92,,,,	37.9,,,,	Devastating hurricane.

Changes, however, in the barometer, amounting to five or six times these differences, are observed to take place at sea without producing winds exceeding in velocity the rates above. This is because the places of high and low barometer at sea are far apart, and because, also, of the obstructions of the winds afforded by the inequalities of the earth's surface.

790. Predicting storms.—But, in this view of the subject, the importance of a daily system of weather reports by telegraph on shore, and across the water between Europe and America when the sub-Atlantic cable is well laid, looms up and assumes all the proportions of one of the great practical questions of the age. We may conjecture, as the probable result of observation, that the greater the distance between the place of high and low barometer, the less the velocity of wind for a given barometric difference would be. Professor Buys Ballot has discovered, practically, the numerical relation between the force of the wind and given barometric differences for certain places in Holland. With the view of ascertaining like relations for this country, it has been proposed to establish a cordon of meteorological stations over the United States, each station being required to report daily to the Observatory in Washington, by telegraph, the height of the barometer, force of wind, etc. By such a plan, properly organized, we might expect soon to be able to give the ships not only on the great lakes, but in our sea-port towns also, timely warning of many a gale, and to send by telegraph to Europe—when one shall be paid—warning of many a one long before it could traverse the Atlantic. The contributions which the magnetic telegraph is capable of making for the advancement of meteorology will enable us to warn the ships in our Gulf ports, as well as those of Cuba, of the approach of every hurricane or tornado that visits those regions.

791. The changing of the wind in a cyclone.—But, returning to the cyclone theory: though the wind be blowing around in spirals against the hands of the watch, yet, from the fact that the centre about which it is blowing is also travelling along, the changes of the wind, as observed by a vessel over which the storm is passing, will not, under all circumstances, be against the sun in the northern, or with the sun in the southern hemisphere. The reason is obvious. This point is worth studying, and any one who will resort to "moving diagrams" for illustration will be repaid with edification.

Piddington's horn cards are the best; but let those who have them not cut a disc of paper of any convenient diameter, say 2½ inches, and then cut out a circle of 2 inches from the middle; this will leave a ring half an inch broad upon which to draw arrows representing the course of the wind. Suppose them to be drawn for the northern hemisphere, as in the annexed diagram; lay the paper ring on the chart: suppose the ship to be in the N.E. quadrant of the storm, which is travelling north, the centre of the storm will pass to the west, but the wind will change from S.E. to S., and so on to the west, with the hands of a watch, though it be revolving about the centre against the hands of a watch; still the rule for finding the direction of the centre holds good: Face the wind, and the centre in the northern hemisphere will be to the right; in the southern, to the left.

792. The wind stronger on one side than the other.—Suppose that in the case before us the storm is travelling to the north at the rate of 20 miles the hour, and that the wind is revolving around the centre also at the rate of 20 miles the hour: when the vortex bears west of the ship, the wind will be south. It is going 20 miles to the north with the body of the storm, and 20 miles around the centre; total force of the wind, 40 miles an hour on the east side. Now imagine yourself on the other side, that is, that you are in the north-west quadrant, and that the storm is travelling due north as before; the vortex will pass east of you, when the wind should have changed from N.E. to north, turning against the hands of a watch; but when the wind is north, it is, in the case supposed, travelling south at the rate of 20 miles an hour around the storm, while the progressive movement of the storm itself is north at the rate of 20 miles an hour. One motion exactly cancels the other, and there is, therefore, a line of calm and light, or moderate or not so heavy winds on one side of the centre, while on the other side there is a line of maximum violence; in other words, in every travelling cyclone the wind blows harder on one side than the other. This is the case in both hemispheres; and by handling these moving diagrams for illustration, the navigator will soon become familiar with the various problems for determining theoretically the direction of the vortex, the course it is travelling, its distance, etc. Therefore, when it is optional with the navigator to pass the storm on either side, he should avoid the heavy side. These remarks apply to both hemispheres.

793. The rainy quadrant of a cyclone. Captain Toynbee asks if it rains more in one quarter of a cyclone than another? In cyclones that travel fast, I suppose there would be most rain in the after quarter; with those that have little or no progressive motion, I conjecture that the rainy quarter, if there be one, would depend upon the quarter whence that wind comes that brings most rain. The rain in a "cyclone is supposed to come from the moisture of that air which has blown its round and gone up in the vortex; then it expands, grows cool, and condenses its vapour, which spreads out at top like a great mushroom in the air, the liberated heat adding fury to the storm.

794. Erroneous theories.—Such, briefly stated, are the two theories. They appear to me, from such observation and study as I have been able to bestow, to be neither of them wholly right or altogether wrong. Both are instructive, and the suggestions of one will, in many instances, throw light upon the facts of the other. That rotary storms do frequently occur at sea we know, for vessels have sometimes, while scudding before the wind in them, sailed round and round. The United States brig "Perry" did this a few years ago in the West Indies; and so did the "Charles Heddle" in the East Indies: she went round and round a cyclone five times.

795. The wind in a true cyclone blows in spirals.—From such observations as I have been able to obtain upon the subject, I am induced to believe, with Thorn, that the wind in a cyclone does not blow round in a circle, but around in spirals. Nay, I go farther, and conjecture, that it is only within a certain distance of the vortex that the wind gyrates, and that the gyrating column is never hundreds of miles in diameter, as the advocates of this theory make it: I shall allude to this again. The low barometer at the centre is owing, in part, to two causes; one is the condensation of vapour, with its liberated heat, as maintained by Espy; the other is the action of a real centrifugal force, which applies to all revolving bodies. In weighing the effect of this centrifugal force upon the low barometer, care should be taken not to give it an undue weight. It is not sufficient to cause the air to fly off in a tangent. The lateral atmospheric pressure would prevent that, if the centrifugal force were never so great; and the lower the barometer in the centre, the greater would be the pressure of the surrounding air. The proper weight, therefore, due to the centrifugal force I hold to be not very great, though it is appreciable to this extent: The storm having commenced revolving, the flow of air into the vortex is retarded, not prevented by centrifugal tendency and this retardation assists in causing the barometer to stand lower than it would if there were no revolution. Any one who has watched the little whirl-winds so often seen during summer and fall, or who can call to mind the whirls or "sucks" in a mill pond, or at the lock in a canal when the water is drawn off at the bottom, may appreciate the extent to which the centrifugal tendency will help to make a low barometer at the centre of a cyclone.

796. An illustration.—The low barometer, the revolving storm, and the ascending column require for a postulate the approach by spirals of the wind from circumference to centre. The wind (§665) blows towards the place of low barometer; that is admitted by all. It can only reach that place by a direct or by a curvilinear motion. If by the former, then there can be no revolution; but if there be revolution, then the air, while as wind it is revolving around the centre in the gyrations of the storm, is approaching the centre also. Hence we derive the elements of a spiral curve; and the physical necessity for spiral motion is demonstrated from the fact that there is circular motion and an uprising in the centre. This spiral movement and the uprising may be illustrated by familiar examples : The angles and corners of the Observatory, and its wings, are so arranged that at a certain place there is, with westerly winds, always a whirlwind. This whirl wind is six to eight feet in diameter; and when there is snow, there is a pile, of it in the centre, with a naked path, in the shape of a ring, three or four feet broad about it. It is the spiral motion which brings the drift-snow to the centre or vortex, and the upward motion not being strong enough to carry the snow up, it is left behind, forming a sort of cone, which serves as a cast for the base of the vortex. If you throw chips or trashy matter into the lock of a canal and watch them, you will see that as they come within the influence of the "suck" they will approach the whirl by a spiral until they reach the centre, when, notwithstanding they may be lighter than the water, they will be "sucked" down. Here we see the effects of centrifugal force upon a fluid revolving within itself. The "suck" is funnel-shaped. As it goes down, the lateral pressure of the water increases; it counteracts more and more the effect of centrifugal force, and diminishes, by its increase, the size of the "suck."

797. Dust whirlwinds and water-spouts.—So, too, with the little autumnal whirlwinds in the road and on the lawn: the dust, leaves, and trash will be swept in towards the centre at the bottom, whirl round and round, go up in the middle, and be scattered or spread out at the top. I recollect seeing one of these whirl-winds pass across the Potomac, raising from the river a regular water-spout, and, when it reached the land, it appeared as a common whirlwind, its course being marked, as usual, by a whirling column of leaves and dust. These little whirlwinds are, I take it, the great storms of the sea in miniature; and a proper study of the miniature on land may give us an idea of the great original on the ocean.

798. A vera causa.—The unequally heated plain is thought to be the cause of the one. But there are no unequally heated plains at sea; nevertheless, the primum mobile there is said, and rightly said, to be heat. Electricity, or some other imponderable, may be concerned in the birth of the whirlwind both ashore and afloat. But that is conjecture; the presence of heat is a fact. In the middle of the cyclone there is generally rain, or hail, or snow; and the amount of heat set free during the process of condensing the vapour for this rain, or hail, or snow, is sufficient to raise from the freezing to the boiling point more than five times the whole amount of water that falls. This vast amount of heat is set free, not at the surface of the sea, it is true, but in the cloud-region, and where the upward tendency of the indraught is still farther promoted. What sets the whirlwind a-brewing is another question; but its elements being put in motion, there is a diminished barometric pressure, first, on account of centrifugal tendency; next, on account of the ascending column of air, which expands and ascends,—ascends and expands on account of such diminished pressure;—and next, though not least, on account of the heat which is set free by the condensation of the vapour which forms the clouds and makes the rain. This heat expands and pushes aside the upper air still more.

799. Objections to the theory.—After much study, I find some difficulties about the cyclone theory that I cannot overcome. They are of this sort : I cannot conceive it possible to have a cyclone with a revolving and travelling disc 1000, or 500 miles in diameter, as the expounders of the theory have it. Is it possible for a disc of such an attenuated fluid as common air, having 1000 miles of diameter with its less than wafer-like thickness in comparison, to go travelling over the earth's surface and revolving about a centre with tornado violence? With the log-books of several vessels before me that are supposed to have been in different parts of the same cyclone I have a number of times attempted to project its path, but I always failed to bring out exactly such a storm as the theory calls for.^[3] I make a distinction between the hauling of the wind in consequence of diurnal rotation of the earth, and the rotation of the wind in the cyclone in consequence of its centripetal force. For the sake of illustrating my difficulties a little farther, let us suppose a low barometer with a revolving storm to occur at A in the southern hemisphere. Let the storm be travelling towards B. Let observers be at c'', d and e, and let c'' and d be each several hundred miles from A, and so far as to be clearly without the reach of the whirl. Now, then, will not the air at c'' and d blow north and east as directly for the place of low barometer as it would were that place an oblong, N A, instead of a disc, as per the arrows? And w4iy should it be a disc in preference to an ellipse a square, an oblong, or any figure, be it never so irregular?

The trade-winds answer, showing the equatorial calm belt—an oblong—as their place of meeting. They neither revolve nor blow at right angles, with the line of their direction from the place of low barometer; but they blow as directly for it as the forces of diurnal rotation will allow. But the cyclonologists, instead of permitting the wind at the distance c'' sometimes to blow to the east, and at d to blow to the north, merely because there is a low barometer east of c'' and north of d, require it always so to blow because, by their theory, there is a low barometer east of d and south of c''! Thus, to reach its theoretical place of destination, the wind must blow in a direction at right angles to that destination! It would require a rush of inconceivable rapidity so to deflect currents of air while they are yet several hundred miles from the centre of gyration. Moreover, the two cyclonologists, c'' and d, would differ with each other as to the centre of the storm. Each at first would assume that the wind was blowing about him in the direction of the curved arrow c'' and d. As the place of low barometer travels towards B, the hauling of the wind would be according to the theory at c'', but against it at d. The cyclonologist in d would place the centre of the storm to the eastward of him, as in the direction of d, but the other would place it to the southward of him, as in the direction c' . By the rule, ship d would be led towards the real track of the storm, i. e.,

into danger, and ship c'' away from it.^[4] From all this it would appear that the cyclone theory is defective in this: when the wind hauls in the storm, the sailor who is contending with it, lacks a rule by which he may know the influence which causes it to haul. The gyrating disc of a cyclone can never, I apprehend, exceed a few miles in diameter. On shore we seldom find it exceeding in breadth as many rods, in most cases not of as many fathoms, as its advocates give it miles at sea. I think the dust-whirl in the street is a true type of the tornado (cyclone) at sea.

800. The three forces.—There are in the various parts of the storm at least three forces at work in effecting a change of wind, as observed on board ship at sea. (1.) One is diurnal rotation: it alone can never work a change of direction exceeding 90°; (2.) another is the varying position or travelling motion of the place of barometric depression; the change effected by it cannot exceed 180°; (3.) and the third is the whirling motion imparted by the rush to a common centre—as the whirl of water at the flood-gate of the mill, the whirlwind in the street, for example.

801. The effect of each.—Hence it appears that in a storm the wind may shift from any one of three causes, and we are not entitled to call it a cyclone unless the wind shift more than 180°. If the change of direction be less than 90°, the shifting may be due to diurnal rotation alone; if it be less than 180°, the shifting may be, and is probably, due to (1) and (2). The sailor has therefore no proof to show that he has been in a cyclone unless the wind during the storm changed its directions more than 180°. Cyclones, there is reason to believe, are often whirlwinds in a storm. This may be illustrated by referring again to our miniature whirlwinds on the land; there we often see a number of them at one time and about the same place; and they often appear to skip, raging here, then disappearing for a moment, then touching the ground again, and pursuing the former direction.

802. A storm within a storm.—Observations have proved that this is the case on land, and observations have not established that this is not the case at sea; observations are wanting upon this subject. Tornadoes on the land often divide themselves, sending out branches, as it were. It remains to be seen whether cyclones do not do the same at sea; and whether, in those widespread and devastating storms that now and then sweep over the ocean, there be only one vortex or several; and if only one, whether the whole storm partake of the cyclone character. In other words, may there not be a storm within a storm—that is, a cyclone travelling with the storm and revolving in it? I ask the question because the theory, as at present expounded, does not satisfy all the facts observed; and because the existence of storms or whirlwinds within a storm would.

803. The Black Sea storm of 1854.—The celebrated Black Sea storm of 1854, which did so much damage to the allied fleet, is still maintained by some to be a true cyclone; and by the observations of some of the vessels a cyclone may be made out. But if we take the observations of all of them, and discuss them upon the supposition that the whole storm was a cyclone, it will puzzle any one to make anything of them. Admiral Fitzroy, in the Meteorological Papers of the Board of Trade, published diagrams of the winds as observed during that storm on -board of various vessels in various parts of that sea. I have not been able to reconcile them with the cyclone theory. Espy maintains that they confirm his theory; and his (§ 787) is anti-cyclonic.

804. Cyclones of the North Atlantic.—The cyclones of the North Atlantic take their rise generally (§ 785) somewhere between the parallels of 10° and 20° north. They take a westerly course until they fall in with the Gulf Stream, when they turn about and run along upon it until their force is expended. The atmosphere over the Gulf Stream is generally well charged with moisture, and in this fact perhaps will be found the reason why (§ 176) the path of the storm is laid along the Gulf Stream.

805. The hurricane season.—The following table is from Birt's Handbook of Storms:

Average. Number of Cyclones or Hurricanes which have occurred in different Months of the Year, and in various Regions.

Locality.	Jan.	Feb.	Mar.	April	May	June	July	Aug.	Sept.	Oct.	Nov.	Dec.	Total
West Indies	..	1	2	..	..	4	15	36	25	27	1	2	113
S. Indian Ocean	9	13	10	8	4	..	..	..	1	1	4	3	53
Mauritius	9	15	15	8	..	..	..	..	..	..	..	6	53
Bay of Bengal	1	..	1	1	7	3	..	1	..	7	6	3	30
China Sea	..	..	..	..	..	2	5	5	18	10	6	..	46

803. Cyclones in the Mississippi Valley.—The vortex of a cyclone is often and aptly compared to a meteor. I have often observed the paths of such through the forests of the Mississippi Valley, and the path of one of these "whirlwinds" as they are there called has in no instance that has fallen under my observation been more than a few hundred yards broad. There the track of these tornadoes is called a "wind-road," because they make an avenue through the wood straight along, and as clear of trees as if the old denizens of the forest had been felled with an axe. I have seen trees three or four feet in diameter torn up by the roots, and the top, with its limbs, lying; next the hole whence the roots came. Nevertheless, the passage of the meteor, whose narrow path was marked by devastation, would create a great commotion in the air, and there would be high winds raging for several miles on either side of the "wind-road." But (§ 799) let US consider for a moment the effect of the diurnal rotation of the earth upon one of these revolving discs 1000 miles in diameter: its height would scarcely be two miles, and its thickness would not be as great, in proportion to its diameter, as half the thickness of this leaf is to the length of an inch. Now the difference in rate of the diurnal rotation between the northern and southern limbs of the disc would be sufficient, irrespective of any other power, to break it up. Suppose its southern limb to be in 20° N., its northern limb would be 1000 miles, say 17°, farther north, that is, in 37°. Diurnal rotation would carry to the east the air in the southern limb at the rate of 845 miles an hour; but when this same air comes round on the northern limb, diurnal rotation would carry it eastward at the rate only of 720 miles. Because the wind hauls in a particular way, it does not follow, as by diagram (§ 799) it has been shown, that it is blowing in a circle, or that the centre of the storm is at right angles to its line of direction.

807. Extra-tropical gales.—In the extra-tropical regions of each hemisphere furious gales of wind also occur. One of these, remarkable for its violent effects, was encountered on the 24th of December, 1853, about three hundred miles from Sandy Hook, latitude 39° north, longitude 70° west, by the "San Francisco," steam-ship. That ship was made a complete wreck in a few moments, and she was abandoned by the survivors, after incredible hardships, exertions, and sufferings. Some months after this disaster I received by the California mail the abstract log of the fine clipper ship "Eagle Wing" (Ebenezer H. Linnell), from Boston to San Francisco. She encountered the ill-fated steamer's gale, and thus describes it: "December 24th. Latitude 39° 15' north, longitude 62° 32' west. First part threatening weather; shortened sail; at 4 p.m. close-reefed the top-sails and furled the courses. At 8 p.m. took in fore and mizen top-sails; hove to under close-reefed main top-sail and spencer, the ship lying with her lee rail under water, nearly on her beam-ends. At 1 30 a.m. the fore and main top-gallant-masts went over the side, it blowing a perfect hurricane. At 8 a.m., moderated; a sea took away jib-boom and bowsprit cap. In my thirty-one years' experience at sea, I have never seen a typhoon or hurricane so severe. Lost two men overboard—saved one. Stove skylight, broke my barometer, etc., etc." Severe gales in this part of the Atlantic—i. e., on the polar side of the calm belt of Cancer—rarely occur during the months of June, July, August, and September. This appears to be the time when the fiends of the storm are most busily at work in the "West Indies. During the remainder of the year, these extra-tropical gales, for the most part, come from the north-west. But the winter is the most famous season for these gales. That is the time when the Gulf Stream has brought the heat of summer and placed it (§ 172) in closest proximity to the extremest cold of the north. And there should, therefore, it would seem, be a conflict between these extremes; consequently, great disturbances in the air, and a violent rush from the cold to the warm. In like manner, the gales that most prevail in the extra-tropics of the southern hemisphere come from the pole and the west, i. e., south-west.

808. Storm and Rain alerts.—Storm and Rain Charts for the Atlantic Ocean have already been published by the Observatory, and others for the other oceans are in process of construction. The object of such charts is to show the directions and relative frequency of calms, fogs, rain, thunder, and lightning. These charts are very instructive. They show that that half of the atmospherical coating of the earth which covers the northern hemisphere—if we may take as a type of the whole what occurs on either side of the equator in the Atlantic Ocean—is in a much less stable condition than that which covers the southern.

1. The Fourth Meteorological Report of Prof. James P. Espy; Senate Doc. 65, 34th Congress, 3rd session.
2. See article Meteorology, Encyclopædia Britannica, 1857.
3. Since this was written, I have had the privilege of examining at the Meteorological Department of the Board of Trade, Admiral Fitzroy's admirable diagrams, in MS., of the "Royal Charter" storm of October, 1859. It was a true cyclone—the best type of one, and the most perfectly developed on a large scale that has ever fallen under my observation. Its largest diameter did not measure less than 300 miles.
4. Sec letter to Commodore Wullerstorf, p. 457, vol. ii., 8th ed., Maury's Sailing Directions.