Chapter IX (1)

Before proceeding to do this, however, it may be well to look at some theories which have been advanced, and to a greater or less extent adopted, and at their bearing upon the question.

The calorific theory is at present the prevailing one in Europe and in this country. Meteorologists there and here refer all atmospheric conditions and phenomena to the influence of heat. The principal applications of that theory have been considered. But within the last few years the elasticity and tension of the aqueous vapor of the atmosphere have received much attention, as exerting an auxiliary or modifying influence. Professor Dove, of Berlin, who ranks perhaps as the most distinguished meteorologist of that continent, attributes barometric variations to _lateral overflows_, and, in the upper regions, resulting from the elevation of the atmosphere by expansion; and in this view meteorologists of Europe seem generally to acquiesce. In an article sent to Colonel Sabine, and recently republished in the American Journal of Science, January, 1855, in thus attempting to account for the annual variation of barometric pressure, which occurs in Europe and Asia, and, indeed, over the entire hemisphere. He says:

"From the combined action or the variations of aqueous vapor, and of
the dry air, we derive immediately the periodical variations of the
whole atmospheric pressure. As the dry air and the aqueous vapor
mixed with it, press in common on the barometer, so that the up-borne
column of mercury consists of two parts, one borne by the dry air,
the other by the aqueous vapor, we may well understand that as with
increasing temperature the air expands, and by reason of its
augmented volume rises higher, and _its upper portion overflows
laterally_," etc.

And in another place he says:

"From the magnitude of the variations in the northern hemisphere, and
the extent of the region over which it prevails, we must infer that
_at the time of diminished pressure a lateral overflow probably takes
place_," etc.

Doubtless, the mean pressure of the atmosphere, in summer, in the northern hemisphere, is less than in winter, in some localities, and greater in others, and it differs in different countries of equal temperature. And this is all very intelligible. The mean of the pressure for the month is made up by _averaging_ all the _elevations_ and _depressions_. During a month, showing a very low mean, the barometer may, at times, attain its _highest altitude_, if the depressions below the mean are great or more frequent. The barometer is depressed during storms, and ranges high during _set fair_ weather. Ordinarily, therefore, the more stormy the season the more diminished the mean pressure; and it is a mistake to look to an overflow to account for the fact. The changes in the location of the atmospheric machinery, and consequent change in the amount and severity of falling weather, and the periodic frequency and character of storms, and consequent _periodic_ depressions and elevations of the barometer, explain the annual mean variations, as they do the other phenomena. But it is perfectly consistent with the calorific theory to attempt to account for these differences by another of those ever-necessary modifications, viz.: the different tension and elasticity of aqueous vapor in different countries of equal temperature; and then to _suppose_ an expansion of the whole body of the atmosphere and a lateral overflow from the place where the air is expanded, on to some other, where it is not; and thus _suppose_ all necessary currents in the upper regions, setting hither and yon, by the force of gravity alone. And apparently he who is best at supposition becomes the most distinguished meteorologist. Perhaps I have already said all that I ought to be pardoned for saying, in relation to the utter absurdity of attributing all meteorological phenomena to the agency of heat; but when I find such views as those which that article contains, emanating from so distinguished a man, sanctioned by the President of the British Association, and copied into the leading journal of science in this country, I can not forbear a further and a somewhat critical examination of them. There is more error of supposition and less truth in it, than in any other article regarding the science, of equal length, which has fallen under my notice.

What is the height of this expansion? The moisture of evaporation ascends, ordinarily, but a few thousand feet. The atmosphere grows regularly cooler, from the earth to the trade, and _the increased warmth that is felt at the surface extends but little way_. Currents of warm air do not ascend. The strata maintain, substantially, their relative positions; and this is a most beneficent provision. In northern latitudes of the temperate zone, all the warmth derived from a few hours' sunshine is needed at the surface; and, deplorable, indeed, would be our condition, if the atmosphere, as fast as warmed by the rays of the sun, were to hasten up, and the frigid strata descend in its place. The earth would not be habitable. All the warm air on its surface would be rising as soon as it became warmed, and the cold air above be descending, and enveloping us with the chilling strata which are ever floating within two or three miles above us. No. Infinite wisdom has ordered it otherwise. The laws of magnetism and of static-electric induction and attraction keep the strata in their places, and preserve to us the warmth which the solar rays afford or produce. The inhabitant of the valley, in a high northern latitude, in summer, can plant, and sow, and reap, at the base of the mountain whose summit penetrates the stratum of continual congelation, and up its sides, almost to the line of perpetual snow; and, as he looks upon the fruits of his labor, and up to the snow-clad peak that towers above him, can thank his Maker for placing a warm equatorial current, a perpetual barrier, between the fertility and warmth which surround him, and the cold destructive strata above; and thank Him for not creating such a state of things, as certain meteorologists insist we shall believe He has created. Again, where are the _upper regions_, from which the lateral overflow takes place? The atmosphere is differently estimated, at from thirty to forty-five miles, or more, in height. Whatever its height may be, it is exceedingly attenuated in its "upper regions."

Gay-Lussac marked the barometer at 12-95/100 inches at the height of 23,040 feet. Two thirds of the atmospheric density, then, is within five miles of the earth. Air, too, is _compressible_. Allowing for the latter and the attenuation, how many miles in vertical depth, of its "_upper regions_," must move from one portion to another, to depress the barometer two inches--its range sometimes in twenty-four hours--or even half an inch? Let the computation be made, and see how startling the proposition, how utterly impossible that the theory can be true.

The distinguished Professor, in the paper referred to, introduces his theory of the formation of hurricanes, and we quote--

"If we suppose the upper portions of the air ascending over Asia and
Africa to flow off laterally, and if this takes place suddenly, it
will check the course of the upper or counter-current above the
trade-wind, and force it to break into the lower current.

"An east wind coming into a S. W. current must necessarily occasion a
rotatory movement, turning in the opposite direction to the hands of
a watch. A rotatory storm, moving from S. E. to N. W., in the lower
current or trade, would, in this view, be the result of the encounter
of two masses of air, impelled toward each other at many places in
succession, the further cause of the rotation (originating primarily
in this manner) being that described by me in detail in a memoir 'On
the Law of Storms,' translated in the 'Scientific Memoirs,' vol. iii.
art. 7. Thus, it happens that the West India hurricanes, and the
Chinese typhoons occur near the lateral confines on either side of
the great region of atmospheric expansion, the typhoons being
probably occasioned by the direct pressure of the air from the region
of the trade-winds over the Pacific, into the more expanded air of
the monsoon region, and being distinct from the storms appropriately
called by the Portuguese 'temporales,' which accompany the out-burst
of the monsoon when the direction of the wind is reversed."

The analogy between this, and a theory of Mr. Redfield's, will be noticed further on. But I remark, in passing, that there is not a fact or inference in this paragraph which will bear examination.

1. There is no such regular S. W. wind over the surface trade, as he supposes. Doubtless, there are, occasionally, secondary S. W. currents between the counter-trade and the surface one, with partial condensation, for much of both becomes depolarized by their reciprocal action and precipitation, and these induced S. W. currents are sometimes so strong as to usurp the place of the surface-trade, and become very violent in the latter part of hurricanes; but such is not the usual course of the upper currents of the West Indies, as the progress of storms there, and observation, prove.

2. There can not be any _periods_ of extensive and _sudden_ expansion over Africa. If there is any place on the earth which has a more uniformly progressive temperature, either way, and is more free from _sudden_ extremes, or which is more arid and destitute of aqueous vapor, and sudden aqueous expansions, than another, it is Africa. No such occasional sudden expansions are there possible.

3. Winds do not, and can not, "_encounter_." They stratify upon each other. They are produced by the action of opposite electricity, and are _connected together_ in their origin and action. The atmosphere is never free from the regular and irregular currents, however invisible for the want of condensation. Aeronauts find them in the most serene days. They exist without encounter or tendency to rotation, every where, and at all times; even over the head of the distinguished Professor, whether he sleeps or is awake. We can all see them when there is condensation, and it is rarely the case that there is not some degree of it in some of them.

4. That "Great region of expansion" is a chimera. It does not exist. It is a region of _lower temperature_, and of _condensation_, instead of _expansion_ of _aqueous vapor_. The trade does not rise in it, or the S. W. wind overflow from it. See the table cited page 165.

5. The hurricanes do not originate _in the surface trades_, as he supposes. They originate in the belt of rains, the supposed "region of expansion," and issue out of it; or in the counter-trade, where volcanic elevations rise far into or above the surface trade.

6. This hypothesis can not be sustained upon his own principles. The distance between Africa and the West India Islands, where most of the hurricanes originate, is from 2,500 to 3,000 miles. These gales are small when they commence, not ordinarily over one or two hundred miles in diameter, and often less. There are trades all the way over from Africa, and S. W. winds also, if they exist, as he supposes, in the West Indies. How can it happen that this lateral overflow should pass _without effect_, over 2,500 miles of S. W. wind and trade, and concentrating the overflow of a continent over one small and chosen spot of the West Indies, _pitch down_ there, and there only, and crowd the S. W. wind into the trade below? This is too much for sensible men to believe.

What does Professor Dove mean by the term _impulsion_, as applied to the winds? How are they _impelled_? It is the fundamental idea of his calorific theory, that they are _drawn_ by the _suction_ caused by a _vacuum_, and the vacuum created by expansion and overflow above, in obedience to the law of gravity; that the S. E. trade is drawn to the great region of expansion, and the S. W. runs from it as an overflow. But if the S. W. is driven down into the plane and place of the surface-trades, how does it continue to be impelled, and why is it not then subject to the suction of the vacuum which draws the trade? Does that vacuum _select its air_, and so attract the trade, in preference to the depressed portion of the S. W. current, that the former runs around the latter to get to the vacuum, and the latter around the former to get away from it? And does the trade, when it has got around the S. W. current, instead of going to the vacuum, continue to gyrate, and the S. W. current, instead of pursuing its regular course, gyrate also about the trade, and both move off together, regardless of the vacuum of the great region of expansion, in a new direction to the N. W., in an independent, self-sustaining, cyclonic movement, increasing in power and extent, involving extended and increasing condensation, producing the most violent electrical phenomena, and thus continuing up, even to the Arctic circle? Yes, says Professor Dove. No, say all fact, all analogy, and his own principles.

7. His theory relative to the typhoons is unintelligible. If they originate near the lateral confines of the great region of atmospheric expansion, they originate in the region of the trade-winds, for the two are identical. How the direct pressure of the air from the trade-wind over the Pacific, in the more expanded air of the monsoon region, can occasion a typhoon upon any principles, passes my comprehension. If, as Lieutenant Maury supposes, the monsoons are reversed trades, then the trade-wind and monsoon region are identical. If the monsoons are found in the belt of rains, then, the trades, upon Professor Dove's principles, pass into the monsoon region by attraction or suction, without pressure. Either way the theory is undeserving of consideration.

A new theory has recently been started by Mr. Thomas Dobson, and, although it is (like all other efforts to get the _upper strata down_ to produce condensation, or those below _up_, that they may be condensed), without foundation, his collection of facts is brief and interesting. I copy his article from the London, Edinburgh, and Dublin Phil. Mag., for December, 1853. It adds to the collection of facts in relation to the connection between volcanic action and storms for the seventeenth century, made by Dr. Webster:

The following appear to be the main facts which are available as a
basis for a theory which shall comprehend all the meteors in
question:

1st. The eruption of a submarine volcano has produced water-spouts.

"During these bursts the most vivid flashes of lightning continually
issued from the densest part of the volcano, and the volumes of smoke
rolled off in large masses of fleecy clouds, gradually expanding
themselves before the wind in a direction nearly horizontal, and
drawing up _a quantity of water-spouts_."--(Captain Tilland's
description of the upheaval of Sabrina Island in June, 1811, Phil.
Trans.)

With this significant fact may be compared the following analogous
ones:

"In the Aleutian Archipelago a new island was formed in 1795. It was
first observed _after a storm_, at a point in the sea from which a
column of smoke had been seen to rise."--(Lyell, Principles of
Geology.)

"Among the Aleutian Islands a new volcanic island appeared in the
midst of _a storm_, attended with flames and smoke. After the sea was
calm, a boat was sent from Unalaska with twenty Russian hunters, who
landed on this island on June 1st, 1814."--(Journal of Science, vol.
vii.)

"On July 24th, 1848, a submarine eruption broke out between the
mainland of Orkney and the island of Strousa. Amid thunder and
lightning, a very dense jet black cloud was seen to rise from the
sea, at a distance of five or six miles, which _traveled toward the
north-east_. On passing over Strousa, the wind from a slight air
became _a hurricane_, and a thick, well-defined belt of large
hailstones was left on the island. The barometer fell two
inches."--(Transactions Royal Society, Edinburg, vol. ix.)

2d. Hurricanes, whirlwinds, and hailstones accompany the paroxysms of
volcanos.

"1730. A great volcanic eruption at Lancerote Island, and _a storm_,
which was equally new and terrifying to the inhabitants, as they had
never known one in the country before."--(Lyell, Principles of
Geology, vol. ii.)

"1754. In the Philippine Islands a terrible volcanic eruption
destroyed the town of Taal and several villages. Darkness,
hurricanes, thunder, lightning, and earthquakes, alternated in
frightful succession."--(Edinburgh Philosophical Journal.)

"In 1805, 1811, 1813, and 1830, during eruptions of Etna, caravans in
the deserts of Africa perished by violent whirlwinds. In 1807, while
Vesuvius was in eruption, a whirlwind destroyed a caravan."--(Rev. W.
B. Clarke in Tasw. Journal.)

"1815, Java. A tremendous eruption of Tombow Mountain. Between nine
and ten P.M., ashes began to fall, and soon after _a violent
whirlwind_ took up into the air the largest trees, men, horses,
cattle, etc."--(Raffles' History of Java.)

"1817, Dec. Vesuvius in eruption. In the evening _a hail storm_,
accompanied with red sand."--(Journal of Science, vol. v.)

"1820, Banda. A frightful volcanic eruption, and in the evening an
earthquake and a violent hurricane."--(Annales de Chimie.)

"1822, Oct. Eruption of Vesuvius. Toward its close the volcanic
thunder-storm produced an exceedingly violent and abundant fall of
rain."--(Humboldt, Aspects of Nature.)

"1843, Jan. Etna in eruption. Violent hurricanes at Genoa, in the Bay
of Biscay, and in Great Britain.

"1843, Feb. Destructive earthquakes in the West Indies, a volcanic
eruption at Guadaloupe, followed by hurricanes in the Atlantic."

"1846, June 26. Volcano of White Island, New Zealand, in eruption.
Heavy squalls of wind and hail; it blew as hard as in a
typhoon."--(Commodore Hayes, R.N., in Naut. Mag., 1847.)

"1847, March 20. Volcanic eruption and earthquake in Java; and on the
21st of March, and 3d of April, violent hurricanes."--(Java Courant.)

"1851, Aug. 5. A frightful eruption of the long dormant volcano of
the Pelee Mountain, Martinique. Aug. 17. Hurricane at St. Thomas,
etc.; earthquake at Jamaica, etc.

"1852, April 14. Earthquake at Hawaii, and on the 15th a great
volcanic eruption. On the 18th _a gale of unusual violence_ lasted
thirty-six hours, and did great damage."--(The Polynesian, April 22,
1852.)

3d. In volcanic regions, earthquakes and hurricanes often occur
almost simultaneously, but in no certain order, and without any
volcanic eruption being observed.

In 1712, 1722, 1815, and 1851, earthquakes and hurricanes occurred
together at Jamaica; in 1762 at Carthagena; in 1780 at Barbadoes; in
1811 at Charleston; in 1847 at Tobago; in 1837 and 1848 at Antigua;
in 1819, an awful storm at Montreal, rain of a dark inky color, and a
slight earthquake. People conjectured that a volcano had broken out.
In 1766 the great Martinique hurricane, a _waterspout_ burst on Mount
Pelee and overwhelmed the place. Same night, an earthquake.

1843, Oct. 30. Manilla.--Twenty four hours' rain and two heavy
earthquakes. 10 P.M., a severe hurricane.

"1852, Sept. 16. Manilla--An earthquake destroyed a great part of the
city; many vessels wrecked by a great hurricane in the adjacent seas,
between the 18th and 26th of September."--(Singapore Times.)

"1731, Oct. Calcutta.--Furious hurricane and violent earthquake;
300,000 lives lost."

"1618, May 26. Bombay.--Hurricane and earthquakes; 2,000 lives
lost."--(Madras Lit. Tran., 1837.)

"1800. Ongole, India, and in 1815, at Ceylon, a hurricane and
earthquake shocks."--(Piddington.)

"1348. Cyprus.--An earthquake and a frightful hurricane."--(Hecker.)

"1819. Bagdad.--An earthquake and _a storm_--an event quite
unprecedented.

"1820, Dec. Zante.--Great earthquake and hurricane, with
manifestations of a submarine eruption."--(Edinburg Phil. Journal.)

"1831, Dec. Navigator's Islands.--Hurricane and
earthquakes."--(Williams' Missionary Enterprise.)

"1848, Oct., Nov. New Zealand.--Succession of earthquake shocks, and
several tempests.

"1836, Oct. At Valparaiso, a destructive tempest and severe
earthquakes."--(Nautical Magazine, 1848.)

When an earthquake of excessive intensity occurs, as at Lisbon, in
1755, the volcanic craters, which act as the safety-valves of the
regions in which they are placed, are supposed to be sealed up; and
it is a remarkable and highly-suggestive fact, that _no hurricane
follows such an earthquake_. The number of instances of the
concurrence of ordinary earthquakes and hurricanes might easily be
increased, but the preceding suffice to show the _generality_ of
their coincidence, both as _to time_ and place.

4th. The breaking of water-spouts on mountains sometimes accompanies
hurricanes.

In 1766, during the great Martinique hurricane, before cited.

"1826, Nov. At Teneriffe, enormous and most destructive water-spouts
fell on the culminating tops of the mountains, and a furious cyclone
raged around the island. The same occurred in 1812 and in
1837."--(Espy and Grey's Western Australia.)

"1829. Moray.--Floods and earthquakes, preceded by water-spouts and a
tremendous storm."--(Sir T. D. Lander.)

"1826, June. Hurricanes, accompanied by water-spouts and fall of
avalanches, in the White Mountains."--(Silliman's American Journal,
vol. xv.)

5th. The fall of an avalanche sometimes produces a hurricane.

"1819, Dec. A part (360,000,000 cubic feet) of the glacier fell from
the Weisshorn (9,000 feet). At the instant, when the snow and ice
struck the inferior mass of the glacier, the pastor of the village of
Randa, the sacristan, and some other persons, _observed a light_. A
frightful hurricane immediately succeeded."--(Edinburg Philosophical
Journal, 1820.)

6th. Water-spouts occur frequently near active volcanos.

This is well known with regard to the West Indies and the
Mediterranean. The following notices refer to the Malay Archipelago
and the Sandwich Islands:

"Water-spouts are often seen in the seas and straits adjacent to
Singapore. In Oct., 1841, I saw _six_ in action, attached to one
cloud. In August, 1838, one passed over the harbor and town of
Singapore, dismasting one ship, sinking another, and carrying off the
corner of the roof of a house, in its passage landward."--(Journal of
Indian Archipelago.)

"1809. An immense water-spout broke over the harbor of Honolulu. A
few years before, one broke on the north side of the island (Oahu),
washed away a number of houses, and drowned several
inhabitants."--(Jarves' History of Sandwich Islands.)

7th. Cyclones begin in the immediate neighborhood of active volcanos.

The Mauritius cyclones begin near Java; the West Indian, near the
volcanic series of the Caribbean Islands; those of the Bay of Bengal,
near the volcanic islands, on its eastern shores; the typhoons of the
China Sea, near the Philippine Islands, etc.

8th. Within the tropics, cyclones move toward the west; and, in
middle latitudes, cyclones and water-spouts move toward the N. E., in
the northern hemisphere, and toward the S. E. in the southern
hemisphere.

9th. In the northern hemisphere, cyclones rotate in a horizontal
plane, in the order N. W., S. E.; and in the southern hemisphere, in
the order N. E., S. W.

By applying the principles of electro-dynamics to the electricity of
the atmosphere, I shall endeavor to connect and explain the preceding
well-defined facts. The continuous observations of Quetelet, on the
electricity of the atmosphere, from 1844 to 1849 (Literary Journal,
February, 1850), show that it is always positive, and increases as
the temperature diminishes. It therefore increases rapidly with the
height above the earth's surface. We may, consequently, regard the
upper and colder regions of the atmosphere as an immense reservoir of
electric fluid enveloping the earth, which is insulated by the
intermediate spherical shell formed by the lower and denser
atmosphere. Now, whenever a vertical column of this atmosphere is
suddenly displaced, the surrounding aqueous vapor will be immediately
condensed and aggregated, and the cold rarefied air and moisture will
form a vertical conductor for the descent of the electrical fluid.
This descent will take place down a spiral, gyrating in the order N.
W., S. E., in the northern hemisphere, since the electric current is
under the same influence as that of the south pole of a magnet; and
in the order N. E., S. W., in the southern hemisphere. The air
exterior to the conducting cylinder will partake of the violent
revolving motion, and a tornado or cyclone will be produced.

Upon the foregoing facts I shall comment in another place.

Three theories have been advanced by meteorologists of this country, two of which profess to explain all the phenomena of the weather. Professor Espy attributed the production of storms and rain to an ascending column of air, rarefied by heat, and the rarefaction increased by the latent heat of vapor given out during condensation, and an inward tendency of the air, from all directions, toward the ascending vortex, constituting the prevailing winds. Thus, Professor Espy conceived, and to some extent proved, that the wind blew inward, from all sides, toward the center of a storm, either as a circle, or having a long central line, and he conceived that it ascended in the middle, and spread out above; and that clouds, rain, hail, and snow, were formed by condensation consequent upon the expansion and cooling of the atmosphere, as it attained an increased elevation.

_This ascent_ was not, in fact, _proved_ by Professor Espy, _has not been found by others_, and _is not discoverable, according to my observations_. The theory was ingenious, founded on the theory of Dalton, that the vapor was maintained in the atmosphere by reason of a large quantity of latent heat, which was given out when condensation took place. This theory is also unsound. No such elevation of temperature is found in clouds or fogs when they form near the earth, however dense. Thus the two principal elements of Professor Espy's theory are found to be untrue, and the theory untenable. But it was sustained with great ability and research, and the distinguished theorist deserves much for the discovery and record of important facts in relation to the weather. Aside from its theoretical views, his book contains a great mass of valuable information, and will well repay the cost of purchase and perusal.

Another theory, by Mr. Bassnett, is of recent date, founded on the influence of the moon, and the supposed creation of vortices in the ether above, whose influence extends to the earth, producing storms and other phenomena. No one can peruse his book without conceding to him great ability and scientific attainment; and if his theory was true, the periods of fair and foul weather could be calculated with great mathematical certainty. But it contains inherent and insuperable objections. I will only add that all herein before contained is in direct opposition to it.

Mr. W. C. Redfield, of New York, as early as 1831, first advanced in this country the theory of gyration in storms, and investigated their lines of progress on our coast and continent. His theory is limited in its character, and does not profess, except indirectly, to explain all, or indeed any, of the other phenomena of the weather. As far as it goes, however, it is generally received in this country and Europe, and has been adopted by Reed, Piddington, and others, who have written on the law of storms. The position of Mr. Redfield is honorable to himself and his country. Science and navigation are much indebted to him for his industry in the collection of facts. Nevertheless, his theory is not in accordance with my observation, and I deem it unsound. Although expressed disbelief of the theory has been characterized as an "attack" upon its author, I propose, with that _respect_ which is due to him, but with that _freedom_ and _independence_ which a search for _truth_ warrants, to examine it with some particularity. It is a part of the subject, and I can not avoid it.

When the theory was first announced, I adopted it as probably true; and being then engaged in a different profession, which took me much into the open air by night and day, I watched with renewed care the clouds and currents for evidence to confirm it. I discovered none; on the contrary, I found much, very much, absolutely and utterly inconsistent with its truth. The substance only of these observations will be adduced.

Mr. Redfield admits that the progression of our storms in the vicinity of New York, is from some point between S. S. W. and W. S. W., to some point between N. N. E. and E. N. E. According to my observation, except perhaps in occasional autumnal gales, they are not often, if ever, from S. of S. W., and the great majority of them, including, I believe, all N. E. storms, are between S. W. and W. S. W. Now, the card of Mr. Redfield, moving over any place from any point between S. W. and W. S. W., calls for a S. E. wind at its axis, an E. wind at its north front, and a S. wind at its south front, and does not call _for a N. E. wind on its front at all, except at the north extreme_, where it could _not continue for any considerable period_.

In relation to this, I observe, 1st. _About one-half of our N. E. storms, including some of the most severe ones, not only set in N. E., but continue in that quarter without veering at all, during the entire period that the storm cloud is over us_; usually for twenty-four hours; not unfrequently for forty-eight hours, sometimes for seventy-two or more hours. This every one can observe for himself, and it can not, of course, be reconciled with his theory.

2d. N. E. storms, whether they set in from that quarter in the commencement, or veer to it afterward, when they do "change" round, more frequently veer by the S. to the S. W. in clearing off, than back through the N. into the N. W. The former, in accordance with his theory, they can not do, as the reader can see by passing the left side of the card over his place of residence on the map from S. W. to N. E.

3d. N. E. storms often pass off without hauling by S. or backing by N., and with or without a clearing off shower, the _wind shifting and coming out suddenly at S. W._ This they could not do in accordance with his theory, as slipping the card will show.

4th. From June to February it is _exceedingly uncommon_ for a N. E. storm to back into the N. W. They do so more frequently from February to May, especially about the time of the vernal equinox and after; and then, because the focus of precipitation and storm intensity of the extra tropical zone of rains is S. of 42 deg. east of the Alleghanies. His theory requires them to back by N. into N. W. _in all cases, when they set in N. E._

5th. When they do back from the N. E. into the N. W., it rarely indeed continues to storm after the wind leaves the point of N. E. by N., and generally, if it does continue stormy, _the wind is light_, and not a gale, how violent soever the gale from the eastward may have been. Usually, by the time the wind gets N. W., it has cleared off. This, Mr. Redfield, as we shall see, evades by embracing the N. W. fair wind as a part of the same gale. According to my observation, therefore, a _very large proportion_ of the _N. E. storms_, and they are a majority of the most violent ones of our climate east of the Alleghanies, do not _commence, continue_, or _veer_ in accordance with his theory, but the _reverse_; and so long as this is so, I can not receive his theory as true.

6th. S. E. storms do not always, or indeed often, conform to the requirements of his card. When they set in violently at S. E., and continue so for hours without veering, the axis of the storm should be over us, and the wind should change _suddenly_ to N. W. This did not occur in the storm of Sept. 3, 1821, nor does it often, if ever, occur in the summer or early gales of the autumnal months. In the later storms of autumn, and as often in those which are very gentle as any, and in the winter months when S. E. gales are rare, it does sometimes so change after the storm cloud has passed. But in the winter months, as in the storm investigated by Professor Loomis, the storms are frequently long from S. E. to N. W., and the S. E. wind blows nearly in coincidence with its long axis, for a thousand or fifteen hundred miles, till the barometric minimum is passed, and the inducing and attracting force of this part of the storm cloud is spent, and then the N. W. wind follows; sometimes blowing in under the storm cloud, turning the rain to snow; but oftener following the storm within a few hours, or the next day. The storm of Professor Loomis, when over Texas, was not probably more than four or five hundred miles in length. As it curved more, and passed north and east, it extended laterally, its center traveling with most rapidity, and when it reached the eastern coast was about fifteen hundred miles long, and not more than six hundred broad. Along the eastern part of that storm, except when by its more rapid progress the front projected much further eastward over New England than its previously existing line, the S. E. winds blew. When it bulged out, so to speak, by reason of the increased progress of the center, the wind veered to the N. E. The center of the storm passed near St. Louis and south of Quebec, as the _fall of rain_, the _bulging_ of the _rapidly-moving center_, and the _line of subsequent cold_, attest. It is utterly impossible for any unbiased mind to look at the description of that storm, and attribute to it a rotary character. With all the data before him, Mr. Redfield himself has not attempted it directly.[8]

The September storm of 1821 was more violent in character than any which have since occurred. My recollection of it is as distinct as if it occurred yesterday. Peculiar circumstances, not important in this connection, fixed my attention upon the weather during that day and night. There were cirro-stratus clouds passing all day, from about S. W. to N. E., thickening toward night with fresh S. S. W. wind and flocculent scud, such as I have since seen at the setting-in of S. E. autumnal gales. In the evening the wind (in the immediate neighborhood of Hartford, Ct.), veered to S. E., the cloud floated low, it became very dark, and the wind blew a most violent gale. The trees were falling about the house where I then resided, the windows were burst in, and I was up and observant. When the cloud passed off to the east, it was suddenly light, and almost calm. The western edge of the storm cloud was as perpendicular as a steep mountain side, and was enormously elevated, and very black. I have sometimes seen the western side of a summer thunder cloud, which had drawn a violent gust along beneath it, as elevated and perpendicular, but never a storm cloud. No cloud of that _depth_, or _intensity_ as exhibited by its peculiar blackness, ever floated or will float so near the earth, without inducing a devastating current beneath. After it had passed the ridges east of the Connecticut valley, its top could be seen for a long and unusual period over the elevated ranges.

Now that storm was but an _intense portion_ of an extensive stratus-rain cloud. Such portions frequently exist, and Mr. Redfield admits the fact. Another like portion, in the same storm, passed over Norfolk, Virginia, and the adjacent section, where the wind was N. E., and veered round by N. W. to S. W. Baltimore, and some vessels at sea, were between the two intense portions of the storm, and were not affected by either. Its northern limit was bounded by a line, drawn from some point not far north of Trenton, New Jersey, north-eastward, and north of Worcester, Massachusetts. I was about forty miles south of its northern limit, and north of its center. During that day, and the next, there was wind from S. W. to S. E., inclusive, including the gale, and _from no other quarter_. It did not at any time veer to the W. or N. W. After the passage of the storm-cloud, the wind was very light. When this intense portion of the storm passed over the valley of the Connecticut, its longest axis was from S. S. E. to N. N. W., and the _wind was S. E. the whole length of it_. In its passage from the longitude of Trenton to Boston, there was N. W. wind at one point, and but one, and that was in the iron region, at the N. W. corner of Connecticut, at the northern limit of the intense cloud, and owing, doubtless, to some local cause. The direction of the wind in that storm was in accordance with what is generally true of our storms. The wind on the front of the storm depends upon its shape. If the storm is long in proportion to its width (and no other _violent_ autumnal or winter storm has been investigated, to my knowledge), the wind blows axially, or obliquely, on its front. Thus, if long from S. E. to N. W., the wind on its front will blow from the S. E. So, if the storm is long from S. W. to N. E., and has a south-eastern lateral extension, with an easterly progression, the wind will blow axially in the center, and obliquely at the edges. Instances might be multiplied, but I refer to one of recent date and striking character. All of us remember the drought of 1854. It ended in drenching rain on the 9th of September. This rain fell from a belt, half showery and half stormy in character, which had a S. E. lateral extension.

The evening of the previous day there was some lightning visible at the north, and the usual S. S. W. afternoon wind _continued fresh after nightfall_. The next day we had a brisk wind from the same quarter, and, after noon, the clouds appeared to pile up in the far north, seeming very elevated. They continued to do so, extending southerly during the afternoon, _with a high wind from S. S. W._, the cumulus clouds moving E. N. E. At 5 P.M., gentlemen who left New York at 3 P.M., reported that a dispatch had been received from Albany, dated 1 P.M., stating that it was raining very heavily there. About 7 P.M., the belt reached us, and it rained heavily from that time till morning. Not far from 8 P.M., and during the heaviest rain, the wind shifted from the S. S. W. to N. E., and blew fresh and cold from that quarter during the night, and till the belt had passed south, and then from N. E. by N., cool, with heavy scud, during the forenoon, veering gradually to the N. N. E., and dying away. After the rain ceased, the northern edge of the belt was distinctly visible in the S. and S. E., its stratus-cloud moving E. N. E., and its scud to the westward.

The front of that storm did not pass over us. It was long and narrow. The wind blew somewhat obliquely inward, along its southern border, to the eastward, and, in like manner, to the westward, on its northern border, but from the N. E. axially along its central portions.

In the last instance, the wind changed from S. W. to N. E. This, too, is impossible, according to Mr. Redfield's theory. Similar instances, in summer, and early autumn, are not uncommon. But I shall recur to this in connection with the different _classes_ of storms.

Again, the manner in which these S. E. winds co-exist with the N. E., and become the prevailing wind, toward the close of the storm, is instructive, and inconsistent with the theory of Mr. Redfield. In the West Indies, the first effect of the storm is to increase the N. E. trade; the wind then becomes baffling, but settles in the N. W. or N. N. W., _in direct opposition to the admitted progress of the storm_. At this point, or at S. W., it blows with most force. Sometimes it veers gradually, and sometimes falls calm, and comes out from the S. W., blowing violently. It ends by veering to the S. E., following gently the course of the storm. Thus, Mr. Edwards, in the third volume of his History of Jamaica, as herein before cited, "_all hurricanes begin from the north, veer back to W. N. W., W., and S. S. W., and when they get round to S. E. the foul weather breaks up_."

A short, sudden gale, resembling those of our summer thunder-showers, is sometimes met with from the S. E.; but the violent hurricanes of any considerable continuance are, in almost every case, as just stated.

Now, there is, in our latitudes, an obvious law on the subject, and it is this:--If the storm is not disproportionately long, northerly and southerly, there is a general tendency to induce and attract a surface current, in opposition to the course of the storm on its front, and especially its north front. At the same time, there is a tendency to induce a lateral current on its side, particularly the southerly side, and sometimes its south front: that the latter current is, in the first part of the storm, above the former; in the middle and latter part, it becomes the prevailing current at the surface, and the wind changes accordingly, with or without a calm--that this lateral change sometimes takes place on either side, but usually occurs on the side where the water is warmest, or there is, for other and local reasons, a _greater susceptibility in the atmosphere to inductive and attractive influence_. Thus, our N. E. storms very frequently have a southerly current also, drawn from the ocean, south of us, which forms the middle current, and, in the middle and latter part of it, becomes the prevailing one. _I have seen more than a hundred such instances, clearly and distinctly marked._ Since I have been writing this chapter, January 29th, 1855, such an instance has occurred. On Sunday, the 28th, the cirro-stratus were all day passing from the S. W. to N. E., and gradually thickening with light air from the E. N. E., in the afternoon. During the evening the wind set in _violently_ from the N. E., with a deluging rain. During the night, and after a brief calm, it changed suddenly to the southward, and blew in like manner. This morning the storm was gone, and with it, six inches of hard, frozen icy snow; the trade was clear, with the exception of here and there a broken, melting piece of stratus, but scud were still running from the southward, and the wind has been from the south, veering to S. W., all day, with sunshine. As I have before remarked, this middle current is always present, in this locality, in stratus storms, when there is a heavy fall of rain or snow, although, when the latter happens, the middle current is sometimes from the northward; if it be from the southward, it turns the snow first into very large flakes, and then to rain in our part of the storm.

Doubtless, the same thing occurs every where. In the West Indies, and especially over the Leeward Islands, the middle current is most commonly from the stream of warm water which runs off to the westward into the Caribbean Sea; as the S. W. moonsoon is from the same current below the Cape de Verdes. The S. W. winds, which come from those south polar waters, in the West Indies, appear to be the most violent. But it may be on either or both sides.

The hurricane cloud of the West Indies moves confessedly N. W. in most instances, and undoubtedly it does in all. There is an immutable law that requires it. The seeming exceptions are not such; they are but instances imperfectly investigated. Now, a circular storm moving N. W. can set in N. W. only on the left front, and _can not change to S. W. on that side of the axis_. Nor can the wind blow at the axis from N. W. at all. It should be N. E. in first half, and S. W. in last half. Strange as it may seem, the axis of a West India hurricane in conformity with Mr. Redfield's theory, and a N. W. progression, has never been found, with perhaps a single exception, in any one of which I have seen a description. On the west coast of Europe, the gale is commonly from the Atlantic, either following under the storm from the S. W., or blowing in diagonally from the W. or N. W.; the N. E. wind of western Europe being a cold, dry wind, which there is reason to believe has been around the Siberian pole and is returning, as the cold northerly winds of the North Pacific have around the North American magnetic pole. "If the N. E. winds always prevailed," says Kaemtz, speaking of Berlin, "even at a considerable height it would never rain." This was based on an observation of showers, and not fully reliable. But the dry and cool character of the N. E. wind of western Europe is unquestionable. The S. E. wind is also a storm wind, but owing to the character of the surface from which it is attracted, it is not as violent as the westerly winds are.

Such, too, is the general course and character of the side wind in the southern hemisphere. There gales are less frequent, the magnetic intensity is less, the counter-trades are less; it is not in "the order of Providence" that as much rain shall fall there. Nevertheless, gales occur, although rarely, if ever, with equal violence. About New Holland, where storms are pursuing a S. E. course, they have the wind N. E., corresponding to our S. E., veering from thence, _by the north_, to the westward, clearing off from S. W., with a rising barometer, as ours do from N. W.

In the Bay of Bengal, the Indian Ocean, and the Arabian Sea, there is more irregularity.

But the law of progress and lateral winds can be distinctly traced as _present_ and prevailing, notwithstanding the irregularities. Our limits do not permit an analysis. In the celebrated case of the Charles Heddle, there was much evidence to show that she was driven across the front of the storm by one lateral wind, and back by another. (Diagram of Colonel Reid, p. 206.)

The waters of the Indian Ocean are hot and confined. Storms there are often composed of detached masses, move slower--sometimes not more than three or four miles an hour--and they curve over the ocean, where it is hotter than in any similar latitude. Yet, notwithstanding all peculiarities and irregularities, the law we have been considering is probably the _prevailing_ law there.

No man knows better the existence of these different currents than Mr. Redfield. Doubtless it has escaped his attention that the upper of two, after the passage of a considerable proportion of the storm, becomes the lower, and causes a seeming change of the same wind.

In a series of elaborate articles, substantially reviewing the whole subject, published in the American Journal of Science, for 1846, he says:

"In nearly all great storms which are accompanied with rain, there
appear two distinct classes of clouds, one of which, comprising the
storm scuds in the active portion of the gale, has already been
noticed. Above this is an extended stratum of stratus cloud, which is
found moving with the general or local current of the lower
atmosphere which overlies the storm. It covers not only the area of
rain, but often extends greatly beyond this limit, over a part of the
dry portion of the storm, partly in a broken or detached state. This
stratus cloud is often concealed from view by the nimbus, and scud
clouds in the rainy portion of the storm, but by careful
observations, may be sufficiently noticed to determine the general
uniformity of its specific course, and, approximately, its general
elevation.

"The more usual course of this extended cloud stratum, in the United
States, is from some point in the horizon between S. S. W. and W. S.
W. Its course and velocity do not appear influenced in any
perceptible degree by the activity or direction of the storm-wind
which prevails beneath it. On the posterior or dry side of the gale,
it often disappears before the arrival of the newly condensed cumuli
and cumulo-stratus which not unfrequently float in the colder winds,
on this side of the gale."

"The general height of the great stratus cloud which covers a storm,
in those parts of the United States which are near the Atlantic, can
not differ greatly from one mile; and perhaps is oftener below than
above this elevation. This estimate, which is founded on much
observation and comparison, appears to comprise, at the least, the
limit or thickness of the proper storm-wind, which constitutes the
revolving gale.

"It is not supposed, however, that this disk-like stratum of
revolving wind is of equal height or thickness throughout its extent,
nor that it always reaches near to the main canopy of stratus cloud.
It is probably higher in the more central portions of the gale than
near its borders, in the low latitudes, than in the higher, and may
thin out entirely at the extremes, except in those directions where
it coincides with an ordinary current. Moreover, in large portions of
its area, there may be, and often is, more than one storm-wind
overlying another, and severally pertaining to contiguous storms. In
the present case, we see, from the observations of Professor Snell
and Mr. Herrick, at Amherst, Massachusetts, and at Hamden, Maine (115
and 135 b.), that the true storm wind, at those places, was
super-imposed on another wind; and various facts and observations may
be adduced to show that brisk winds, of great horizontal extent, are
often limited, vertically to a very thin sheet or stratum."

Much of the foregoing is graphically described, and unquestionably true. But it may well be asked how he, or others, distinguish which of two or more currents (for there are frequently three, and sometimes four visible), are the true currents of the storm, and which interlopers from another storm? Is the true one always the upper one, and why? If the upper one, why is the interloper at the surface noted and quoted to prove what a storm is? How does he know what proportions of the winds he has recorded to show the revolving motion of gales, were the true storm winds of the particular storm? or, that every one of them was not an interloping wind on which the true storm wind was superimposed?

These inquiries are pertinent, for obviously, unless some rule for distinguishing between the currents is given, and there be evidence of direct observation to show that the surface wind, whose direction is noted, is the true wind of the storm, and that the _latter_ is not _superimposed_, no reliance can be placed upon logs, or newspaper accounts, or registers. There is another element besides direction, viz.: superimposition, a determination of which _is_ essential to _truth_. It will be difficult for Mr. Redfield to say that a determination of that element has been made, with certainty, in a single storm he has investigated; and in relation to the convergence of storms, and blending, and superimposition of their winds, I think he is mistaken.

Mr. Redfield is right in saying (American Journal of Science, vol. ii., new series, p. 321) that "too much reliance may be placed upon mere observations of the surface winds in meteorological inquiries," and yet _they_ only have thus far been regarded, and he has proved gyration in no other way. I have frequently, with a vane in sight, asked intelligent men how the wind was, and been amused and instructed by their inability to state it correctly. Mr. Redfield, in his inquiries, often found two reports of the weather at the _same time_, from the _same place_, materially different; and I have known, from my own observation, newspapers and meteorological registers to be several points out of the way; and this, because the vanes are influenced by local elevations, and change several points, and very often; because few know the exact points of the compass in their own localities, and because entire accuracy has not been deemed essential. For these reasons, newspaper and telegraphic reports are not always reliable; and therefore, and because, also, storm-winds are easterly and fair winds westerly, and the former veer from east around to west, on one or both sides in many cases, there are few storms which can not be represented as whirlwinds, by a proper _selection_ of _reports_, a corresponding _location_ of the _center_, and an _extension_ of the lines of supposed gyration, so as to include the _preceding_ winds, the actual winds of the storm, and the _lateral_, and _succeeding_ fair weather ones.