EFFECT OF THE STORM-WIND ON THE BAROMETER.-The unfailing mechanical effect of the cyclonic wind in producing a fall of the barometer within the area of its circuit, and greatest in the axial region of the cyclone, is clearly seen in this storm. following are the cases reported in which the barometer fell below 29 inches: to which I have annexed a rough estimate of the probable distance of the vessel from the axis of the storm at the time of nearest approach.

Hermann (5)

Minimum of Barometer.

Supposed distance from gale's axis.

near.

45 miles.

Inches.

27.30

30 66

Independent (49)

Robert Kelly (68)

Southerner (74)

Larpool (85)

Avalanche (90)

Andes (96 a)

From observations of the barometer and winds taken at various points in the United States near the Atlantic coast, and from those made at the signal station in Bermuda, it appears that the storm was but little felt at the latter place, except as exhibiting. the true cyclonic wind, from 6th to 8th, from S. E., veering gradually from S. E. to S. W., as the bearing and progression of the storm became changed; with a force of wind marked from 2 to 4; the barometer at 30 10, at noon of 5th, 6th and 7th and 30-07 at noon of Sth. The left side of the storm encroached to some extent upon the eastern borders of North Carolina and Virginia. At Scuppernong, N. C., lat. 35° 50′, lon. 76° 20', the cyclonic wind blew from N. E., with a force marked 3 and 2, with rain; and the lower stratum of clouds [the true storm-scud or cyclonic stratum] flew rapidly from N. E.; the upper clouds quite still. At 9 P. M., the wind had veered to S. W. No observations of barometer. At Fort Monroe, Va., the reported direction of the storm-wind and cloud, are the same as at Scuppernong, with rain from S. W. at 9 P. M., with thunder and lightning.

At Savannah, on the 5th and 6th, in front of the storm, the maximum of the barometer was 30-20, and 30-19; on the 7th, the minimum of the report is 30-06, and on the 8th, under the rear portion of its annular wave, 30 21. At Jacksonville, East above is apparently required. The normal course of the circulation, or of the “current of rotation" in the basin of the North Atlantic, between latitudes 10° and 50°, as well as the routes taken by some storms which have recurvated in low latitudes, clearly indicate that a part of the out-moving atmosphere, from the inter-tropical latitudes, in the western Atlantic and the Mexican Gulf, moves in an elliptical circuit, and returns to the trade wind in the eastern Atlantic. This apparent tendency in storm-routes is seen in Col. Reid's track of the great hurricane of Barbadoes in 1780, and in my track xv, on the present Chart; also in track xIII of my former Chart, which is better seen on Chart III in vol, i of this Journal, New Series.

Florida, nearly the same: as also at Charleston, S. C., nearer to the path of the storm; and scarcely falling below 30 inches at New York and Nantucket, as the storm passed. But at Camden, S. C., 140 miles N. N. W. of Charleston and 280 miles W. S. W. of Hatteras, the successive maxima on the 6th and 8th were only 30-04, and 30 inches; showing this place to have been beyond the crest of the external barometric wave. When storm-tracks recurvate on the interior meridians of the United States, the minimum depression of the barometer frequently moves nearly parallel to the direction from Camden either to Hatteras or to Chesapeak Bay.*

The annexed diagram shows the barometric curve at Washington, Fort Monroe, steamer Georgia, and Bermuda, while the storm was passing between the latter and Washington. These two places are distant from each other about 840 miles; which perhaps may be considered as an approximate measure of the barometric diameter of the storm on the 7th of September. The barometric curve of the Georgia, if increased so as to reach the minimum of the Swan, may represent a section through the center of the cyclone, in the direction of the storm's progression. Storm-Curves of barometric progression.

1. Curve at Washington: 2. at Fort Monroe: 3. Steamer Georgia: 4. Brig Swan : 5. Ship Eagle, crossing in front of storm: 6. Bermuda.

* I am indebted, on this occasion, to the officers of the Smithsonian Institution, and to Gen. Lawson, chief of the medical bureau at Washington, for observations from various parts of the United States; also to Lieut. Maury, Supt. of the Naval Observatory, for abstracts of the logbooks of Ship Eagle, and steamer Northern Light. Surgeons Williamson and Harrison of the Navy, and many ship-masters, and others, have kindly aided my inquiries.

It will be seen that the above diagram includes a period of thirty-three hours; and if we rate the progression at twenty-five miles an hour, it will comprise a distance of 825 miles. It contains the curve derived from observations on board the ship Eagle. I add the following condensed statement which is derived from the abstract of the ship's log sent me by Lieut. Maury.

The clipper ship Eagle, Warren, from Rio, crossed the centerpath on the morning of Sept. 7th, perhaps 350 miles in front of the axis of the gale, while running in the direction of Cape May. It is interesting to find that this vessel, which crossed the equator Aug. 17th, was overtaken by the external barometric wave of the storm as early as 4th-5th of Sept., and by a long swell coming up from S. E.; being then from 100 miles to 60 miles southward of Bermuda. Through the 6th, winds from southeast quarter, veering to South, with a heavy swell from S. S. W.; latitude at noon 34° 39', lon. 69° 32'; bar. 30-11. Sept. 7th, a very heavy swell from S. S. W.; wind fresh, from S. S. E. to S. S. W.; and from S. S. W. back to E. N. E.-bar. at 8 a. m. 29.90, at noon, 29.84 in. lat. 37° 17', lon. 72° 28': P. M. very threatening appearances from S. E. by S. to S. W., with a very heavy swell from S. W.; at 4 P. M. bar. 2970,—at 5 P. M., 29.77 wind fresh from E. N. E. to N.; 8 P. M., lightning at N. W.; at 11 P. M., in a heavy squall, wind shifted to N. N. W.; no rain; heavy sea still.-Sept. 8th, cloudy; no sea; lat. at noon, 38° 38', lon. 74° 13'; bar. 30 inches.

The steamer Northern Light, bound for the Isthmus, was several hours ahead of the Georgia, and on a more eastern track. She crossed the center-path in front of the gale, and ran through its eastern side.-Sept. 7th, lat. 34° 30', lon. 73° 25'; through the day, strong gales from the South; with a heavy sea from S.W.- -Sept. 8th, lat. 32° 01', lon. 73°, strong gales from S.W., with heavy'squalls, and a large sea from W. N. W.: Clear in the S. E., with stormy appearances in the N. and W. :—found the weather improving as we made south.- -This account is probably in nautical time.

The succeeding diagram represents, in its horizontal scale, the distance of 840 miles between Washington and Bermuda. The full line (1) represents, approximately, the barometric curve through the center of the storm, transversely to its path. The comparison of this transverse curve with the central curve of progression, indicated on p. 18, is of some interest; although we have no observations intermediate to the Swan and Bermuda. The resemblance of the two central cross-curves may show that the storm was of nearly equal extent and force on all of its sides, at that time.

I have been apprehensive of a clerical error in the barometric report from Fort Monroe for 3 P. M. and 9 P. M. of Sept. 7th; and that 29-063 and 29-087, should have read 29-63 and 29-87, respectively. In preparing the second diagram I became convinced

that the correction is required; and have accordingly applied it, in tracing the transverse curve: but have drawn a short trace line to show the observations as found in the report.

I have also inserted in this diagram, in broken lines, the transverse barometric curve through the center of the Cuba hurricane of October, 1844; when in nearly the same geographical position. This curve is approximated from twenty-eight observations in the path of that storm.

1. Transverse centre-curve of Cape Verde and Hatteras hurricane, Sept. 7, 1853. 2. Transverse center-curve of Cuba hurricane of 1844, Oct. 6th.

VORTICAL ROTATION OF THE GALE.-The true character of this gale as a cyclone, is made evident by the foregoing series of observations. This is most extensively shown by the various observations made on all sides of the storm during its passage between Bermuda and the nearer portions of the United States. The absolute whirlwind movement of the storm stratum, and the increasing rapidity of its leftwise rotation which is found in approaching the axial area, from whatever side of the storm, as well as the direct effects of this increased rotation on the fall of the barometer, in the interior portions of the gale, are made manifest by direct observation. This I might point out in full detail, were it at all necessary in the present stage of the inquiry. Nor can these results be evaded by denominating any one portion of the cyclonic wind, on either side of the cyclone, as another or distinct gale. The local variations and inequalities of the cyclonic action and the errors, imperfections or defects which may exist in the reports, are alike overborne by the amount of evidence SECOND SERIES, Vol. XVIII, No. 52.-July, 1854.

3

which serves to show the extent and general entireness of the vortical rotation in the gale.*

It would be an error to suppose that the gales and hurricanes which have been traced on our storm charts, were but exceptional cases of cyclonic action and progression in the winds of our globe. For there is a constant succession of rotary movements, greatly variant in their activity and their visible effects, and to which I shall further allude. It is the more violent cyclones, however, that afford us complete evidence of their geographic routes and their continued movement of rotation.

Of this active class, designated as hurricanes, gales, and storms, it is believed that the tracks or routes of several hundred might be added to our storm maps, by carefully collating the records which already exist. It is certain that a large number might be traced from the records and notices now in my possession or otherwise at hand; of which, the case I have now presented is but a single example. But the storms noticed in the succeeding portions of this article are selected in reference to their peculiar localities, as showing the uniform extension of the cyclonic system in equal latitudes around the globe, rather than for the amount of information possessed, regarding their extent and progression. (To be continued.)

ART. II. Account of a Rainbow caused by Light reflected from Water; by Prof. E. S. SNELL, of Amherst College.

I HAVE received from my friend and former pupil, Mr. H. M. Adams, of the Theological Seminary in East Windsor, Conn., a very interesting description of a brilliant rainbow scene, witnessed by himself and others, on the 24th of Sept. last. After a slight shower, the sun shone out, about 5 P. M., and produced the usual primary and secondary bows, except that they were of uncommon brightness. Four or five supernumeraries, exceedingly vivid and beautiful, underlined the upper part of the primary, the usual attendant of a very brilliant rainbow. In addition to these, there was seen an excentric bow, quite as luminous as the secondary, but in angular size and order of colors, just like the primary, and

While the printing of these pages was in progress I received from the government of Denmark, through Consul Bech, observations made at Oefjord, on Skage Strands Bay in the north of Iceland, in lat. 65° 40′ N., lon. 20° 40′ W.; which show the maximum pressure in advance of the cyclone to have been 29.75 in., at 1 P. M. Sept. 10th, under an east wind, the force of which is marked 2. The fall of the barometer under the cyclone continued till the night of 12th; the lowest observations being 29.12 in. at 10 P. M. of 12th, and 29:11 in. at 6 A. M. of 13th. The wind was S. E. on the 11th and 12th, and, on the rising of the barometer, was followed by the N. W. wind of the cyclone on the 13th, which afterwards changed to S. W., its force being marked 1. The normal effect of the cyclone at Oefjord, a position which is remarkably sheltered from the force of the cyclonic winds, by the peculiar outline and the extensive elevations of Iceland, is deserving of notice; not taking into account the modified direction and the abatement of force which appears to occur in the left quadrants of the gales in this highly northern portion of the Atlantic.