These lofty clouds, whether formed over land or water, when the dew point is very near to the temperature of the air, appear to let down from their bases a tongue of vapor in the form of an inverted cone, which has been called a spout. Mr. Espy, in his explanation of this phenomenon, says: "If, however, the air is very hot below, with a high dew point, and no cross currents of air above to a great height, then, when an upmoving current is once formed, it will go on and increase in violence as it acquires perpendicular elevation, especially after the cloud begins to form. At first the base of the cloud will be flat; but after the cloud becomes of great perpendicular diameter, and the barometer begins to fall considerably, as it will do from the specific levity of the air in the cloud, then the air will not have to rise so far as it did at the moment when the cloud began to form, before it reaches high enough to form cloud from the

cold of diminished pressure. The cloud will now be convex below, and its parts will be seen spreading outwards in all directions, especially on that side towards which the upper current is moving, assuming something of the shape of a mushroom. In the mean time, the action of the in-moving current below, and upmoving current in the middle, will become very violent, and if the barometer falls two inches under the centre of the cloud, the air, on coming in under the cloud, will cool by diminished pressure about ten degrees, and the base of the cloud will reach the earth, if the dew point was only eight degrees below the temperature of the air at the time the cloud began to form. The shape of the lower part of the cloud will now be that of an inverted cone with its apex on the ground, and when a little more prolonged and fully developed, it will be what is called a tornado if it is on land, and a water-spout if at sea."

Mr. Espy observes that there is a tendency in one of these clouds to form another, and the second has a tendency to form a third, and so on, till a number are in operation at the same time. The cause of this he very happily explains, but our limits will not allow us to follow him. Lieut. Ogden gives an account of seven of these spouts seen at one time, in the edge of the Gulf Stream, in May, 1820, which we copy, together with the annexed cut.

He says:-"The atmosphere was filled with low, ashy-colored clouds, some of which were darker underneath than others, and from these the water-spouts were generally formed, each one from a separate cloud. In some instances, they were perfectly formed before we observed them, but, in others, we could see a small portion of the cloud, at first extend downward, in the shape of an inverted cone, and then continue to descend, not very rapidly, until it reached the water. In other instances, however, we observed that this conical appearance of a portion

of the cloud did not always result in the perfect formation of a waterspout. Several times we saw the cone project, continue for a short time stationary, then rise again slowly, and disappear in the clouds. This would, in some cases, occur two or three times to the same cloud; but, eventually, a larger and darker cloud would descend, and result in forming the visible spout, as above mentioned."

It will be seen at a glance, that the principle on which Mr. Espy explains the phenomena of nature in the production and development of storms, requires the convergence' of the winds towards a common centre or line at the base of the cloud. In this he differs materially from Mr. Redfield, who has been at great pains to show that all storms are whirlwinds, and that the air moves around from right to left, or contrary to the hands of a watch. On this point there is still much controversy, but we have no room to enter on the merits of the discussion in this article, and shall content ourselves with exhibiting some of the facts on which Mr. Espy relies to establish this, one of the main pillars of his theory.

As the violent action which attends tornadoes is generally confined to very narrow limits, these storms seem to furnish the best means for testing the truth of these different theories. It is, we think, clear that if the wind moves around a common axis in the form of a whirl, that the trees which are thrown down on the borders of the storm should lie parallel to its path, while those which fall in the centre should be left in a transverse position, or at least be thrown outwards and forwards on one side, and outwards and backwards on the other. Now it would seem from a great variety of testimony that the trees in these violent storms are not prostrated in the above named direction.

President Bache, of Girard College, after having carefully taken the direction in which the trees fell in the New Brunswick tornado with a mariners' compass, says :-"I think it entirely made out, that there was a rush of air in all directions at the surface of the ground towards the moving meteor; this rush of air carrying objects with it. The effects all indicate a moving column of rarefied air, without any whirling motion at or near the surface of the earth."

Professor Loomis,* of the Western Reserve College, after drawing a map of the trees and buildings which fell in a hurricane that passed over Stowe in Ohio, comes to a similar conclusion. "It will," he says, "then appear from an inspection of the diagram, that in the midst of some disorder there was a degree of uniformity. Thus upon either border of the track the trees all incline towards some point in the centre of the track. There is not an example of a tree being turned outwards from the track, nor even one which lies in a direction parallel to it." He afterward adds,—“ We have now established, by a fair deduction, that there was a powerful current of air from the opposite sides of the track towards some point in the centre of the track, and that here there was also a powerful current upward."

Professor Olmsted,† of Yale College, in his account of the New Haven tornado, which occurred on the 31st of July, 1839, says :-"The first great fact that strikes us, is, that all the trees and other objects that mark the direction of the wind which prostrated them, are, with a very few ex

• Professor Loomis is not an advocate of Mr. Espy's theory. + Professor Olmsted is not a believer in Mr. Espy's theory.

ceptions, turned inwards on both sides towards the centre of the track; while near the centre, the direction of the prostrate bodies is coincident with that of the storm.'

Professor Forshay in his account of the Natchez tornado is equally in point. He declares that "the nearer the axis of the tornado, the nearer were their bearings parallel with that axis, and the more remote, the nearer perpendicular, while those that point to the direction from which the storm came, or cross a line perpendicular to the axis, lie beneath those that point in the forward direction of the same."

We may mention also, that the storm which occurred in France, of which we have given a chart in the former part of this article, could not possibly have been of the whirlwind character. Had the wind moved in a whirl, the hail which fell during its progress, must have been scattered over the whole area of the storm, and not been deposited in two veins for many miles as we have seen. If the whirlwind theory is correct, therefore, this storm at least must have been a wonderful exception to the general law.

But Mr. Espy does not rely alone on the direction of fallen trees in tornadoes to prove the centripetal course of the wind in storms. By means of observers in different sections, he has been enabled to surround some of our great northern storms, and has satisfied himself that the same law uniformly prevails. We can only make a few selections from the great number which we find recorded in the volume before us.

The following diagram represents a destructive storm which swept along our southern coast in the middle of August, 1837. The facts respecting it were collected by Col. Reid, but Mr. Espy finds that they maintain his own views, although recorded by an advocate of the whirlwind theory. The map represents the position of the storm as it was at noon, on the 18th of August, and the arrows are intended to show the direction of the wind at that time.

Mr. Espy observes: "I have culled out of this storm, that portion of time in which I find the greatest number of simultaneous observations, and I have exhibited on the annexed wood-cut the localities of all the ships within the boundaries of the storm, whose latitudes and longitudes could be ascertained with any degree of certainty, with arrows, exhibiting the

course of the wind. The time is noon of the 18th of August, 1837. At this time, the Duke of Manchester was only a few miles N. E. of the centre of this storm; for some time in the afternoon, the centre of the storm passed nearly over her, when the wind changed pretty suddenly S. W. At this time, and for some seven or eight hours both before and after, all those ships which were laboring in the most violent part of the storm, had the wind blowing towards a central space of no great magnitude. This settles the question of a violent centripetal motion of the wind in this storm, in conformity with the five previously examined, and also with the twelve investigated by the Joint Committee of the American Philosophical Society and Franklin Institute, and with not less than fourteen land-spouts which have already been examined, in all of which the trees were thrown with their tops inwards-and when any are thrown across each other, those which are underneath, are uniformly found to be thrown inwards and backwards, and those on the top, to be thrown inwards and forwards, just as they should be, if the wind blows inwards. Whereas, if the wind is centrifugal, many of the trees should have the tops thrown outwards on both sides of the path."

The following chart represents the course of the wind in the storm which occurred in Great Britain on the 17th of August, 1840.

0. Workington, changed at 10, A. M., from S. S. E. to N. N. W.-1. Plymouth, W. on 17th, S. W. on 16th.-2. Pill-Bristol, S. W., A. M.3. London, southwardly, on 17th.-4. Lynn, heavy S. till noon, then S. W., more moderate.-5. Hull, S. S. W., strong.-6. Leeds, S. E. or S. S. E., strong from 8, A. M., to 1, P. M., clouds at this time moving from S. W.-7. Sheffield, S. S. E. all day, next day, E. Strong on 17th. -8. Hyde, near Manchester, S. W., in the morning; west in P. M.; strong gale all day. -9. Liverpool, S. W., A. M., N. Westerly, P. M., strong.10. Belfast, N. by W. strong gale.-11. Point of Ayre Light, N. W. gale.-12.

Corsewell Light, N. N. W., storm.-13. Dublin, W. N. W.-14. Largs, heavy from Ñ. N. W. from 7, A. M. till 8, P. M.-15. Kyntire Light, N. W. gale.-16. Pladda Light, N. W. breeze.-17. Greenock, N. W. and N.-18. Lismore Light, N. W. gale.-19. Dumferline, N. and N. E. till 2, P. M. increasing to a gale.-20. Edinburgh, N. N. E. strong. -21. Berwick, S. by E. to S. E., strong.-22. Aberdeen, E. all day, strong.-23. Middle line of the storm on morning of 17th.

To the mariner it is of immense importance to discover the true law of storms. As his life and property will often depend upon the theory which