In the city of New York, the damage done by the hail was not very great; for the stones were not numerous, although of prodigious size. The ship-yard of Mr. Thomas Collyer at the Dry Dock, was covered with singularly shaped pieces of ice,-one of which was measured and found to be 64 inches in circumference -another seven inches, and a third measured three inches long, and two inches thick.

In Williamsburgh the hail appears to have destroyed more glass than in New York. In many houses nearly half the glass was broken in windows which were unprotected on the north side. Over 400 panes of glass were broken from the north side of a single school house.

On the same day with the preceding storm, large hail is said to have fallen at several places in Pennsylvania. About three o'clock in the afternoon, a terrific hail-storm passed over Northumberland doing great damage. Hail-stones are said to have been picked up measuring 73 inches in circumference; and several thousand panes of glass were broken in that town.

At 53 o'clock, P. M., a severe hail-storm passed about 20 miles. north of Philadelphia. At Upper Dublin, the storm was very destructive. Several barns were unroofed, many fruit and forest trees were blown down-and many fields of wheat and oats so badly damaged, as scarcely to pay for harvesting. One hailstone was measured, and its greatest circumference found to be 61 inches, and its smallest five inches,-and this was half an hour after the storm had abated. At Norristown and Doylestown the crops were much injured by the hail, and at Burlington, N. J., the wind was exceedingly violent.

It is not probable that either of these storms was the same as that which passed over New York. The hail-storm near Philadelphia, was about simultaneous with that at New York. The storm at Northumberland may have been identical with that at Upper Dublin, the distance of the places being 100 miles-the interval of time 24 hours-and the direction nearly parallel with the track of the New York storm.

It would appear that a violent wave of great extent set in from the northwest, which rolled over both New York and Philadelphia, and within this wave were formed about simultaneously several distinct veins of hail.

Was the storm which passed over New York a whirlwind? I have surveyed every part of the track of the storm where I have heard of any violent effects, especially with reference to the decision of this question. Throughout Williamsburgh, I could find no unequivocal evidence of rotation. The steeples which were prostrated, fell in a direction coinciding very nearly with that of the storm's progress, that is, towards the southeast. In the case of one of the churches whose steeple was blown down SECOND SERIES, Vol. XVII, No. 49.--Jan., 1854.

6

(the first Presbyterian Church) I noticed a phenomenon which I considered worth recording. The track crossed the ridge of the church at an angle of about 45°. On the leeward side, the tin. roof was started from the boards (but not broken) and puffed up forming a wrinkle about twenty feet long, two or three feet wide, and ten inches high. This appears to me to indicate the operation of a force beneath, pushing up the tin; but not being able to tear the tin open, bulged it up and left it in a ridge.

This phenomenon appears to be analogous to what often occurs in tornadoes, and I ascribe it to a rarefaction of the air on the leeward roof. A current of air, forcibly impelled over an obstacle like the roof of a building, by friction drags along with it the air lying upon the leeward side of the roof, producing a partial rarefaction, and the air beneath by its expansion, tends to lift the roof. Thus the leeward roof is often carried away, while the windward roof remains. In the present case, this upward pressure lifted the tin about ten inches, stretching but not tearing it. This force appeared to be insufficient to tear the tin from its fastenings-perhaps because from the carrying away of the steeple, and the ripping up of the adjacent edge of the tin, the air beneath found a ready escape.

In the neighborhood of the Crystal Palace, occurred a phenomenon which appeared to indicate the existence of currents blowing nearly in opposite directions. The wooden buildings, which have already been mentioned, were blown toward the southeast -but the brick wall, the line of which run from N. 28° E. to S. 28° W., fell toward the west; that is, in a direction nearly contrary to that of the storm's progress.

The following appears to me to be the explanation of this phenomenon. The Latting Observatory is an octagonal tower, 300 feet high and 75 in diameter at base, sloping uniformly to the top. In the annexed figure,

the octagon represents the base of the tower, and the line N S represents a meridian. On the west side of the tower, was erected a brick wall A B, 25 feet high, and only three feet from the side of the tower. At the south end, it was connected with another wall B H, but at the north end it was entirely free, and had no support except an iron clamp projecting from the side of the tower. The direction of B the storm's progress is indicated by the arrow. It might have been anticipated that the wall A B would have been thrown towards the east upon the tower; whereas in fact it was thrown outwards towards the west. But we know

S

from the testimony of spectators, that this wall fell at the first onset of the blast, when, as we shall presently see, the wind blew nearly from the north, or perhaps a little east of north. Now a violent current from the north, driven into the triangular space ADC, would necessarily become condensed between the wall and the tower, exerting a force to push the wall outward, and the wall had little strength to resist the pressure, being weak not only from its great height, but also being unsupported at the north end. The bricks also had been recently laid, and the mortar was not yet dry. On the east side of the tower, was a similar wall EH, but only 14 feet high, which was not prostrated. Its security is probably to be ascribed to its inferior height.

The following facts at first seemed to me a little puzzling. Near Paterson, a large oak limb, a foot thick, was twisted off at the height of fifty feet, and prostrated in a direction pointing S. 20 E., the top of the limb lying towards the base of the tree. Within a short distance I found another large oak limb, torn off at a great height and thrown towards S. 40° E., with its top also turned towards the base of the tree. Not far off, I found a third limb in a similar position. Broken limbs were generally found to have been carried eastward, with the top pointing to the S. E., and the base towards the N. W. The three cases I have here specified were exceptions to the general rule, and it appears to me that they are to be explained by supposing that the branches turned a somerset in falling.

A like case occurred with the steeple of the first Presbyterian Church in Williamsburgh. The spire fell across the street-the top struck a brick house on the opposite side of the street and broke off, while the upper portion of the spire remained imbedded in the roof of the house which was crushed in by the blow. The remainder of the spire, which was now the frustum of a pyramid, fell down into the street; but it is probable that the smaller end of the frustum struck the pavement first, for the steeple turned a somerset along the street towards the east, and lay after the storm with the smaller end of the frustum turned towards the church. A similar supposition will satisfactorily account for the observed position of the three limbs above mentioned.

In the woods between Acquackanonck and Paterson, I measured with a pocket compass the direction of a large number of prostrate trees. The following list shows the entire range of the bearings which I measured; not arranged in the order in which they were measured, but classified according to the points of the compass. They were

S. 70° W.; S. 20° W.; S. 15° W.; S. 10° W.; south; S. 10° E.; S. 20° E.; S. 25° E.; S. 30° E.; S. 35° E.; S. 40° E.; S. 45° E.; S. 50° E.; S. 60° E.; S. 70° E.; east.

These bearings were measured at various points upon a portion of the track about two miles in length; and it will be noticed that there is not a single instance of a tree which was prostrated towards any point between East, North and West. The bearings extend from east, through south, to S. 70° W., a range of 160 degrees; but I found only one instance of a bearing approaching nearly so close to the west point. With but one exception, the bearings were all confined between east and S. 20° W. The wind did not then blow from every point of the compass indifferently, at least not with sufficient force to prostrate trees, but it blew only from the northward, including northeast and northwest. Neither was the wind a simple rectilinear current. What law then did the wind observe? Was its motion merely centripetal ? Did it revolve in a whirl? Or did it follow some other law?

In order to decide these questions, I attempted to apply the method which I had successfully practised in the Mayfield tornado of Feb., 1842. This method consisted in selecting groups of prostrate trees which lay upon each other, and measuring successively the bearings of the bottom tree, of the next in order, and so on to the top, and regarding these bearings as indicating the successive directions of the wind at the point of observation, as the storm passed by. At Mayfield I had no difficulty in finding groups of trees piled upon each other, frequently four or five in a group; but at Paterson I found in no case more than two trees crossing at a considerable angle, and only five instances of this description. The following are the observations in the cases referred to, the bearing first mentioned in each case being that of the bottom tree.

The first four cases present no angle greater than 30°; the fifth case presents an angle of 130°; that is, the two trees were turned in nearly opposite directions.

From a comparison of all the facts, I conclude that the wind. blew first from the northeast, and that this current was succeeded by a north and presently a northwest wind. The following are my reasons for this conclusion.

1. The fifth case of interfering trees, just mentioned, taken in connection with all the bearings observed, points to this conclusion. We find that one large tree was prostrated with its top turned towards a point S. 70° W. Upon it lay another large tree with its top turned S. 60° F. We may safely infer that these directions corresponded nearly with the direction of the wind when the trees were prostrated, and that the wind came first from a point N. 70° E., and was succeeded by a curreut from N. 60° W. In each of the other cases of interfering trees, the angle of crossing was so small, as to convey no very distinct information

on this question. In three cases out of the four, the first blast appears to have been a little more westerly than the final one; but all the trees were prostrated by a northwest wind.

2. A very intelligent farmer, whose house was close upon the northeast margin of the track, about four miles from Paterson, gave the following testimony. He first took refuge from the hail under a shed on the southwest side of his barn, the wind then blowing from the N. E. After a short time, the hail began to beat upon him, the wind having veered to the N. W., and he was obliged to seek a shelter on the southeast side of his barn in order to escape the hail.

3. It is known from the testimony of several individuals, that the wind at New York was easterly on the first approach of the

storm.

Upon comparing these facts, it appears to me that the direction of the wind at the time of its most destructive violence may be tolerably well represented by the annexed diagram, showing a

South

current from the N. E., on the front of the storm; and from the N. W. in the rear, the whole having a progressive motion towards the S. E., which would give to each place in succession (unless near the southwest margin of the track) first a N. E. wind, and afterwards a N. W. wind.

I do not then find in this case that evidence of a complete rotation which I have found in some other tornadoes; but it is possible that at a little elevation above the earth's surface, the rotary motion may have been more decided.