in terrible commotion; in fact, while the hail was falling, a sort of tumbling in the clouds was noticed as they came up from the north-west and south-west, and about where they appeared to meet was the point from which the funnel was seen to descend. There was but one funnel at first, which was soon accompanied by several smaller ones, dangling down from the overhanging clouds like whip-lashes, and for some minutes they were appearing and disappearing like fairies at a play. Finally one of them seemed to expand and extend downwards more steadily than the others, resulting at length in what appeared to be their complete absorption. This funnelshaped cloud now moved onward, growing in power and size, whirling rapidly from right to left, rising and descending, and swaying from side to side. When within a distance of three or four miles, its terrible roar could be heard, striking terror into the hearts of the bravest." The eye-witness judged that the funnel itself would reach a height of about five hundred feet from the ground. As the storm crossed a river, a cone-shaped mass came up from the earth to meet it, carrying mud, débris, and a large volume of water (Fig. 57). The cloud then passed the observer's house very near to 4 p.m. The progressive velocity at the time was considered to be about thirty miles per hour, although at Delphos, three and a half miles distant, it had slackened down to near twenty miles. A few minutes previous to and during the passage of the funnel, the air was very oppressive; but ten minutes after, the wind was so cold from the north-west that it became necessary to wear an overcoat when outside. As the tornado struck the house, another member of

the family says, "We think it is coming near us. We can now see its fury. Shall we leave the house? No; for we are not certain on which side it will pass. We are apparently as safe here as elsewhere. The windows are nailed fast. Three of us lean against the door which is nearest the storm; the rest go into the cellar. It is about 4 p.m. A moment of breathless suspense, and the storm strikes us. The timbers creak, the sides of the house sway in and out; surely they cannot outlast it. We hear no well-defined roar now, for on the outside boards and other débris are fiercely crashing. All is dark within. In about fifteen minutes the storm is over. We leave the house. The centre of the storm has passed to the west of us, and we can see its dark form moving away in a north-east direction."

The actual diameter of this storm appears to have been only forty-three yards. On the right of the track, destructive winds extended to a further distance of from one to two miles, sensible deflected winds for another mile and a half, beyond which only the usual wind of the day was experienced. On the left or northern side of the tornado path, the damage did not extend quite so far, for the width of the belt of destructive winds was not more than twenty-eight yards across, and that of sensibly deflected winds one mile and a quarter.

As a specimen of the damage done, a large two-horse sulky plough, weighing about seven hundred pounds, was carried a distance of twenty yards, breaking off one of the iron wheels attached to an iron axle one and threequarter inches in diameter. A woman was carried to the north-west two hundred yards, lodged against a barbed

wire fence, and instantly killed. Her clothing was entirely stripped from her body, which was found covered with black mud, and her hair matted with it. A cat was found half a mile to the north-west of the house, in which she had been seen just before the storm, with every bone broken. Chickens were stripped of their feathers, and one was found three miles to the north-west.

A few miles further on, another eye-witness says, "The dark, inky, funnel-shaped cloud rapidly descended to the earth, which reaching, it destroyed everything within its grasp. Everything was taken up and carried round and round in the mighty whirl of the terrible monster. The surrounding clouds seemed to roll and tumble towards the vortex.

"The funnel, now extending from the earth upwards to a great height, was black as ink, excepting the cloud near the top, which resembled smoke of a light colour. Immediately after passing the town, there came a wave of hot air, like the wind blowing from a burning building. It lasted but a short time. Following this peculiar feature, there came a stiff gale from the north-west, cold and bleak, so much so that during the night frost occurred, and water in some low places was frozen.”

RELATION OF WHIRLWINDS TO CYCLONES.

Before concluding this chapter, we may make a few remarks on a very interesting question which here presents itself. Commencing with a whirlwind only two or three feet across, we find every gradation of size till we come to the destructive tornado. From the small

secondary which deflects the wind in connection with a thunderstorm, there seems to be every gradation of size into the secondary which is so large that we can hardly say whether it should not be called a primary cyclone.

In both the whirlwind and cyclone series we have certain common features-a horizontal rotation, and more or less uptake near the centre of the gyration. But is there any intermediate series between the whirlwinds and the cyclones, or can the former ever develop into the latter? We believe not, though the opposite opinion has often been propounded.

In the first place, we are unable to find any connecting link between the two types of rotation. Under certain conditions, the wind seems to have a tendency to form little eddies, which under favourable circumstances may grow into complete cylinders of rotating dust. In our chapter on Prognostics we have shown that in England whirling dust is a well-known precursor of showers of rain, but not of the true cyclone-rain. In other countries, such as the Punjab and on the Isthmus of Suez, regular whirlwinds are of daily occurrence at certain seasons of the but these never by any year; chance grow into even the smallest secondary. We have just seen that the terrific tornadoes of the Western States of the American Union are merely an episode in the conflict of opposing currents near the trough of larger depressions, but the whirlwinds never give rise to any larger disturbance. In every one of the eleven tornadoes which occurred on the day we have just described in some detail, it was found that rain and hail invariably preceded the tornado-cloud from ten to thirty minutes, and that the tornado was

only, as it were, a local accident in a very large disturbance.

In our chapter on Weather-types, we shall give abundant illustrations of the manner in which both primary and secondary cyclones are formed without the presence of two such opposing currents as are found in front and rear of the trough of a V-depression, and we shall see how the two kinds of cyclones may either develop or degrade from one into the other.

We have also already seen the very small circulation of the wind which accompanies some kinds of thunderstorms, but in no case do we find any transitional link between the whirlwind and cyclone types of rotatory motion.

At the same time, we may note that the inner core and very deep central depression of a tropical hurricane approximates more nearly to the tornado type than the cyclones of temperate regions; but the absence of transitional forms seems conclusive against the identity of tornadoes and cyclones. In both the destructive fury is out of all keeping with any forces that we are acquainted with; and their true nature remains to be discovered by future research.