never did rain fall much thicker than the meteors fell towards the earth; east, west, north, and south, it was the same."

Notwithstanding, says Humboldt, the great quantity of falling stars and fire-balls of the most various dimensions which were seen to fall at Potsdam on the night of the 12th and 13th November, 1822, and on the same night of the year 1832, throughout the whole of Europe from Portsmouth to Orenburg on the Ural river, and even in the Southern hemisphere at the Isle of France, no one seemed to remark the coincidence of so many of these displays happening on the same day of the month. Olmsted and Palmer were the two principal scientific men who described the great meteoric shower of 1833 in America. The latter, calling to mind that the date of the shower described by Humboldt and Ellicott in 1799, was the 13th of November, first suspected the periodicity of these showers, a fact fully confirmed by a historical investigation into the dates of extraordinary showers of

meteors.

On the 9th and 10th of November, 1787, many falling stars were observed at Manheim, in southern Germany. Besides the manifestations already mentioned on the 13th November in the years 1799, 1822, 1832 and 1833, on the same day of the month. in 1831, at four o'clock in the morning, a great shower of falling stars was seen by Captain Bérard on the Spanish coast near Carthagena. On the same date in 1834 a similar shower, though not so great as that of 1833, was seen. Olmsted was the first to remark that nearly all the falling stars on the 13th of November, 1833, seemed to radiate from one point in the heavens, namely near the star y in the constellation Leo. The point of radiation did not change, but followed the apparent height and azimuth of the star during the continuance of the shower. This remarkable fact has been confirmed by observation of all the showers witnessed on this date since 1833. According to Enke's computation, this radiant point in space marks the direction in which the earth was moving on the 13th of November in its annual course round the sun.

The periodical appearance of falling stars on the same day of the year, all radiating from a point in the direction of the earth's motion, led Humboldt to conjecture that at that particular period the earth was passing through a ring or belt of minute planetary bodies, which were then drawn within the sphere of the earth's attraction,-a conjecture since pretty generally adopted. He also conjectured that there was, owing to the earth's or other planetary disturbance, a gradual retardation of the November phenomenon, owing to the change of the points where the ring of meteoric bodies intersected the earth's course.

He sought for records of falling stars in an

cient histories. On the night when king Ibrahim Ben Ahmed died, in October, 902, there fell a heavy shower of shooting stars "like a fiery rain," on which account that year was called the year of stars. On the 19th of October, 1202, the stars were in motion all night, and "fell like locusts." On the 21st of October, O.S., 1366, stars fell in such multitudes that they could not be counted. On the night between the 9th and 10th of November, 1787, many falling stars were observed at Manheim. Adopting the conjecture of Humboldt as to the gradual retardation of the November shower, others, more than twenty-three years since, ventured to predict that the great November shower of shooting stars and fire-balls intermixed, falling like flakes of snow, would not recur till between the 12th and 14th of November, 1867, taking for granted that the great November star-showers occurred once in thirtythree years, when the earth intersected the hypothetical ring of minute planetary bodies.

These showers are not equally visible from all parts of the earth's surface. The shower of 1799 was only seen in America; those of 1831 and 1832 were only visible in Europe; those of 1833 and 1834 only in the United States of America; and while a very splendid meteoric shower was seen in England in the year 1837, a most attentive observer at Braunsberg, in Prussia, on the same night, which was there uninterruptedly clear, only saw a few shooting stars, radiating from no particular point of the heavens, between the hours of seven in the evening and sunrise the next morning.

Though such occurrences as the great star-shower on the 19th of October, 1202, and 21st of October, 1366, seem to indicate a gradual retardation of the November shower, the relation of Theophanes, one of the Byzantine historians, that in November of the year 472 the sky appeared to be on fire over the city of Constantinople with the coruscations of flying meteors, may make us pause before assuming the November shower to be the retardation of the October phenomenon. Again, in the year 1766, just before the fearful earthquake at Quito, Humboldt states that the volcano of Cayambe was so enveloped with falling stars for the space of an hour, that the inhabitants fancied the mountain on fire, and endeavoured to appease Heaven by religious processions. The year corresponds with the 33-year period; but as the earthquake occurred on the 21st of October, the shower would seem to belong rather to the October manifestations of the 19th of October, 1202, and 21st of October, 1366. This should caution us not to generalize too hastily on a few recurrences of similar dates. Again, not to speak of the November showers for two years pre

viously and for three years successively to that of 1833, the showers of 1787 and 1822 cannot be brought into the 33year period of maximum manifestations of falling stars.

Besides the November period, there is another well-marked periodical fall of stars between the 9th and 14th of August. The 10th of August (St. Lawrence's day) was traditionally famous for "the fiery tears of St. Lawrence." As early as 1762, Muschenbroek remarked the large number of falling stars in the month of August. But the periodic return on St. Lawrence's day was first shown by Quetelet, Olbers, and Benzenberg. Bessel and Erman pointed out that the radiant point for the August shower was in Perseus. In April it is probable that there may be another period. On the 25th of April, 1095, "innumerable eyes in France saw stars falling from heaven as thickly as hail;" and on the 25th of April, 1800, a great fall of stars was observed in Virginia and Massachusetts; it was a "fire of rockets that lasted two hours." On the night of the 6th and 7th of December, 1798, Brandes counted 2,000 falling stars. At Quito on the 4th of February, 1797, shortly before the terrible earthquake of Riobamba, stars were seen to fall in swarms.

It may be well to remark that the train seen to follow a shooting star is no mere optical delusion, produced by the impression of light remaining impressed on the retina. It sometimes continues visible for a minute, or even longer, and even changes its shape. The falling stars which ordinarily occur, that is, which cannot be traced to any periodic display, and do not seem to emanate from any particular point of the heavens, are termed "sporadic." Eight is supposed to be the mean number to be observed in the course of an hour on any night. Perhaps they are more abundantly seen at some places than others. Burnes, describing the clear atmosphere of Bokhara, says, "At night the stars have uncommon lustre, and the Milky Way shines gloriously in the firmament. There is also a never-ceasing display of the most brilliant meteors, which dart like rockets in the sky: ten or twelve of them are sometimes seen in an hour, assuming every colour,-fiery red, blue, pale and faint." Jansen, again, describing the night scenes of the Java Sea, says, "The starlight, which is reflected by the mirrored waters, causes the nights to vie in clearness with the early twilight in high latitudes. Numerous shooting stars weary the eye, although they break the monotony of the sparkling firmament. Their unceasing motion in the unfathomable ocean affords a great contrast to the seeming quiet of the gently-flowing aërial current of the land breeze. But at times, when 30° or 40° above the horizon, a fire-ball

arises which suddenly illumines the whole horizon, appearing to the eye the size of the fist, and fading away as suddenly as it appeared, falling into fiery nodules; then we perceive that, in the apparent calm of nature, various forces are constantly active, in order to cause, even in the invisible air, such combinations and combustions, the appearance of which amuses the crews of ships."

The Reports of the British Association on Meteors from 1848 to 1853, from observations so zealously collected by the late Mr. Baden Powell, show how very frequent are the phenomena of fiery meteors, and what remarkable appearances some of them present. Considering the difficulty of determining the height of bodies presented so suddenly, and so transitory in their appearance, it is no wonder there should be considerable discrepancy among calculators. Olmsted thought the radiant point in Leo of the November meteors could not have been less than 2,238 miles above the earth's surface. Humboldt considers that the heights at the points of which shooting stars begin and cease to be visible fluctuate between 16 and 140 miles. Professor Brandes gives from 240 to 400 miles, and Olbers considers all determinations for elevations beyond 120 miles doubtful, owing to the smallness of the parallax. Brandes ascribes a diameter varying from 80 to 120 feet for shooting stars, and a luminous train extending from twelve to sixteen miles. The relative velocity of their motion he computes to be from 18 to 36 miles per second, their motion being frequently opposite to that of the earth.

Having now described the phenomena of the falling stars, and the larger meteors that accompany them, we proceed to the consideration of another class.

The falling stars, and the larger fire-balls sometimes associated with them, make their appearance suddenly, and, after describing an arc in the heavens, are as suddenly extinguished without passing below the horizon. No sound is heard to accompany this phenomenon. There is another class of fireballs which are seen to traverse the whole vault of heaven, seen often simultaneously over a large extent of country, whose course can sometimes be traced as passing, for instance, from one end of Great Britain to the other,-of large apparent magnitude, and of such brightness as sometimes to emit a light dazzling even at midday, and superior to the light of the sun. They are sometimes seen to burst into fragments with an explosion heard over a large area.

In the year 1676, on the 21st of March, O.S., about two

hours after sunset, a large meteor passed over Italy: it came over the Adriatic Sea as if from Dalmatia, crossing the country in the direction of Rimini and Leghorn, disappearing out at sea towards Corsica.

It was heard to make a hissing sound as it passed, like that of artificial fireworks. At Bologna the head of the meteor appeared larger than the moon in one diameter, and above half as large again in the other. At Leghorn it was heard to give a very loud report, like a great cannon; immediately after which another sort of sound was heard, like the rattling of a great cart running over stones. Its velocity was estimated at 160 miles per minute, its height about 38 miles, and its lesser diameter about half a mile. Another meteor was observed to pass over all England on the 19th of March, 1718. It was seen by Sir Hans Sloane in London at about a quarter after 8 at night. He was surprised by a sudden light far exceeding that of the moon. Turning towards it, he observed a large spherical meteor not so large as the moon, near the Pleiades, whence it moved after the manner of a falling star, but more slowly, in a seeming direct line, descending beyond and below the stars in Orion's belt. Its brightness was so dazzling that he was obliged to turn his eyes several times from it, as well when it appeared as a stream as when it became pear-fashioned and a globe. It left behind a track of a faint reddish-yellow colour, that continued for more than a minute. He heard no noise. Through Devon and Cornwall, and along the opposite coast of Bretagne the meteor was heard to explode. The report was like that of a very great cannon, or, rather, a broadside at a distance, followed by a rattling noise, as if many small-arms had been promiscuously discharged. Halley estimated the height of this meteor at 60 miles, and its rate at 300 miles per minute.

A similar meteor was observed in England on the 11th of December, 1741, while the sun was shining brightly in a serene sky, and was heard to explode in Sussex. A friend of Humboldt, in the year 1788, at Popayan, found his room lighted up at noonday by a meteor, while the sun was shining brightly in a cloudless sky.

Hundreds of such meteors have been described, and though explosions have been heard over parts of the country favourable for recovering anything solid if it fell from a meteor, in but some four or five cases has anything been picked up likely to have fallen from a meteor; and out of these few cases some are considered doubtful.

We now proceed to consider another phenomenon,--the fall