in height. Water-spouts are analogous phenomena on a much larger scale. The whirlwind not only exists in the cloud, but even in the water, which rises and joins the cloud that descends to meet it.

Water-spouts are not equally common in all parts of the ocean. In the middle of the equatorial sea we only find them when the trade-winds do not blow in a regular manner; they are only manifested in the region of calms. They are commonly met with near the coasts, or in straits, and they are more frequently formed at the time when the monsoons change: something analogous occurs in the high latitudes, where they frequently coincide with storms.

From the facts collected by myself and others, waterspouts almost always occur when two opposite winds pass side by side, or when a very brisk wind prevails in the higher regions of the atmosphere, while it is calm below: I saw a remarkable example on the Rigi in 1832. Being on the summit of the mountain, I examined masses of fog which were proceeding towards each other in the valley of Goldan, whilst around me the air was calm and the sky serene. At the end of a few moments the masses united, and I observed a gyratory movement in the midst of them; the fog extended with inconceivable rapidity, and violent gusts drew from it rain and hail. In the meantime the temperature had fallen, so that the water between the teeth of the wheels of my Woltmann's anemometer was congealed. A physician of Dantzig, who arrived in the evening, told me, that on the lake of Quatre-Cantons he had experienced a violent storm, during which the clouds had been driven in different directions; at the same time he saw a waterspout.

If the currents that meet in the high regions of the atmosphere are violent, if their temperature and the quantity of vapour of water with which they are charged are very different, then the vapour is rapidly condensed. In proportion as the whirlwind increases, it descends, and the diameter of the column decreases; we cannot decide whether the vesicles of water are drawn from above downwards, or whether the condensation occurs in the opposite direction. Finally, the whirlwind reaches the surface of the earth, the latter is agitated, and rises resembling a boiling caldron. While the sea rises the cloud descends, and finally they both unite; it also sometimes happens, that the sea rises in the form of a cone, while an inverted cone descends from the cloud without their both uniting. In the majority of cases' the column is thinner in the middle than it is above or

below; under other circumstances, the first trace of the waterspout is evidenced on the sea, a cone rises from the surface of the waters, and it is not till after some time that the vapours from above condense in their turn.

A fact, which proves that the water-spout is in a great measure formed of condensed vapours, is, that the water escaping from it is never salt, not even in the open sea.

If the air is very dry, then these whirlwinds do not always determine the condensation of vapours, and the violence of the wind is the only thing remarkable. Thus, two of my friends were one cloudy day going from Halle to Giebichenstein; on a sudden they were separated by a gust of wind, and one was driven against a wall and the other thrown into a field, while the people who were near had not perceived the least disturbance in the atmosphere.

Almost all observers say, that the water-spout moves slowly onward, turning on its axis; if the current rises, as we see in the whirlwinds of sand, it may draw up enormous masses. Wolke relates, that a shepherd near the Jever saw, near Repsolt, a place situated three myriametres from the sea, a water-spout pass before him, and instantly dry up a pond and throw the fishes in all directions. All descriptions prove that in these cases the wind possesses extraordinary force. Thus, Dr. Mercer observed two or three water-spouts in the port of St. Jean of Antigua on the surface of the sea; he saw a circle of about sixty metres in diameter, within which the water was agitated and darted towards the sky. A small wooden house was carried away entire, and transported to the distance of thirteen metres without being overturned or destroyed. It is remarkable, that the house was carried from east to west, although the water-spout travelled from west to east. I borrow from the public newspapers another example of the same kind :-Oct. 25, 1820, a large quantity of linen had just been spread out in a meadow in Silesia; the workmen were at dinner, when the tempest announced itself a few minutes after noon, and raised such thick clouds of dust that the day was converted into thick darkness. The doors and blinds of the bleaching establishment were torn down with a terrible crash, the doors were carried off their hinges, and the wind overturned a heavy cart, so that the wheels were placed upwards. The linen was seized upon, and rolled up, and the largest mass was carried fifteen metres above the house, and deposited at 150 paces distant in a ditch in the midst of some bushes. They had to work for several hours in order to disentangle this immense skein; it consisted of twenty-seven pieces, each of

which weighed eleven kilogrammes; and in the middle was found a post two metres long, thirty centimetres wide, and six thick, which was used as a bridge for crossing a ditch at a little distance. The water-spout had carried it away with the linen, which it had rolled round and carried over the house, although its weight, without calculating that of the plank, was about 297 kilogrammes.*

When we call to mind the force with which small waterspouts raise water, we are not astonished that a great one can produce such effects. Some authors have attributed these effects to electricity; but if we merely rest on the fact, that this fluid determines such effects as these on the surface of water, we should not forget that purely mechanical forces could produce it as well. Other philosophers have thought that a partial vacuum was formed, in which the water rises as in the body of a pump; but, to suppose the existence of this vacuum, the water would only rise to the height of ten metres, and the helical motion would not exist. It has also been said, that gases suddenly escape from the earth where the water-spout is formed, and raise the water; such an hypothesis as this requires no refutation.†

The water-spout that devastated the village of Châtenay, near Paris, June 18th, 1839, brake, near their base, elm-trees 1,50 in circumference. M. L. LALANNE, civil engineer, who drew up the plan of the place after the disaster, estimates at 456 kilogrammes per metre the effort exercised against certain parts of the walls that were overturned.

According to M. RENAUX, architect, the water-spout that passed over the town of Courthezon (Vaucluse), May 30th, 1841, overturned a face of a rampart, twelve metres long by eight high and one thick. A great part of the materials were carried to the other side of the Seille, to the distance of about eight metres. In the Faubourg d'Orange, a newly constructed façade was demolished.-(Comptes rendus de l'Acad. des Sciences, t. ix. p. 219, and t. xiii. p. 223.)

In the Traité des Trombes, which M. PELTIFR published in 1840, we find an account of 137 water-spouts. In this number we notice thirty-three that existed in the midst of calm, twenty-five which had no migratory movement, thirty-seven which occurred with this motion. The silence of the accounts upon the rest of the water-spouts is a presumption in favour of the negative, because, as M. PELTIER says, a relation is the indication of that which is, and not of that which is not. Ten occurred in a sky without clouds; seven are multiples, that is, there were several branches from the same trunk; three were formed within the clouds, &c.

In addition to this, we find fifty-two relations of stormy phenomena which produced effects analogous to those of water-spouts; finally, this treatise contains the details of experiments by which the different parts of this meteor were produced by means of electricity. The sum of the facts does not appear favourable to the theory that attributes this order of phenomena to whirlwinds.-(l'ide also, on water-spouts, the article by M. OERSTEDT, in M. SCHUMACHER'S Annuaire for 1838.)-M.

VII.

OPTICAL PHENOMENA

OF

THE ATMOSPHERE.

HITHERTO We have only studied the heating power of the sun; we have seen that the diurnal and annual range of temperature depends on the revolution of the earth around this body. The unequal heating of the different regions of the earth has disclosed to us the origin of winds, the changes in the aspect of the sky, and the differences in the pressure of the atmosphere, changes which are interwoven with a rupture of equilibrium in atmospheric electricity. We are now about to examine the phenomena to which the sun gives rise, when regarded as a luminous body, after having prefaced them by a few general observations upon light.

NATURE OF LIGHT.-Light and heat are so intimately connected with solar rays, that it is difficult to separate these two manifestations; we shall not enter deeply into this subject, which belongs to the domain of experimental philosophy; nor shall we seek to penetrate farther into the intimate nature of light. Two hypotheses have been put forth. In the first, light is composed of material molecules sent forth by the luminous body; by acting on the retina, they enable us to see the body from which they emanate. In the other supposition, which appears much more probable, the luminous impressions are due to the undulations of a very subtle fluid spread throughout the

universe and in the interior of all bodies, a fluid known by the name of ether. There exists, therefore, in luminous bodies a force which puts the ether in motion, as the string of a violin, when put in motion, makes the air vibrate, and produces a sound. The undulations of light are transmitted through the ether, as those of sound are through air; and they excite in the eye a sensation analogous to that produced by the others on the ear. In passing from vacuum into the air, these undulations undergo the several modifications that we are about to study.

REFLECTION AND REFRACTION OF LIGHT.

The ether being only a hypothetical fluid, the properties of which cannot be recognised by aid of experiment, we can say nothing positive as to its nature in the interior of bodies; however, if we conclude from analogy with that existing in the air, we may admit that its density in the interior of bodies is greater than in vacuo, and that this density is different in the different bodies in nature. Take, for example, pieces of glowing carbon and allow them to cool in vacuo, then place them in a small capsule swimming on a surface of mercury, and cover them with a small bell-glass; in a short time, the quantity of air contained in this bellglass will diminish, it will be absorbed by the carbon; and, as the mass of air that has disappeared is greater than that of the carbon, we must conclude that the air has been condensed within the carbon. This absorption of air is so considerable, that fifty kilogrammes of newly prepared carbon, when exposed to the air, will weigh, after a certain time, 52,5 kilogrammes. Almost all bodies exercise an analogous influence, but with a different intensity, according to their nature: it is probable that the same occurs with respect to the ether.

When light passes from vacuum to a transparent body, the undulations change with the density of the ether; their nature remains the same, but the amplitude of the oscillations is less, as a wave diminishing in height. Light is divided into two parts; one is reflected, the other penetrates into the interior of the body. Sound presents analogous phenomena: if sonorous waves strike against a wall, one part of the sound is propagated through the wall, and we hear it behind the wall; the other is reflected and produces echo.

Experiment proves that the reflection of a ray is not made in the same direction as that of its incidence. Indeed, let it be made to fall upon a flat and perfectly polished surface, and suppose a perpendicular to be raised at the