vapours which liberate electricity: hence we can understand how complicated the phenomenon is.

All storms furnish a proof of these successive condensations. A flash of lightning passes the zenith, and before the clap of thunder, but rarely afterwards, the rain or hail escapes in torrents from the cloud; the drops at first fall in a line inclined to the horizon, and then return to a vertical direction. It is commonly stated, that the rain is an effect of the lightning tearing the clouds; 19 but it is the gust of wind condensing the vapours into large drops, having first driven them into an almost horizontal direction: hence, the escape of electricity and the claps of thunder. As a proof that this condensation precedes the lightning, the rain often falls before the noise of the thunder is heard: now, the latter travels 333 metres per second; if, therefore, the rain were an effect of lightning, it would follow that the drops of water would have fallen with a velocity at least equal,―a velocity which they never have, even at the end of their fall.

Add to this, that storms are frequently heard over a surface of many thousand square myriametres, and that the electricity of each of their parts reacts on the other. The observer, who is situated on the plain, has not a sufficiently extensive view to embrace the whole, and he who is on a mountain is most commonly enveloped in clouds. In a storm that I observed on the Faulhorn, August 13, 1833, the lower clouds did not exist, and I was able to contemplate the phenomenon in all its grandeur. Many times during the day it had rained in the distance, and also near me. About seven in the evening, the mass of clouds, composed of several strata, presented a stormy appearance; their lower surface was at an elevation of about 3300 metres. Beyond the Diablerets in the Bas-Vallais, and Glaernisch in the canton of Glarus, nothing was visible. In this storm, which had an extent of more than 150 kilometres, the lightnings distinctly came from five different points: beyond the Diablerets, in the country of Vaud; to the right of Rinderhorn, perhaps in Simmenthal; in the direction of Berne; in that of Lucerne, behind the summit of Mount-Pilate, and in that of Schwitz. Many hours of observation convinced me that the electricities from these five points acted and reacted on each other. Out of at least one-third of the lightnings I drew the following conclusions: in the Pays de Vaud, a flash passed between two strata of clouds, for the lower stratum was very little illuminated; immediately afterwards, frequently

19 Vide Note s, Appendix, No. II.

at the same time, a zigzag flash, directed from above downwards, was seen in the neighbourhood of Rinderhorn. Some instants afterwards, electric lights shone above Berne, and a zigzag flash replied to them in the neighbourhood of Lucerne, and then in that of Schwitz. When the atmosphere became very dark, I also saw lightnings in the east; but they were too far off for me to be able to study them. It is evident that the first flash, which passed in the neighbourhood of Vaud, disturbed the equilibrium of the whole system: an observer, placed at Schwitz, would, therefore, have observed oscillations in the electrometer, the prime cause of which depended on a flash of lightning that had occurred in the neighbourhood of Lake Leman.

RETURN STROKE.—It is not uncommon to see two storms separated by a part of the sky almost entirely serene; a flash of lightning in the one is followed by a flash in the other. But, the earth being by induction in a state opposite to that of the cloud, the electricity may recombine with that of the cloud and produce a violent shock. Few events of this kind have made so much stir as that of July 19, 1785; of which Brydone has preserved all the details. After a fine morning, clouds appeared in the north-west at 11 o'clock; between 12 and 1 o'clock they exchanged lightnings, which were succeeded by claps of thunder after intervals of twenty or thirty seconds. Suddenly Brydone heard a loud detonation, as if several guns had been discharged at short intervals; this detonation was not preceded by any lightning. At a short distance from the house a man named Lauder, driving a cart of coals, was killed, as were also his horses; another carman, seated on a cart that followed the former, saw the horses fall without perceiving any lightning or feeling any shock. Several pieces of coal were scattered about. At five decimetres behind each of the cart-wheels, there was a hole in the earth five centimetres in diameter, the middle of which corresponded with the tire of the wheel. These details were confirmed by ocular testimony. In the neighbourhood, a shepherd, who was feeding his sheep, saw a lamb fall dead, and he himself felt a flame pass before his face. This accident preceded that of Lauder's by about a quarter of an hour, and it occurred about 2700 metres from the place where the latter was killed. A woman, who was cutting grass at a short distance, experienced a violent shock in the feet, and fell. Bell, the shepherd, asserts that he felt the ground of his garden tremble beneath his feet.

These phenomena result from the action of clouds on each other and on the earth; we may imitate them by

means of our machines. Electrise positively a conductor which I will call A, then arrange in its neighbourhood two small cylinders, B and C, placed one behind the other; if A and B are sufficiently far apart so that the spark cannot pass between them, B will be electrised by induction, the extremity nearer to A will be negative, the other positive, and a great number of sparks will pass from B to C. The same thing happens after a flash of lightning among several clouds, or between a cloud and the earth. Suppose that a large cloud electrises the earth by induction; if, at one of its extremities, a flash of lightning falls on the ground, the electricity of the opposite side, becoming free, unites with that of the earth. If the latter is moist, the passage is easily made; if not, there is a shock, because the earth conducts the electricity badly.

LINES OF THE SEPARATION OF STORMS.-In mountainous countries storms are generally more frequent and more violent than in the plain, because the winds produce a more rapid condensation of vapours; at the same time, the mountains oppose the movement of the clouds, and the electricity produced accumulates, as it were, in a single point. In some countries, mountains are actual lines of separation; often, indeed, a storm, formed in the plain or in a valley, is driven by the wind toward a chain of mountains; it stops there, whence it is afterwards drawn in another direction, and ramified in various ways. In every village, they will shew you the spot whence storms come; however, these operations must be subjected to scrutiny, which does not generally confirm the received opinion.

Mountains oppose to storms a purely mechanical obstacle; the storm is frequently drawn onward by a wind of moderate power, but the cold air below the cloud is precipitated with extreme rapidity, and passes away in all directions; while the warm air above moves in all directions toward the cloud. If the current of cold air meets a chain of mountains, it experiences a resistance, and arrests the movement of the cloud by reacting on it; if the direction of the progress of the cloud is perpendicular to that of the chain of mountains, it may remain clinging to them for a long time. If its direction makes an acute angle with that of the chain, it then follows it until it finds a valley whose direction is parallel to that which it had in the outset, it penetrates this valley, and then discharges itself. Isolated summits frequently separate storms into two parts, each of which pursues a separate course.

STORMS IN WINTER.-The formation of storms, as I have said already, is accompanied by a slow and continued

fall of the barometer, which proves that south winds prevail in the higher regions of the atmosphere. When the ascending current elevates the vapours to a great height, they are rapidly condensed. The wind draws them along with it; wherefore, storms in winter always come with a S.W. wind. In other cases, the storm is formed at the point where two opposite winds meet; it is then very violent, and the state of the atmosphere is troubled for a long time. When east winds have been constantly blowing, and the S.W. obtains the pre-eminence, the weather becomes rainy. During these storms, there is so much confusion between the aërial currents, that it requires the most attentive observation to unravel them. I have rarely been able to observe them so well as during the storm of July 21, 1834; east winds had been prevailing for a long time, the sky was serene, and the temperature high, but the barometer fell slowly. On the morning of the day of the storm there were interlaced cirri in a sky with a heavy aspect, especially in the west. The cirri gradually condensed, the brightness of the sun became paler and paler, whilst in the east the sky remained serene. After four o'clock its azure hue disappeared behind thick cirri; and bluish clouds, the precursors of a storm, ascended from the west toward the zenith, which they passed. Thunder was soon heard, rain and hail fell in abundance, but the clouds moved with variable velocity. It was for a long time evident that clouds were moving from west to east, although the principal mass moved toward the west. Vapours coming from the west mingled with air coming from the east. Those which were lower were condensed, but were always driven back toward the west. Although the entire phenomenon was a consequence of the contest of opposite winds, yet it was easy to see that the combat was more violent in certain spots, and was accompanied with a condensation of vapours and a developement of electricity. Indeed, the clouds moved with velocity in one point; they turned on themselves, and became more and more opaque. The lightnings in this point succeeded each other with rapidity, and occurred less frequently in the other parts of the sky. This phenomenon was soon produced in another part of the sky, very distant from the former, and the lightnings ceased in the first point.

Thus it is that all the winter storms are formed, when two opposite winds contend together, and especially when a storm arriving, from the west, is driven back by an east wind. At the moment when the storm breaks out, the barometer generally begins to rise. But storms are seldom

very violent in our countries, for the air is not sufficiently charged with vapours for a notable quantity of electricity to be developed. They are very common in the neighbourhood of the coasts, where the temperature in winter is higher, and the evaporation more abundant than in the interior of the country; but their duration is not so long, for the electricity produced is soon exhausted, and equilibrium is immediately re-established.

LIGHTNINGS WITHOUT THUNDER.- When a storm is situated below the horizon, we observe in the evening, and during the night, very brilliant flashes of lightning, while no thunder is heard, because the storm is too far distant from the observer for the noise of the thunder to be able to reach his ear. But, when the lightnings attain an angular height of 20°, it may sometimes happen that the thunder is scarcely heard. This is particularly the case when they are very high in the atmosphere; for then the sound produced in a highly rarified air is weakened, more and more as it traverses strata of greater density.

On a serene evening we often see after sunset intermittent lights that illumine a great portion of the sky; these are called heat-lightnings. They are observed within the tropics, as well as with us. At Demarara, they occur at the commencement of the rainy season, for then it is that storms are very common among the mountains in the interior, whilst the sky is serene all along the coast. We regard these lights as reflections of the lightnings of distant storms. Every one may convince himself that lightnings are reflected through the air with great intensity on a dark night. When a storm is in the west, and the rest of the sky remains serene, we have only to turn our back to the storm to see the lightning reflected in the east part of the heavens; and yet, in this case, the conditions for reflection are far less favourable than in the preceding example.

We may in this way perceive storms at enormous distances; but, as the observer has not always the opportunity of obtaining evidence of the existence of these storms, it has followed that various hypotheses have been given in order to explain these lights. Some have thought them a phosphorescence of the atmosphere; others have admitted electric sparks in a serene sky. But we commit here the same fault as for storms; they are observed at the moment they break forth, and all that precede them is neglected. In almost all cases when I have observed violent heat-lightning, the sky has been dull throughout the day, interlacing cirri have been perceived here and there; every thing has