diffused, not only by radiations from their upper surfaces to the canopy of the sky above them; but by horizontal radiations, to the air surrounding their sides. A volume of the atmosphere therefore, between two neighbouring masses of ice, will necessarily have its middle portion of a higher temperature than that of either of the portions of air between it and the icebergs *; and the consequence of such an unequal distribution of temperature must be, to cause the cold air to mingle with that of a higher temperature, and thus to produce mist or fog. The density of such mist or fog will depend on the difference between the temperatures of the mingling volumes, and on the quantity of vapour contained in the air.

The elevation of those mists above the surface of the sea will also be regulated by that of the icebergs, near which they form; since the cooling influence of the frozen mass, by rapidly diminishing above its summit, will as rapidly destroy all tendency in the portion of the atmosphere above the level of the iceberg, to assume a condition favourable to the formation of mist; thus prescribing to the mist an elevation dependent on that of the iceberg near which it forms. Captain Ross accordingly remarks, in his Account of the Polar Voyage," that the fog was extremely thick on the surface of the sea, but at the mast-head, and at the top of the iceberg, it was perfectly clear." Captain Scoresby also, in his Paper on the Fogs of the Polar Seas, read before the Wernerian Society †, alludes to their definite elevation, and to the sky above them being perfectly clear.

It is possible, however, that two icebergs may be situated so near each other, that their reciprocal horizontal radiations will so cool the volume of air between them, as to reduce it to a tem

If the water in the vicinity of icebergs presents considerable inequalities of temperature, the air which reposes on it must be subject to like variations; and numerous examples of the former are to be met with in the Accounts of the Polar Voyages. Thus Captain Franklin remarks, "the temperature of the surface water was 35° when among the ice, 38° when just clear of it, and 41°.3 at two miles distant."

+ Edinburgh Philosophical Journal, Vol. VI.

perature nearly uniform; and thereby prevent the formation of mist. The cold volume of air so formed, may, however, pass from between the icy masses, and by mingling with the air reposing on the warmer water, beyond the icebergs, produce mist at a distance from them. Nor is it absolutely necessary that two icebergs should exist, in order to form mist; since the horizontal radiations of one, by cooling the portion of air in contact with it, will cause it to mingle with the warmer air beyond the last-mentioned stratum, and thus create fog. The density of a mist when formed under the latter conditions, will be of a more variable character, than when it is formed between adjacent icebergs.

The general diffusion of fogs over the Northern Seas may also be satisfactorily accounted for, from the scattered icebergs separating the water into detached portions; and thereby creating, in innumerable directions, volumes of air, possessed of unequal temperatures. The cold air near the icebergs being blended, therefore, with the warmer air reposing on the middle portions of the broken intervals of water, must form, between most of the floating masses of ice, visible volumes of vapour, having their density dependent on the relative difference of heat between the mingling portions of air, and on the degree of humidity possessed by each.

The cause here referred to, for the production of the Polar fogs, is also one likely to promote their continuance for a considerable time; it being known, that the sea continues for many months relatively warmer than the icebergs; and therefore capable, in conjunction with the constant radiation of the ice, of producing that almost constant succession of fogs which cover the Arctic Seas during the greater part of the summer months; and which increase, in so considerable a degree, the difficulties of Polar navigation.

Plymouth, July 19th, 1825.

ART. III.-On One of the Causes of the Movements of the Barometer, and of the South and West Winds. By Marshall Hall, M.D., F.R.S.E., &c. &c.

[Communicated by the Author.]

OUR inquiries into the nature and causes of the changes in the atmosphere will be greatly facilitated, by having first apprehended its more usual and quiescent state. The following observations will therefore be properly introduced by a very short description of what may be termed the natural state of the atmosphere.

If the surface of the globe were even and uniform, unchequered by mountains and valleys, and unintersected by rivers and seas, the waters of which continually evaporate and recondense, the atmosphere would remain in a comparatively tranquil state; and those movements which did take place in it, being only excited by the influences of the sun and moon, would be regular and periodical, and would consist chiefly in a constant wind from the north-east in the northern hemisphere, and in a diurnal oscillation or atmospheric tide. This movement of the atmosphere from the north-east is produced in the following manner. The sun, acting powerfully on the surface of earth within the torrid zone, heats and rarefies the superincumbent air, and causes it to rise into the upper region of the atmosphere; the place of this air is supplied from the poles, and thus, in the northern and southern latitudes, a north and south wind is produced respectively: as these portions of air have, however, little or no other motion except that in the direction towards the equator, and as the surface of the earth as we approach the equator has more and more motion from west to east, from its rotation on its axis, it follows that the air so brought from the poles, must, as it approaches the equator, have more and more of an apparent motion westward. These two motions being combined, the air will pursue the course of the diagonal, and to the inhabitants of the northern hemisphere of the globe, will constitute a perpetual north-east wind.

Such is the principal effect of the sun's influence on our atmosphere. A further influence of this celestial body, and of the

moon, would be to produce atmospheric tides, similar to those observed in the sea, varying the weight of different columns of the atmosphere by augmenting and diminishing their height.. These tides would doubtless, too, be measured by our barometers. They are obscured, in the existing state of things, by causes modifying the elasticity of the lower region of our atmosphere, and so cutting off, for a time, the influence of the height and weight of the superincumbent columns.

These regular movements of the atmosphere are disturbed chiefly by the irregularities on the earth's surface, by which the winds are diverted from their original and natural course,-being divided asunder by mountains, and made to clash violently together in the course of valleys and rivers. These effects are followed by others which arise out of the admixture of portions of air of different temperatures, and containing different portions of aqueous vapour; and it is not improbable that there are other phenomena of an electric nature, not hitherto fully appreciated and understood, which act silently and obscurely in general, though they are occasionally manifested in the phenomena of thunder and lightning.

The object of this paper is not to treat fully of the movements of the Barometer, but to point out ONE of the causes of these movements more particularly than has hitherto been done. This cause is the transition of vapour in the atmosphere from the transparent to the opaque and fluid form,-or the FORMATION of clouds, and of rain. We shall now, therefore, proceed to trace the effects of this change upon the elasticity of the atmosphere in the region of the clouds, and consequently upon the barometer.

It is well known how much the presence of water adds to the expansibility of the air in contact with it, by heat. In the experiments of Guyton and Duvernois, the object of which was to ascertain the comparative expansibilities of the different kinds of gas, the presence of an inappreciable quantity of moisture occasioned an error of such magnitude, that, of gases which are now known to expand alike, some appeared to dilate five times as much as others by the same application of heat.

* Journal de l'Ecole Polytechnique, Cap. 2. † Essai sur l'Hygrometrie. ~

The rate of the expansibility of air in contact with water, is also greater as the temperature rises; for, from the experiments of Saussure†, we learn that the power of air for retaining water in a transparent state increases in a geometric progression, whilst the temperature rises in an arithmetic progression only.

From these considerations we are enabled to conceive how great the contraction of a given volume of atmosphere must be by the mere admixture of different portions of air charged with vapour and of different temperatures, even without the precipitation of water, but especially if, whilst its temperature is diminished, its transparent vapour be condensed and withdrawn, in the form of clouds and rain.

To these considerations must be added, that of the vast extent of country over which the changes which take place in the atmosphere are observed to diffuse themselves,-generally more than one thousand miles,-that of the extent and height of that region of the atmosphere occupied by the clouds-upwards of three miles, and that of the immense quantity of water which must at some periods pass, in a very short space of time, from the state of transparent vapour to that of clouds or rain.

If we fully consider the extent and magnitude of these phenomena, we must conceive that the degree of contraction in volume in that part of the atmosphere which constitutes the region of the clouds, must on some occasions be great indeed. Let us now consider what must be the further effects of this contraction in the immediate and adjacent region:-a portion of the atmosphere being condensed and withdrawn, the elasticity of the remaining portion occupying the same space must be diminished; and, as further consequences, the barometer must fall, and the air of the adjacent regions must be attracted, and wind be produced ;-and these effects and phenomena will continue as long as the atmospheric process of the formation of clouds or rain continues. When the deposition of atmospheric moisture ceases, the equilibrium in the elasticity of the air is gradually restored, the barometer rises to its natural level, and the wind or influx of air from the adjacent regions subsides.