Passing over this theory as neither consistent with the known laws regulating the motions of elastic fluids, nor sufficient even if the consideration of those laws were neglected, we come to the theory suggested by Captain Maury-a theory deserving of much more attentive consideration. I shall quote his own words, as the fairest method of presenting his theory; after stating the observed difference of barometric pressure in the two hemispheres, and mentioning the expulsion of air from the northern hemisphere as the cause of this difference, he writes:-'To explain the great and grand phenomena of nature, by illustrations drawn from the puny contrivances of human device, is often a feeble resort, but nevertheless we may, in order to explain this expulsion of air from the watery south, where all is sea, be pardoned for the homely reference. We all know, that, as the steam or vapour begins to form in the tea-kettle, it expels air thence, and itself occupies the space which the air occupied. If still more heat be applied', as to the boiler of a steam-engine, the air will be entirely expelled, and we have nothing but steam above the water in the boiler. Now at the south over this great waste of circumfluent waters, we do not have as much heat for evaporation as in the boiler or the teakettle; but, as far as it goes, it forms vapour, which has 'proportionately precisely the same tendency that the vapour in the tea-kettle has to drive off the air above, and occupy the space it held. Nor is this all. This austral vapour, rising up, is cooled and condensed. Thus a vast amount of heat is liberated in the upper regions, which goes to heat the air there, expand it, and thus, by altering the level, causes it to flow off.'

The theory thus divides itself into two parts : we have first the expulsive effects due to the vapour raised from southern oceans; and, secondly, the expansive effects due to the liberation of heat as the vapour is condensed. Now I would, in the first place, submit that we cannot assign to the second cause the effects here considered. The amount of heat liberated as the vapours rising from southern ocean are condensed is undoubtedly great, but it cannot be more than the equivalent of the amount of heat rendered latent as the vapours are formed, and therefore the expansive effects due to the liberation of heat cannot be greater than the contrary effects due to the prior imprisonment of heat. It is quite true, and has been accepted as the undoubted explanation of many climatic effects, that if vapour be raised in one place and condensed over another, then the temperature of the air over the latter place is raised. But when we have to consider a phenomenon extending over a zone twenty or thirty degrees in width, we cannot argue in this manner. Nay, it is necessary to the force of Maury's second cause that the condensation of vapour should take place over the very zone in which the vaporisation is proceeding. To assign similar effects to both processes, is to require that the winding up and the loosening of the spring should take place in the same direction.

Whatever effects, then, are due to the constant evaporation going on in the southern hemisphere, must not be derived from changes of temperature. So far as these are effective at all, they must depend on the excess of evaporation over condensation (since the excess cannot possibly lie the other way), and therefore represent diminution of heat or increase of pressure, the contrary effect to that we have to account for. We have, therefore, only to consider the first cause mentioned by Maury; that is the expulsive effects due to the formation of aqueous vapour.

At first sight, this process of expulsion appears simple enough, and seems further to coincide with many well-known phenomena. The theory supposes that over a wide zone of the southern hemisphere aqueous vapour is continually rising; that as it rises it displaces in part the heavier air over these regions; and that equilibrium being thus disturbed, the excess of air flows off continually towards the- equator. Now we know that the prevailing surface-winds over that zone of the southern hemisphere in which the barometer exhibits the peculiarity we are considering, blow from the equator; that is, they tend to sweep the lower strata of the atmosphere towards the south pole. They therefore tend to increase the quantity of humid air in high southern latitudes. We know also that the prevailing upper currents over the southern zone we are considering blow towards the equator. They tend, therefore, to carry the drier portion of the air towards the equator. All this seems in accordance with Maury's theory, and indeed if the prevailing upper and lower currents flowed in directions contrary to those indicated, the theory would fall at once.

Again, although we find no evidence in barometric pressure over the south tropical zone of that increase which Maury's theory would lead us to expect (since the surplus air is carried first to this zone), yet it might be argued that the surplus is so distributed as to appear in another way. It is evident that if the atmospheric envelope normally appertaining to the southern hemisphere were, through the effects of the causes assigned by Maury, increased in extent, this increase might show itself, not in an increase of pressure over the south tropical zone-that is, not in an increase of height there -but in the extension of the surplus atmosphere into the northern hemisphere. This would be shown by the extension of the southern trade-winds to or beyond the equator, so that the (so-called) equatorial zone of calms should lie north of the equator. As this is really the position occupied by the belt of calms, Maury's theory appears to gain additional force by the coincidence.

Another argument may be drawn from the analogy of the low barometer in moist weather. In fact, it is well known that Deluc explained this phenomenon in a manner precisely accordant with the views expressed by Maury.