Despite the apparent force of these arguments, and others that might be adduced, it will not be difficult, I think, to show that neither is Maury's theory consistent with known physical laws, nor (passing over this objection) is the theory sufficient to account for the grand phenomenon under consideration.

It is quite true that a volume of aqueous vapour weighs less than an equal volume of air; it is equally true that a volume of moist air weighs less than an equal volume of dry air at the same tension.- But water, quietly evaporating in the open air, does not displace the air, but penetrates into its interstices, according to the well-established law regulating the mixture of vapours. The aqueous vapour which thus intimately mixes itself with the air produces no effect whatever, either by its weight or by its elasticity, on the movements of the atmosphere. The experiments of Gay-Lussac, Dalton, and others, have long since proved that the actual effects of the quiet evaporation of water are those here described. It is on this account that Deluc's hypothesis in explanation of the fall of the barometer when the air is moist is now no longer accepted. It has been shown that the observed fall is not due to the moistness of the air, but to increase of temperature. Hot winds bring (in Europe) moist air, and thus moist air and a low barometer are found to be coexistent phenomena. But they are not in the relation of cause and effect. In fact, in New Holland, where hot winds bring dry air, we find the barometer low when the air is dry.

It follows from what has just been said of the manner in which aqueous vapour associates itself with air, that atmospheric pressure is increased instead of diminished by the process of quiet evaporation, since the weight of the vapour is added to that of the air. Therefore, all things being equal, we should expect to find the barometer higher in the southern or watery hemisphere than in the northern.

It might seem unnecessary to consider Maury's theory further, but as some doubts may still remain whether some process of the kind conceived by him may not take place,1 I proceed to consider the efficiency of such southern hemisphere, nor (2) do they prevail over the whole of that hemisphere.

1 In fact, Sir J. Herschel, in his work on Meteorology, assigns as a cause of the low barometric pressure near the equator, compared with that near the tropics, a process similar to that conceived by Maury, only depending on the excess of heat near the equator. I cannot but agree a process to account for the great phenomenon we are dealing with.

It must be remembered, in the first place, that the theory requires that there should be a greater volume of mixed air and vapour over the southern temperate zone than there is in the corresponding northern zone, otherwise there would not be that continual overflow towards the equator which is required by the theory. So far as it goes, this increment of volume implies an increment of weight. The increase of volume is more than compensated (in theory) by diminution of specific gravity, but it must be held in mind that the increase of volume has to be accounted for by the theory as well as the difference in barometric pressure.

Again, the theory requires that the upper regions of air should be dry, for it is the upper air that is carried towards the equator; and if this air were moist, we should no longer have the different proportions of moist and dry air which are required by the theory. We must have an aggregation of moist air in high southern latitudes, and of dry air towards the equator.

Again, we must call to mind that one-half of the northern hemisphere is covered by water, and a part of the southern hemisphere is not so covered, so that the effects suggested by Maury are (1) not peculiar to the with those metereologists who consider that the notion of any appreciable uplifting of the air by the rising vapour of water is a mistaken one. But whether it be so or not, it is evident that Herschel's view would require a regular increase of pressure from the equator to the antarctic pole, and therefore is opposed to Maury's explanation.

Lastly, we must remember that the process conceived by Maury must be wholly or principally a diurnal process, and so can only take place (on an average) over one half of the southern zone at any one time.

All these considerations tend to diminish very importantly the efficiency of the cause assigned by Maury, Let us, however, consider what is the maximum value that efficiency could have if all these circumstances were neglected. We shall see that even in this case, which assigns an efficiency at least three or four times as great as would be consistent with actual facts, we shall still find the cause assigned by Maury inadequate to the production of the phenomenon under consideration.

The greatest weight of aqueous vapour which is ever present in a given volume of air is equivalent to about one-sixtieth part of the weight of the air. Now, if we suppose the barometer at thirty inches, and the whole column of air above the barometer to be impregnated with air in the above-named proportion-a view very favourable to the theory, since the cold of the upper regions of air largely diminishes the proportionate weight of aqueous vapour - it is clear that one-sixtieth part - or half an inch - of the barometer's height is due to the presence of aqueous vapour. Now, at mean tensions the specific gravity of aqueous vapour is about three-fifths of the specific gravity of air, so that the proportion of one-sixtieth part of weight corresponds to a proportion of one-thirty-sixth part of volume; in other words, our column of air owes one-thirty-sixth part of its height to the presence of aqueous vapour. If we suppose this thirty-sixth part to flow off-not from the upper regions only, but in such a manner that one complete thirty-sixth part of the volume of the column should pass off - then, instead of standing at a height of thirty inches, the barometer would stand at a height of 29 1/6 inches, less by only one-third of an inch than the height of 29- 1/2 inches due to the dry air alone. Now we cannot, in accordance with Maury's theory, legitimately add the five-sixths of an inch of barometric pressure to the height of the barometer under a neighbouring column. For we have no evidence to show that the air assumed to be expelled from the southern temperate zone is heaped over the southern tropical zone; on the contrary, we have a barometer in the latter zone not quite so high even as the barometer in the corresponding northern zone. Therefore if air is expelled in the manner supposed by Maury, it must be distributed over a very much greater portion of the globe's surface than it had been expelled from. Hence, returning to our imaginary column of air, but a small fraction of the five-sixths of an inch due to overflow must be added to the barometer under a neighbouring air-column. The latter barometer originally at 29 1/2 may be fairly assumed to rise at most to about 29 5/6 inches. We have, then, a difference of 29 5/6-29 1/6 inches, or two-thirds of an inch; so that despite all the opposing considerations we have neglected, we still have a difference less by one-third than that for which we have to account; and. indeed, so far as the comparison between the northern and southern temperate zones is concerned (and this is the true question at issue), we are only entitled to consider the third part of an inch lost by overflow, as the true measure of the efficiency of this cause.