quisite to form the alkalies. Potassa contains th, and soda 4th of the weight of oxygen; ths of the potassa and ths of the soda, will be equal in weight to five times the oxygen, the weight of which has just been obtained. But potassa contains ths of its weight of potassium, and soda ths of its weight of sodium, and the united weights of these metals has also been obtained as above. Taking this latter sum from the former we find that one half the weight of the soda is equal to five times the oxygen less the weight of the metals. One sixth of the weight of the potassa is equal to the weight of the oxygen, less one fourth the weight of the soda.

Resuming the example before employed, in which the weight of the mixed chlorides is 24.5 grains, and of the chloride of silver 54.75 grains. The chlorine in 54.75 grains of chloride of silver is 13.5 grains. Subtracting this from 24.5 grains, we have 11 for the weight of the metallic bases of the alkalies. The oxygen equivalent to 13.5 grains of chlorine is 3 grains.

By the rule, 5x3 or 15, less 11, the weight of the metals, is half the weight of the soda; the soda is, therefore, 8 grains. The weight of oxygen, 3 grains, less one fourth of the soda, 2 grains, gives one sixth of the weight of the potassa. The potassa is therefore 6 grains.

It is readily seen that the general principle to which these results refer themselves, may be used with great effect in avoiding a difficult step in chemical analysis, by the substitution of a less direct, but more simple one, aided by easy calculations.

This problem is only a particular case of a general one, well deserving the attention of analytical chemists.

Essays on Meteorology. By JAMES P. ESPY, Mem. Am. Philos. Soc., &c. No. II.

Theory of Hail.

On the principles established in the first essay, the spout, which is nothing but visible condensed vapour, may sometimes not reach entirely down to the surface of the earth, or sea, when the dew-point is too low for such an effect; in this case, it will appear as an inverted cone, reaching down from a cloud already formed.

It may here be observed, that a spout will always begin to be formed at a considerable elevation above the surface of the earth, because the vapour will always begin to condense there, from a law too well understood by meteorologists to need elucidation here. When, however, it begins to condense, it begins, also, by its diminished specific gravity, to rise, and then, if all circumstances are favourable, the cloud will increase as it ascends, and finally become of so great perpendicular depth, that, by its less specific gravity, the air below it, and contiguous to it, in consequence of diminished pressure, will so expand, and cool by expansion, as to condense the vapour in it; and then the air below this again, will, in its turn, experience a greater and greater expansion and refrigeration, and, consequently, condensation of vapour; and this process may go on so rapidly, that the visible cone may appear to descend to the surface of the sea, or earth, from the place where it first appears, in about one or two seconds.

The terms here employed must not be understood to mean that the vapour, or cloud, actually descends; it appears, to the spectator, to descend, but this is an optical deception, arising from new portions of invisible va

pour constantly becoming condensed, while, all the time, the individual particles are in rapid motion upwards.

For the sake of illustrating the principle, without aiming at absolute numerical accuracy, let us suppose a dew-point ten degrees below the temperature of the air. Now, a diminished pressure of one pound to the square inch, will cause a fall of temperature of only six degrees, so that, in this case, the visible cone would not reach down to the surface of the earth, or sea, and the air would have to ascend a little more than 400 yards, before condensation would commence. I say more than 400, because, though the temperature sinks one degree for every hundred yards of elevation, the dew-point also sinks a little from the expansion of the air, and the same quantity of vapour occupying a larger space. But, if the dew-point in the above case had been only six degrees below the temperature of the air, then the spout, or visible cone of vapour, would have reached the earth.

Now, it is highly probable that a spout, in passing over the surface of the earth, would meet with slight variations in the dew-point, and, if so, it would rise as the dew-point fell, and fall as the dew-point rose; and thus the theoretical deductions correspond exactly with the facts.

Again, the direction of the two spouts, as also of the great storm with two veins of hail, mentioned in a former essay, was from the south-west to the north-east; and Pouillet says, that a large majority of these storms are known to move in this direction. I presume he means those which occur in France. Now, it is manifest that these storms, according to the theory, must move in the direction of the upper current into which they may ascend, for the top of the vortex will lean in that direction; and, as theory demonstrates, and observation agrees with that demonstration, that the uppermost current of air in the temperate zone moves constantly from the south of west, towards the north of east, this will satisfactorily account for the general tendency of these storms in that direction, all over the northern temperate zone, or, at least, above lat. 30°. For, from that latitude, down to the tropic of Cancer, the uppermost current of air moves nearly towards the north, and, within the tropic, it moves towards the north-west; and so the theory would lead us to presume that, in these regions, the storms will be found to move in these directions. Such is shown to be the fact by Mr. Redfield, as to all great storms which travel any considerable distance in the West Indies. And in the Philosophical Transactions, Lathrop's Abridgement, vol. 2, p. 107, it is said that hurricanes in the West Indies begin from the north-west, and terminate with a south-east wind.

It is quite reasonable to suppose that these spouts sometimes meet with a middle current, moving in a different direction from the uppermost, which will account for the exceptions to the general rule; for the spouts will, in such case, certainly lean, and, of course, move, in the direction of the middle current.

These three storms all occurred in the day, and two of them in the afternoon; and M. Pouillet says that many more occur in the day than in the night. Now, this is precisely what the theory would lead us to suppose, and the explanation of this fact affords me an opportunity of explaining the very commencement of those spouts which occur during the day. The sun, during the day, and especially in the afternoon, heats up the surface of the earth, and the air in contact with that surface, many degrees above the air, a few hundred feet above the earth. This heated air below, and cold air above, will form an unstable equilibrium, and a very slight agitation will cause to be formed upward vortices of the light air below, and

downward vortices of the dense air above. Now, if the dew-point is not more than ten degrees below the temperature of the air in contact with the soil, the air of the upward vortex will not ascend much above one thousand yards, before the refrigeration, caused by expansion, will cause a beginning of condensation of vapour; and the moment this occurs, the velocity of upward motion is rapidly increased, from the expanding effect caused by the evolution of latent caloric, as before explained.

If the dew-point of the air at this elevation should be almost identical with its temperature, the cloud of the upward vortex will go on increasing in size and perpendicular height, until the air immediately below it, being pressed downwards with less and less weight as the cloud above increases in height and levity, will, by expanding more than the air which preceded it in the vortex, be cooled down to the point of deposition, before it reaches the elevation of one thousand yards. And if, in this case, the vortex should rise to a height sufficient to produce a diminution of pressure under it of one pound to the square inch, the cone of visible vapour, or cloud, will reach down to a point four hundred yards from the earth's surface. And, in general, the nearer the dew-point is to the temperature of the air, the lower will the visible spout descend; so that, if they had been assumed only six degrees apart, in the above case, the apex of the spout would have descended to the earth. And, if they had been assumed still nearer together, the spout would not only have descended to the earth, but it would have been of some considerable size there. Thus we find that this mode of calculation not only enables us to account for the more frequent appearance of these spouts in the day than in the night, but also to assign a reasonable, hypothetical cause, why these spouts, or storms, are sometimes broad, and sometimes narrow, and sometimes even do not reach down to the surface of the earth.

It is known, also, that spouts, and violent storms, are always preceded by calms. This fact, also, is easily explained by the theory. For, in the first place, it is known that a calm favours the production of a high dewpoint, which is an essential ingredient in these storms; and, second, a vortex of great strength cannot be formed, unless it can rise nearly perpendicular to a great elevation, which never can happen where there is a strong wind. This will readily be admitted, when it is considered that the wind is always stronger at some distance above the surface of the earth, than at the surface itself; and, therefore, no vortex of any great height, in these circumstances, could be formed, for the upper part would be blown away from the lower.

I have frequently seen those large columnar clouds, which form in mid air during a warm summer's day, have their tops blown off by an upper current, when the lower air was almost still, and thus a vortex of great strength prevented from forming. That these clouds are actually formed by rising vortices, occasioned by the disturbance of the equilibrium of the air during the day, is rendered almost certain by the following facts. First. When the supply is cut off in the evening, by the air near the surface of the earth becoming cold, these clouds cease to form, and not unfrequently disappear, and a day with many clouds is followed by a cloudless night. On the supposition of upward vortices, this phenomenon is very simple and natural; but on any other supposition, it is utterly paradoxical, (especially when it is now known that depression of temperature is the only cause of the condensation of aqueous vapour,) how clouds can be formed under a

meridian sun, which will be dissipated under the refrigerating influences of a nocturnal sky.

Second. I once saw, during a profound calm, those columnar clouds, in all parts of the heavens, appear to be coming slowly towards me, which I think can only be accounted for by supposing that they were all rising perpendicularly. These clouds, however, were gradually dissipated after they had increased to a considerable size, which proves beyond doubt that they were surrounded by air, at that elevation, whose dew-point was below the temperature of the air; and it may be added, that this is one of the cases where a spout cannot be formed, for the ascending air of the vortex will always, more or less, be mingled with the air through which it passes.

Again, nothing but an upward or a downward vortex will account for the well known fact, that, in these storms, the clouds are frequently seen to rush together with great rapidity, for some time, without overlapping each other, and crossing, which proves that they are on the same horizontal plane, and so demonstrates the existence of a vortex. I need hardly add, that other phenomena show that the motion, after meeting, is upwards, and not downwards.

Clouds have also frequently been seen to ascend, by spectators on mountains, and æronauts have found their temperature much higher than the surrounding air. Thus, it is demonstrated, beyond all doubt, that there is an upward current in these storms, whether the latent caloric given out by the condensation of the vapour, is the cause of that current, or not. And, as no fact in physics is better established than that precipitation will in-. stantly take place, if saturated air is suddenly rarefied, we are sure, also, that this upward motion of saturated air will, by causing expansion, produce precipitation.

I had long been desirous to ascertain, by actual observation, how high these vortices carry the condensed vapour, or cloud, into the upper air, and a fine opportunity was afforded me, on the 31st of July, 1834. This morning, says my journal of that day, it begain to rain early, with the wind and lower clouds north-east, middle clouds south, and upper clouds west. Several showers occurred during the morning, and the wind gradually shifted round to the south-east. At 5 o'clock, P. M., a most violent shower, which lasted about fifteen minutes, came up from the north-west, and at the moment of the hardest rain, the lower wind being strong from the north-west, the lowest visible clouds in a south-east direction, were seen to move with great velocity in the opposite direction, towards the north-west.

As soon, however, as the shower passed off to the south-east, the lower clouds changed their course, and followed the shower towards the southeast, exposing to view, near the zenith, a most magnificent columnar cloud, with its summit and western side as white as snow, being exposed to a western sun, in a perfectly clear sky. This cloud seemed nearly stationary for some time in its upper snowy part, while the scattering clouds in its lower parts were seen to rush under it, towards the south-east, with great velocity. The principal cloud moved slowly and majestically towards the ESE.; the sun's rays gradually climbing up this mountain of snow, fourteen minutes after he set, his last beams ceased to illuminate its summit.

The altitude of this summit being taken by a sextant, was found to be nine and a half degrees. The line which bounded light and darkness, as it rose up the sides of this columnar cloud, was well defined, the western horizon being entirely free from clouds, so that I think I could not be mistaken one-quarter of a minute in the time when the sun's rays ceased to

shine on the top of the cloud. Calculating from these data, I find the cloud reached to the amazing height of eleven miles, and that it travelled eastsouth-east, with a velocity of about forty-eight miles an hour.

A much more violent storm than this had occurred at Wilmington, (Del.) about twenty-eight miles south-west of Philadelphia, two days before this, as appears from Dr. Gibbon's Journal. He says it commenced raining with a thunder gust, at 5 o'clock in the morning, and poured down in torrents till half-past 7, when it ceased. In this short time, two and a half hours, 5 inches of water fell. This rain, he says, did not extend further than ten or fifteen miles from Wilmington, in any direction, except, perhaps, in an easterly course, in New Jersey.

On that evening, my journal says, "The upper clouds from the WSW. were tinged with pink, thirty-one minutes after seven o'clock, mean time. These clouds, being in the zenith, must have been the astonishing height of fourteen miles.

The angular velocity of one of these upper clouds was taken; it was found to rise from 25° to 32°, in two and a half minutes. Its absolute velocity, therefore, at this great height, was about two miles a minute. This great velocity is not at all inconsistent with the velocity with which storms are known generally to travel towards the north-east, in our latitude, even on supposition that this direction is given to the upward vortices of these storms, by this uppermost current, as explained before; for the inertia of the air in the vortices must be overcome, and, therefore, the velocity of the storm, at least the hinder part of it, cannot be so great as the velocity of this uppermost current.

There are many well authenticated accounts of showers of dust, and bloody, or, as I imagine, reddish rain, having fallen, and also of hail, with earthy or stony matter contained in the stones, and some with green leaves of forest trees; all these facts are mere corollaries from the theory. Prof. Zimmermann analyzed the sediment of some red rain which fell on the 3d of May, 1821, near Geissen, and found it to contain chrome, oxide of iron, silex, lime, carbon, and a trace of magnesia, but no nickel. On the 13th of August, 1824, in the city of Mendoza, in Buenos Ayres, dust fell from a black cloud, and at the same time, in another place, distant forty leagues, the same phenomenon occurred.

In Persia, near Mount Ararat, there fell, in the month of April, 1827, a shower of seeds, which, in some places, covered the earth to the depth of six inches. The sheep ate of it, and men made a tolerable bread of it. The French ambassador in Russia obtained some specimens of this grain, and sent them to Paris, where they were analyzed and examined by MM. Desfontaines and Thenard, and determined to be lichens of the genus Lecidea. Now, as neither leaves of forest trees, nor seeds of lichens, can grow in the upper regions of the atmosphere, or be precipitated to the earth from any other planet, if these accounts are believed, and M. Pouillet doubts not the truth of them, then the existence of upward vortices, however these vortices may be formed, is established.-[Pouillet, p. 770.

The theory will also account for the water spout. Indeed, a spout at sea, and a spout on land, are identically the same thing, and many have been known to pass from water to land, exhibiting the same appearance in both situations. To show their identity, I will copy from Silliman's Journal, vol. xiv., p. 171, an account of a water spout seen off the coast of Florida, in the spring of 1826, by Benjamin Lincoln, M. D., of Boston. "April 5th.-At 6 o'clock, A. M., an order was heard from the deck to VOL. XVII.-No. 5.-MAY, 1836.

30