whole, thus condensed, falls at once (which is what I suppose is vulgarly called the breaking of the spout); and in the interval between this period and that of the next set of particles being ready to unite, the spout shuts up. So that, if this reasoning is just, these phenomena agree with my hypothesis.

The usual temper of the air, at the time of their appearance, if I have a right information, is for me too; it being then pretty cool for the season and climate. And this is worth remark, because cool air is weighty, and will not ascend; besides, when the air grows cool, it shows, that the upper region descends, and conveys this temper down; and, when the tempers are equal, no whirlwind can take place. But spouts have been known, when the lower region has been really cold. Gordon's spout in the Downs is an instance of this, (See Philosophical Transactions), where the upper region was probably not at all cooler, if so cold as the lower; it was a cold day, in the month of March; hail followed, but not snow; and it is observable, that not so much as hail follows or accompanies them in moderate seasons or climes, when and where they are most frequent. However, it is not improbable, that just about the place of descent may be cooler than the neighbouring parts, and so favor the wonderful celerity of condensation. But, after all, should we allow the under region to be ever so much the hottest, and a whirlwind to take place in it; suppose then the sea-water to ascend, it would certainly cool the spout; and then query, whether it would not very much, if not wholly, obstruct its progress.

It commonly rains when spouts disappear, if it did not before, which it frequently does not, by the best accounts I have had; but the cloud in creases much faster after they disappear, and it soon rains. The first shows the spout to be a contracted rain, instead of the diffused one that follows; and the latter, that the cloud

was not formed by ascending water, for then it would have ceased growing when the spout vanished.

However, it seems that spouts have sometimes appeared after it began to rain; but this is one way a proof of my hypothesis, viz. as whirlwinds do not come under a cloud.

I forgot to mention, that the increase of cloud, while the spout subsists, is no argument of an ascent of water by the spout; since thunder-clouds sometimes increase greatly while it rains very hard.

Divers effects of spouts seem not so well accounted for any other way as by descent.

The bush round the feet of them seems to be a great spray of water made by the violence of descent, like that in great falls of water from high precipices.

The great roar, like some vast inland falls, is so different from the roar of whirlwinds, by all accounts, as to be no ways compatible.

The throwing things from it with great force, instead of carrying them up into the air, is another difference.

There seems some probability that the sailors' traditionary belief, that spouts may break in their decks, and so destroy vessels, might originate from some facts of that sort in former times. This danger is apparent on my hypothesis, but it seems not so on the other; and my reason for it is, that the whole column of a spout from the sea to the clouds cannot, in a natural way, even upon the largest supposition, support more than about three feet water, and, from truly supposable causes, not above one foot, as may appear more plainly by and by. Supposing now the largest of these quantities to rise, it must be disseminated into drops, from the surface of the sea to the regio of the clouds, or higher; for this reason, it is quite unlikely to be collected into masses, or a body, upon its falling; but

would descend in progression, according to the several degrees of altitude the different portions had arrived at when it received this new determination.

Now that there cannot more rise upon the common hypothesis, than I have mentioned, may appear probable, if we attend to the only efficient cause in supposed ascending spouts, viz. whirlwinds.

We know that the rarefaction of the lower, and the condensation of the upper region of air, are the only natural causes of whirlwinds. Let us, then, suppose the former as hot as their greatest summer heat in England, and the latter as cold as the extent of their winter. These extremes have been found there to alter the weight of the air one tenth, which is equal to a little more than three feet water. Were this case possible, and a whirlwind take place in it, it might act with a force equal to the mentioned difference. But, as this is the whole strength, so much water could not rise; therefore, to allow it due motion upwards, we must abate at least one fourth part, perhaps more, to give it such a swift ascension as some think usual. But here several difficulties occur; at least they are so to me; as, whether this quantity would render the spout opake, since it is plain that in drops it could not do so; how, or by what means, it may be reduced small enough; or, if the water be not reduced into vapor, what will suspend it in the region of the clouds when exonerated there? And, if vaporized while ascending, how can it be dangerous by what they call the breaking? For it is difficult to conceive how a condensative power should instantaneously take place of a rarefying and disseminating one.

The sudden fall of the spout, or rather the sudden ceasing of it, I accounted for, in my way, before. But it seems necessary to mention something I then forgot. Should it be said to do so (that is, to fall), because all

the lower rarefied air is ascended, whence the whirlwind must cease, and its burden drop; I cannot agree to this, unless the air be observed on a sudden to have grown much colder, which I cannot learn has been the case. Or, should it be supposed that the spout was, on a sudden, obstructed at the top, and this the cause of the fall, however plausible this might appear, yet no more water would fall than what was at the same time contained in the column, which is often, by many and satisfactory accounts to me, again far from being the

case.

We are, I think, sufficiently assured, that not only tons, but scores of hundreds of tons, descend in one spout; scores of tons more than can be contained in the trunk of it, should we suppose water to ascend.

But, after all, it does not appear, that the abovementioned different degrees of heat and cold concur in any region where spouts usually happen, nor, indeed, in any other.

Observations on the Meteorological Paper; by a Gentleman in Connecticut.*

READ AT THE ROYAL SOCIETY, NOVEMBER 4TH, 1756.

“AIR and water mutually attract each other," (saith Mr. Franklin,) "hence water will dissolve in air, as salt in water." I think that he hath demonstrated, that the supporting of salt in water is not owing to its superficies being increased, because "the specific gravity of salt is not altered by dividing of it, any more than that of lead, sixteen bullets of which, of an ounce each,

This gentleman was Mr. Jonathan Todd, who wrote the "Observations" in a letter to Mr. Eliot, by whom they were communicated to Franklin. - EDITOR

weigh as much in water as one of a pound." But yet, when this came to be applied to the supporting of water in air, I found an objection rising in my mind.

In the first place, I have always been loath to seek for any new hypothesis, or particular law of nature, to account for any thing that may be accounted for from the known, general, and universal law of nature; it being an argument of the infinite wisdom of the Author of the world, to effect so many things by one general law. Now I had thought that the rising and support of water in air, might be accounted for from the general law of gravitation, by only supposing the spaces occupied by the same quantity of water increased.

And with respect to the lead, I queried thus in my own mind; whether, if the superficies of a bullet of lead should be increased four or five fold by an internal vacuity, it would weigh the same in water as before. I mean, if a pound of lead should be formed into a hollow globe, empty within, whose superficies should be four or five times as big as that of the same lead when a solid lump, it would weigh as much in water as before. I supposed it would not. If this concavity was filled with water, perhaps it might; if with air, it would weigh at least as much less, as the difference between the weight of that included air and that of water.

Now, although this would do nothing to account for the dissolution of salt in water, the smallest lumps of salt being no more hollow spheres, or any thing of the like nature, than the greatest; yet, perhaps, it might account for water's rising and being supported in air. For you know that such hellow globules, or bubbles, abound upon the surface of the water, which, even by the breath of our mouths, we can cause to quit the water, and rise in the air.

These bubbles I used to suppose to be coats of water, containing within them air rarefied and expand