It seems to me, that the manner of their appearance and procedure, favour the notion of a descent.
More or less of a cloud, as I am informed, always appears over the place first; then a spattering on the surface of the water below; and when this is advanced to a considerable degree, the spout emerges from the cloud, and descends, and that, if the causes are sufficient, down to the places of spattering, with a roaring in proportion to the quantity of the discharge; then it abates, or stops, sometimes more gradually, sometimes more suddenly.
I must observe a few things on these particulars, to shew how I think they agree with my hypothesis.
The preceding cloud over the place shews condensation, and, consequently, tendency downwards, which therefore must naturally prevent any ascent. Besides that, so far as I can learn, a whirlwind never comes under a cloud, but in a clear sky.
The spattering may be easily conceived to be caused by a stream of drops, falling with great force on the place, imagining the spout to begin so, when a sudden and great condensation happens in a contracted space, as the Ox-Eye on the coast of Guinea.
The spout appearing to descend from the cloud seems to be, by the stream of nearly contiguous drops bringing the air into consent, so as to carry down a quantity of the vapour of the cloud; and the pointed appearance it makes may be from the descending course being swiftest in the middle, or centre of the spout: this naturally drawing the outer parts inward, and the centre to a point; and that will appear foremost that moves swiftest. The phenomenon of retiring and advancing, I think may be accounted for, by supposing the progressive motion to exceed or not equal the consumption of the vapour by condensation. Or more plainly thus: the descending vapour which forms the apparent spout, if it be slow in its progress downwards, is condensed as fast as it advances, and so appears at a stand; when it is condensed faster than it advances, it appears to retire; and _vice versa_.
Its duration, and manner of ending, are as the causes, and may vary by several accidents.
The cloud itself may be so circumstanced as to stop it; as when, extending wide, it weighs down at a distance round about, while a small circle at the spout being exonerated by the discharge ascends and shuts up the passage. A new determination of wind may, perhaps, stop it too. Places liable to these appearances are very liable to frequent and sudden alterations of it.