water was probably owing to our suddenly passing from that current into the waters of our own climate.

On the 14th of August the following experiment was made. The weather being perfectly calm, an empty bottle, corked very tight, was sent down 20 fathoms, and it was drawn up still empty. It was then sent down again 35 fathoms, when the weight of the water having forced in the cork, it was drawn up full; the water it contained was immediately tried by the thermometer, and found to be 70, which was six degrees colder than at the surface; the lead and bottle were visible, but not very distinctly so, at the depth of 12 fathoms; but, when only 7 fathoms deep, they were perfectly seen from the ship. This experiment was thus repeated September 11th, when we were in soundings of 18 fathoms. A keg was previously prepared with a valve at each end, one opening inward, the other outward; this was sent to the bottom in expectation that by the valves being both open when going down, and both shut when coming up, it would keep within it the water received at bottom. The upper valve performed its office well, but the under one did not shut quite close, so that much of the water was lost in hauling it up the ship's side. As the water in the keg's passage upwards could not enter at the top, it was concluded that what water remained in it was of that near the ground; and, on trying this by the thermometer, it was found to be at 58, which was 12 degrees colder than at the surface.

[This last Journal was obligingly kept for me by Mr. J. Williams, my fellow-passenger in the London Packet, who made all the experiments with great exactness.]*

*The chart given in this edition has been constructed with a view to give a more comprehensive idea of the course of the Gulf Stream. Volney very plausibly suggests, that the earth, deposited by the Gulf Stream southeast of Newfoundland, has formed the great banks; and that the accumu lation there has given the stream a new or more eastwardly direction. This chart also serves to illustrate the long-received ideas of the progress of the shoals of fish. May not the glutinous matter seen on the water, and which all persons who have been across the line must have noticed, be another cause of the phenomena of fish shoals. May they not come in search of the food, which the matter seen on the water in such abundance affords? The writer of this note has observed, that, on entering the trade winds, the seamen have judged of the change of wind approaching by the direction of the bonetta and other fish, which pass in shoals in the South Atlantic and southeastern Seas, in a direction opposite to the wind; and when not opposite to the prevailing wind, they conclude a change to be at hand from the direction towards which the fish go. The appearance of luminous floating matter at night is often followed by shoals of fish; the spawn or gluten, which the writer has had taken up in a bucket, has been often found as large as two inches diameter, and frequently induced an opinion that it was a species of maritime cocoon or egg of an animal. Fragments of irregular shaped gluten have been also often seen. An inquiry into the

TO JOHN INGENHOUSZ, AT VIENNA.*

On the Causes and Cure of Smoky Chimneys.

DEAR FRIEND,

At Sea, 28 August, 1785.

In one of your letters, a little before I left France, you desire me to give you in writing my thoughts upon the construction and use of chimneys, a subject you had sometimes heard me touch upon in conversation. I embrace willingly this leisure afforded by my present situation to comply with your request, as it will not only show my regard to the desires of a friend, but may at the same time be of some utility to others; the doctrine of chimneys appearing not to be as yet generally well understood, and mistakes respecting them being attended with constant inconvenience, if not remedied, and with fruitless expense, if the true remedies are mistaken.

Those, who would be acquainted with this subject, should begin by considering on what principle smoke ascends in any chimney. At first, many are apt to think, that smoke is in its nature and of itself specifically lighter than air, and rises in it for the same reason that cork rises in water. These see no cause why smoke

periodical appearance of these luminous substances on voyages to the southward, and remarks on the usual direction of the shoals of bonetta and other fish, might perhaps lead to very interesting discoveries. It might be assumed as a question worthy of examination, whether the direction of shoals of fish is not towards those points from which periodical winds or currents move the waters; and if the shoals of fish, which move from the north poles, and by the British Isles across the Atlantic, are not led by their instinct in search of these periodical supplies of food; and if the deposits made by the Gulf Stream on the banks of Newfoundland are not the true cause of the great abundance of fish found there. - DUANE.

This letter, which has been published in a separate pamphlet, both in England and America, first appeared in the Transactions of the American Philosophical Society, in which it was read October 21st, 1785. -EDITOR.

VOL. VI.

64

Q Q

should not rise in the chimney, though the room be ever so close. Others think there is a power in chimneys to draw up the smoke, and that there are different forms of chimneys, which afford more or less of this power. These amuse themselves with searching for the best form. The equal dimensions of a funnel in its whole length are not thought artificial enough, and it is made, for fancied reasons, sometimes tapering and narrowing from below upwards, and sometimes the contrary, &c. &c. A simple experiment or two may serve to give more correct ideas. Having lit a pipe of tobacco, plunge the stem to the bottom of a decanter half filled with cold water; then putting a rag over the bowl, blow through it and make the smoke descend in the stem of the pipe, from the end of which it will rise in bubbles through the water; and, being thus cooled, will not afterwards rise to go out through the neck of the decanter, but remain spreading itself and resting on the surface of the water. This shows that smoke is really heavier than air, and that it is carried upwards only when attached to, or acted upon, by air that is heated, and thereby rarefied and rendered specifically lighter than the air in its neighbourhood.

Smoke being rarely seen but in company with heated air, and its upward motion being visible, though that of the rarefied air that drives it is not so, has naturally given rise to the error.

I need not explain to you, my learned friend, what is meant by rarefied air; but, if you make the public use you propose of this letter, it may fall into the hands of some who are unacquainted with the term and with the thing. These then may be told, that air is a fluid which has weight as well as others, though about eight hundred times lighter than water. That heat makes the particles of air recede from each other and take up