fore it again sets downwards for the day. This is the general course, with some occasional variation of hours, according to the length of days or changes of weather.

Now when the air of any vent is in this descending state, if the smoke issuing from a neighbouring vent happens to be carried over it by the wind, part will be drawn in and brought down into the room. The proper remedy then is, to close the opening of the chimney in the room by a board so fitted that little or no air can pass, whereby the currents abovementioned will be prevented. This board to remain during the summer, and when fires are not made in the chimney. Chimneys that have fires in them daily are not subject to this inconvenience, the walls of their vents being kept too warm to occasion any downward current during the hours between the going out of one fire and the kindling of another. And indeed, in summer, those vents that happen to go up close joined with the kitchen vent, are generally kept so warm by that as to be free from the downward current, and therefore free from what you call neighbour smoke.

The Philadelphia grate which you mention is a very good thing, if you could get one that is rightly made, and a workman skilful in putting them up. Those generally made and used here are much hurt by fancied improvements in their construction, and I cannot recommend them. As fuel with you is cheap and plenty, a saving in it is scarce an object. The sliding plates (of which I sent a model to Sir Alexander Dick) are, in my opinion, the most convenient for your purpose, as they keep a room sufficiently warm, are simple machines, easily fixed, and their management easily conceived and understood by servants.

I shall leave Europe with much greater regret, if I cannot first visit you and my other friends in Scotland.

I promise myself this happiness, but am not yet clear that I shall have time for it. Your kind invitation is extremely obliging.

With sincere esteem I am, my dear friend,
Yours most affectionately,

boat touched the muddy

You may remember, that when we were travelling together in Holland, you remarked, that the trackschuit in one of the stages went slower than usual, and inquired of the boatman, what might be the reason; who answered, that it had been a dry season, and the water in the canal was low. On being again asked if it was so low as that the bottom; he said, no, not so low as that, but so low as to make it harder for the horse to draw the boat. We neither of us at first could conceive, that, if there was water enough for the boat to swim clear of the bottom, its being deeper would make any difference. But, as the man affirmed it seriously as a thing well known among them, and as the punctuality required in their stages was likely to make such difference, if any there were, more readily observed by them, than by other watermen who did not pass so regularly and constantly backwards and forwards in the same track, I began to apprehend there might be something in it, and attempted to account for it from this consideration, that the boat, in proceeding along the canal, must in every

boat's length of her course, move out of her way a body of water equal in bulk to the room her bottom took up in the water; that the water so moved must pass on each side of her and under her bottom to get behind her; that, if the passage under her bottom was straitened by the shallows, more of that water must pass by her sides, and with a swifter motion, which would retard her, as moving the contrary way; or that, the water becoming lower behind the boat than before, she was pressed back by the weight of its difference in height, and her motion retarded by having that weight constantly to overcome. But, as it is often lost time to attempt accounting for uncertain facts, I determined to make an experiment of this, when I should have convenient time and opportunity.

After our return to England, as often as I happened to be on the Thames, I inquired of our watermen whether they were sensible of any difference in rowing over shallow or deep water. I found them all agreeing in the fact, that there was a very great difference, but they differed widely in expressing the quantity of the difference; some supposing it was equal to a mile in six, others to a mile in three, &c. As I did not recollect to have met with any mention of this matter in our philosophical books, and conceiving, that, if the difference should really be great, it might be an object of consideration in the many projects now on foot for digging new navigable canals in this island, I lately put my design of making the experiment in execution, in the following manner.

I provided a trough of planed boards fourteen feet long, six inches wide, and six inches deep, in the clear, filled with water within half an inch of the edge, to represent a canal. I had a loose board of nearly the same length and breadth, that, being put into the water, might be sunk to any depth, and fixed by little wedges

where I would choose to have it stay, in order to make different depths of water, leaving the surface at the same height with regard to the sides of the trough. I had a little boat in form of a lighter or boat of burthen, six inches long, two inches and a quarter wide, and one inch and a quarter deep. When swimming, it drew one inch water. To give motion to the boat, I fixed one end of a long silk thread to its bow, just even with the water's edge, the other end passed over a well made brass pulley, of about an inch diameter, turning freely on a small axis; and a shilling was the weight. Then placing the boat at one end of the trough, the weight would draw it through the water to the other.

Not having a watch that shows seconds, in order to measure the time taken up by the boat in passing from end to end, I counted as fast as I could count to ten repeatedly, keeping an account of the number of tens on my fingers. And, as much as possible to correct any little inequalities in my counting, I repeated the experiment a number of times at each depth of water, that I might take the medium. And the following are the results.

I made many other experiments, but the above are those in which I was most exact; and they serve sufficiently to show that the difference is considerable. Between the deepest and shallowest it appears to be somewhat more than one fifth. So that, supposing large canals and boats and depths of water to bear the same proportions, and that four men or horses would draw a boat in deep water four leagues in four hours, it would require five to draw the same boat in the same time as far in shallow water; or four would require five hours.

Whether this difference is of consequence enough to justify a greater expense in deepening canals, is a matter of calculation, which our ingenious engineers in that way will readily determine.

I cannot be of opinion with you that it is too late in life for you to learn to swim. The river near the bottom of your garden affords a most convenient place for the purpose. And as your new employment requires your being often on the water, of which you have such a dread, I think you would do well to make the trial; nothing being so likely to remove those apprehensions as the consciousness of an ability to swim to the shore, in case of an accident, or of supporting yourself in the water till a boat could come to take you up.