a dense elastic vapour, arising from burning bodies. As this vapour is extremely disagreeable to the senses; and often prejudicial to the health, mankind have fallen upon several contrivances to enjoy the benefit of fire, without being annoyed by smoke. The most universal of these contrivances is a tube leading from the chamber in which the fire is kindled, to the top of the building, through which the smoke ascends, and is dispersed into the atmosphere. These tubes are called chimneys; which, when constructed in a proper manner, carry off the smoke entirely; but, when improperly constructed, they carry off the smoke imperfectly, to the great annoyance of the inhabitants. As our masons at present seem to have a very imperfect knowledge of the manner in which chimneys ought to be built, we can hardly perform a more acceptable service to the public than to point out the manner in which they ought to be constructed so as to carry off the smoke entirely; as well as to explain the causes from which the defects so often complained of generally proceed, and the method of removing them.

Although we would naturally imagine, that the causes which occasion smoke in rooms are exceedingly various; yet, upon examination, it will be found that they may all be reduced to one of these three general heads, each of which will admit of several varieties.

I. To a fault in the form of the tube, or chimney itself.

II. To some fault in the other parts of the building, and a wrong position of the chimney with respect to these.

Or,

III. To an improper situation of the house with respect to external objects. And it is of the utmost consequence, in attempting a cure, accurately to distinguish from which of these defects the smoke proceeds, it will be necessary to point out with care the several phenomena which are peculiar to each.

1. Of smoke occasioned by a fault in the form of the chimney itself. But, before we proceed, it will be necessary to premise something with regard to the general cause of the ascent of smoke in chimneys.

The earth is everywhere surrounded with a great body of air, called the atmosphere. This air is an elastic fluid subjected to many particular laws, as hath been fully explained under the article Pneumatics; where it hath been sufficiently demonstrated, that, like other fluids, it hath a constant tendency to preserve an equilibrium in all its parts; so that, if at any time the weight of it at one place is diminished, the heavier air rushes from all sides towards that point, till the equilibrium be again restored. We therefore likewise saw, that heat was one of the most powerful means of disturbing this general equilibrium of the air, by expanding it to a great degree, and making the same quantity occupy a much greater space than before, and consequently become lighter. Hence it necessarily follows, that wherever a fire is kindled, the air immediately contiguous to it will be heated, and of consequence rarified and made light; which must ascend into the higher regions of the atmosphere, till it becomes of the same gravity with the air contiguous to it; while the denser cold air below rushes toward the point from which it departed, is there heated and rarefied in its turn, and ascends in the same manner, carrying the smoke or vapour arising from the burning body along with it. In this manner that constant suction of air towards every fire is produced, and from this cause proceeds the constant tendency of smoke to ascend upwards from the surface of the earth. But as the body of our atmosphere is often agitated with wind, &c. and as it is an elastic fluid, it endeavours to spread itself every way; from which causes the warm air would quickly be diffused among the cold air before it could arise to any considerable height; so that the smoke would always remain low, and be tossed about near the surface of the earth; all of which inconveniencies are avoided by confining this heated air in a tube, which prevents it from mixing with the external air, till it arrive at the height to which we desire it should ascend.

To render this still more clear, see Plate CLVI. where AB (fig. 1.) represents the tube of a chimney, having a fire at the bottom at A. It is obvious, that, in this situation, the air which was heated by the fire at A, will ascend directly upwards, without mixing with the external air, till it arrives at B, beyond which it will be at liberty to disperse in the atmosphere; and the more weighty air which presses in to supply its place can have no access to it but at the opening between A and E, where it also is heated by the fire, and in its turn ascends to the top of the chimney, thereby occasioning a constant stream of air to ascend up the chimney, which carries the smoke along with it. This is the manner in which fuliginous vapours are made to ascend in chimneys; and by attending to it, we may draw the following corollaries with regard to the construction of this useful part of our habitations.

1st. The higher the chimney, that is, the greater the distance between the fire-place and the top of the chimney, the greater will be the difference between the weight of the column of heated air in the tube, and another column of the atmosphere of the same diameter without the chimney, and consequently the air will enter with the greater force at the opening AE, and carry up the smoke more readily along with it: for as the warm air within the tube continues rarefied to a high degree till it issues from the top of the chimney, and is, in every part of its length, lighter than the same bulk of external air marked by the dotted line CD, it follows, that the longer these two columns of unequal gravity are, the greater must be the difference of their weight. Hence it is, that high chimneys (caeteris paribus) have a greater suction of air, and are less liable to vent ill, than low ones. A smoky chimney may therefore sometimes be cured by raising it higher. It is likewise obvious, that if any opening is made into the chimney, as at F, the air will enter with less force at E, and carry up the smoke with less velocity, and by that means be in danger of producing smoke in the room; for this opening, as it admits the fresh air into the tube, has nearly the same effect as shortening the tube so much would have.

2d. As the smoke is forced up the chimney merely by the rarefaction rarefaction of the air in consequence of heat, it is evident, that the more the air is heated, with the greater force (ce- teris paribus) will it ascend, because the difference between the weight of the external and internal air will be greater; and as the air will be the more heated the nearer it is made to pass by the fire in its entry into the chimney, it is evi- dent, that the smaller the opening at AE is, or, in other words, the lower the mantle of the chimney is, the air will be for- ced to pass the nearer the fire, and therefore be more rarefied, and ascend with the greater velocity; so that lowering the mantle of the chimney will often cure smoke.

But it is frequently inconvenient to have the mantle of the chimney too low. However, the same effect may often be produced by another contrivance. For as the fire-place is usually made wider than the length of the grate, a great deal of cool air passes at the two sides of the grate without being much heated. This greatly diminishes the suction of the chimney: but it may easily be prevented by building up the vacancies at each side of the grate, so as to allow no air to enter from below, except what comes immediately through, or before the fire. For this purpose, grates consisting of a neat hewed stone at each end, with a breast and bottom of iron fitted to them, as represented at fig. 2, are extremely convenient. But the aperture of the chimney is often not suddenly contracted above the mantle, but goes up tapering slowly, as in the same fig. 2. This structure allows a quan- tity of cool air to enter at the two corners of the mantle, and steal up the tube without coming near the fire. The most easy and effectual method of remedying this defect, is to place a sheet of milled iron within the mantle on each side, as low down as possible, making them slant a little up- wards towards the middle of the chimney; as at A, fig. 2, the mantle being represented by the dotted line. By this contrivance, the air, which enters at the side of the mantle, before it can ascend into the chimney, is forced to pass very near the fire, and of course is much rarified. The good ef- fects of this would be still more strongly felt, if one of these plates were placed a little lower than the other, and made so long, that the ends should cross each other, as at AB, fig. 2, by which means every particle of air that went up the chim- ney behoved to pass immediately above the fire. It is al- most unnecessary to observe, that these plates ought to be so contrived as to be taken out at pleasure to allow the chimney to be cleaned.

A chimney may not only be defective by having the man- tle too high, or by being too wide from side to side, but also by being too deep between the fore side and the back, as is often the case in very old houses. In this case, the distance between the fire and the mantle is so great, that much air passes up without being sufficiently rarified, as is represented at fig. 3. This may be sometimes cured by bringing the grate a little forward, which, by making the fire act more powerfully upon the mantle, rarifies the air more in its passage. But this can seldom produce the desired effect, and it often does harm: for when the grate is brought for- ward, there is a great vacancy left between it and the back of the chimney, so that the air passes under the grate, and ascends behind it very little rarified; so that, if the feet of grate are not very low, there will be as much lost in this way as will be gained in the other; and as there is not enough of heated air in the chimneys of this kind to make the va- pour ascend with rapidity, they are often choked with thick fuliginous vapours hanging in them, almost in equili- brium with the rest of the atmosphere, so that the least puff of wind beats them down the chimney, and pushes the smoke into the room; whereas, when it is far back, it is dri- ven down upon the hearth, and rises upwards again when the gust is over, and a great deal of it is caught within the mantle as it rises, which in the other case would have been dispersed through the room. When this is the case, the most effectual method of cure is, to bring the grate forward till the forepart of it is immediately under the inner edge of the mantle; then build up the vacancy at the back of it, the whole width of the fire-place from side to side, raising it per- pendicularly till it is as high as the back of the grate, and then bending it forward towards the mantle, as is repre- sented at fig. 4. When it is as high as the workman can reach, let it be suddenly turned backward again, sloping a little upward, as in the figure; then fit a sheet of milled iron to the inside of the mantle, making it slant a little up- ward toward the back part, at a small distance above the new erected masonry, and extending within a few inches of the back wall, as at A, fig. 4. By this construction all the air that enters into the chimney is made to pass immediately above the fire, between it and the heated iron, upon which the flame acts with the greater force, as the back of the fire-place is bent a little forward above the grate, and the heat is likewise reflected into the room with the greater force: at the same time, if the smoke is at any time beat down the chimney by a sudden gust of wind, it will be caught by the sheet of iron, and prevented from coming into the room. If the fire-place be very wide between the one side and the other, the new masonry may be carried quite up to the sheet of iron on each of the sides.

3d. As every fire requires a constant succession of fresh air, the tube for conveying this rarified air to the higher regions of the atmosphere must be of a sufficient size to contain the whole of it, and allow it a ready passage; other- wise a part of it will be forced to seek some other passage; by which means, the apartment in which the fire is placed will be constantly filled with smoke. Every chimney there- fore ought to have a degree of wideness sufficient to carry off the whole of the smoke arising from the fire usually burnt in it, otherwise the apartment will be almost continually filled with smoke.—This is a fault more common at pre- sent than any of those already mentioned, especially in large towns, where the number of chimneys in one wall is often so great that it is difficult to get a sufficient space for each. The most obvious cure, where the situation admits of it, is to widen the chimney, by opening a hole a little above the grate thro' the back-wall of the chimney, slanting a little upward, and building on the outside of the wall a small chimney open from that hole to the top of the building, as in fig. 5. where AB represents the new tube going through the wall at the opening at A, which will receive the superfluous smoke, and carry it off. This additional chimney must always be carried as high as the other. But as there are many situations in which this method of cure would be impracticable, we must try e- very method for accelerating the ascent of the smoke; (for the more quickly it ascends, the more narrow may the tube be;) and with that view, the chimney may be heightened at top, and contracted at bottom, in any or all the various ways we have mentioned. But if none of these methods prove effectual, let the chimney be built quite close at the under part, leaving only as much room as is sufficient to contain the grate, having a cover of metal fitted to that o- pening pening, which can be taken off or put on at pleasure; by which the whole air that enters into the chimney is made to pass through the fire like a furnace, and carries the smoke up it with great velocity. These are well known in large towns by the name smoke-chimneys: but as they occasion a prodigious waste of fuel without warming the room, and, unless attended with very great care, are in danger of setting the building on fire, they ought to be as much avoided as possible. But if neither this, nor any of the other methods prove effectual, the wall must either be taken down and rebuilt in a proper manner, or the chimney abandoned as incurable. As this is a defect more difficult to be remedied than any other, we would strongly recommend it to every builder to build his chimneys of a sufficient width throughout; there is no danger of erring on this extreme, as it is easy to remedy any defect that might arise from it.

4th. As the air which ascends through the chimney continues nearly of an equal degree of heat to the top, the tube should be of an equal degree of width at the top as at the bottom, as well as through the whole of its length. It ought not therefore to be made tapering gradually from the fire-place to the top, but to be suddenly contracted above the grate, as in fig. 6, from which it ought to be continued of an equal wideness throughout its whole length: but if it is narrower at any one place than another, it ought to be at the under part, immediately above the fire, for a very short space; because, as this is within reach of the hand, the foot can be cleaned from it as often as is necessary, so that when the other parts of the chimney are full and clogged with soot, they will not be narrower than this place is at that time.

5th. It seldom happens that a chimney can be carried quite straight upwards; and it is an advantage that it is so, as they ought always to be bent a little. For if a chimney be straight, and of a proper width to transmit the whole of the smoke and no more, it will not be sufficient for that purpose, when there is a heavy fall of rain, or snow, or hail, with little wind; for the great drops will fall perpendicularly from the top to the bottom of the chimney; and as they occupy a considerable space, the smoke will not have room to ascend, but must be forced down with the shower, and dispersed in the apartment: whereas, if the chimney is bent, the rain falls upon some of the sides, and glides gently down without disturbing the ascent of the smoke. The same inconvenience will be felt in a straight chimney, where it is so placed as to be exposed to winds which sometimes enter the top and blow down with a sudden puff: for, if it be straight, the air meets with no interruption till it descends into the chamber, and there disperses the smoke; but if it be crooked, the descent of the wind will be obstructed its force broken, and the bad effects of it in a great measure prevented. Upon the whole, bent chimneys are always preferable to straight ones. However, a perpendicular chimney may be easily cured, by some of the contrivances after mentioned.

These are the most general defects arising from the structure of the chimney itself, which are all reducible to the following causes: 1. Too little height; 2. Too great wideness of the bottom of the chimney; 3. Too little width; 4. Unequal wideness between the top and bottom; and, 5. Straightness of the tube. We have pointed out the best methods of curing each of these defects; and to finish our remarks on this general head, and give the reader a more perfect idea of the best form of construction for a chimney, we have drawn two different sections of one constructed on the justest principles, in figures 6. and 7. the several parts of which appear so plain from the figures, and the reasons for this construction have been already so clearly assigned, that a very short explanation will be sufficient. Figure 6. represents a front-view of the fire-place; supposing the fore-part of the wall taken down, and the chimney laid bare from top to bottom; AB, and DC, representing the two sides of the fire-place; and BC, the mantle, being cut through, to shew the manner in which the aperture is suddenly contracted, immediately above the fire within the mantle. The tube from the point E, to the top, ought to be of an equal wideness, and bent in any direction that may be convenient. The two planes, FF, represent the two sides of the fire-place, which ought to be as much flopped inward towards the back as the form of the grate will admit of; for the more they are inclined, the more powerfully will they reflect the heat into the apartment. Fig. 7. represents a perpendicular section of the wall, through the middle of the chimney, to discover its shape, if viewed from a side. And here it is to be observed, that it ought ever to be a rule to bring the building at the under part of the chimney immediately behind the grate, as far forward as possible, because this throws more heat into the chamber than if it were placed farther back: but as the fore-part of the grate ought never to project beyond the inner edge of the mantle, care should be taken to have the under part of the mantle wrought as thin as the nature of the materials will admit of, making it thicker towards the upper part, so as to slope inwards above the fire, as represented at B, fig. 7. To throw the heat outward, let the upper part of the back of the fire-place be a little inclined outward, as at C; but, after it is carried up in this manner a little higher than the mantle, let it be suddenly turned back, as in the figure, the projection above the mantle inclining backward in the same direction, being carried up to the top at an equal wideness the whole way.

Although it is necessary to have all chimneys pretty wide; yet this, on many occasions, is attended with inconveniences: for as they transmit a great quantity of heated air, too large a portion of that warm air which ought to heat the chamber is carried off; and as it is only when the fire is first kindled that the great quantity of gross vapour is exhaled which fills the chimney, and maketh a large tube necessary, if it were so contrived as to contract or dilate at pleasure, we might have our chimney of a sufficient width to convey away the greatest quantity of smoke that could ever have occasion to pass through it, at the same time that we might never allow more air to pass off at other times than was necessary to carry away the whole of the smoke, by which means a much smaller quantity of fuel would keep our apartment equally warm. This we apprehend might be accomplished by the following simple apparatus. Let a sheet of milled iron be fixed at the upper part of the mantle, on the inside, at B, in such a manner, as that, by means of a small wire de, passing through a small hole left for that purpose in the fore-part of the chimney, it might be let down at pleasure towards C, or drawn up towards B, so as to apply quite close to the upper edge of the chimney between B and C. This would leave the tube of its full wideness when necessary, or close it to any degree at pleasure, with the greatest ease*. It is unnecessary to add, that the plate at A fig. 4 might be employed in the same manner, when it should be found convenient. So much for what relates to the construction of the chimney itself.

We now proceed to consider the second general cause of smoke, viz.

II. Of smoky houses proceeding from faults of other parts of the building, altogether independent of the structure of the chimney itself.

1st. The first we shall mention is too great closeness of the room. Smoke, as shown above, is impelled up the chimney by the pressure of the air entering at the fire-place and ascending upwards; but, if fresh air is not admitted into the apartment in sufficient quantities to supply the consumption by the fire, the room will be quickly exhausted, and the air in it become as light as the external air at the top of the chimney, so that the smoke will as readily be dispersed into the chamber as through the chimney. But if any door or window is opened so as to admit plenty of free air, the smoke will be quickly dispelled, and the proper circulation established: the same effect will be produced by making a small hole in some of the sides of the room; but unless this be done with some judgment, it may frequently add to the disease, as it may concur with some of the other causes of smoky houses, to be afterwards mentioned. A better method of remedying this evil would be to have a small hole made in the wall at the back of the chimney, and immediately underneath it: or a small perforation, made in the wall in any other convenient manner; the one end of which should communicate with the external air, and the other communicate with the chamber in any place near the grate, and as low down as possible, through which a constant supply of air would be administered to the fire without the smallest inconvenience or trouble. If this were practised, doors and windows might with safety be made much closer than at present, and our apartments rendered equally warm and comfortable with a much smaller quantity of fuel than we use at present. For as the fire, in the present mode of constructing chambers, is kept alive by a constant succession of cold air from the doors, windows, and other crannies of the room rushing towards the chimney in all directions, the air of the room, which, if not cooled by this means, would be quickly heated to a great degree, is constantly kept cold in spite of the strong heat of a blazing fire; which, at the same time that it scorches the parts of our body which are most exposed to it, does not warm the parts which are turned from it; and we experience at the same time a burning heat and piercing cold, which is often productive of the most disagreeable effects. But if the fire were supplied with air in the manner abovementioned, there would be less air drawn in through the crannies of the room, so that the air within would be soon warmed, and continue long so even with a small degree of heat. However improper this might be for people in perfect health, it might surely be of great use for those who are in a weakly habit of body; especially if care were taken to carry off the foul air, by having a small tube leading from the upper part of the room to the top of the house, through which the air which had been rendered noxious by the smoke of candles or perspiration would be conveyed away, and a succession of fresh air admitted from the tube near the fire-place to supply that want. That the reader may more readily comprehend what is here meant, we have represented in fig. 6, a view of two small tubes for this purpose supposed to be laid open by taking away the boxing or inner coating of the wall which ought to cover them. These are of wood, and must not be above one inch in diameter. One end, g g, goes quite through the outer wall of the house, and communicates with the open air, having a small grate upon it to prevent vermin from entering. The other end, h h, passes behind the two slabs at the side of the fire-place, and opens in the inside of the fire-place at i i, at which place they have each of them a small bit of brass fitted to them, being closed with two sliding doors exactly like those that are used to cover the end of telescopes, by means of which more or less air may be admitted at pleasure.

2d. A second cause of smoke, is the wrong position of doors and windows, with respect to the fire-places.

As the smoke is impelled up the chimney by the pressure of the air, if that air is driven away from the fire-place by any cause more powerful than the suction occasioned by the fire, the smoke must also be drawn away with it, and follow the same directions with that current of air; so that whatever tends to draw a current of air from the under part of the chimney, will also tend to produce smoke in the house; from whence it is easy to conceive how doors or windows may occasion smoke when the wind is in certain directions. Thus, suppose a chamber, A, B, C, D, fig. 8, having a door or window at E, another at F, and a fire-place at G; when the wind is in the direction D A or C B, the general current of air will occasion a sort of suction at the opening E, so that the air will be drawn from the chimney G towards E; and if the current be strong, and the opening at E large, it will become more powerful than the suction of the chimney, and produce smoke in the apartment. If the window at F should be opened in this case, it would not mend the matter; for any wind which should enter at F, would be carried straight out at the opening E, and the current of air would be drawn from the chimney as strong as ever. If the window at E were shut, and that at F left open, and the wind still continued in the same direction as before, the current of air rushing past the window would have a tendency to draw the air of the room along with it,

* If any one should think, that the wire d would be a disagreeable object in the middle of a chimney-piece, it might easily be hid by a picture of any kind. The wire might be fixed to a small brass-handle moving freely upward and downward like that for a bell; only this should have a long slit in the middle of it, with notches on each side, to receive a pin placed in the middle of the slit, by means of which the wire might be lengthened or shortened at pleasure. The whole of this apparatus is represented at fig. 16, where (a) represents the wire fastened to the brass-plate; (b) a piece of brass, raised a little, to serve as a handle. The slit in the middle is represented by the dark line, having notches c c c at convenient distances. The pin d is fixed into the wall, but left at liberty to turn about with ease; and its head stands up a little, so as to be easily turned with the finger and thumb. The body of this nail is made so thin in one direction, that when it is turned half round, it easily passes through the slit in the plate; but in the other direction its diameter is greater, so that when the plate is brought so as to have one of the notches opposite to the nail, and it is then turned half round, it catches the plate so that it cannot be moved till the nail is again turned about. and occasion smoke, but not so powerfully as if the window at E were open; but if the wind were in the direction DB, it would be very bad; but if it blew in the direction CA, the case would be very much altered; for then a quantity of air being forced in at the opening F, and finding no ready passage, it would be pent up in the chamber, and force itself up the chimney with violence. We omit mentioning what would be the effect if the wind were in other directions, as it is imagined these will be sufficiently obvious to every attentive observer. It is only necessary here to observe, that as doors or windows are seldom so exactly made, but they produce some effect, as they always admit some air even when shut, and often occasion smoke when the wind blows from a particular quarter; and as workmen and others generally apprehend, when houses are troubled with smoke in this manner, that it is occasioned by some external cause, and apply their attention to cure it by altering the top of the chimney, which never can produce the smallest service in this case; we would recommend a more particular attention to be paid to the situation of doors and windows than is generally bestowed; especially in such situations where they are exposed to any violent current of air in a particular direction, as in narrow lanes or defiles of any sort, where the wind, when in particular directions, is hurried along with a prodigious rapidity. And, that the effects of different positions may be still more obvious, we shall produce several other examples.

Suppose a chamber, fig. 9, having a door at A, and two windows BC, with a fire place D. If the wind came in the direction DA, and if the door transmitted as much or more air than was admitted at both the windows, a current of air would run from all parts of the chamber towards A, and therefore would have a tendency to occasion smoke; but if as much or more air came in at the windows than could get out at the door, there could be no such current; but, on the contrary, it would be forced up the chimney, and carry the smoke along with it: wherefore in this situation, a room might sometimes be cured of smoke, by making the door as close as possible; nothing could be more hurtful in this case than boring a hole in the door. But if the house was in such a situation as to be more frequently exposed to a wind which came in the direction of AD, it would run little risk of being troubled with smoke.

Suppose a room, fig. 10, having a door at A, and two windows B and C, with a fire-place D. If the wind came in the direction CB or BC, and both the windows were open, it is evident that the smoke would be drawn from the chimney by the strong current of air passing through the room; or if the window upon which the wind came were closed, and the opposite one open, nearly the same effect would be produced: but if the window upon which the wind blew were open, and the opposite one and the door shut, the room would be immediately cleared of smoke entirely. In this situation, it is evident, that if the windows were badly made, so as to admit much air, it would tend to occasion smoke, especially if the door were in the same situation; it is therefore of consequence to attend to this circumstance in a situation similar to this.

Having premised so much with regard to single rooms, we shall now proceed to consider a more compound structure. Thus, let fig. 11 represent a building consisting of two chambers, KL, joined by a passage. The chamber K having a door B communicating with the passage, a window F, and fire-place G; and that at L having a door C, another door or window D, the window E, and fire-place H, the entry to the whole being by the door A. Let us now consider what would be the effect of the wind coming from different directions upon this building. And first, suppose the wind blew in the direction AM: If all the doors and windows were close shut, and very little air were admitted, there would be little risk of smoke; but as there would surely be some admitted through these different openings, there would be some chance that the chamber L would be troubled with smoke, because of some air which forced its passage through the chinks of the door A would pass through the doors C and D, which might produce smoke in a small degree. There would be little chance that the chamber K would smoke in this case; because although there is a general suction through the passage from B to D, yet as it is in some measure interrupted by the close door at C, it will be but small; and as the wind is interrupted in its course by the wall of the passage, some of it will be forced through the chinks of the window F, which would more than counterbalance the effects of the other suction. But if the door D were open, both the chimneys would smoke; especially if the doors B and C were open also, as the current would be then very strong towards that point. But in all cases the smoke of this house would be prevented by keeping the door at D shut, and that at A open; but if the house was in such a situation as to be more exposed to that wind than any other, it would be better to close up the door D altogether. If the wind more commonly came from M towards A, it is more than probable that a house situated like this would be quite free of smoke, as the general current of air would be towards the chimneys; but the chamber K would run greater risk than L, as the suction might sometimes be drawn towards the window F; but if the door A were in the opposite side of the passage, that inconvenience would be avoided also. If the more general current of air were from K towards L, this house behoved to be troubled with smoke unless the windows were very close: but there would not be the smallest chance for that, when it came from L towards K.

We might now proceed to give more examples of this sort: but as it would be impossible to enumerate all the variety of cases that might occur, it is imagined that these will be sufficient to give the reader an idea of the manner in which any building ought to be examined in this respect; and he must be left to his own discretion to apply the principles above explained to all the variety of cases that may occur. In large complicated buildings, it no doubt requires a greater extent of thought to combine all the various circumstances together, and draw a general conclusion, than in smaller and more simple ones; but if the following general rules are attended to, the complaints arising from this cause would be but few. 1st, Avoid as much as may be long passages leading to very distant parts of a building, as there is often a strong current of air in these which helps to disturb the free circulation of air up the chimneys. 2d, Place the chimneys in general on that side of the apartment towards which the wind which in general prevails most in the situation where the house is placed blows: And, 3d, make as many, if not more, doors and windows (especially such as have occasion to be most frequently open) on that side of the building from whence the most prevalent wind does come.

IIId. The third general cause of smoke in houses is the wrong position of the house with regard to external objects, which, by interrupting the course of the air, makes it assume various directions, and wheel about in eddies, so as to prevent it from ascending with ease from the chimney top, or beats it down into the room with violence. This is more seldom the cause of smoky houses than either of the two before mentioned; although it seems to be almost the only one attended to by the persons who pretend to cure smoky houses at present, as most of their remedies are adapted to remove the disorders arising from this cause alone. We shall briefly point out the several cases in which this can occur, that every one may be enabled to judge for himself when these cures are proper or not.

The air (as has been said) is a fluid, and wind a current of that fluid; which, when driven along the surface of the earth, flows with a smooth and equal stream, unless when opposed by some object which interrupts its course; but when it meets with any object which directly opposes its course, it is in some measure pushed back again, and made to spread on every side, till it meets with some open side, towards which it flows with great impetuosity. It is likewise a fluid of considerable gravity, and therefore presses upon the surface of the earth with great force; so that, when a current of it flows along the surface of our globe, it has a tendency to move forward and press downward at the same time: from whence it happens, that when a current of air is forced over the top of any high object, the side of which descends perpendicularly downward, the velocity of the current at first overcomes the gravity, and it flies a short way over in that direction; but the power of gravity acting upon the under surface, draws it downward, and in a short time overcomes the impetus that it had to rush forward, and occasions a sort of eddy nearly similar to what we see among water behind a stone which interrupts the violence of its currents.

To illustrate this more plainly, let AB, fig. 12, represent a part of a high building, near to which is a smaller one CD; and let the dotted line EF represent a current of air flowing with considerable force in the direction FE. It is plain that it will flow straight forward over the top of the small building; but when it meets with the large object, it will be interrupted in its course, and spread itself on every side, as represented by the dotted lines GG &c. at last it will flow towards that place through which it can escape with the greatest ease. If the opposing object be large, and has no opening through which it can issue near the ground, then it will ascend to the top of it, and flow off in that direction, carrying the smoke which ascends from the small chimney C along with it: but if there is any opening below, either a street or lane, or any other passage that will admit the wind to pass, then will the natural gravity of the air draw the general current downward to flow off through the lower passage; in which case, the smoke which ought to ascend through the chimney C, meeting with a current of air opposing its passage, will not be at liberty to issue forth, but be forced back again into the room from whence it proceed, unless some contrivance is fallen upon to prevent it.

Again, let A, fig. 13, represent a small building at the side of a great rock B, and the wind coming in the direction CD; when the current of air comes to the point D, being hurried forward with great velocity, it goes a little forward, but soon descends downward, and gradually is reflected more and more inward, as represented by the dotted lines EE, &c. so that, descending downwards upon the top of the chimney A, the smoke is beat back again into the apartments. Thus it is that low houses, when contiguous to high objects, are in danger of being disturbed with smoke. If the contiguous object be not very high, the disorder may be cured by heightening the chimney of the low house; but if it is very high, it will be necessary to cover the top of the chimney in such a manner as to prevent the wind from entering it, at the same time that a passage is left at some of the sides through which the smoke may issue with freedom. Many are the contrivances which have been invented for this purpose, which are to be met with everywhere; and as there is no difficulty in accomplishing the desired end by an infinite variety of methods, every one who needs such a thing may please his own fancy in the choice. We have thought it unnecessary to add any more but one kind of these, fig. 14, which will answer the end effectually.

It is evident that houses situated near high hills, or thick woods, will be in some measure exposed to the same inconvenience; but it is likewise plain, that if a house be situated upon the slope of a hill, as at F, fig. 13, it will not be in any danger of smoke when the wind blows towards that side of the hill upon which it is situated; for the current of air coming over the house-top in the direction GH, is immediately changed by the slope of the hill to the direction HC, which powerfully draws the smoke upward from the top of the chimney. But it is also evident, that a house in this situation will be liable to smoke when the wind blows from the hill; for the current of air coming downward in the direction CH, will beat downward on the chimney F, and prevent the smoke from ascending with freedom. But the effect will be much heightened, if the doors and windows are chiefly in the lowermost side of the house.

There are some of the most general circumstances which prevent the free ascent of smoke, arising from external objects: but there are many other lesser causes which may at times occasion smoke, all of which it would be tedious here to enumerate; such as, blasts of air, reflected from the sides of mountains, and coming down valleys with great impetuosity, occasioning, in particular situations, eddies or whirlwinds of different sorts. In short, whatever in any measure disturbs the free motion of the air, is in danger of producing sudden gusts, which may occasion smoke. Therefore, whoever builds in a situation which is not altogether free, may lay his account with having some sudden gusts of smoke, unless he forms the top of his chimney so as to obviate it. And there are some situations so much exposed to sudden gusts of wind, sometimes whirling round, sometimes beat suddenly downward, or as suddenly carried up again, that it is difficult to guard against every danger. In these situations we would recommend something of the form of what is represented at fig. 15, which would be proof against every wind whatever.

Having thus traced the causes of smoky houses, and reduced them to distinct classes for the sake of distinctness; it is necessary, before we quit this subject, to observe, that in many cases, two or more of these may be combined to augment the malady, and therefore it is necessary to have all these circumstances in view in every particular case. It now only remains that we point out the several phenomena which may lead us to distinguish from which of these general causes the disorder complained of may proceed. And, 1st. If it is owing to a fault in the construction of the chimney itself, it will smoke almost continually, especially in calm weather.

2nd. If it does not smoke in calm weather, or only when the wind comes from some particular quarter, and can then be cured by opening some door or window, the fault may be looked for in the distribution of the doors or windows of the house. The only case in which there is a difficulty to distinguish whether it is owing to the fault of the chimney or the house, is when it proceeds from too much closeness of the apartment. But this may be easily known by trying it in a calm: for if it proceeds from this cause, there will be no smoke in a perfect calm, if the doors are left open; whereas, if the defect proceeds from a fault in the chimney itself, it will still continue to smoke when calm, even when the doors are open.

3rd. When the smoke is occasioned by external causes, these can be generally seen; but it may be likewise known by this, that it descends in sudden puffs with great violence at times, even when the doors and windows are not altered. By attending to these few rules with care, there will be little danger of mistaking the cause from whence this disorder proceeds.

We shall conclude these observations with a few remarks on some particular cases which can hardly be reduced to any of the foregoing heads. And,

1st. It sometimes happens, that the smoke is prevented from ascending with freedom, by having a small part of the top of the chimney broken down, so that some parts of it remain higher than others, which in some measure reduces it to the state of a chimney at the side of a higher one. To prevent this, it is always proper to have the top of the chimney finished with stones neatly cut, and firmly built. It is not to be doubted but that those things which are placed upon particular chimneys with a view to cure them of smoke, do often, from the same cause, hurt the neighbouring chimneys built in the same wall.

2nd. A chamber is sometimes filled with smoke, when a fire is kindled in a neighbouring chimney, and none in it, although there is no appearance of smoke when it has a fire burning in its own grate. This may sometimes proceed from a small hole breaking through the thin partition that divides the two chimneys from one another, and as smoke is a weighty body, which is only buoyed up by the warm air which passes through the fire, when it penetrates into the cold chimney it naturally subsides, and comes down to the chamber with which the chimney communicates, when there is no fire to carry it off. But this disease is generally produced by the smoke entering at the top of the chimney, and descending downwards: if this last is the case, it may be cured, on many occasions, by setting a pretty high stone at the top of the chimney, as a division between each two: but the surest method, in all cases, is to have a smoke-board exactly fitted into the chimney above the grate, which on all occasions effectually prevents it.

3rd. It frequently happens, that a chimney does not carry off the smoke well at first when the fire is kindled, although there is not the smallest tendency to it at other times. This proceeds from the narrowness of the chimney; for when the fire is kindled, the whole tube is filled with cold air, as weighty as that in the apartment; and being expanded by the fire at the bottom; it endeavours to ascend upward, but being pent in by the narrowness of the tube, and pressed by the column of cold air above it, it is some time before it can wholly overcome that resistance, and some of it is forced into the chamber, till by degrees the whole chimney is heated, and then it vents quite well. If the smoke produced by this means is not very troublesome, it may be borne with; but if it be extremely disagreeable, it may be cured by having a large sheet of milled iron, large enough to reach between the two sides of the fireplace, and as deep as to reach from the mantle to the grate, or lower, which might by any contrivance be hung up before the fire at that time to act in some measure as a smoke-chimney. This would quickly make the fire burn, and carry off the smoke entirely. After that is effected, it might then be removed, till another occasion.