Sometimes the smoke from a chimney is driven out into the room, even when the chimney is not commanded by a superior elevation, it being driven down by strong winds passing over the top of the chimney. Dr. Franklin mentioned one or two instances of this kind which he had met with:--“I once lodged at a house in London, which in a little room had a single chimney and funnel. The opening was very small, yet it did not keep in the smoke, and all attempts to have a fire in this room were fruitless. I could not imagine the reason, till at length observing that the chamber over it, which had no fire-place in it, was always filled with smoke when a fire was kindled below, and that the smoke came through the cracks and crevices of the wainscot, I had the wainscot taken down, and discovered that the funnel which went up behind it had a crack many feet in length, and wide enough to admit my arm; a breach very dangerous with regard to fire, and occasioned probably by an apparent irregular settling of one side of the house.” This does not at first thought seem to be an illustration of the effect of wind passing over the top of a chimney; but the explanation is to be sought for in a similar way; the air, by entering this fractured part freely, destroyed the drawing-force of the chimney. The manner in which the passing of a current of wind over the top of a chimney may produce a “smoky room” is this:--the warm air which rises from the fire, in order to obtain a free issue from the chimney, must repel the air that is hovering over the chimney-pot. In a time of calm or of little wind, this is done easily; but when a violent current is passing over the top of the chimney, its particles have such a strong horizontal velocity, that the heated air in ascending has not power to displace it, and thus the smoke, not finding a ready exit by that path, is driven back into the room.