the bend, and is therefore soon lost as a heating factor, the heat being wasted in the chimney.
From the commencement of my study of the fireplace question the value of the "lean-over" has not only taken firm hold of my fancy, but my sense of its importance has been growing in intensity, until I saw that the best construction must show the greatest possible extent of "lean-over" that could be obtained without sacrifice of other important details of construction. How to accomplish this will appear in considering the fifth rule.
Rule IV. "The bottom of the fire, or grating, should be deep from before backward, probably not less than nine inches for a small room, nor more than eleven inches for a large room." This is a corollary to Rule III. We can not possibly have the back of the fireplace over-hanging the fire when there is a shallow grid. If for no other reason than the demands of the "lean-over," depth of fire-space is essential. But there is gain, thereby, in another direction. It affords plenty of room for the burning fuel to lie down close to the grid, and away from swift air-currents, and prevents the tendency of the fire to burn hollow.
On this point Rumford has a word to say: "But as many of the grates now in common use will be found too large when the fireplaces are altered and improved, it will be necessary to diminish their capacities by filling up with pieces of fire-brick. But, in diminishing the capacities of grates, care must be taken not to make them too narrow, i. e., too shallow.
"The proper depth for grates for rooms of middling size will be from six to eight inches. But, where the width (i. e., depth) is not more than five inches, it will be very difficult to prevent the fire going out."
"Where grates designed for rooms of middling size are longer (and broader) than fourteen or fifteen inches, it will always be best to diminish their length by filling them up at their two ends by fire-brick."
Rule V. "The sides or 'covings' of the fireplace should be inclined to one another as the sides of an equilateral triangle" (Fig. 2). The working out of this rule has cost me much thought and experiment. It was worked out more or less empirically with a view to attain certain objects, and, having attained them, I discovered that I had unwittingly selected the sides of an equilateral triangle. It is of some importance, and may be of interest, to tell how the question arose. In my earlier fireplaces the sides or "covings" were parallel to each other, and had the defect that they radiated most of their heat from one to the other, not into the room, with the probable result that much of such heat would eventually escape up the chimney.
It was clear, then, that the sides must be set at an angle with the back, so as to face toward the room. But at what angle? My first
