the same goodly logs as fuel, and a supply which will enable us to afford the blaze that alone would suffice to rarefy a cavern. The ordinary coal-fires of our apartments do not need a larger flue than a pipe of 9 in. in diameter. Mr. Richardson, in his work, states that the houses built by Cubitt in Belgravia have flues 9 in. by 9 in. only, while others erected later have them 14 in. by 9 in., and that these are distinguishable outside by the absence or presence of objectionable cowls respectively. Kitchen-flues must be larger, in proportion to their fires, or better, perhaps, doubled—a practice for which old precedents may be found, and which seems calculated to avoid down-draughts.
For the avoidance of that particular nuisance, however, special provision should be made in every flue. This may be done by an enlarged space, wherein the force of gusts of wind may expend itself upon, as it were, a cushion of air. If the first pipe above the chimney-mantel be a 9-inch pipe, let the next be a 15-inch one, and the flue above continued with 9-inch ones. A somewhat similar arrangement has been proposed by Mr. Boyd for brick flues. He discontinues the vertical flue a few feet above the mantel with an enlarged space or pocket, and carries an inclined one from the fireplace into this on one side, and the down-draught, thus meeting resistance at the bottom, eddies round the space, without being able to check the upward draught from below. Mr. Cubitt's continuation of the flue to the basement also obviously affords a resisting column of air to accomplish the same purpose. It may be impossible to make such cavities large enough to overcome the effect of every down-draught, but these provisions against them will generally secure this desired end if combined with ample provision of air to the fireplace.
The use of pipes for the lining of flues has the advantage of compelling a good and non-porous finish, which would otherwise be neglected by careless workmen, who often will not take the trouble to properly parget and core the flues in stone and brick walls. The interior of the pipes, however, should be rough, and by no means glazed, or their inability to give any means of adherence to soot will be found a nuisance, in consequence of its continual dropping. The old funnel-shaped throat left a large space above the grate filled with cold air, which checked the draughts. This depends much upon the grate itself; but, generally speaking, the flue should be contracted to its smallest size as soon as possible above the mantel. Iron frames for this purpose, serving as mantel-bars as well, such as Gibbs's registered fireplace-lintel, are useful appliances. A concrete block may be made of the shape required at perhaps the least cost. . . .
The construction of the fireplace itself is of the most importance. The contraction of the flue immediately over it is the first point to be looked to, and next the provision of a proper supply of air for the combustion of the fuel. To illustrate this in the simplest manner, I may refer to a small room with a large fireplace in it, belonging to a friend,