For the same reason, that a high column of warm air will overpower a low one, if the supply is not sufficient for both, it is very common, where the cold-air box of the furnace is of insufficient capacity, to find one or more registers in the house, usually in the first story, and most frequently, perhaps, in the entrance hall or staircase hall, which deliver little or no warm air; and, by testing them with smoke, a downward current may often be detected in them. This is so much a matter of course in most houses that it attracts little attention; but it is by no means necessary, and if an adequate supply of fresh air is given to the furnace, and the dampers in the hot-air pipes are adjusted, so as to cut off a part of the flow into the rooms which have more than their share of warm air, and throw it into those which have less than their share, all the registers can, with a furnace of sufficient capacity, be made to deliver warm air.

Unequal heating.

Cold register.

The reason why hall registers are more apt to fail in doing their duty than others in the same story is that the cold currents entering around or through the door, or descending a staircase, press upon the surface of the register, slightly, but sufficiently to counteract the very feeble ascending force of the current in a first-story register. In consequence of this, a hall register should never be placed under the well of a staircase, but under the staircase itself, or in some other position where it will be sheltered from the descending current; nor should it be placed where the wind from the door will blow directly upon it. In cases where hall registers have already been badly placed in this respect, their efficiency may often be improved by putting a table, or other piece of furniture, over them, so as to shelter them from cold currents.

Hall registers.

The current from a register may, however, be greatly affected by other causes than that of a want of balance between it and other currents. The most serious of these is, perhaps, the exposure of the pipe supplying it to cold currents of air, which, added to the friction in a long pipe, particularly in a pipe which has not a decided and continuous ascent from the furnace, will often completely annihilate the flow. Much may be done in such cases by wrappings of asbestos paper, in several layers, to protect the pipe from the cold, using in preference the corrugated paper, which is a much better non-conductor than the plain paper; and it is often possible, by changing the position of a register, to shorten very materially a long pipe, without any inconvenience whatever, and to the great improvement of its heating power. In general, the registers in rooms at a distance from the furnace should be in the side or corner nearest the furnace. The rule for steam or hot-water radiators, that they should be placed under the windows, or in the coldest part of the room, is totally inapplicable to hot-air registers, the efficiency of which mainly depends on the shortness and directness of their communication with the furnace. . Vertical pipes, especially those carried up in partitions, or behind furrings, are often exposed to cold air, and can have their working improved by covering them, or, if that is impracticable, by admitting warm air from a kitchen, or some other warm room, to the space in which they run; or, in extreme cases, a new pipe may be carried up inside the old one, or in a different place.

If all these precautions fail to improve the warming of a room, the fault may be in the furnace, which is very likely to be incapable of heating a sufficient quantity of air to supply all the rooms. Many furnaces, particularly those of the simplest wrought-iron type, in order to throw the air passing through their air-chambers against the comparatively limited heating surface of the furnace, have a flat ring fastened to the inside of the casing, leaving a narrow annular space, sometimes not more than two inches wide, between this ring and the drum of the furnace, through which must pass all the air to supply all the registers in the house. With all reasonable allowance for the increased rapidity of the currents next to the hot surface, it is obvious that the quantity of air which could pass through so small an opening would be very inadequate to supply any considerable number of registers; yet it is not uncommon to see houses, with furnaces of this sort, in which the register area is far beyond the capacity of the furnace to supply. Where there is reason to suppose that a furnace is overtaxed in this way, something may be done to improve its working by nearly closing the dampers in the hot-air pipes, so that each room can get a small share of the hot air. Then, by driving the furnace, the air may be made so hot that the small quantity delivered to each room, mixed with the other air in the room, will give a tolerable temperature to the whole; but the only remedy which is really satisfactory is to put in a furnace of a different kind, with sufficient heating surface to warm the amount of air required for all the pipes.

It may be observed that warm air, either from a furnace or an indirect steam or hot-water apparatus, will not enter freely a room which has no outlet, either in the shape of a fireplace, or an open stove, or a ventilating register, or even a "stove hole" into a flue; and the better the house is built, and the more tightly the windows fit, the more difficult it will be to heat such a room with a register. In cases of this kind, an escape for the air should be contrived by an opening into a flue, or even by means of a transom sash, or a hole into a hall or another room, making the opening as high above the floor as practicable. It often happens that nothing more is needed to enable a furnace to keep a room comfortably warm which could not previously be heated.