It is desirable, however, now to discuss one point connected with solid carbide about which some misconception prevails. Calcium carbide is a body which evolves an inflammable, or on occasion an explosive, gas when treated with water; and therefore its presence in a building has been said to cause a sensible increase in the fire risk because attempts to extinguish a fire in the ordinary manner with water may cause evolution of acetylene which should determine a further production of flame and heat. In the absence of water, calcium carbide is absolutely inert as regards fire; and on several occasions drums of it have been recovered uninjured from the basement of a house which has been totally destroyed by fire. With the exception of small 1-lb. tins of carbide, used only by cyclists, &c., the material is always put into drums of stout sheet-iron with riveted or folded seams. Provided the original lid has not been removed, the drums are air- and water-tight, so that the fireman's hose may be directed upon them with impunity. When a drum has once been opened, and not all of its contents have been put into the generator, ordinary caution--not merely as regards fire, but as regards the deterioration of carbide when exposed to the atmosphere--suggests either that the lid must be made air-tight again (not by soldering it), [Footnote: Carbide drums are not uncommonly fitted with self-sealing or lever-top lids, which are readily replaced hermetically tight after opening and partial removal of the contents of the drum.] or preferably that the rest of the carbide shall be transferred to some convenient receptacle which can be perfectly closed. [Footnote: It would be a refinement of caution, though hardly necessary in practice, to fit such a receptacle with a safety-valve. If then the vessel were subjected to sudden or severe heating, the expansion of the air and acetylene in it could not possibly exert a disruptive effect upon the walls of the receptacle, which, in the absence of the safety-valve, is imaginable.] Now, assuming this done, the drums are not dependent upon soft solder to keep them sound, and so they cannot open with heat. Fire and water, accordingly, cannot affect them, and only two risks remain: if stored in the basement of a tall building, falling girders, beams or brickwork may burst them; or if stored on an upper floor, they may fall into the basement and be burst with the shock--in either event water then having free access to the contents. But drums of carbide would never be stored in such positions: a single one would be kept in the generator-house; several would be stored in a separate room therein, or in some similar isolated shed. The generator-house or shed would be of one story only; the drums could neither fall nor have heavy weights fall on them during a fire; and therefore there is no reason why, if a fire should occur, the firemen should not be permitted to use their hose in the ordinary fashion.