A fire given fifteen minutes' start is often hard to overtake. Imagine a warehouse alight on three floors before the alarm is raised! Engines may come one after another and pour deluges of water on the flames, yet as likely as not we read next morning of "total destruction." No stitch in time has saved nine!
The sad part about fires is that they represent so much absolute waste. In commercial transactions, if one party loses the other gains; wealth is merely transferred, and still remains in the community. But in the matter of fire this is not the case. Supposing that a huge cotton mill is burnt down. The re-erection will, it is true, cause a lot of money to change hands; but what has resulted from the money that has _already_ been put into the mill? Nothing. So many hundred thousands of pounds have been dematerialised and left nothing behind to represent them. The great Ottawa fire of a few years ago may be remembered as a terrible example of such total loss of human effort.
The first recorded specification for an automatic detecting device bears the date 1763. In that year a Mr. John Greene patented an arrangement of cords, weights, and pulleys, which, when the cord burnt through, caused the movement of an indicating semaphore arm. As this action appealed only to the eye, it might easily pass unnoticed, and we can imagine that Mr. Greene did not find a gold mine in his invention.
Twenty-four years later an advance was made when William Stedman introduced a "philosophical fire alarum." "His apparatus consisted of a pivoted bulb having an open neck, and containing mercury, spirit or other liquid. As the heat of the room increased, the expansion of the fluid caused it to spill over, release a trigger, and allow a mechanical gong to run down. This arrangement, whilst an advance upon the first referred to, is quite impracticable. Evaporation of fluid, expansion of mercury, a stiff crank, or other causes which will readily occur to you, and the thing is useless."[14]
In 1806 an automatic method for sprinkling water over a fire appeared. The idea was simplicity itself: a network of water mains, with taps controlled by cords, which burnt through and turned on the water. William Congreve patented, three years later, a sprinkler which was an improvement, in that it indicated the position of the fire in a building by dropping one of a number of weights. But string is not to be relied upon. It may "perish" and break when no fire is about, and any system of extinction depending on it might prove a double-edged weapon.
