When the fire has broken through a floor, the supply of air along that floor is not to be depended on--the fire drawing the principal supply of air from the apartments below. When the two first firemen have gained a favourable position, they should keep it as long as they are able; and when they feel exhausted, the men behind them should take their place. The great point to which everything ought to be made subservient is, _that the water on its discharge from the branch-pipe should actually strike the burning materials_. This cannot be too often or too anxiously inculcated on every one connected with a fire-engine establishment. Every other method not having this for its grand object, will, in nine cases out of ten, utterly fail; and upon the degree of attention paid to this point, depends almost entirely the question as to the amount of damage the fire will occasion. When approaching a fire, it should always be done by the door, if possible. When this is attended to, it is much easier to shift the hose from one apartment to another; and the current of fresh air, entering by the door and proceeding along the passages, makes respiration easier and safer than elsewhere. When entrance by the door is impracticable, and access is to be gained by a window, the flames frequently burst through in such a manner as to render advance in the first instance impossible. In that case, the branch should be pointed against the window, nearly in a perpendicular direction; the water striking the lintel, and falling all round inside the window, will soon extinguish the fire at that point sufficiently to render an entrance practicable. The old plan of standing with the branch pipe in the street, and throwing the water into the windows is a very random way of going to work; and for my own part, although I have seen it repeatedly tried, I never saw it attended with success. Indeed it is hardly to be expected that water, thrown from the street into a room three or four storeys high, can have any impression on closets, presses, or passages, divided probably with brick partitions in the centre of the house. The circumstance of having engines at work on both sides of the house does not alter the case. The fire very often burns up through the centre, and frequently, when the space between the windows is large, along the front or back wall, till it arrives at the roof, which the water cannot touch on account of the slates or tiles. On the other hand, when the firemen enter the house, the fire is almost wholly under their command. And when it happens that there is any corner which the water cannot directly strike, the fire in it may often be extinguished by throwing the water against an opposite wall or partition, and trusting to the recoil to throw it to the point required. When the water is thrown from the street, it is impossible to say whether it touches the parts on fire or not. No one can tell anything about it, except when the flame appears at the windows.