_Third._ The steam, which at each stroke of the piston fills the passages between the steam-valves and the piston, at the [Pg291] moment the latter commences the stroke will be inefficient. If it were possible for the piston to come into steam-tight contact with each end of the cylinder, and that the steam-valve should be in immediate contact with the side or top of the piston, then the whole of the steam which would pass through the steam-valve would be efficient; but as some space, however small, must remain between the piston and the ends of the cylinder, and between the side of the cylinder and the steam-valve, there will always be a volume of steam bearing a sensible proportion to the magnitude of the cylinder, which at each stroke of the piston will be inefficient. This volume of steam is called the _clearance_. _Fourth._ Since the piston must move in steam-tight contact with the cylinder, it must have a definite amount of friction with the sides of the cylinder by whatever means it may be packed. This friction will produce a corresponding resistance to the moving power. _Fifth._ The various joints of the machinery where steam is contained are subject to leakage, and whatever amount of steam shall thus escape must be placed to the account of power lost. _Sixth._ When the eduction-valve is opened to admit the steam to the condenser, a certain force is required to expel the steam from the cylinder. This force reacts upon the piston, and counteracts to a proportional extent the moving power of the steam on the other side. Besides this the water in the condenser cannot be conveniently reduced below the temperature of about 100°, and at this temperature steam has a pressure of about 1 lb. per square inch. This vapour will continue to fill the cylinder, and will resist the moving power which impels the piston. _Seventh._ Power must be provided for opening and closing the valves or slides, for working the air-pump, hot-water pump, and cold-water pump, and finally to overcome the friction on the journals and centres of the parts of the parallel motion, the main axle of the beam, the connecting rod, crank, and fly-wheel axle. It will be apparent how very much these sources of resistances must vary in different engines, and how rough [Pg292] an approximation any general estimate must be of their gross amount.