Surface-friction resistance, especially for vessels moving at moderate or slow speeds, is much greater than the resistance due to other causes—that is if the hull is ordinarily well formed. Its amount depends upon the area of the immersed surface, upon its length, upon its degree of roughness, and upon the velocity with which the water glides over it—_i.e._, upon the speed of the vessel. Eddy-making resistance only acquires importance in exceptional cases, _e.g._, in ships having unusually full sterns. In ordinary well-formed ships it is of small amount, and is caused mainly by blunt projections such as shaft tubes, propeller brackets, and stern-posts. Wave-making resistance is much more variable than surface-friction resistance. Its amount depends on the form and proportions of vessels, and on the speed at which they move: being greatest, of course, in ships of full form and in those moving at high speeds. The sum of these three main elements of resistance constitutes the total resistance experienced by a vessel if “towed” through the water, that is, the resistance considered apart from the action or influence of the propelling instrument. In the case of a steamship, however, propelled by a screw or paddle-wheels, the resistance is augmented, more or less considerably, according to the form, surface, and disposition of the propelling instrument. By the employment of various formulæ deduced by scientific authorities from theory and experiment, an approximation can be made before-hand to the total resistance of a proposed vessel, and from this an estimate of the power required to drive her at a certain speed. Moreover, through the law of comparison propounded by Mr Froude, the resistance of a ship can at all times be deduced with fair accuracy from the resistance of her model, certain corrections well determined by experiment having to be made. The power of marine engines is expressed either in “nominal” or “indicated” horse-power. Nominal horse-power is a term practically obsolete so far as being a measure of the efficiency of engines, and only exists as a conventional method of commercially measuring the sizes of engines. Indicated horse-power measures the work done by the steam in the cylinders during a unit of time, and 33,000 units of work per minute, or 550 units of work per second, constitute one horse-power. The effective mean pressure of the steam is ascertained from diagrams drawn by means of the instrument known as the “Steam Engine Indicator,” and hence the term “indicated” horse-power.