Friction (Lat.fricare, to rub), in mechanics, the resistance caused by the moving of the surfaces of bodies over each other. It is usual to distinguish two kinds of friction, that which is produced when bodies slide one upon another, and that which takes place when they roll one upon another. The term rolling friction is not, however, regarded as strictly correct, and that of resistance to rolling is used instead. The first experiments upon the friction of sliding were made by Amontons, and are described in the memoirs of the academy of sciences, 1099; but his estimates were much higher than those which have since been made. Eu-ler, Desaguliers, and Vince also paid considerable attention to the subject, but the first complete set of experiments were made by Coulomb at Rochefort about 1780. His results, although in some respects since modified, have been of inestimable value to the science of engineering. He employed a bench made of two horizontal timbers (3 ft. long, upon which a loaded sledge was drawn by a weight acting by a cord running over a pulley. The resistance bodies offer to motion after they have been for some time in contact he called the friction of departure.

The general conclusions at which he arrived are as follows: 1. Friction is greatest between rough bodies. 2. It is greater between the surfaces of like than of unlike material. 3. The rubbing surfaces remaining the same, friction is proportional to the pressure, and is not increased or diminished by increase or diminution of surface. Some uncertainties in the observations of Coulomb, and the introduction of many new materials in machinery, made it desirable to make a more extended series of experiments. Such were made at Metz in the years 18.31, '2, '3, and '4, by M. Morin. The values obtained by him differed in some particulars from those of Coulomb, but the general conclusions at which he arrived were the same. He however established one important fact scarcely to be anticipated, viz., that friction is independent of the velocity of motion. The ratio which the resistance offered to sliding between two surfaces bears to the force with which they are pressed together is called the coefficient of friction, and has greatly differing values between different surfaces, and different conditions of surfaces as to whether they are highly or partially polished, moistened, or lubricated.

It has various values between different kinds of wood, depending upon whether the motion is made across or with the fibres, and the condition of the wood; and also between different kinds of metals, and with these depends upon whether they are rolled, hammered, cast, or tempered. Thus the coefficient of friction of motion between oak and oak in a direction parallel with the fibres was found by M. Morin to be, without lubrication, about 8/17; lubricated with tallow, about 1/10; with lard, about 1/15. When the fibres of one surface were perpendicular to the line of motion, the coefficient was, without lubrication, about 13/40; lubricated with tallow, about 1/12; with lard, about 1/14; with water, about -1/4. The coefficient of friction between common wrought and cast iron is about 1/5; of iron on brass, 1/7; that of an iron axle in a brass box, lubricated, about 1/40. The least possible friction is found in the use of lubricated steel moving upon hard gems. Coulomb found: 1, that resistance to rolling varies in an inverse ratio with the diameter of the rolling body theoretically, but that in practice small rollers of wood caused more resistance, because of the greater indentation produced, the coefficient ranging from 6/1000 to 17/1000; 2, that it is less between heterogeneous than between homogeneous surfaces; 3, that it is directly proportional to pressure; 4, that it has no relation to surface.

Upon this principle depends the advantage of using friction wheels and friction rollers in machinery. The application of friction wheels is said to have been first made by Henry Sully in 1716. The friction caused by water in moving over surfaces in conduits is called hydraulic friction. It has been found to be independent of the material of the surface of the conduit, provided it be smooth, but depends considerably on the viscosity of the liquid; thus, ice-cold water offers greater resistance to the passage of a body through it than warm water, and conversely, produces a correspondingly greater degree of friction in moving over surfaces. Friction always develops heat, and precisely in proportion to its amount, as has been established by the experiments of Count Rumford, Davy, Thomson, Mayer, and Joule. By rubbing two pieces of ice together in a vacuum, Sir Humphry Davy partially melted them. Count Rumford found the heat developed in boring a brass cannon sufficient in the course of 2 1/2 hours to raise 26 1/2 lbs. of water from zero to 212° F. At the Paris exhibition in 1855 MM. Beaumont and Mayer exhibited a machine in which a wooden cone covered with hemp made 400 revolutions per minute inside of a hollow copper cone immersed in a tightly closed boiler.

With this apparatus 88 gallons of water were raised from 50° to 226° F. in a few hours. In all cases the quantity of heat evolved by friction is exactly sufficient to reproduce the power expended in overcoming the friction; and although in mechanics friction is said to cause a loss of power, there is really no loss of energy, but simply its transformation. Another kind of energy is developed by friction, viz., electricity; and in this case also it has been found that the force produced is precisely proportional to that which was expended in producing it.