As an example of the converse proposition of the transformation of heat into mechanical work, the steam-engine affords the aptest illustration. The original power came from the sun millions of years ago, and did work by enabling the leaves of plants to overcome the strong mutual affinity of carbon and oxygen in the carbonic dioxide in the air, and store up the carbon in the plant, where it remained since the coal era in the form of energy of position. By lighting the coal, or in other words separating its molecules more widely by heat, we enable them to exert once more their natural affinity for oxygen, and burn, that is recombine into carbonic dioxide. The heat thus produced turns water into steam, which passes through a cylinder, either into a condenser if the steam is at low pressure, or into the outer air if it has been superheated and brought to a higher pressure than that of the atmosphere. The difference of the pressure or elasticity of the steam in the boiler, and of the same steam when it is condensed or liberated, is available for doing work, and, being admitted and released alternately at the two ends of the cylinder, drives a piston up and down, which, by means of cranks and shafts, turns a wheel or does whatever work is required of it. In doing this, heat disappears, being converted into work, and the amount of heat would exactly equal that into which the work would be converted according to Joule’s law, if it could all be utilised without the loss necessarily incurred by friction, radiation, and the still more important absorption of latent heat required to convert water at boiling-point into vapour of the same temperature. This latter is not really an annihilation of the heat, but its conversion into work done in separating the molecules against the force of cohesion. The whole heat, therefore, is transformed into work, mainly molecular work in tearing molecules asunder, and the residue into mechanical work turning spindles and driving locomotives and steamboats.