It is almost impossible to determine with accuracy, in regard to isolated muscles, the amount of food taken up from the blood, and the return in work by the muscles of the energy potential in the food. Calculations have to be based upon observations of food consumed, gain or loss of body-weight, work done by a man or an animal during a period lasting for several days. We shall consider the evidence obtained in this way in a subsequent section (p. 149). But whether we study isolated muscles or the body as a whole, the relation between work and heat varies within wide limits. So wide, indeed, are the variations as to justify the conclusion that there is no necessary relation between the two phenomena. Muscles develop heat when they are quiescent. Activity is accompanied with an increased evolution of heat; but, if it be desirable, the evolution of heat is reduced until it is, relatively to the output of work, much smaller than in the case of any engine which has yet been made. It is sufficient in this connection to state that, under certain conditions, the return in work may amount to about one-half. The comparison with an artificial motor, of whatever kind, breaks down. In an engine combustion develops heat, heat causes steam or gas to expand, the expanding gas pushes a piston. In muscle certain of the carbon, hydrogen, and oxygen atoms contained in protoplasm combine to form water and carbonic acid—compounds too stable to be reassociated with the remaining atoms of the protoplasm-molecule. They are replaced by complex, energy-yielding substances—foods—and by oxygen, carried in the blood. Their displacement brings about a change in the form of the molecules which involves, owing to their peculiar orientation, a change in shape of the muscle as a whole. Such an explanation is, perhaps, more exact than our knowledge at present warrants; or rather let us say, since we do not know what the expression “the form of a molecule” means, it has an appearance of an exactitude which does not characterize it. It is merely intended to help the reader to realize the hopelessness of attempting to compare muscle with any mechanical contrivance. In the boiler of a steam-engine heat is applied to water until its molecules cannot remain in so close a state of aggregation. Their orbits are greatly increased. The cause of the thrust given to the piston of an engine is the increased amplitude of movement of the molecules of steam behind it. In a combustion engine a mixture of petrol and air is ignited. Energy is set free by the resolution of unstable petrol into stable water and carbonic acid. This energy heats the gases, causing them to expand. Waste of energy as heat is inevitable in a machine which depends for its motive-power upon the translation of molecules. The source of muscular force (if it be not intramolecular change) is certainly not, directly, increased amplitude of molecular swing.