IN his Guthrie Lecture on the subject, Prof. A. V. Hill pointed out that the study of the heat given out by muscles in relation to the work done by them is one of the classics of physiology. Until recently, however, the matter appeared much more complicated than it really is, owing to technical difficulties. These have been overcome by the use of a very rapid recording system and an insulated thermopile only 0.002 inch thick. Some very simple relationships have now emerged. An active muscle liberates energy in three forms: in maintaining a contraction, as heat ; in shortening, as heat ; in shortening against load, as work ; its behaviour in any circumstances is deduced from the resultant of these three. Rate of total energy liberation of a muscle is determined by the load upon it, increasing as the load decreases. This allows a simple equation to be deduced for the relation between speed and load. The constants of the equation are the same whether they are obtained by thermal or by mechanical measurements. The fact that a muscle does less external work when shortening at a higher speed has led to the hypothesis that muscle is endowed with 'viscosity', attributed to a lag in the rearrangement of its molecules, as the external form of the contractile elements changes. This viscosity hypothesis is, however, altogether unnecessary ; for the decrease of force and work with increased speed can be deduced from the manner in which the energy liberation is regulated. Some applications were also described. The maximum power developed by a muscle is with a load about three tenths of the maximum load it can bear. The highest efficiency (work/total energy) is with a load of about 0.45 of the maximum. These are near enough for maximum power and maximum efficiency to occur very nearly at about 37 per cent of the maximum load. These results obtained with frog's muscle almost certainly apply, though possibly with different constants, to man, and it would be very important to find out and to determine the constants of human muscle. The technique required would be a very different one.