MUSCLE 665 



muscle acts on the periphery of the wheel, and by rotating it raises 

 the weight a little at each contraction. At the end of the con- 

 traction the wheel is prevented from moving back by a catch. 

 The work done in a series of contractions is calculated from the 

 total height to which the weight has been raised. Suppose a frog's 

 gastrocnemius 'is made to contract a certain number of times while 

 attached to the work-adder, and that simultaneously the heat pro- 

 duction is measured by means of a thermopile. Let H represent 

 the heat actually produced, and h the heat equivalent of the work 

 done. Now let the muscle be disconnected from the adder and 

 made to raise the same weight, directly attached to it, by a series 

 of contractions elicited in precisely the same way as the previous 

 ones, except that the weight is allowed to fall with the muscle when 

 it relaxes after each contraction. Here heat corresponding to the 

 external work disappears from the muscle during the contraction 

 just as in the first experiment, but this heat is returned to the muscle 

 during the relaxation, since on the whole no external work is done. 

 The heat produced in the second experiment is found, as a matter 

 of fact, allowing for unavoidable errors, to be equal to H+ h. 



Here the assumption is made that the difference in the mechanical 

 conditions during the relaxation (the muscle in the first experiment 

 relaxing without load but being stretched by the weight as it relaxes 

 in the second) does not affect the heat-production. This assumption 

 has been shown to be correct, although it was at one time supposed 

 that changes in the tension of the muscle produced during the 

 relaxation did cause changes in the amount of heat produced. On 

 the other hand, it has been clearly proved that the total energy 

 transformed during the period of contraction, and the fraction of 

 it which appears as external muscular work, are greatly influenced 

 by the mechanical conditions under which the contraction takes 

 place. A stretched muscle, when caused to contract, produces 

 more heat than if it had started without tension, and still more 

 heat when it is fixed so that it cannot shorten during stimulation. 

 A muscle, starting without tension, produces more heat when it 

 contracts isometrically than when it contracts isotonically. This 

 fact does not, however, prove that the heat-production is greater 

 when no work is done, because the tension increases during excita- 

 tion when contraction is prevented, and, as has been said, increase 

 of tension alone causes more heat to be given out by an active muscle. 



When a muscle, excited by maximal stimuli, is made to lift 

 continuously increasing weights, both the work done and the heat 

 given out increase up to a certain limit. The muscle, as it were, 

 burns the candle at both ends. This would be of itself enough to 

 show that there is no fixed relation between the work and the heat- 

 production ; although the latter reaches its maximum somewhat 

 sooner than the former. 



Although a loaded muscle kept in steady tetanic contraction is doing 

 no work, it produces heat, but far less than would be produced if the 

 muscle could fully contract and relax at each excitation. The amount 

 of energy liberated by an excitation of given strength depends, there- 

 fore, on the mechanical condition of the muscle into which it falls. 



The fraction of the total energy transformed which appears 

 as muscular work varies with the conditions of the contraction. 

 The greater the resistance, so long as the muscle can overcome 

 it so as to do its utmost amount of external work,* the larger 



* This statement, based on experiments with excised frog's muscles, is 

 not, of course, inconsistent with the fact mentioned on p. $83, that in the 



