104 Mr. J. S. Macdonald. Studies in the 



chemical changes underlying this transformation of energy. This is true of 

 such experiments in general, and in this particular case has been found to 

 be true by Dr. Duffield, investigating the carbonic output in similar experi- 

 ments in the calorimeter. There is little doubt that the curve of trans- 

 formation of chemical energy has a much greater parallelism to that of work 

 performance, and would, therefore, in this case be represented by a straight 

 line. It follows then that my experimental curve of " heat-production " has 

 only a secondary relation to the real curve of transformation of energy, and 

 that I am losing sight of some part of the energy transformed, large at 

 first, but subsequently diminishing. I find myself therefore obliged in dealing 

 with the events of the first hour to consider the following possibilities : — 



A. That a storage of heat in the disc of the cycle, representing some large 

 fraction of the mechanical work clone, is of great importance, and has been 

 neglected. 



B. That there is a storage of heat within the subject, as, for instance, 

 within his musculature, to which the rectal temperature is no adequate index. 



C. That some of the energy transformed in the oxidation processes 

 accompanying the performance of work is stored in the body in some form 

 other than heat. 



Now, the first of these possibilities, dealing with the rising temperature of 

 the copper disc of the cycle, is certainly a source of some error. It cannot, 

 however, be considered as a source of errors of the magnitude present in this 

 case, since at the most it could hardly account for more heat than is due to 

 the conversion of the mechanical work performed, and this value is no more 

 than a small fraction of the missing quantity. Obviously again, the third, 

 and extremely interesting, possibility is best left on one side until the 

 second one is adequately considered, namely, that under working conditions 

 the rectal temperature fails to represent the real average temperature of the 

 body. It has been noted that, during conditions of rest, the rectal tem- 

 perature may be taken as an index to variations in the real average 

 temperature, even if it is not on absolutely the same level as that average 

 temperature, since the variations in the two are probably similar. Thus, it 

 has been observed that temperature observations taken under such conditions 

 at several different sites (e.g. skin and rectum) give parallel curves. In this 

 case there is no such parallelism, as may be seen from the curves of rectal 

 (uppermost) and surface (middle) given in the figures. 



It is, indeed, probably the case that the rectal temperature represents no 

 more than the mean temperature of the mixed blood sent from the heart to 

 one part of the general circulatory system, a particular part in which, owing 

 to an absence of much local performance of work, this mean temperature is 



