ON CALORIMETRIC OBSERVATIONS ON MAN. 287 
In addition, the weight of the water leaving the internal radiator 
system of the calorimeter was measured at intervals of time determined 
by the rate of water-flow and the comple%e filling of a pan placed upon 
one arm of a balance. 
From observations (a), (b), (c), (d) the heat-output of the subject 
(plus any heat arising from a subsequent conversion of mechanical 
work into heat) was. calculated for each five-minute period and the 
results plotted as curves. By correction from observation (g) these 
curves, altered by allowances for the storage of heat in the subject, 
were converted into heat-production curves—that is to say, curves 
representing the total transformation of energy within the calorimeter. 
This endeavour to obtain results capable of erection into curves 
showing variations corresponding with such short time intervals has 
been unexpectedly justified by coincident observations of the tempera- 
ture of the subject’s surface from which the heat is mainly passing into 
the calorimeter. The curves of heat output obtained by the complicated 
addition of data from observations (a), (b), and (c), and by subtraction 
of data from observations (d) are parallel to the surface temperature 
curves obtained simply by one set of observations during the first half- 
hour of each ‘ work experiment,’ that is to say, so long as the observa- 
tions of surface temperature are not complicated by the accumulated 
presence of surface moisture, and in some of the extremely light ‘ work 
experiments ’ continue in parallel fashion to the end of the experiment 
whilst’ showing corresponding variations at nodal points. 
This confirmation of the value of these curves brings into greater 
prominence the fact that the ‘ heat-output’ curves, and even the 
‘ heat-production ’ curves—which latter represent an attempt to 
describe the total transformation of energy within the calorimeter— 
rise gradually for the best part of an hour to summits that are then 
more or less maintained during any further period of continuance in 
the performance of uniformly maintained mechanical work—a fact upon 
which I have elsewhere commented. It would thus seem as if the 
transformation of energy per unit of mechanical work performed was 
a quantity that increased up to a certain value which was then main- 
tained, and that; the ‘ efficiency’ of man as a machine varied in this 
fashion with the time spent in work. 
This may be the case, but is improbable as not supported by 
evidence obtained by other observers in different ways—as by the quan- 
titative examination of the carbonic-acid output. If it is not the case, 
then two other lines of explanation have in addition to be examined. 
Thus it may be that the ‘ deep temperature ’ (rectal) is not a satisfac- 
tory criterion of the mean temperature of the human body and does 
not therefore provide a proper basis for corrections representing its 
average storage of heat during any short period (five minutes) of time. 
It might, on the other hand, be the case that energy liberated during 
the performance of mechanical work as the outcome of oxidation 
processes developed as fully at the commencement as at the end of 
the experiment might be stored within the body, possibly within the 
musculature, in some form other than heat, as, for example, in the 
