800 



REPORT 1893. 



not therefore, strictly speaking-, give the value of heat-emission as constant per 

 degree T.D., nor make, as above, the assumption that cal. = T.D. x k. But the dis- 

 crepancy is not very serious within a range of 5° or even of 10°, and it is easily 

 eliminated. Thus we may treat the discharge at 10° T.D. as a constant, and give 

 the variations per degree above this value. Or, better, we may draw up a table 

 of the values T.D.'''^^^ x k. The following table gives in three parallel columns the 

 calorimetric values per minute per 1,000 cm.- of temperature-differences such as 

 are ordinarily observed on the human forearm ; 1° on the least accurate assumption 

 cal. = T.D. X K ; 2° on the more accurate assumption cal. = C + (T.D. — 10) k ; 3° on 



the least inaccurate assumption cal. = k T.D. 



have been treated as negligible, 

 pex minute per cm.- :^ 



Variations of barometric pressure 



Calories per minute per 1,000 cm.'-; or millicalories 



2. The coefficient of radiation varies with the nature of the radiating surface. — 

 Heat is emitted from a warm surface, such as the human skin, by conduction, by 

 evaporation, and by radiation. We may not assume that loss by radiation is 

 identical from the surface of the indiarubber sphere and from an equal area of 

 human skin. Moreover, we must admit as theoretically possible that the radiating 

 power of the skin, apart from alterations of temperature, may vary. 



I have not attempted to determine what is the percentage of the total loss borne 

 by radiation alone, nor the variations of that percentage with varying character of 

 surface. I have (1) taken simultaneous and separate estimates of the loss (a) by 

 evaporation, (U) by conduction and radiation conjoined ; (2) made experiments to 

 see whether the loss by radiation cum conduction was sensibly altered by gross 

 differences of the radiant surface ; (.3) compared roughly by thermopile and galvano- 

 meter the radiation of a warm indiarubber surface with that of the human skin ; 

 and (4) tried whether radiation from the human skin varied parallel with variations 

 of T.D., or in such wise as to suggest the intervention of a distorting factor, such 

 as an alteration of the radiating power. 



All these were comparatively rough experiments, made not to determine actual 

 alterations of heat-emission with alterations of radiant power, but to determine 

 whether the latter could be regarded as markedly influencing the result, and thus 

 forbidding the translation of surface thermometer readings of T.D. into calorimetric 

 values of heat-emission. 



I found (1) that two otherwise similar spheres with respectively white and 

 black covers showed no difference in the rate of cooling outside the range of 

 experimental error, apart from variations due to differences of T.D., or of thickness 

 of cover, or of moisture of cover, or of air ; (2) that practically the amount of 

 radiation was proportional to the T.D., within a range of variation of 10° C. ; 

 (.3) that the radiant power of the indiarubber surface did not sensibly differ from 

 that of the skin of the forearm, with identical values of the T.D. 



From which I concluded (1) that variations of radiant power are negligible in 

 observations of this order ; ' (2) that the absolute calorimetric standard obtained 

 from observations on the cooling sphere of water might without gross error be 

 applied to the heat-emission of the human skin. 



3. The evaporation factor in heat-emission has to be separately dealt with. — To 



' Dr. Stewart, in the course of experiments on the radiation from the animal body 

 (^Studies from the Physiological Laboratory of Omens College, Manckester, 1891), has 

 anticipated me in this first conclusion, to the effect that heat-emiss'.on depends upon 

 the T.D. and not upon variations of radiating power. 



