216 BIOLOGY OF DEATH 



It would seem, at first thought, that one should be 

 able to test the theory here suggested, that rate of energy 

 expenditure in the business of living is negatively corre- 

 lated with the total duration of life, by an examination of 

 the mortality rates for persons in different occupations 

 as set forth, for example, in the well known paper of 

 Bertillon. When one endeavors to make such a test, 

 however, he is at once confronted with, a series of diffi- 

 culties which presently convince him that the project 

 is virtually an impossible one, if he wishes critical results. 

 In the first place, mean age at death will not do as a 

 criterion, because of the great differences in the age dis- 

 tributions of those engaged in different occupations. 

 This point has lately been thoroughly discussed by Collis 

 and Greenwood, in their book l i The Health of the Indus- 

 trial Worker. ' Indeed, their whole treatment of the prob- 

 lem of occupational mortality is by far the most sound 

 and critical which the present writer has yet seen. One 

 must deal with age and sex specific death rates, or mor- 

 tality indices based upon them. 



In the second place, there are specific hazards, direct 

 or indirect, in various occupations, quite apart from any 

 question of energy expenditure involved in the case. 

 These hazards will, obviously, tend to obscure any direct 

 effects of the energy relations involved. 



In the third place, we have only the merest suggestion 

 of quantitatively accurate knowledge as to the average 

 energy output involved indifferent trades and occupations. 



On the last point, a beginning to collect information 

 has been made by Waller and his co-workers. In a re- 

 cent paper Waller and De Decker have given the mean 

 calory output, per hour, per square meter of body surface 

 for a small sample of workers in a few trades. But the re- 



