514 VITALITY AND EFFICIENCY WITH RESTRICTED DIET. 



decrease about November 24 but no appreciable subsequent change 

 about January 26; with Moy a continuous decrease throughout the 

 entire period; with Pea a decided fall with practically constancy on 

 November 24 and January 26; with Pec a continuous fall on the three 

 dates; with Spe a considerable fall on November 24 but no record for 

 January 26; with Tom a progressive decrease; with Vea a marked fall on 



Table 129. — Heat per kilogram per S4 hours {computed from individual respiration experi- 

 ments) at different diet levels — Squad A, subjects post-absorptive and in lying position. 



^ See table 133 (footnotes) for explanation of period averages. 



' Minimum during sleep, after standard supper (700 calories) at restaurant from 6 to 12 hours 



before the experiment. Computed from values in table 131, for experiments on Sept. 



29-30, Nov. 24-25, and Jan. 26-27. 



November 24 but no subsequent change. The picture of the squad 

 as a whole is shown by the average values, which are for October 1, 



25.2 calories, for November 24, 23.1 calories, and for January 26, 



22.3 calories. Here we have a clear picture of a continuous reduction 

 in heat production per kilogram of body-weight from the beginning to 

 the end of this long test, a picture fully confirmed by the data computed 

 from the results of the experiments with the group chamber on the 

 nights of September 29-30, November 24-25, and January 26-27. 



Heat Output per Square Meter of Body-Surface with Low Diet, Squad A. 



A second popular method for comparing different individuals and 

 the same individual under different conditions is on the basis of the 

 heat production per square meter of body-surface. Employing the 

 measurements of body-surface obtained by the Du Bois linear formula, 

 which have been verified both by the height-weight chart and by the 

 photographic method (see p. 241), we have to deal here with true 

 body-surfaces rather than with the erroneous values computed by the 



