162 METABOLISM AND GROWTH FROM BIRTH TO PUBERTY. 



the total calories were referred to weight in figure 26 (page 140). 

 Since there is a close relationship between surface area and weight, 

 this is not surprising, but it is important to note the scatter of the 

 various points about the hypothetical smoothed curve. If the heat 

 production is more closely proportional to the surface area than it is 

 to the body-weight, we should logically expect a closer grouping of the 

 points around the curve. A superficial inspection of the curves in 

 figures 38 and 26 does not indicate that the scatter varies much in 

 the two curves. If anything, it would appear that the points lie 

 closer to the curve in the body-weight chart (fig. 26) than in the 

 body-surface chart (fig. 38). The question of absolute scale values 

 for ordinates and abscissa enters here, however, and since subse- 

 quently both of these charts and smoothed curves are put to special 

 use in an attempt to predict the heat production of the various children, 

 further discussion at this point is unnecessary. 



The smoothed curve indicates a decidedly rapid increase in metab- 

 olism with increasing surface area up to about 0.6 square meter. 

 Thereafter, although there is an increase in heat with larger surface, 

 it is somewhat less in degree, as indicated by the sketched curve. 

 Considering the variations in total calories for boys of like surface, 

 we find some striking differences. For example, with the surface of 

 1.3 square meters, one boy has a heat production of 1,096 calories and 

 another of 1,401 calories, a difference of 28 per cent. Again, at about 

 0.80 to 0.85 square meter, we have variations from 716 to 984 calories, a 

 difference of 268 calories, or about 37 per cent; while at approximately 

 0.53 square meter we have a range of 456 to 684 calories, a difference 

 of 228 calories, or 50 per cent. 



These percentage differences are essentially of the same order as 

 that noted in discussing the chart showing the relationship between 

 weight and total calories (fig. 26, page 140). Indeed, by reference 

 to the discussion on page 141, it will be seen that the differences there 

 were, if anything, somewhat less. So far as this picture thus far goes, 

 the evidence is slightly in favor of a greater degree of regularity in the 

 relationship between heat and weight than between heat and body- 

 surface, even though we are now referring to body-surface accurately 

 measured and not approximately computed. A more rigid test of 

 this, however, will be made when we come to consider in a subsequent 

 section the utilization of the general smoothed curves in figures 26 

 and 38 as a basis for predicting the heat production of unknown 

 subjects. What is of special physiological significance, however, is 

 that the general picture presented by the chart in figure 38 indicates 

 that the heat production referred to surface area is almost identical 

 with that in which reference is made to body-weight. 



Extremely few children studied by earlier investigators can be 

 compared to our measurements. They are confined exclusively to the 



