METABOLISM AS AFFECTED BY GROWTH. Ill 



of the metabolism values from day to day is best made by some 

 graphic form of representation, such as a chart. This comparison has 

 been made in figure 15, in which, to avoid confusion from a multi- 

 plicity of points, the most important (19 in all) have been selected 

 for plotting. 



In selecting the points for a chart to show the changes in the metab- 

 olism of an individual as the age increases, values for single days 

 could not advantageously be used. They were accordingly not 

 infrequently averaged, most of the points in this chart being made 

 up of one or more days. Since a selection of data was necessary, which 

 is always undesirable in the preparation of scientific reports, we have 

 given in table 26 a summary of the actual minimum values and averages 

 used in figure 15, to show exactly how these points were derived. 

 Many of these values were likewise used on the general charts or 

 group summaries (see figures 23 to 47, pages 135 to 175), those not so 

 used being indicated by an asterisk. Thus, this child (No. 145) 

 appears on the general charts nine times, or as nine different indi- 

 viduals, since at all of these nine points there was sufficient differ- 

 ence in either age or weight to meet our requirements for the designation 

 of children as new individuals. We believe such points may properly 

 be included in any general chart in which values for a number of 

 children are represented. 



For all of the individual charts (see figures 15 to 21, pages 114 to 

 130), which were prepared primarily to indicate the basal metabolism 

 at different ages of the same child, well-substantiated figures were 

 invariably sought. In some instances it may be a little difficult for 

 the reader to see why certain values are not selected or included in an 

 average. For instance, the values obtained with the child No. 145 

 on June 21, 1917, appear in figure 15, while those for June 20, 1917 

 (see table 25), do not, although one might naturally expect that an 

 average for these two days would be used. Since, as poined out on 

 page 106, the possibility of an analytical error is always present, it 

 can be seen that the value of 565 calories for the first period on June 

 21 is not substantiated, except that it agrees essentially with the low 

 values on May 18. The minimum value on June 20 (627 calories) is 

 noticeably higher than that found in two periods on June 21. Conse- 

 quently, it was necessary here to make a selection of material, and it 

 seemed logical to choose for charting an average of the three periods 

 on June 21, i. e., 600 calories, rather than to use the unsupported 

 minimum. 



On March 5, 1918, a very low value of 539 calories is noted in the 

 first period, with 620 calories in the following period. Here again we 

 have an unsupported minimum figure which it seems unwise to include, 

 as there was a change in age of only 4 months, i. e., January to March. 

 The value of 620 calories, although fitting well into the general curve, 



