The first two columns have already been considered in 

 Table 8 (page 16). They give the average weights for each inch 

 of height. The average weights as given may be considered as 

 norms for our purpose, except at the extremes of height, where 

 the number of cases observed is small. These figures may be 

 replaced by the smoothed "values (also in Table 8) to which 

 reference has already been made. 



The problem now resolves itself into this: What departures 

 or deviations may be permitted from the average weights in 

 granting certificates? The children who are above the averages 

 may be excluded from consideration. They are, if anything, 

 better physical specimens of their class. Where overweight is 

 so marked as to be a menace to the child, the physical exam- 

 ination will be likely to disclose defects of heart, kidney or other 

 vital organ which should bar the applicant. The under- 

 weights are therefore our sole consideration, and the question 

 still further reduces itself to the determination of the minimum 

 weights which are consistent with physical fitness. 



An interesting suggestion was made in this connection by 

 Porter*, who urged that a child whose physical development 

 departs more than the probable deviation from the average 

 weight for his height, should be considered sub-normal. This 

 suggestion attracted our attention, for it seemed to give the 

 very value which would serve as a first approximation for the 

 determination of the minimum weights to be used by health 

 oflicers. We have used this suggestion in Table 16. The 

 standard deviations ( <* ) are given for the average weights for 

 each inch of height. If we follow the method of Porter, the 

 children whose weights were below the average by a number 

 of pounds equal to the standard deviation, would be excluded. 

 Column 4 gives the minimum weight values on this basis. 

 Column 5 gives the number of children who would fall below 

 this limit (A-<r). We have also included the percentage 

 deviation from the average which is allowed according to this 

 method, namely, . This is shown in Column 6. Among 

 boys, age 14 to 15, there would be 437, or 12.8%, who would 

 have to be excluded on this basis. In like manner, 247 boys, 

 or 12.4%, would be eliminated at age 15 to 16; 378 girls, or 

 13.4%, at age 14 to 15; and 254 girls,, or 13.9%, at age 15 to 16. 



*W. Townsend Porter, "On the Application to Individual School 

 Children of the Mean Values Derived from Anthropological Measurements 

 by the Generalizing Method." Publications of The American Statistical 

 Association. Vol III, Boston, 1893. 



38 



