522 HEIGHT AND WEIGHT IN RELATION TO BUILD 



age but who vary in stature. For each subgroup the average or mean weight 

 is divided by the cube of the stature characteristic of the subgroup. From the 

 empirical data thus obtained it is possible to calculate curves to illustrate the 

 weight and index typical of a given stature at a given age. Curves of this kind 

 are shown in charts B, C, D, and H. Weight-for-height curves are hyperbolic in 

 character. 



We may now consider some of the characteristic features of curves based on the 

 height-weight index of build in infancy, childhood, adolescence, and maturity. 



For infancy, data on height in relation to weight are scanty except for the new- 

 born. Such data as exist are somewhat contradictory. After plotting the available 

 data I have come to the conclusion that for the present it is best to consider the 

 height-weight index for this period as a straight line with a uniform value of 0.918 

 (0.915 for females) for any given stature from 20 inches to 27 inches (26.5 inches for 

 females) . The choice of 0.918 as the value of the index is purely empirical and can be 

 considered only as a rough approximation. As such, however, it is of value, because 

 wide departures from it enable us to judge whether a child is abnormally thin or 

 abnormally fat, and, within limits, to what extent. In the use of 0.918 as a standard 

 height-weight index for this period (0.02543 is the corresponding centimeter-gram 

 ratio) , the following facts should be taken into consideration : 



1. At the time of birth full-term short infants are relatively heavy; long infants 

 are relatively light. 



2. Growth, during infancy, even in healthy babies, is seldom uniform. Periods 

 of rapid growth in length are apt to alternate with periods of less rapid growth in 

 length and a relatively more rapid growth in weight. 



3. Statistics relating to groups of healthy infants as a rule show a relatively 

 large body-weight when the measured length is relatively short for age and vice 

 versa. 



4. To some extent differences in height-weight indices based on the obser- 

 vations of different investigators are to be attributed to variation in method of 

 measuring length. 



The index 0.918 appears to be approximately normal for an infant 20.3 inches 

 (51.56 cm.) long at birth, a length which may be taken as normal for American male 

 children. Holt (1916) gives a slightly greater length (20.61 inches, 52.3 cm.), 

 Taylor (1918) a slightly less length (51.18 cm.). Infants shorter than this, as a 

 rule, have a higher index, i. e., are relatively heavier; those longer are relatively 

 lighter. Thus the male infants of the British Anthropometric Report 1883 with the 

 average length of 19.52 inches (49.58 cm.) have an index of 0.957 (metric 0.0265), 

 while the male infants of the Pearson series (1899), with a mean length of 20.503 

 inches (52.08 cm.), have an index of 0.847 (metric 0.02344). 



In chart B several curves are given to illustrate the relative proportions of the 

 body in the new-born. The general construction of this chart is like that of chart A 

 and therefore need not be described. 



The heavy line which extends upwards at the left side of the 0.900 column as 

 far as the 27-inch-stature line and then swings toward the right represents the 



