the curves are somewhat irregular, especially in the older age groups, the mode is fairly well defined for each measurement. In regard to weight, the modal weight, like the average weight, is in each age group distinctly less for the "fair" or "poor" nutrition group than for the children of "good" or "excellent" nutrition.

But it is important to note that, although the children in the "fair" or "poor" nutrition group weigh less on the average than those in the "good" or "excellent" group, a large proportion of the children in the well-nourished group weigh less than some in the "fair" or "poor" group and a considerable proportion weigh less than the average weight of the children of "fair" or "poor" nutrition. Figure 2 brings out this fact graphically. While all the children in the lower nutrition groups are not classed as poorly nourished, these curves seem to indicate that weight measurements alone are not sufficient for determining nutrition, and a check by a physical examination is necessary to detect at least certain cases of malnutrition. As regards standing and sitting height, the differences are not so clear; the modal heights of the two nutrition groups appear to be practically the same in the majority of the age groups, but the curve for the less well nourished is in practically all cases a little nearer the left-that is, heights tend to be a little less in the "fair" or "poor" nutrition group. This tendency seems to increase as age increases, since as regards both standing and sitting height the mode in the last two age groups (12-13 and 14-16 years) seems to be appreciably less for the "fair" or "poor" group than for the wellnourished children.

In order to show more clearly this relation of weight to nutritional status, an attempt was made to construct weight curves representing each of the four nutrition groups used. In order to get sufficient numbers of children for the two extreme groups ("excellent" and "poor"), it was necessary to consider the weights of children of all ages at the same time. To do this legitimately the following statistical procedure was used: (1) Average weights were calculated for children of each sex for each year of age and the percentage deviation in the weight of each child from the average weight of all children of the same sex and age was computed; (2) a distribution according to these deviations in weight was then made for each nutrition group by computing the number of children who were not over 3 per cent above or 3 per cent below the average weight for their own sex and age, then the number who were 4 to 10 per cent above, 11 to 17 per cent above, etc., in 7 per cent classes; in a similar manner the children who were below the average were divided into classes of the same interval; (3) these numbers were combined for both sexes and all ages, but each nutrition group was kept separate. The result was a frequency distribution for children of each of the four nutrition grades according to the percentage deviation from the average weight of all children of the same sex and age. Table V shows this distribution in actual numbers of children in each class.

Number of children who are each given per cent above or below the average weight of all children of the same sex and age.

Excellent.

152

3

Good.

5

19

8, 153

25

26 |

2

Fair

1,534

7

33

31 136
83

Poor.

1,391

861

485

141

47

134

19

3

1

6

14

31

36

24

13

6

BOYS.

3

3

8

3

2

Percentage of children who are each given per cent above or below the average weight of all children of the same sex and age.

RELATIVE DEVIATION FROM AVERAGE WEIGHT ACCORDING

TO NUTRITIONAL STATUS

PERCENTAGE DISTRIBUTIONS OF CHILDREN IN DIFFERENT NUTRITIONAL GROUPS ACCORD-
ING TO THE PERCENTAGE DEVIATION FROM THE AVERAGE WEIGHT OF ALL CHILDREN
OF THE SAME SEX AND AGE

NUTRITION AS JUDGED FROM CLINICAL EVIDENCE WAS:
EXCELLENT

-GOOD

----- FAIR

............POOR

Since the number of children of "good" nutrition is so much larger than any of the other grades, comparison is difficult. In order to make the different nutrition groups comparable, each distribution was reduced to a percentage basis, 100 per cent representing the total children in any given nutrition group. Table VI shows these percentage distributions and Figure 3 shows them graphically.

Although there are irregularities, the curves in Figure 3, particularly the curves for both sexes combined, show clearly that while the modal weight is greater in the better nourished groups there is a definite overlapping. Some of the "excellent" and many of the "good" group are as much under the average weight as some of the "poor" group. Although height was undoubtedly a factor in this, it seems hardly likely that it would explain the whole difference. Whatever these other factors may have been, it seems clear that weight alone is not sufficient for determining nutrition for the individual child, if the diagnosis of nutrition here used is at all accurate. The classification of children according to nutrition by weight alone would therefore class as undernourished some children who would be judged as well nourished as the result of a physical examination. It would also fail to detect some children who are poorly nourished but who weigh as much as the average. The situation seems to be that undernourished children on the average weigh less than well-nourished children, but the fact that an individual child weighs less or more than the average is not conclusive proof that he is undernourished or overnourished; he may weigh less and yet be well nourished and he may weigh more than the average and yet be poorly nourished. NUTRITION, UNDERWEIGHT, AND OVERWEIGHT IN RELATION TO SEX AND AGE.

It was found that the percentage of children who were judged as not well nourished varies quite definitely with age. Table VII shows by sex and age the percentage of children whose nutrition was set down as "fair" or "poor." These percentages are plotted in Figure 4-A. The proportion of children who were "fair" or "poor" in nutrition as judged from clinical evidence is highest at 8 years of age— 20.7 per cent and falls off as age increases to 7.1 per cent among 16-year-old children. No consistent difference appears between boys and girls as regards the percentage who were not well nourished.