The growth of children is governed by their internal metabolic and hormonal state, which is in turn dependent on hereditary, environmental, nutritional and psychological factors and any severe illness they may experience. The attained height and weight of children are the product of the accumu lated effects of these factors acting since conception, and any study of the reasons for differences in height and weight among children of the same age involves an examination of one or more of these factors. Genetic differences in growth rate have been suggested by Jenss (1940), who found that, within groups matched for socio-economic background, children aged 6 to 8 years of American-born parents and ancestry showed a greater rate of weight gain than American children of the same age of Italian parents. In a comparison of Nilotic negroes and whites living in the Sudan, Roberts (1960) found that Nilotics had a considerably lower weight for height ratio at all ages and heights, though they were not undernourished. The same author pre viously examined statistics throughout the world, showing a close negative correlation between weight per unit stature at ages 6 to 14 years and average annual temperature (Roberts, 1953). The probable explanation given was that natural selec tion, acting over many generations, favours a slender physique in hotter climates, though other factors which vary between climatic regions, such as economic and nutritional conditions, may also be involved. The relationship between the growth of children and their socio-economic background has been extensively studied. Bransby, Burn, Magee, and MacKecknie (1946) found, for example, that children from poor homes or large families were consistently smaller and lighter than those from good homes or smaller families. The authors' assessment of the children's homes was based on both the quality of care and economic conditions. Families were divided into three categories?good, fair and bad?accord ing to reports made by the children's school teachers. These reports were checked by a health visitor, and though both assessments are subjective, they are probably adequate in view of the broad scope of the inquiry. The findings of this study were confirmed by those of the Oxford Child Health Survey (Acheson and Hewitt, 1954), in which chil dren of both sexes from the lower social class groups were found to be smaller than those of the same age from the higher social class groups. Other studies have been concerned with growth in relation to father's occupational status. Douglas and Blom field (1958) found that the more skilled the father's occupation, the greater was the child's growth rate and stature. They also showed that children whose parents had risen in the social scale tended to be taller than those whose parents remained static, and that the children of parents who fell in the social scale were smaller than the average for their original social class group. In a study in Newcastle upon Tyne, Miller, Court, Walton, and Knox (1960) found that 3-year-old children of the pro fessional and salaried classes were on average 2-5 cm. taller and 0-5 kg. heavier than those of semi skilled and unskilled workers. The studies discussed above use very general 154