The growth of children follows an annual rhythm. Wahl, the director of an educational establishment for homeless girls in Denmark, who investigated this question, found that the increase of weight for all the ages investigated was constantly about 33 per cent. greater in the summer half-year than in the winter half-year. It was noteworthy that even the children who had not reached school-age, and therefore could not be influenced by school-life, showed a similar, though slighter, difference in the same direction. It is, however, Malling-Hansen, the director of an institution for deaf-mutes in Copenhagen, who has most thoroughly investigated this matter over a great many years. He finds that there are three periods of growth throughout the year, marked off in a fairly sharp manner, and that during each of these periods the growth in weight and height shows constant characteristics. From about the end of November up to about the end of March is a period when growth, both in height and weight, proceeds at a medium rate, reaching neither a maximum nor a minimum; increase in weight is slight, the increase in height, although trifling, preponderating. After this follows a period during which the children show a marked increase in height, while increase in weight is reduced to a minimum. The children constantly lose in weight during this period of growth in height almost as much as they gain in the preceding period. This period lasts from March and April to July and August. Then follows the third period, which continues until November and December. During this period increase in height is very slight, being at its early minimum; increase in weight, on the other hand, at the beginning of the period (in September and October), is rapid and to the middle of December very considerable, daily increase in weight being three times as great as during the winter months. Thus it may be said that the spring sexual climax corresponds, roughly, with growth in height and arrest of growth in weight, while the autumn climax corresponds roughly with a period of growth in weight and arrest of growth in height. Malling-Hansen found that slight variations in the growth of the children were often dependent on changes in temperature, in such a way that a rise of temperature, even lasting for only a few days, caused an increase of growth, and a fall of temperature a decrease in growth. At Halle, Schmid-Monnard found that nearly all growth in weight took place in the second half of the year, and that the holidays made little difference. In America, Peckham has shown that increase of growth is chiefly from the 1st of May to the 1st of September.[168] Among young girls in St. Petersburg, Jenjko found that increase in weight takes place in summer. Goepel found that increase in height takes place mostly during the first eight months of the year, reaching a maximum in August, declining during the autumn and winter, in