Several such enquiries have, however, been made in the past, the best known being that made by Dr. Ralph Crowley at Bradford in 1907.[489] The results of this experiment have been often quoted, but they are so important that they will bear repetition. Forty children were selected from two of the poorest schools in the city, the children being mainly those who appeared to be most in need of food, though a few were included primarily on the ground of their particularly poor home circumstances.[490] To these children from April 17 to July 24 two meals a day were given--breakfast, consisting of oatmeal porridge with milk and treacle followed by bread and margarine or dripping, with hot or cold milk to drink; and a dinner comprising in rotation one of seventeen different menus specially drawn up so as to contain the amounts of fat and proteid necessary for a child's nourishment.[491] Every effort was made to render the meals of as much educational value as possible, and special attention was given to such matters as the provision of table-cloths and flowers and the inculcation of good manners. Footnote 489: Bradford Education Committee, Report on a Course of Meals given to necessitous children from April to July, 1907. Footnote 490: _Ibid._, p. 3. Footnote 491: _Ibid._, pp. 4, 5. The children experimented on were weighed three times during the five weeks preceding the starting of the meals, and every week while they were receiving them. For the purpose of making comparative observations 69 children were selected who were being fed at home, and who in other respects were as comparable as possible with those who were receiving the breakfasts and dinners. These "control children" were also weighed weekly. During the four weeks, March 12 to April 9, before the feeding began, the forty children gained on an average .17 kilos, and during the week previous to feeding .008 kilos. At the end of the first week of feeding the average increase was found to be .58 kilos (1 lb. 4 oz.).[492] During the next week, there was a slight loss of .001 kilos, followed by a gain during the next two weeks of .15 and .13 kilos respectively. During the ensuing eleven days, the Whitsuntide holiday, no meals were given. At the end of this period it was found that the "control children" who, during the three weeks preceding the holiday, had lost .003 kilos on the average, had during these eleven days gained an average of .23 kilos; in the case, however, of the children fed at school, not only had the lack of food neutralised the benefits of fresh air and exercise, but they had actually lost an average of .48 kilos, a loss which it took them nearly a fortnight to make up, after the meals had been started again. During the eleven days after the holiday the "control children" only gained .02 kilos. A group of "control children" from another school similarly gained .21 kilos during the holiday, and only .04 kilos during the subsequent fortnight. The same result was