18 



Table 3. — Proportion and amount of nutrients in articles used in making bread 



pudding. 



The coinpo.sition of the cooked pudding was computed as protein 

 2.6 per cent, fat 2.3 per cent, and carbohydrates 22.1 per cent, by 

 dividing the total quantity of each initrient given in the table above 

 by 228.5, the weight of the jjudding when cooked and multiplying by 

 100, the assumption being that there would be no appreciable loss of 

 nutrients in cooking. 



Obviousl}' considerable labor was involved in making weighings of 

 the raw foods used in preparation of the different dishes. In dietary 

 studies Nos. 361, 365, and 371 these weighings were made for all foods 

 served at each meal, but in the other .studies, which were made in 

 dining rooms supplied from the larger kitchens, this was not practi- 

 cable, for the reason that the cooking was done for a large number of 

 dining rooms at the same time, and the food for one dining room could 

 not be separated from that for the others. In order to obtain data 

 for computing the composition of the cooked foods under such circum- 

 stances it was necessary to weigh the raw ingredients u.sed in preparing 

 food for all the wards supplied from the kitchen, and the number of 

 weighings involved for such a simple dish as boiled cabbage, for 

 example, was from 40 to 60, so that one observer could not collect data 

 for all the foods used at each meal, in addition to gathering those for 

 food served, returned, and wasted in the dining room. It was there- 

 fore necessary in all other studies than the three just mentioned to 

 reduce to a minimum the labor of collecting statistics in the kitchen. 



It was observed that for any given dish the cooks would use practi- 

 cally the same quantities of raw ingredients each time, and that the 

 other conditions, namely, the amount of water added and the time of 

 cooking, were generally the same; under such conditions any given dish 

 made in the same kitchen at different times was quite uniform in char- 

 acter. The composition as computed at different times was likewise 

 quite uniform, the variations being generally no greater than in the 

 analyses of different samples of the same kind of food material. It 

 was therefore believed to be sufficiently accurate to compute the com- 

 position of each cooked food in most cases but once for each kitchen, 

 and use the computed value for all studies in which the particular food 



