14 



iminbcr of meals taken b}' each. From these statistics, and data 

 regarding the composition of the food materials, as determined l)}^ 

 analyses of samples of materials used or as assumed from previous 

 anah'ses of similar materials, the total amounts of protein, fats, and 

 carbohj^drates in the diet and the average amounts consumed per man 

 per day are computed. 



In carr3nng out the studies here reported some moditications of 

 this method were necessary. For instance, separate studies were 

 made with dijEferent groups of the hospital population. Obviously, 

 this could not be done by taking account of the amounts of food 

 materials brought into and issued from the hospital storeroom from 

 which all the kitchens were supplied, which would correspond to the 

 method usually followed in a study with a family. Nor was it practi- 

 cable to stud}^ the food consumption of a given group by determining 

 the quantities of food l^rought into the kitchen in which the meals for 

 the group were prepared, because in each kitchen food was prepared 

 for several groups at once, whereas commonlv but one group could 

 be studied at a time. Data regarding the food consumption of each 

 group were therefore obtained by taking account of the food used in 

 the dining room in which the group was fed. 



In each stud}' all food sent from the kitchen to the dining room was 

 weighed, as well as all not served which was returned to the kitchen 

 after meals. After each meal the wasted food, Avhich comprised 

 that remaining upon the plates, which was never served again, and in 

 some cases also that left in the serving dishes, was carefully scraped 

 into receptacles and also weighed, each kind of food being kept l)y 

 itself. The separation of the waste into the different kinds of food 

 proved to be a matter of some considerable difficult}', for the reason 

 that where a num])er of articles of food are served on the same plate 

 the uneaten portions are apt to become more or less mixed and hence 

 difficult to separate satisfactorily. 



The figures obtained by the above-mentioned weighings give for 

 each article the amount served, the amount returned, if any, and the 

 amount wasted, thereby furnishing the data for determining the 

 amount consumed. These statistics are recorded for the difierent 

 studies in Table 35 in the Appendix. 



The waste just referred to consisted of actually edible material that 

 was rejected. In addition some foods contained inedible material or 

 refuse, such foods being fish and meat containing bones, prunes con- 

 taining pits, etc. The amount of such refuse was determined in each 

 case, because such data were necessary in the computations of the 

 quantities of nutrients in the food consumed, as hereafter explained. 



The next step, that of ascertaining the amount of protein, fat, and 

 carbohydrates in the amount of food consumed, demands particular 

 consideration, as it differs from the methods which have been com- 



