The empirical methods are not nearly so satisfactory, although they may under certain circumstances yield very accurate results. Much knowledge may be acquired from the statistics published by the financial department of the Government, but this is necessarily very inexact. It is much better to institute an active dietary study, either where full liberty is accorded in the selection of the diet, or in establishments where definite restrictions are placed upon the quantity or quality of the food consumed. Or, better still, deliberate feeding experiments may be contrived for testing the amounts of the various food substances required by different classes of people and the energy evolved thereby, as the true food requirements of the body are not likely to be ascertained with any degree of accuracy by observations of what people are in the habit of eating, and the customs and habits are not a safe index of the true physiological needs. Many researches of this character have been undertaken, and the general conclusion has been arrived at that in every case, where possible, the tendency was to indulge in liberal quantities of food, which were increased in direct proportion to the amount of work done. The fuel value of the food varied from 6,000 calories for excessively hard workers to about 2,000 calories for those employed in sedentary occupations.

These results are of immense importance when we reflect that the body absorbs the nutrients from food in certain proportions, no matter how much food has been consumed, although the proteins and fats of vegetable foods are less easily digested and absorbed than those of animal foods. This is probably entirely owing to the different forms in which they occur, the former being enclosed in a stout covering of cellulose, although one is justified in assuming that the more nearly the structure of a protein in the food approximates to that of the proteins of the body, the more chance it has of being absorbed with little or no change. It has been calculated that 97 per cent. of the protein in animal foods is absorbed, as against 84 per cent. of vegetable protein, and that in a mixed diet containing both kinds about 95 per cent. of the total protein is digested. Roughly speaking, 95 per cent. of the ingested fats is absorbed and 98 per cent. of carbohydrates, and these proportions appear to hold good even although excessive quantities of food are indulged in.