the body, is oxidised at once, or holds its place as fat. There are no observations to show that it can become or give rise to either proteid or carbohydrate in the body. All observers are agreed that fat does not assist in forming glycogen in the liver; Seegen, indeed, states that fat may directly form sugar in the liver, but his observations are unconfirmed. The normal diet of man. It is best, and, indeed, absolutely necessary, that human food should include proteids, fats, and carbohydrates, to which must be added salts, oxygen, and water. The absence of any one of these proximate principles kills by starvation. Absence of oxygen kills in two or three minutes, absence of water in a few days, absence of proteid, or fat, or carbohydrates in two to three weeks, and the absence of salts cannot be endured for a much longer period. That common salt is one of the necessaries of life is attested by the risks which men have run to obtain it when scarce, and by the 'salt-licks' frequented by wandering troops of animals. How little can a healthy man endure life upon, and how much must be given to him if he is to be kept efficient? The answer is afforded by the dietaries of armies, prisons, and gangs of navvies. In round numbers, and omitting circumstantial considerations, the minimum daily income of a man in full work must be about 5 per cent. of his body-weight, viz., That is to say, a man of 70 kilogrammes must get Or, to reduce those numbers to an easily-remembered form, from which the adequate diet of men of different weights can be estimated, it should contain per one kilogramme body-weight, 2 grms. proteid; 1.5 grm. fat; 6 grms. carbohydrate; 0.5 grm. salt-in all 10 grammes, or one per cent. of solid food. These numbers are applicable to normal adults doing ordinary work. The estimate is a liberal one, especially as regards the proteid. Voit gives as adequate :-Proteid, 118 to 137; fat, 56 to 117; carbohydrate, 352 to 500. Ranke's normal diet' is Proteid, 100; fat, 100; carbohydrate, 240 grms. : The chief conditions that will modify the amounts of an adequate diet are work, health, and age. Effect of work.-A person doing no work at all will subsist upon amounts somewhat smaller than those given above; a person engaged in exceptionally heavy work (soldier in the field, prizefighter, navvy, &c.) will require larger amounts: Effect of food upon excretion of urea.-The amount of nitrogen excreted in urea is directly dependent upon the amount of nitrogen taken in proteid. A dog taking 500 or 1,000 grammes of meat eliminates 35 or 70 grammes of urea, and if one day he takes and digests double the amount of proteid, he will during the next day excrete about double the amount of urea. The condition of the dog or man is one of nitrogenous equilibrium, or, in other words, the amount of urea excreted is in direct relation with the amount of eaten and digested proteid. Effect of muscular exercise upon excretion of urea.-It might be expected that great muscular exertion should be attended with a disintegration of muscle and a corresponding increase in the amount of urea excreted. The expectation has not been verified by experiment. Muscular exertion per se gives rise to no marked increase of urea excreted; in other words, there is no demonstrable relation between amount of work done and amount of urea excreted. The best-known observations in justification of this statement are those of Fick and Wislicenus upon themselves, and those of Flint and of Pavy upon a pedestrian. But while we must admit that the discarded doctrine of a ratio between work and wear of muscle, as evidenced by urea, is incorrect, we must also admit that the contradictory doctrine has been expressed too absolutely. Recent experiments show that urea excretion does vary a little with muscular action, although the variations are so small as to be easily swamped and hidden by the larger fluctua tions due to variations of diet; Zuntz's observations are to the effect that muscular exercise, if excessive-i.e. if pushed to the point of producing dyspnoea-causes disintegration of organproteid, and increased discharge of urea. Although muscular exercise may produce little or no increase of the nitrogen discharge if the nitrogen income is maintained constant, it may, under ordinary conditions of life, give rise to a very considerably increased discharge, by provoking a greater ingestion of proteid. In an animal fed upon proteid alone the total energy evolved as heat and as work must ultimately have its source from proteid, and whether or not that proteid forms carbohydrate in its passage through the body is a matter of secondary importance. Pflüger's dog, e.g., weighing 30 kilos, was in equilibrium on a pure flesh diet of 2 kilos per diem when doing no work, and required a daily supplement of 5 kilo when doing work estimated at 109.608 kgm. per diem; in this case the work must have been sustained from the proteid, and, comparing the heat-values of the two quantities, the yield of work comes out as equivalent to between 40 and 50 per cent. of the supply of potential energy afforded by the supplementary food. Diet during infancy.-A child or infant requires absolutely less food than a full-grown man, but relatively to their respective body-weights a much larger quantity. Thus, it would not be sufficient to give to a child of 7 or 14 kilos only or the amount required by a man of 70 kilos, and we should find in drawing up a diet-table for children at different ages, that the area of the body-surface is a far better proportional indicator than the body-weight. This is especially the case as regards proteid, and will be evident on examination of the table below. The last two columns of this table show that, in proportion to body-weight, the amount of proteid required by an infant is greater than that required by an adult, while in proportion to bodysurface it is approximately constant (viz. between 80 and 90 FIG. 111.-WEIGHT AND SURFACE OF BODY AT DIFFERENT AGES. PROTEID REQUIRED. SS= body-surface in sq. meters (1 cm. = 1 sq. meter). = 8 s = body-surface per 1 kilo body-weight in sq. cms. (1 mm. = 10 sq. cms.). NN proteid per kilo body-weight in grammes (1 cm. 1 grm.). The surface measurements are calculated upon the assumption that the surface of an animal bears a constant relation to its weight, such that The ratio generally adopted is rather higher, viz.: S12 W2. grammes per square meter). As regards other constituents, the best guide to their amounts is the composition of milk, which is normally the only food required by infants during the first months of life. In round numbers human milk contains per 100 c.c. An infant at the breast must therefore take per diem 500 to 1,000 c.c. of milk to get W If artificial feeding is required, cow's milk is usually first resorted to, and seeing that cow's milk is, as compared with human milk, richer in fat and proteid, poorer in sugar, it is necessary to dilute it and to add sugar to it. A common error in the feeding of infants is the premature augmentation of the carbohydrate element, 60 to 90 grammes carbohydrate being the amount commonly given in diet-tables to correspond with 20 to 36 grammes proteid. The composition of milk is, however, such as to indicate that the carbohydrate should not largely exceed the proteid constituent, during, at any rate, the first 9 months of life. After this period the requirements may be regarded as approximating towards those of the adult, in whom, as we have seen, the amount of carbohydrate required is three times that of proteid. A growing infant excretes less carbon and less nitrogen than it takes in; it gains in weight and puts on both flesh and fat; during the first year of its life it gains about 7 kilos. Taking half of this amount to be flesh, this would mean a gain of about 120 grammes nitrogen, i.e.gramme per diem; the average proteid income, 30 grammes, contains about 4 grammes nitrogen; the amount of nitrogen kept back in the flesh of a growing infant up to the end of the first year of life is thus about of the nitrogen contained in the total proteid consumed; during the second year it is about, during the third about Diet during lactation.-Consider the case of the mother, supposing her to weigh 50 kilos, and to receive an average proteid allowance of 100 grammes. If she is to supply a child with one litre of milk containing 30 grammes of proteid per diem, she should receive an addition of about 90 grammes to her food, i.e. her day's proteid should amount to nearly 200 grammes, a requirement that is recognised in the instructions issued in France to persons employing a wet-nurse, to the effect that she must receive nearly twice as much food as an ordinary woman. While nursing she will also excrete considerably more urea than in her normal state. That so large an increase in the proteid allowance should be requisite during lactation is no less apparent if we consider the facts already alluded to concerning the relation between proteid and fat. We have seen that proteid can fatten, that 100 grammes proteid contain about 44 grammes of surplus carbon, which may go to form fat or carbohydrate; now it is shown both by experience and by experiment that not fat but proteid is the substance from which milk-fat is formed. The day's milk yielded by |