Old and feeble persons and infants can not usually abstain from food as long as is theoretically indicated and sometimes can, at most, lengthen the usual periods between meals. In such cases, the rule is to make a considerable change in diet, to administer food as perfectly digestible and as nearly sterile as possible. Peptonized milk, boiled milk, toast water, artificial cereal-milk mixtures, sometimes buttermilk, koumiss etc., are employed. Boston crackers, toasted and moistened with tea, dry toast with salt instead of butter, zwieback - which has no obvious advantages over any other slowly and slightly toasted bread that has been kept for a few weeks - or other soft but not doughy cereal food is to be preferred to milk when milk has already been used to any great extent by adults. Sugar, fat, meat and coarse residue are the principal ingredients of diet to be avoided. Clear tea may be used rather freely provided the patient is not already too much under its influence, its value being as a. heart tonic, astringent and appetizer.

Breast-fed infants may become infected through water, fruit, filthy breasts and articles handled and carried to the mouth. The autochthonous germs of the alimentary canal may also develop a special virulence if digestion is disturbed by frequent, irregular nursing. If possible, several nursings should be omitted, the child being given boiled water with a little salt, at the regular feeding times. If the child is very feeble, it requires considerable courage to discontinue its nourishment, or even to change from the accustomed, normal food, although it is a question whether this would not be the best routine treatment. At any rate, the intervals between nursings can be lengthened to four or six hours and relative asepsis of the maternal breast and the child's mouth can be secured.

Unfortunately the decision for or against substitute feeding or suspension of nutrition, has to be made promptly and without waiting for the result of experience. If there is any probability of deleterious change in the maternal milk, as on account of fever, notable digestive disturbance, Bright's disease, or emotional shock, nursing should l>e discontinued.

Peptonized, sterilized cow's milk, prepared according to the usual formulas, excepting that cream should usually be omitted, may suffice as a change of diet, but it is probable that the heightened virulence of intestinal bacteria will be equally well maintained by milk of a different species. We are further hampered in the treatment of infants by their inability to take starch and cereals generally. The non-starchy proprietary infant's foods sometimes act very satisfactorily in cholera infantum. White of egg with a little salt, in water, is often well borne and it may be supplemented with milk sugar. During the first week, the child normally takes about 300 c.c. of milk a day while, at the close of the first year, fully four times as much is taken. While the milk varies considerably in composition in different women and while the first milk secreted is not up to the full standard, Leeds has shown by a large series of analyses that there is no great progressive change in composition and the average is about 4% of fats, 7 of lactose and 2 of protein. Thus one white of egg and 21 grams of lactose would represent nearly the full ration, except fat for the first week. Another white of egg would almost exactly supply the calories normally derived from fats. At the close of lactation 4 whites of eggs and 80 grams of lactose would represent nearly the entire ration except fat. and four more whites of eggs or 2 more whites of eggs and 10 additional grams of lactose would nearly compensate for the lacking fat.

Expressed beef juice, gelatin etc., may also be employed. If the saprophytosis is mainly of a putrefactive kind, as it usually is, buttermilk may be employed. Sometimes, as in more chronic intestinal disturbances, the substitution of goat's milk for cow's milk, works well. Fresh cocoanut milk is theoretically available.

Both in adults and in infants, alcoholic beverages, mainly the best brandy, or champagne if there is great gastric irritability, is of advantage, subject to the general limitations to the use of such beverages. For infants, a drop or two of brandy is sufficient and even this dose, if frequently repeated, will cause intoxication which may be confused with febrile symptoms.

Dysentery is a term often loosely used to designate any form or grade of colitis. It is generally conceded that it should be limited to colitis of specific, infectious etiology but some confusion remains since both the amoeba coli and Shiga's bacillus have been found in cases considered clinically specific. Moreover, at least two varieties of the bacillus have been deseril>ed and one of these is now considered by many as the cause of ordinary cholera nostras, including cholera infantum. Then, too, it is claimed that one form of amoeba is merely an intestinal commensal parasite, not pathogenic unless by causing a mild grade of irritation, while another form is strictly pathogenic.

In the eastern tropics, Shiga's bacillus is generally found in acute cases of dysentery, while amoebae are more frequently found in chronic cases and even in persons in good or fair health. Obviously, therefore, both germs may be found in the same case. Whether the American dysentery is identical with either form, is not fully established. Thus, it is impossible to speak with certainty as to the etiologic relations of food and drink to the development of the disease, although the wisdom of applying the same general principles of prophylaxis as for typhoid fever and cholera, is apparent.

As, in dysentery, the infectious lesion is mainly limited to the large intestine, it is usually possible to nourish throughout the disease. Owing to the lesion of the colon, it is advisable to use food with as little indigestible residue as possible. In most cases, peptonized milk and egg, boiled milk, beef juice, whey, milk and lime water, with or without whisky or other alcoholic, should be used. Raw scraped beef is highly praised by some, though it would seem better to insure some degree of sterilization by toasting it on a hot plate.