The same idea which underlies the " empty bowel theory " is no doubt responsible for the attempts made to treat enteric fever with either no food by the mouth at all, or at the most with very small quantities. Thus Queirolo has recommended that feeding should be entirely rectal, a lemonade made up with a little hydrochloric acid being the only drink allowed. Provided that the bowel of a patient so treated was first emptied by a dose of calomel or other suitable purgative, such a method of dieting should secure complete rest for the affected parts and absolutely exclude the possibility of fermenting masses of partially digested material lying in the gut. The nutritive value, however, of rectal feeding in a prolonged disease is so limited that this method may fairly be regarded as a treatment by starvation.

Similar in its object and effects is the method suggested by Williams, who, believing that the exhausting diarrhoea of the fever is due to improper feeding, endeavours to secure that the bowel shall, as far as possible, remain empty. Only water is allowed in severe cases, sometimes for days at a time, and he regards half a pint of milk in the twenty-four hours as a liberal diet, seldom apparently exceeding this amount until the temperature is normal. The method seems drastic, but I have reason to know that the cases do remarkably well. I have often marvelled at the amount of starvation which a typhoid case can safely tolerate after a haemorrhage, and it is only rational to suppose that the patients would support starvation even better before such a depressing complication had occurred. Under such a regime Williams probably more nearly attains the ideal of the " empty bowel " than any other observer. It seems almost incredible that patients so treated should occasionally gain in weight and that they do not in any case waste more than patients more liberally fed; but it is, after all, obvious that, if food is not assimilated there is no benefit to be derived from it, and in many cases of enteric fever assimilation is undoubtedly extremely poor.

The theoretical objection to both these methods of treatment is that, if ulceration has once started, such a remarkably low diet would apparently give the intestinal lesions only a poor chance of repair; on the other hand, it is possible that the absence of irritation would go far to counterbalance this defect, especially as the patients seem to stand their starvation so well. If plenty of water was supplied this would be more easily understood, but some of Williams' patients were limited, for a time at least, to one pint of water per diem, which seems a most inadequate amount.

The above are instances of a treatment which has often been attempted, and which is known in France as "Traitement a vide." While not prepared to recommend its adoption, I think that from it, as from the experiments in feeding with solids, there is much to be learned. It encourages us to starve, for two or three days if necessary, severe cases with marked gastric and intestinal disturbance, probably very much to their advantage. It is, however, unnecessarily severe for the average patient, even while we admit that in enteric fever there is no certainty as to what may happen from day to day.