The important point for the clinician is, Can rectal feeding be relied on as a means of nourishing a patient if gastric rest is indicated as a therapeutic measure, or can it be relied on to improve nutrition in a case where buccal feeding is inefficient or impossible? The caloric value of the food absorbed in the investigations quoted varied from 645 calories as a maximum to 240 calories as a minimum, giving an average of 389 calories. Thus we get from rectal feeding only about a quarter of the nourishment required to maintain equilibrium, even if a reduced standard of nutritive requirements be accepted. In considering these cases it must be remembered that they were all cases in which rectal feeding was well borne, so that the conditions were in fact favourable to food absorption. If the rectal feeding was not well borne, if diarrhoea or vomiting was at all a prominent symptom, or if the patient's nutrition was obviously suffering under the rectal feeding, other methods of treatment were adopted and the observation was not included in the series. From the metabolic work which has been done on rectal feeding it must be concluded that even under the most favourable circumstances it is sub-nutrition of a most pronounced character. Given a patient in poor condition it cannot be relied on alone to produce any material improvement in nutrition. This seems an important point to emphasize. The surgeon frequently wishes to have a patient suffering from oesophageal or pyloric obstruction brought into a condition of improved nutrition before operation. It cannot be too strongly emphasized that rectal feeding cannot be relied on to produce this result. It will at once be urged that patients have been known to gain weight under rectal feeding. In considering this the absorption of water from the intestine must be remembered. After severe haemorrhage or persistent vomiting, as in pyloric obstruction or gastric ulcer, the tissues are suffering markedly from want of water. On placing the patient on rectal feeding, water is absorbed and there may be a gain in weight. In one case which the writer studied, where rectal feeding was begun after a very severe haemorrhage, though the nitrogen balance was markedly negative, yet there was a distinct gain in weight after nine days rectal feeding.

If the physician desires to make use of rectal alimentation, it seems that the best results will be got from the carbo-hydrates and fats. The absorption of protein food from the bowel is so small as to make it of little value as a food-stuff.

Folin has recently enunciated a theory of protein metabolism which would place protein on a much lower plane as a food than it has up to the present occupied. He regards proteins, in fact, as of secondary importance, insomuch as their nitrogen is split off and eliminated, leaving the carbon and hydrogen containing part to be oxidized, and thus to yield energy as it is yielded by the carbo-hydrates and fats. Folin's views have not so far met with general acceptance (Paton), but even without accepting Folin's views, we may evidently with advantage abandon much of the nitrogenous material which is at present used in rectal alimentation.

There remain then the carbo-hydrates and fats. The best carbo-hydrate to use is pure dextrose, but failing it dextrine is always available, and is probably better than a crude sugar, in that the latter may contain impurities which may cause bowel irritation. Fat is best given in the natural state, as in yolk of egg. If prepared fat is used, it should be in the form of a fine emulsion of a fat of low melting point. Pure olive oil can be used, a little being saponified and used to emulsify the whole. The difficulty in regard to fat in rectal feeding seems to be the wide difference in the absorptive power of different individuals for fat. When well absorbed it is a most valuable source of energy, but the absorption cannot be foreseen in any given case.

The conclusion to which a study of the metabolism of rectal feeding leads seems to be that its field of usefulness is more limited than is at present recognized by clinicians. In acute gastric diseases it must interfere with complete gastric rest by inducing gastric secretion. It must, however, be remembered that while the amount of nourishment absorbed is not sufficient to keep the patient in nitrogen balance, it may be sufficient to tide over a critical period and to prevent an undue strain upon the tissues. It is a dietetic means of treatment which, used with intelligence, will prove helpful alike to the patient and to the physician.