When nourishment by the ordinary channel is contra-indicated or impossible, rectal alimentation may be impracticable from disease of the bowel preventing the retention of the nutrient injections, or, as in the case of the insane, from the inability of the patient to aid the physician in carrying out the treatment. Under such circumstances it has been sought to maintain nutrition by the subcutaneous or intravenous injection of food-stuffs. There are a number of early and tentative observations on this method of feeding. Thus in 1850 Hodder made intravenous injections of milk in cholera patients apparently with favourable results. In 1869 Menzel and Perco experimented with the subcutaneous injection of food-stuffs both in dogs and in man. These observers were followed by Karst, Krueg, Whittaker, Pick and others, but it is to Von Leube that we owe the demonstration of the apparent utility and simplicity of the method, both from the experimental and from the clinical aspect. It is at once obvious that certain food-stuffs are incapable of use by this method, either because they are not directly assimilable, or because their solutions are incapable of proper sterilization. Thus though Leigh has claimed good results from the injection of white of egg, there is no proof of its assimilation, and the sterilization of the solution presents great practical difficulties. Albumoses and peptones, while their solution is easily sterilized, are foreign bodies in the blood, are toxic, and are excreted by the kidneys and produce much irritation during their excretion. Leube found that albumoses were assimilable up to a certain point, but beyond this produced toxaemia and albuminuria. The injection of a solution of 270 grains of somatose into a large dog produced a lowering of the general health with albuminuria and nephritis, Alkali-albumin when injected is apparently assimilated, but its injection causes marked irritation, and the solutions are difficult to sterilize. Blood serum has been used by a number of observers, but the nutritive result obtained from the injection has not been great. It was found experimentally that if the serum of one animal were injected into another it proved toxic, but if the serum had previously been treated by prolonged heating at a temperature of 58 to 60° C. it became opalescent, but did not coagulate, and no longer produced toxic symptoms when injected. Clinically serum injections have been made use of by Reinach and by Salter, but while the serum seems to have been metabolized and no ill effects have followed the injection, the nutritive result obtained has never been sufficiently great to justify their extensive use as a method of artificial feeding in clinical work. So far then the use of protein solutions for subcutaneous injection has not proved encouraging. Until we get a protein body which in solution is non-irritating, and is capable of sterilization and of assimilation after injection,protein must be abandoned for subcutaneous nourishment.

The use of carbo-hydrate solution would at first sight appear both promising and rational, but the result obtained has not been much more satisfactory than in the use of protein. Voit found that subcutaneous injections of sugar were slowly absorbed, and hence the tissues were saved from sudden flooding with carbohydrate, and that even with large injections glycosuria was but trifling. Two ounces of dextrose could be injected with only traces of sugar appearing in the urine, while if three ounces were injected only about half a drachm of sugar passed out by the kidneys. The injections of concentrated solutions of sugar are, however, liable to cause marked pain, and necrosis of the skin has resulted. Strong solutions of sugar cannot, therefore, be used. To be free from pain and discomfort a subcutaneous injection must be isotonic with the blood. This in the case of glucose is a 5 per cent solution. A pint of this solution can be injected without subsequent discomfort. The nutrient value of such an injection is approximately 102 C. Barker claims very good results from the use of such injections in exhausted surgical cases as regards the condition of the pulse, the general strength, and the relief of thirst. Considering the nutrient value of the injection, much of the benefit obtained must, however, be attributed to the water infused. A patient's nutrition cannot be kept up on injections the value of which in the twenty-four hours may reach 204 C. Still, such injections may be of use in tiding over a critical period of illness.

The use of protein and of carbo-hydrate for subcutaneous injection having proved of no great nutritive value, it is natural to turn to fats in the hope that better results might be obtained. Fat has a high energy value. In starvation the fat depots are called upon to provide energy and to save tissue protein. By increasing subcutaneous fat by injection good results might be hoped for in cases of subnutrition, where food could not be taken in sufficient quantity by the mouth. Much work on this line has been done by Leube. He found that fat injected under the skin was slowly absorbed in the form of a soap and that fat embolism was not to be feared. Working with dogs, he proved that butter injected under the skin was stored in the omentum as dog fat. Koll, working at the gaseous interchange and nitrogenous excretion after fat injections, showed that the fat acted as a protein sparer and was for the most part used by the animal. Leube has used the method extensively in clinical work, and claims that if not more than 40 grammes (10 drachms) be injected at a time there is no pain, no inflammatory reaction, and if antiseptic precautions be observed, no untoward after-effect. Absorption of the fat takes place in 12 to 24 hours. Jakob injected much larger doses, 7 to 10 oz., without pain, and similar results were got by Du Mesnil de Rochemont. He used large injections of olive oil and satisfied himself that the fat was absorbed, slowly undoubtedly, and was made use of in the body, so that it could be counted on to spare protein metabolism. These larger injections have, however, the disadvantage of causing considerable pain and discomfort to the patient.

Much of the value obtained from fat injections must depend upon the rapidity of absorption, and it is in this that the method seems to fail. Winternitz has recently discussed the question from this aspect. The observations were carried out with a combination of fat and iodine (iodipin). The metabolism of the fat resulted in the liberation of the iodine which could be recovered from the urine. The results seem to show that fat is very slowly metabolized when injected under the skin, and that even after large injections and when absorption is relatively rapid only half a drachm to one drachm can be metabolized per diem. Thus if a large injection be given, months may be required for its metabolism. The most that can be expected is a daily energy value of 20 to 25 calories from fat injections. Winternitz's results if confirmed tell heavily against the usefulness of fat injections in clinical work where rapid results are desired, and prove the uselessness of large injections.

The technique of fat injections is very simple. Elaborate apparatus is unnecessary. The oil, preferably pure olive oil, is sterilized by heat. The point selected for the injection, the subcutaneous tissue of the mammary region, of the abdomen, or thigh, is carefully cleansed. The injection is given very slowly by means of a large syringe, two or three ounces being injected daily. It is of great importance to give the injection slowly; for this reason some advise the use of a pressure bottle and trochar. The trochar is introduced, the pressure is raised in the bottle and the oil slowly flows into the tissues. However the injection is given, it is not free from considerable discomfort to the patient, which may last for some hours, even if the oil is diffused by gentle massage. It must always be remembered that the material injected is a nitrogen-free food, and in spite of the apparent favourable results obtained by some observers accurate metabolic observations show that the injections have but little nutritive value.