Shattuck in the Massachusett's General Hospital feeds his enteric fever patients on the following lines.

1. Milk, hot or cold, with or without salt, diluted with lime water, soda water, Apollinaris, or Vichy; peptogenic and peptonized milk; cream and water; milk with white of egg, buttermilk, koumyss, matzoon, milk whey, milk with tea, coffee and cocoa.

2. Soups : beef, veal, chicken, tomato, potato, oyster, mutton, pea, bean, squash; carefully strained and thickened with rice (powdered), arrowroot, flour, milk or cream, egg, barley.

3. Horlick's food, Mellin's food, malted milk, carnipeptone, bovinine, somatose.

4. Beef juice.

5. Gruels : strained corn meal, crackers, flour; barley water, toast water, albumin water with lemon juice.

6. Ice cream.

7. Eggs, soft boiled or raw, egg-nog.

8. Finely minced lean meat, scraped beef. The soft part of raw oysters. Soft crackers with milk or broth. Soft puddings without raisins. Soft toast without crust. Blancmange, wine jelly, apple sauce and macaroni.

This diet may be accepted as that favoured by American physicians who believe in varied feeding. Manges, Fitz, and Kinnicut, from whose admirable paper on the "Principles of Feeding in Typhoid Fever" I have abstracted it, have all had success with a diet very much on the above lines. Vaquez in France has also had good results with a diet consisting of milk, soups thickened with oat or rice flour, meat jelly, the yolks of three eggs daily, and somatose during the period of advance, and during the lysis grated meat, mutton for preference, instead of the meat jelly. Marsden at the Manchester Fever Hospital allowed bread and milk, custard, and minced meat to those patients who were hungry, and added fish in very many instances before the temperature reached the normal. His relapses numbered 13 per cent and had apparently no connexion with additions to the diet. The haemorrhage rate was low, only 3 per cent, and among the 200 cases there was no instance of perforation.

Marsden's conclusion is that on such a diet there is more rapid recovery, less risk of surreptitious feeding with undesirable substances, and a lessened tendency to asthenic complications.

My own experience of liberal feeding throughout the fever is a somewhat limited one, but I must confess that I saw no harm result from the use during the acute stage of milk puddings, bread and milk, and beef-tea thickened with well-boiled rice, the diet which was given by my predecessor at the Edinburgh City Hospital at the time I worked there as house physician. At present, however, as I have already stated in the earlier part of this article, I prefer to postpone semi-solids and solids to the period of early convalescence. The interesting figures, however, published by Kinnicutt go far to prove that it is quite safe to give a liberal dietary. He has collected records of 4,654 cases treated with fluid and of 733 with liberal diet, and the results are as follows : -

Cases.

Relapses per cent.

Haemorrhages per cent.

Perforations per cent.

Mortality per cent.

On liberal diet

733

11.38

4.77

1.36

9.47

On fluid diet ....

4,654

1089

8.83

240

1055

It is of course difficult to draw conclusions from statistics of this sort, but the greatest care has been taken by Kinnicutt to secure that the test should be a fair one. The most interesting feature is the larger number of ulcerative accidents in the second group. It may fairly be asked does a low diet by starving the bowel walls help to increase rather than limit the process of ulceration?