The mother, previously described as having been investigated by Slemons, had plenty of milk, and the baby gained an average of 30 grams a day during the first forty days of his life.

Slemons remarks that the low protein metabolism, as indicated by the urinary nitrogen of the period of settled lactation, is a proof that there can be no important production of milk fat from protein.

1 Rubner and Langstein: "Archiv fur Physiologie," 1915, p. 39.

In the above experiment it will be noticed that the nitrogen of the milk is small in quantity as compared with the urinary nitrogen. On a strictly vegetarian diet the relation would change. Thus Voit1 found 48.8 grams of nitrogen in the milk of a cow and 93.7 grams of nitrogen in her urine for the same period.

The influence of nutrition on the production of milk has been the object of countless investigations, but unfortunately most of these experiments have been conducted for commercial purposes on cows and goats. These animals, with their fundamental ration consisting of hay, do not allow of the ingestion of simple foods. On the other hand, the milk supply of even a large bitch is very limited in quantity and is with difficulty obtained. The writer is not aware of any systematic observations on the composition of human milk as influenced by food, although such researches would seem of great importance.

Perhaps the most valuable research which can today be used is an old one of Voit2 upon a bitch weighing 34 kilograms. It confirmed the previous work of Kemmerich and of Ssubotin, and has since been verified by Grimmer.3 The animal was given meat alone, meat and starch, meat and fat, starch alone, fat alone, and was also starved. The influence upon the milk secretion was found to be comparatively small. The research is a model of completeness, the plan of which could well be copied in an experiment on a human being.

A part of the results are given on page 392.

1 Voit: "Zeitschrift fur Biologie," 1869, v, 122.

2 Voit: Ibid., 137.

3 Grimmer: "Biochemische Zeitschrift," 1914, lxviii, 311.

Influence Of Diet On The Composition Of The Milk Of A Dog Weighing 34 Kilograms

The largest quantity of milk, as well as the richest in protein, was obtained when meat or meat and fat were ingested. Curiously enough, a diet of 500 grams of meat and 300 grams of fat gave milk of the same amount and quality as did 2000 grams of meat. It is usually said that a large protein diet stimulates the milk secretion; but this may also be due indirectly to the development of the gland cells.

The milk-sugar content was scarcely affected by the diet, although a slight percentage increase was observed after starch ingestion.

The fat content was increased in starvation to its highest percentage. It was not very greatly affected by adding fat to a meat diet and it was greatly reduced by giving carbohydrates.

The action of fasting on the fat content of milk is better shown in the herbivorous goat. The writer1 gave a milch goat a constant diet of hay, cornmeal, and bran, starved the animal for two days, and then continued the former diet. The fat content of the milk was determined. The results were as follows:

1 Lusk: "Zeitschrift fur Biologie," 1901, xlii, 42.

In fasting, therefore, the fat content in the milk of the herbivorous goat approaches that contained in the carnivorous dog. With a return to the normal diet the fat content in goat's milk is reduced to its former level.

Morgen, Beger, and Fingerling1 find that a diet rich in carbohydrate and poor in fat produces in sheep and goats a poor milk containing little fat, although the general condition of the animals remains perfect. Addition of protein increases the quantity of the milk without changing the low fat percentage. Replacement of some of the carbohydrate with isodynamic quantities of fat, up to 0.5 to 1.0 gram per kilogram of animal, largely increases the fat content of the milk and thereby its nutritive value.

Contrary to this is Jordan's2 statement that the amount of fat in the fodder is without influence upon the fat content of a cow's milk. Here the breed of the cow and not the diet is the determining factor. The German agricultural stations have recently reached the same conclusion. Morgen3 states that the principal cause of the difference in the results of the experiments on cows and on sheep and goats lies in the fact that the smaller animals produce more milk for their weight than do cows, and, therefore, the milk production is much more dependent on the food supply.

Newer work by Prausnitz4 concludes that although food does not determine the quantity of protein, lactose, or ash in cows' milk, yet the percentage of fat, and hence the caloric value of the liter of milk, may be considerably influenced by variations in the diet.

1 Morgen, Beger, and Fingerling: "Landwirtschaft. Versuchsstationen," 1904, lxi, 1.

2 Jordan and Jenter: "New York Agricultural Experiment Station," 1897, Bulletin 132; 1901, Bulletin 197.

3 Morgen, Beger, Fingerling, and Westhauser: "Landwirtschaft. Versuchsstationen," 1908, lxix, 295.

4 Helle, Muller, Prausnitz, and Poda: "Zeitschrift fur Biologie," 1912, lvii. 355.

It has long been known that ingested fat may appear in the milk of an animal. Gogitidse1 has shown that after giving linseed oil to sheep their milk fat may contain 33 per cent, of linseed oil. He also finds2 that the fat of linseed oil passes readily into human milk, and that the fat of hempseed, while influencing the composition of the milk, greatly depresses lactation during the period of its ingestion.