In the case of physiological underfeeding there is almost invariably an actual increase in the yield of milk fat, as well as in the percentage in the milk. The character of the milk fat is altered, as shown by the change in the physical and chemical constants. Eckles and Palmer suggest that the production of the normal amount of milk fat is controlled by the activity of the lipases and other enzymes which accelerate this synthetic reaction in the mammary gland, being greatly influenced by changes in the general metabolic activity of the body, particularly by the changes that affect the fat metabolism, and that the composition of the fat produced is controlled by variations in the amount and kind of material presented to the milk glands by the blood stream. Koning,49 on the other hand, found that poor food caused a fall in the fat content of the milk. Behr 50 states that owing to the deficiency of food, the average fat content of milk in Germany fell to 2.98 per cent in 1918, being a gradual decline since 1914. The quantity of milk fell also from 2900 1. to about 1700 1. per year.

Engel and Plaut51 found that the milk of a nursing mother is reduced if the diet is deficient in fat, but was unable to increase the fat in the milk by increasing the fat in the food.

According to Eckles and Palmer (1. c.) a subnormal plane of nutrition at times affects the percentage of protein in the milk. In some of the experiments conducted there was a decline in the percentage of casein, while in others the total protein only was affected. In the latter cases the percentage of ash in the milk was decreased. None of the types of underfeeding influenced the percentage of lactose in the milk.

47 Woll, Agric. Exp. Sta. Wisconsin, Bull. 116. 48 Eckles and Palmer, Agric. Exp. 8ta. Un. Mo. Bull. 25, 1916. 49 Koning, Milchw. Zentr. 6, 675, 1910. 50 Behr, Z. Nahr. Genuwem. 35, 471, 1918; 87, 165, 1919. 51 Engel and Plant, Mun. Med. Wochensch. 55,1158,1906, quoted by Lane-Claypon, Milk and its Hygienic Relations, London, 1916.

Many attempts have been made to increase the mineral content of milk by additions of salts to the food, particularly in the case of iron 52 and calcium,53 but there is no decisive evidence of appreciable success.53a Kohn 54 found that both the ash and the fat content of goats' milk was higher when the animals were grazing in the fields than when they were restricted to dried feed. Fingerling 55 found that when goats were fed on a diet deficient in calcium and phosphorus a decrease both in the total amount of milk and in the proportion of these constituents occurred, but only after a certain period had elapsed, during which the system became gradually depleted. On addition of the missing elements to the diet the secretion became normal. Hess, Ungcr, and Supplee 56 have noted that the percentages of calcium and phosphorus were below normal in milk produced on a ration of bean meal, oil meal, bran, beet pulp, molasses, and straw, which contained an ample supply of minerals, but was entirely deficient in vitamin C. Koning,57 and Allemann 58 found that even large increases of sodium chloride in the food caused no rise in the salt content of the milk.

Babcock 59 fed cows on a ration suitable in every respect except that the sodium chloride supply was limited to that naturally present in the food, until some of the animals died, and others were only saved from death by the administration of salt. In no instance was there notable decrease in the yield of milk until just before general decline set in, but the ash content of the milk under these circumstances was not reported.

52 Lane-Claypon, 1. c, p. 63.

53 Znckmayer, Pflug. Arch. f. d. gee. Physiol. 168, 209, 1914; Dibbelt, Atb. a. d. Geb. d. path. Anat. u. Bakt. Tub. 1908; Schabad, Jahtb. f. Kinderh. 74, 511, 1911.

53a Ramsey (Brit. Dental J. 4a, 49, 1921; Physiol. Abstracts 6, 66) notes that the incidence of caries is complex, and diet a most important factor. He considers milk to be the ideal food for supplying the necessary calcium and should be relied on during the growth-period; bread should be made so as to contain more cellulose and more vitamin; among fats, olein is preferable to the more solid fats; calcium intake should be regulated in adults, and symptoms of thyroid abnormality should be recognised and treated.

54Kohn, Deutsch Tier. Wochenscht. 21, 49, 1913.

55 Fingerling, Landwirtsch. Versuchsstat. 75, 1, 1911.

56 Hess, Unger, and Supplee, J. Biol. Ghem. 45, 299, 1920.

57 Koning, Milchw. Zentr. 6, 675, 1910.

58 Allemann, Schweiser Milchwent. 1911, Nos. 67, 71, 72, 74, quoted by Lane-Claypon, 1. c. p. 21.

59 Babcock, Wis. Exp. Sta. Ann. Rep. 22, 1905.

McCollum and Simmonds 60 made an elaborate study of the relation between the diet of lactating rats and the value of the milk produced. The rats were fed on an adequate diet until the birth of the young, when they were immediately placed upon a ration known to be entirely inadequate for growth owing to deficiency in one or more factors. The litters were in all cases reduced to four, so that the burden imposed upon the mothers was uniform in this respect. Their results show that the mothers were capable of providing milk which was better adapted for growth than the diet which they themselves received; but sooner or later failure ensued. On a diet of rolled oats, deficient in three factors, protein, mineral content, and vitamin A, the young grew very slowly for the first 20 days, then became stunted and died between the fortieth and fiftieth days. When A was added to this diet the young grew faster and growth continued over a longer interval, up to 60 days. Although they began to eat the oat and butter-fat diet of the mother about the twentieth day it is evident that this must have been supplemented by the inorganic content of the milk, since without such supplement rolled oats fails entirely to support growth. Addition of protein was of less benefit than addition of salts.

When the mother was fed on a ration of rolled oats supplemented with protein only, the best growth observed was half-normal. When A was applied in addition the young did slightly better, but addition of mineral salts was more beneficial than addition of A. Only when all three deficiencies, A, salts, and protein, were corrected was it possible to secure entirely normal growth, but in every case after the young were old enough to share in the mothers' diet they grew for a time, although slowly, which would have been impossible on this diet alone, showing that the mother was producing milk capable of supplementing to some extent the diet on which that milk was produced. According to McCollum and Simmonds, the ultimate failure which invariably ensued indicated that the milk produced on these deficient diets was not of normal quality.

60 McCollum and Simmonrit, Am. J. Phyriol. 46, 275, 1918; McCollum, "Newer Knowledge of Nutrition," 1919, pp. 117-186.