From his experiments on rats Drummond7 has been led to believe that certain types of malnutrition common among the lower classes may be the result of diets more or less inadequate in regard to vitamin A, and particularly emphasises the fact that young rats nursed by a mother whose diet was deficient in this respect never exhibit a normal resistance to disease.

When the mother's milk is not available for the infant, cow's milk in some form is almost invariably used as the substitute. This may be fed raw or after pasteurisation, condensing, or drying and it is cpm-monly modified in various ways in the attempt to approximate more closely the content of human milk. The effect of heat upon the various constituents of milk, including the vitamin content, has been fully discussed in Chapter XV (Milk).

In spite of the somewhat conflicting evidence it seems justifiable to conclude that a healthy child can develop normally on an exclusive diet of cow's milk, whether raw, pasteurised, dried, or condensed, provided a sufficient quantity can be fed. Even in raw milk, however, the supply of vitamins is near the minimum, and heating reduces this still further, especially if long continued. For this reason milk which has been pasteurised by the "flash " method, heating to boiling point or a little under for a minute and then cooling rapidly, is preferable to that which has been kept at comparatively low temperatures for longer periods, although this latter method is the one most generally used for pasteurisation in the United States. For the same reason dried milk appears to be preferable to condensed milk for infant feeding, and the sweetened condensed milk is probably to be regarded as superior to the unsweetened "evaporated " variety which is frequently sterilised after condensation to improve its keeping qualities, although the high proportion of sugar in the sweetened product is probably a disadvantage.8 Moreover, since aging appears to result in loss of C it is important to decrease the time between pasteurisation and consumption as far as possible.

7 Drtnnmond, Lane. 1918, ii, 482.

8 This is denied by Lassabliere. See Una 1918, ii, 64.

It seems to be the general impression that where whole milk in any form is the chief component of an infant's food there is little danger of deficiency of A. Bloch's experience in Copenhagen9 indicated that a very small amount of this factor may be sufficient for successful nutrition. Milk fat appears to be the best source of this vitamin for infants; with older children eggs, and where necessary cod liver oil, may be used with advantage.

The feeding of children on separator milk to which vegetable fats have been added in order to bring this up to the fat standard of human milk is in general to be reprobated, owing to the deficiency of most vegetable fats in A. Linseed oil, which has been most commonly used for this purpose, is devoid of A, and most authorities agree that the use of this product as a substitute for butter-fat is likely to result in greatly diminished resistance to disease unless the necessary A is provided from some other source.10 Pritchard and others report good results with an emulsion of soy bean oil, cotton seed oil, and suet, in proportions designed to resemble butter-fat. This is combined with sugar and dried separator milk to the standard of breast milk. Soy bean oil probably contains a small amount of A, and the beef suet is fairly rich in this factor. Hampshire and Hawkerl1 recommend a somewhat similar emulsion of beef suet with a small quantity of olive or arachis oil. This product is known as " university cream " and is said to give good results.

Pritchard supplements such a food with fruit or turnip juice as an additional source of C and an ounce of meat and vegetable broth to increase the supply of B, along with small quantities of either yolk of egg or calcium glycerophosphates, which seem to be equally beneficial.

Daniels, Byfield, and Loughlin12 have reported that addition of B as obtained from fruit or vegetable juice or from wheat embryo to a diet already furnishing an adequate number of calories invariably stimulated the growth of babies, while orange juice from which the B had been removed was of no effect. They also made use of a specially prepared vegetable soup as a milk diluent, with good results which they ascribe to the B present. They consider that the retarded growth of infants and young children is often the result of a deficiency of this vitamin in their food, and suggest that the relatively large food requirement of the artificially fed infant is connected with the poverty of diluted cow's milk in respect to this growth factor.

• See p. 235.

10 Drummond, Lane. 1918, ii, 482; Pritchard, lb., 647.

11 Hampshire and Hawker, Pharm. J. 103, 82, 1919.

12 Daniels, Byfield, and Loughlin, Am. J. Die. Child. 18, 646, 1919; 19, 349, 1920.

Eddy and Roper13 have studied the effect of administering vitamin B in cases of infant malnutrition, and report excellent results in all cases (16 children). In one case they observed an average gain of 0.84 ounces per day over a period of 32 days, as compared with a gain of 0.47 ounces per day for 17 days preceding the use of the vitamin, the diet and food intake being the same through both periods. A three per cent increase in vitamin intake (as measured by the Bach-man test14) produced marked stimulation. Eddy suggests that the pronounced effect of so small an addition indicates that extracted vitamin may be more readily utilised by the child than that present in the diet, and that the way in which the vitamin is held in the food may be an important factor in nutrition.