The data indicate a wide variation in composition. Animals Nos. 20 and 23 were well along in lactation.

A chemical study of goat's milk.—In 1915, Mr. A. W. Bosworth, at that time Associate Chemist at this institution, and L. L. Van Slyke, Chemist, made a somewhat elaborate chemical study of goat's milk in comparison with cow's milk and human milk. The results of this investigation may be seen in Technical Bulletin No. 46, issued by the Station in December, 1915, entitled "The Casein and Salts in Goat's Milk." It is not desirable to reproduce the text of that bulletin in this connection, but a few of the most important conclusions are here reviewed:

1. So far as could be judged from its combinations with bases, the casein of goat's milk does not differ essentially from that prepared from cow's milk.

2. Both goat's milk and cow's milk contain much larger percentages of solids than is the case with human milk.

3. Goat's milk contains practically the same amount of calcium phosphates as cow's milk, excepting that in goat's milk a certain proportion is in the tri-calcium form which is not true of cow's milk. In this respect both differ from human milk which appears to contain no calcium phosphates.

4. Cow's milk and goat's milk contain a much larger proportion of the magnesium phosphates than does human milk.

5. The proportion of mono-potassium phosphate was found to be practically the same in goat's milk and human milk. Cow's milk was found to contain di-potassium phosphate of which goat's milk and human milk were found to contain none.

6. The proportion of potassium citrate was found to be much larger in goat's milk than in either cow's milk or human milk; on the other hand, cow's milk was found to contain considerable sodium citrate, of which goat's milk contains none and human milk only a small proportion.

7. Goat's milk was found to contain both sodium and potassium chlorides, of which cow's milk and human milk contained none.

8. Calcium chloride was found in about equal proportion in cow's milk and goat's milk, being about twice as large in both cases as found in human milk.

As will be seen later, these differences in the composition of the three milks studied do not at present furnish an explanation of the

reason why goat's milk appears to have been better adapted in many instances to infant feeding than cow's milk.

GOAT'S MILK AS FOOD FOR INFANTS AND VERY YOUNG CHILDREN.

Experience seems to have indicated that goat's milk has a peculiar adaptability to feeding infants and young children in cases where they must be given artificial food and neither modified cow's milk nor the proprietary infant foods are fed with success. Because of this fact, it was determined to make a somewhat exhaustive study of the applicability of goat's milk to infant feeding.

An effort was made to get data in two ways: (1) By a somewhat extensive series of observations carried on in cooperation with Doctors DeWitt H. Sherman and Harry R. Lohnes, of Buffalo; and (2) by supplying goat's milk for use with infants in those cases where other foods had been found inapplicable. The physicians mentioned carried on their study in the Saint Mary's Infant Asylum and Maternity Hospital, in Buffalo, to which institution practically 16 quarts of goat's milk was shipped daily in an iced container. This milk was drawn under the best of conditions and was found to be in good condition at the time of arriving at the hospital.

The results of this work have been published in the Journal of the American Medical Association.1 The text of the report is here reproduced:

PRACTICAL STUDY OF GOAT'S MILK IN INFANT FEEDING AS

COMPARED TO COW'S MILK.

DE WITT H. SHERMAN, M. D., AND HARRY R. LOHNES, M. D.
Buffalo, N. Y.

This study was made to learn, if possible, the reason why goat's milk agrees better with some infants than cow's milk.

The babies were selected at random from the inmates of St. Mary's Infant Asylum and Maternity Hospital.

The goat's milk averaged from .5 per ct. to 1.50 per ct. richer in fats than the cow's milk used in this institution. Nevertheless, in making our modifications of the two milks the same amount of either milk was used in the stock formulae. Consequently the caloric value of the goat's milk modifications were greater than of the cow's milk.

1 Jl. Amer. Med. Assn. 62:806, 807. 1914.

I will present our data under different headings, for example:

I. Results of Gastric Analyses.

All test meals were withdrawn one hour from the middle of the feeding. The amount recovered averaged, in fourteen cases, nearly twice as much of cow's milk formulae as goat's milk, the accurate ratio being 41 c.c. to 27 c.c. This points

to the lower digestion of cow's milk.

The curds of the goat's milk formulae were smaller and more flocculent, corresponding to the appearance in test tube digestion.

The analyses in fifteen cases showed in goat's milk an average of:

The analyses in fourteen cases, the same infants, on cow's milk gave an average of:

In a previous paper on Gastric Analyses in Infants" we calculated the average gastric analysis on (a) barley water and (b) proprietary foods containing some proteids, fat, sugar and carbohydrates but no milk. For convenience of comparison we give below a table showing the averages of these and also of modified goat's and cow's milk, as shown above:

This table indicates the greater stimulating effect on the stomach of cow's milk than goat's milk; the greater stimulating effect of both than of the proprietary foods made without milk; and finally the greater stimulating effect of all three than of barley water.

In some vomiting cases this table gives definite information as to the causal factor of this condition in these four kinds of foods (a) through direct gastric stimulation, and (b) because of the size and density of curds, and (c) because of slower digestion.

As regards absorption and gain in weight our statistics are indefinite for three reasons. First, the babies being institution babies were apt to gain slowly: Second, owing to an epidemic of streptococcus infection which had swept through the infant's ward causing gastro-intestinal disturbances so serious that about one-fifth of the babies died, the digestion of those who did recover was more or less impaired; and third, because a certain number who were doing well were removed from the institution through adoption before our experiments were finished.

Of the number tested, sixteen cases in all, on similar formulae, twelve gained more rapidly on cow's milk modifications, and four on goat's milk. The gain of the former group was in the ratio of three to one. The gain of the latter group was in the ratio of nine to one. Consequently those who did gain on goat's milk gained more rapidly than on cow's milk, but fewer gained on goat's milk.

In infantile atrophy and inanition, three cases, goat's milk was no more suitable than any other food.

Taste of goat's milk.-At first the babies did not finish all their bottles, not liking the flavor of the food as well as of cow's milk. This surprised us because there seemed to us to be no unusual taste, as is usually supposed to be the case. If the milk is clean, we think the strong taste so common in Switzerland is unnecessary, and believe that the taste, as well as odor, is due to lack of udder cleanliness or possibly the type or food of the goat. But taken from three sources in this state Geneva, East Aurora, and Buffalo we have never found the strong taste or odor commonly supposed to be characteristic of goat's milk.

Fats. In goat's milk the fat averaged from .5 per ct. to 1.50 per ct. higher than in cow's milk. The more thorough emulsification of the fat in goat's milk, possibly the finer fat globules, prevents the separation of the cream upon standing, as occurs in cow's milk. This fact may be an element in reducing the tendency to regurgitation in goat's milk as compared to cow's milk, and further, may be a very important factor in lessening the liability to sour vomiting due to fatty acid fermentation, so common in high fat mixtures of cow's milk.

Strength of formulae. The babies, as a whole, tolerated equally well similar amounts of goat's or cow's milk with the same diluents, but as the goat's milk contained higher fat than cow's milk, they actually received more fat per feeding, and hence higher caloric value. It is consequently strange, that more babies gained, as stated above, on cow's milk than on goat's milk modifications.

In our formulae fat rarely averaged over 24 per ct. in goat's milk, and a little less than 2 per ct. in cow's milk.

Stools. On goat's milk the stools were as a rule smaller and of a more vivid yellow color.

Age.-Age did not influence materially the ratio of gain on goat's versus cow's milk, but the younger the baby, the more the evidence pointed toward a greater gain on goat's milk.

Vomiting. Of twenty-four children receiving both goat's and cow's milk formulae, five vomited goat's milk some, and none cow's milk. The explanation may be the slightly lower fat percentage in the cow's milk formulae. On the other hand, in certain cases goat's milk is often tolerated in similar amounts due to the fact that there is less gastric stimulation, that is, smaller, lighter curds.

After getting accustomed to goat's milk more children seemed hungry, and cried more at night than when taking cow's milk.

Type of modification commonest used was similar amounts of goats milk or cow's milk diluted with rice water and 2 per ct. cane sugar.

We are very much indebted to Doctor W. H. Jordan, director of the New York State Agricultural Experiment Station, for sending us for six weeks sixteen quarts a day of goat's milk in the best of condition. We are also indebted to Sister Frances and Sister Clare at St. Mary's Infant Asylum and Maternity Hospital for their assistance and cooperation in our work.

Practical results with individual children. During several years. the Station gave goat's milk without charge to all applicants, so far as the supply would permit, where it was desired to try this milk with infants or young children who were not thriving on any other kind of food. The only condition imposed was that reports should be made to the Experiment Station of the results of the experimental feeding. There follows the testimony of parents as to the outcome of the feeding and in a few instances the notes of attending physicians. It will be observed that this testimony is largely from parents, the professional judgment of physicians not being available in many of these cases. It is fair to conclude, however, that evidence so marked, even if given on an unprofessional basis, is not to be disregarded.

No. 1.- Mrs.

I cannot thank you enough for the milk we are getting. Our baby has improved wonderfully on it. He has gained two pounds in the three weeks we have been using it. We tried everything we could find for him but nothing would agree with him until we got the goat's milk.

No. 2.- Mr.

This milk has done more for our babe than any food we have used, as yet. Thanks to you for your kindness, I will report again at week end or according to your orders.

No. 3.- Mrs.

I will give you as nearly as possible the results we had with the goat's milk. The baby had been troubled with constipation and with vomiting, showing large undigested curds. We started on the goat's milk, diluting it about one-fourth with water. She took to it kindly and conditions greatly improved. A little later I added between one and two tablespoons of cream to the goat's milk and that seemed to satisfy her better and still agreed with her. As warm weather come on, she turned against it, and I could not get her to take it any longer.