An eighteen ounce mixture, of two per cent. fat, one-half per cent. proteid, six per cent. sugar, five per cent. limewater was required. The modification (Baby F) was made according to the equation given. The modified milk was then tested for its fat per cent., with the result that in thirteen days the average fat was 1.99 per cent. We could hardly look for more accurate results than these. Our method of separating the upper four ounces was one that every busy mother would probably use. We simply poured off the top milk or cream, our object being to test a home method of modification that would be within the range of ability of any household. We find that the percentage of fat by this method does not differ from that obtained by use of the siphon.

NO.

TABLE C.

PER CENT PER CENT OF DISTANCE OF CREAM LEVEL FROM THE TOP TO THE
OF FAT
BOTTOM OF A STANDARD QUART BOTTLE AFTER STANDING

The average cream depth of those milks approximating 4% fat is a little less than a third of the distance from the top to the bottom of the bottle. Those approximating 5% a little over a third of the distance (36.9%).

The percentage of fat in the whole milk of herds will necessarily vary with the conditions existing at different times of the year, and while it is not necessary to make daily fat estimations they should be made at more or less frequent intervals, particularly if the infant shows signs of digestive derangement.

We have made a few observations on the depth of the cream layer, after standing four hours (Table C), with relation to the fat content of the whole milk. A regulation quart bottle, allowed to stand four hours, after being thoroughly mixed, will show a more or less definite fat per cent. when the cream layer is a certain depth. When the depth was thirty-three and one-third of the bottle depth the per cent. of the fat in the whole milk averaged about four; when the cream depth was thirtyfive to thirty-seven per cent. of the bottle depth, the fat per cent. aver

aged five. This, too, can be used with fairly accurate results in home modification.

To vary the percentage of fat, it is necessary to use different layers of the milk. When the upper four ounces contain approximately sixteen per cent. fat, the second four ounces will yield six per cent. fat, the third four ounces four per cent. fat and the fourth four ounces three per cent. fat, so that by simply pouring off the top in definite amounts almost any desired per cent. of fat content may be obtained. By using the upper eight, twelve or sixteen ounces, we may get fat per cents. of approximately twelve, ten, and seven per cent., which are somewhat lower than the percentages claimed by Holt.

There are a good many reasons why gravity cream is better than centrifugal cream, for milk modifications. There may be something in the objection that in centrifugal cream the natural emulsion is partially destroyed, but there is considerable difference of opinion upon this point. We think that the arguments in favor of gravity cream are that it is usually fresher and the percentage of fat more constant than in the centrifugal cream distributed in this section. If creams of definite percentages were placed on the market, we can see that a great advantage would be secured. There seems to be no effort in this direction in the smaller towns and unless the fat per cent, in cream is estimated very frequently, disordered digestion will surely arise.

Our method of removing different layers of top milk after it has stood a certain length of time, and thus procuring more or less definite fat percentages, seems to us to be far more satisfactory. Holt and others following in his lead, have given examples of top milk feeding and have reported their results of analyses of the different layers of top milk. We have repeated their experiments (Table D) and have

been unable to produce the results obtained by them. Our conditions were probably different. We sought to have them the same as in the home, where the modifying is most often done. A milk that is set immediately after milking will, after four hours, raise a higher per cent. of fat in the cream than milk that has stood for a time, then mixed and reset for four hours. The latter are the conditions that must necessarily exist in the home, when milk is obtained from a milkman some hours after milking, the time of the milking and the amount of jolting unknown. It may be for this reason that our results differ from those of other investigators. At no time were we able to obtain twenty per cent. fat in the upper four ounces of a quart of milk. The highest we encountered was eighteen and four-tenths per cent. Usually an average of sixteen per cent. was obtained. Even with five per cent. whole milk twenty per cent. cream could not be obtained by four or six hours setting.

From our investigations we can say that a milk, after being thoroughly mixed by pouring from one vessel to another and set in a standard quart bottle, will yield in the upper four ounces approximately sixteen per cent. fat instead of twenty per cent., as is claimed by Holt and others.

THE PROTEIDS OF COW'S MILK.

The proteids of cow's milk differ from those of woman's milk in that the percentage of casein is greater and that of soluble albumin less. It is the casein that gives us the greatest trouble in infant feeding. It was never intended that the human animal should consume large amounts of casein. It would seem then that any method of milk modification that has as its object the elimination of excess of casein, must be the most ideal. So far as we know the proteids that remain in cow's milk after the casein has been removed, are identical with those of woman's milk. They are at any rate soluble albumins and require much less digestive work than do the insoluble or solid milk bodies. If, then. we wish to feed an infant soluble albumin, in quantities sufficient to equal the proteid content of mother's milk, it is evident that the use of whey alone which contains but 1.17 per cent. proteid will not accomplish this. To increase the content we should have to evaporate the whey, that is, if we evaporate the water of twenty ounces until it is reduced to ten ounces we will have a whey containing twice the amount of proteid of the twenty ounces, or 2.34 per cent. proteid. This is hardly practical for home modification because of increased time required in the preparation, which is in reality the only objection raised by mothers against modified milk. During the first two months of life is the time when an infant most needs soluble albumin or milk that closely resembles mother's milk. We do not care to raise our proteid content during this period above one per cent.

It would seem that our argument does not hold, when our constant aim has been to modify the milk of the cow to as nearly a counterpart of mother's milk as possible, for it is a well-known fact that throughout

the lacteal period mother's milk never contains less than two per cent. proteid. But we must bear in mind constantly that cow's milk can never be exactly like mother's milk, and practical experience has demonstrated to us that infants thrive on whey modifications and develop bone.

Nothing definite has been worked out with regard to the exact modification of the salts. These are reduced by dilution as are the proteids. Experiment has shown, however, that very little of the salts are lost in the making of whey, so that by the use of whey in our modifications instead of water as a diluent we increase rather than decrease our salt per cent.

Clinical experience has taught us that infants fed on modified milk do better proportionately as the amount of casein or solid food is gradually increased. It may be necessary to begin with simple whey, a method that has proven very satisfactory, where the child is never nursed by the mother. To feed an infant for any considerable length of time on a purely soluble albumin diet or low casein diet might easily lead to serious consequences.

To prepare soluble albumin or whey the casein is removed most conveniently by the use of commercial junket tablet. One tablet dissolved in a little water, added to a quart of skimmed milk, which has previously been heated to 98° Fahrenheit will yield from thirteen to fourteen ounces of whey in fifteen minutes time. The casein separates as a greenish-yellow, semiopaque fluid. In separating the curds from the whey pressure should be avoided. We found a single layer of absorbent cotton worked rapidly and effectually as a filter.

The mother must be especially instructed in regard to the use of the junket tablet as the directions on the box are not for making simple whey. Trouble arises if squeezing is employed, as casein passes through and the whey will have a milky look.

It must be remembered that rennet works rapidly, and like other ferments there is little limit to the duration of its action. It is still actively present in the whey and if it comes in contact with the casein, as the casein of the cream in a milk modification, it will rapidly convert it into curds. Hence the necessity of bringing the whey to a temperature of 150° Fahrenheit before adding the cream. This requires but a few moments time.

HOME MODIFICATION VERSUS LABORATORY MODIFICATION.

We now come to the consideration of the accuracy of percentage milk modifications, modifications produced at milk laboratories, and those produced in the hospitals and in the home. We will consider only fat percentages, as fat is the most variable constituent of cow's milk, and the one that concerns us most in infant feeding. It is generally believed that percentage milk feeding and the term "modified milk" originated with those who were the instigators of the laboratory idea. It will be of interest to many of you to know that percentage feeding

was not only worked out some time before the formation of the WalkerGordon Company, but the ideas were given to the profession in book form by Doctor William Henry Cummings, of Williamstown, Massachusetts, in 1859, in a little book entitled "Food for Babes; or, Artificial Human Milk and Manner of Preparing and Administering It to Young Children." If any of you have access to this interesting little work by Doctor Cummings and will turn to page 58, you will find that he says: "The milk adapted for the newborn calf is not suitable for an infant. It must be modified or else it will do harm rather than good." He goes on to state that "there is too much cheese in cow's milk and the child cannot digest it." He then accurately works out the yearly needs of a child in pounds of butter, proteid and sugar, which he estimates at twenty-five, seventeen, and ninety respectively for a year's feeding for an average child. Then he proves the inadequacy of cow's milk as a substitute, showing that in the same quantity of milk a child would get but sixteen and one-half pounds of fat in a year's feeding, but a great increase in the casein. He further gives specific dilutions for various ages suiting the modification to the changes incident to development in the child.

It is said that Liebig had it so arranged that his prescriptions for infants' food, which consisted in the addition of maltose to cow's milk, were filled at an ordinary pharmacy and that other physicians patronized these places. Liebig's object was to bring about conditions that exist in woman's milk. This is another evidence of the adage, “There is nothing new under the sun," and would seem to detract from the claim of the laboratory people that they were the originators of the percentage method of feeding. However, as they have popularized this method they are entitled to the honor that should come from its general adoption.

Milk laboratories claim for their product accurate percentages and clean milk. It is upon these claims that their success depends. There has been some dissatisfaction with laboratory milk both on the part of the profession and the laity. We think that Doctor Morse is justified when he says that "much of the criticism has come from physicians who have failed to appreciate the purpose and capabilities of the laboratory; that some physicians seem to have found the laboratory a convenient scapegoat to account for the lack of development of certain babies, fed on laboratory milk prepared according to their own directions." "It has been much easier and more condusive to their selfrespect to attribute the failure to the laboratories rather than to their own imperfect methods of prescribing," and that "the laity are inclined to exaggerate every mistake made by the laboratory and minimize their own."

However, knowing much of the above to be true, we cannot help being impressed with certain facts that confront us. We must stand with those who give obeisance to the sentiment, "to the law and to the testimony." Does the Walker-Gordon Company live up to its claims?