supplying them with samples of milk from the various dairy farms which supply the city, and we take this opportunity to thank Doctor Wessinger, the health officer, for much useful information furnished. us at an expense of his valuable time. The milk was brought to the milk laboratory in serial number. The name of the dairy was withheld. It was collected by an officer at an unexpected time, of his own choosing. Except where otherwise stated the milk was bottled at the dairy and not taken from the tanks in the milk wagon. The analyses of this series has to do only with the per cent. of fat in the whole milk, the specific gravity and the total acidity. We have considered standard milk one which has four per cent. fat, a specific gravity of 1028 to 1033, and a total acidity not higher than 25.1 The results of the analyses are as shown in Table A. By reference to this table it will be seen that of the forty-nine dairies examined the fat percentage was four, or above four in twenty-five, and that in nine of these it was five or above; seven were between three and one-half and four; eight were below three and one-half, and two of these were three per cent. One dairy furnished a ten per cent. milk, which was too high for good milk and too low for good cream. It was reported as cream below standard, and subsequently it was found that the high percentage of fat was due to an attempt to deceive the Board of Health, which had had trouble with this dairy before. The specific gravity varied proportionately with the fat in most cases. It is interesting to note that in the milk of herds in which the inspector found mammary and udder disease, the total acidity was high, always 281 or above. Invariably when the acidity was much above 25 it could be traced to dirty stables, dirty cows, and careless milking methods. The trouble has been in the stables and in the milker. It is of very little avail that the housekeeper puts the milk in a cool place and protects it from all contamination, if from the time it leaves the udder, until it reaches the consumer, it has been through a course of dirt collection from unclean hands, unclean pails, unclean tanks, et cetera. This journey through unclean receptacles is too frequently evidenced by the visible sediment that greets the eye as the bottom of the container is reached. We found through the investigation of the Board of Health, that the usual plan of dispensing milk is as follows: All bottled milk is taken from the previous evening's milking, and that which is not bottled is carried in the faucet tank in the front compartment of the wagon. All milk taken from the cows in the morning is put into ordinary cans. It must be evident to every one that the bottled milk, even though it is of the night milking, stands a better chance of reaching the consumer in good condition than milk that is put in the faucet tank or in ordinary cans. It must be a very difficult task to keep the tanks and cans clean, even when the strictest measures of cleanliness are observed. How great the chances of contamination when no special thought is given to them! Picture the bacteria that may lurk in the seams of the can and the bottom where it is impossible to reach with hand and brush! After counseling with the Board of Health we decided that for the present, until more elaborate methods can be carried out to prove the value of these simple methods, that standard milk must contain four per cent. fat, and have an acidity not above 25 per cent. decinormal sodium hydroxide solution. We have found that by notification and instructions to the dairymen that the high acid values in the milk have been lowered. We find most of the milkmen anxious to cooperate as best they can, and a visit to the various dairies in this district demonstrated to us that much of the faulty condition was due to their lack of knowledge of bacterial growth. The condition in the milk rooms was for the most part fairly good, but in the stables no efforts were put forth to secure the cleanliness that is so essential. The amount of good that can be accomplished by local boards of health cannot be overestimated. When the public begin to realize that by far the greater number of deaths in infants and young children, rich and poor alike, are due to diseases of the digestive organs, and that these are caused by impure milk, greater efforts will be put forth to bring about conditions that will compel dairymen to have regulation stables and produce standard milk. If only common cleanliness could be guaranteed, thousands of infants' lives could be saved yearly. Sterilization will not accomplish as much as is expected of it. Germs kill not only by invading the tissues of the body but also by elaborating chemical poisons in the milk, which are unaffected by either Pasteurization or sterilization. We must have clean milk. If health officers were better paid and were obliged to qualify for office by competitive examination, before an unbiased board, much of the difficulty would be solved. Taking all things into consideration the above analyses show that the milk supplied to Ann Arbor is on the whole good so far as its fat content is concerned. Some of the dairies, particularly numbers one and six, we have examined at various times for two or three years and have never found the fat per cent. below four. They have frequently been dirty, and at these times their acidities have been high. The dairies around Ann Arbor represent the prevailing type. They are probably no better and no worse than those in other districts. The milk that is being delivered is physically good, is seldom skimmed or diluted. Lowered fat percentages can usually be traced to improper and poor feeding, in but few instances to diluting. MILK ANALYSIS. Practical milk analysis resolves itself into the estimation of but a few points. Without these no milk modification can be correct. The possession of more minute data will help us but very little. Even methods that are thought to be most ideal are only approximate. We have, then, to estimate the per cent. of fat in the whole milk and in the cream, the specific gravity, and if we would have the alkalinity correct, the total acidity should be known. However, for practical purposes this is not necessary, and the total acid value may be employed as a means of gaining some idea of the cleanliness of the product used. Fat estimation requires about ten minutes time. The most reliable. test is the Babcock test. Small hand-turning testers are manufactured by various creamery manufacturing companies. A whole outfit can be bought for a comparatively small sum, including directions which are too familiar to you all to require repetition here. All fat estimations reported in this paper were made with a large electric centrifuge. The approximate percentage of proteids is estimated by comparison of specific gravity with fat per cent. The proteids of milk are fairly constant, more constant than the fat. They are estimated at four per cent. If the specific gravity is low and the fat per cent. high, the proteids are normal. If the fat per cent. is normal or high and the specific gravity high, the proteids are increased. If the fat per cent. is normal or low and the specific gravity is low, the per cent. of proteids is decreased. It is more necessary to be in possession of definite percentages of fat than of proteid. The symptoms of excess of proteid are much more evident than are those of excess of fat. The stools may be used as an index of proteid digestion, yet it must be remembered that infants may thrive and manifest no complaints and still show curds in the stools. This symptom alone does not necessarily mean that the casein must be decreased. It must be combined with other phenomena which go to make up the symptom complex of proteid indigestion. Sugar in cow's milk for practical purposes may be considered as a constant factor varying little from six per cent. The polariscopic method for sugar estimation is the most practical and is very accurate. Knowing the percentage of fat in the whole milk and in the cream, it is a very simple matter to calculate the quantity necessary to give the required amounts. We find Baner's2 method of great practical value. It is simple and gives fairly accurate results. We think as accurate results can be obtained by home modification with this method as are usually obtained by Walker-Gordon Laboratories, and it is much better from the standpoint of independence as there is no patent upon it. Percentage feeding resolves itself into so simple an algebraic expression that any one mathematically inclined can devise equations of his own, if he cares to. Those who do not wish to spend time working out equations of their own can accept this method and have the assurance that they will obtain good results. Possibly its only point of failure is in the calculation of low proteid percentages, but this applies only when low per cent. top milk fat is employed. The method is as follows: BANER'S METHOD. Let Q represent the quantity of mixture required for twenty-four hours. F represent the desired per cent. of fat. P represent the desired per cent. of proteid. A represent the desired per cent. of limewater. C = Cream, M = Milk, W = Water, LW = Limewater, S = Sugar. It will be found that Example: A mixture of forty-eight ounces containing three per cent fat, one and five-tenths per cent. proteid, six per cent. sugar, and five per cent. limewater is desired. The whole milk is known to contain four per cent. fat, and the cream, upper four ounces, after standing four hours, sixteen per cent. fat. These equations assume that the whole milk contains four per cent. fat and four per cent. proteid. In order to use the formula with exact results we must substitute the actual per cent. of fat in both cream and whole milk. For instance, if milk tests 4.5 or five per cent. fat the cream will test about eighteen per cent. fat (Table D). After having run a number of tests we are satisfied that most milk containing four per cent. fat will not show greater than sixteen per cent. fat in the upper four ounces of cream. The following cases illustrate how near the desired amount of fat is obtained by the equations. TABLE B. |