The instrument employed by Marchand is a tube about thirty centimeters long and twelve in diameter, closed at one end and marked in three portions of ten cubic centimeters each. The upper part is divided in tenths of a cubic centimeter. The superior divisions are subdivided so that the readings can be made to hundredths of a cubic centimeter.
The tube is filled with milk to the first mark and two or three drops of a twenty-five per cent solution of soda lye added thereto. Ether is poured in to the second mark, the tube closed and vigorously shaken. Alcohol of about ninety per cent strength is added to the upper mark, the tube closed, shaken and allowed to stand in a vertical position, with occasional jolting, until the separation of the liquids is complete. In order to promote the separation the tube is placed in a cylinder containing water at 40°.
When the separation is complete the milk serum is found at the bottom, the mixture of alcohol and ether in the middle and the fat at the top. The mixture of ether and alcohol contains 0.126 gram of fat, and each cubic centimeter of the separated ether fat 0.233 gram of fat. The total volume of the separated fat, multiplied by 0.233, and the product increased by 0.126, will give the weight of fat in the ten cubic centimeters of milk employed.
Example.—Milk used, ten cubic centimeters of 1.032 specific gravity = 10.32 grams. The observed volume of the saturated ether fat solution is two cubic centimeters. Then the weight of fat is 2 × 0.233 + 0.126 = 0.592 gram. The percentage of fat in the sample is 0.592 × 100 ÷ 10.32 = 5.74.
In the apparatus used in this laboratory the upper division of the graduation is marked 12.6, because this represents the quantity of fat which remains in the ether-alcohol mixture for one liter of milk. From this point the graduation is extended downward to ninety-five, which, for ten cubic centimeters of milk, represents 0.95 gram. After the fat has separated, enough ninety-five per cent alcohol is added to bring the upper surface exactly to the graduation 12.6. The number of grams per liter of milk is then read directly from the scale.
In respect of applicability, the observation made regarding Soxhlet’s areometric method may be repeated.
In practical work in this country the lactobutyrometer is no longer used, but many of the recorded determinations of fat in milk have been made by this method.
469. Volumetric Methods.—For practical purposes, the volumetric methods of estimating fat in milk have entirely superseded all the other processes. It has been found that the fat readily separates in a pure state from the other constituents of milk whenever the casein is rendered completely soluble; whereas no process has yet been devised whereby the fat can be easily separated in a pure state from milk which has not been treated with some reagent capable of effecting a solution of the casein. The volumetric methods may be divided into two classes; viz., (1) Those in which the fat is separated by the simple action of gravity, and (2) those in which the natural action of gravity is supplanted by centrifugal motion. Each of these classes embraces a large number of variations and some of the typical ones will be described in the following paragraphs. As solvents for the casein a large number of reagents has been used, including alkalies and single and mixed acids. In practice, preference is given to the least complex and most easily prepared solvents.
470. Method Of Patrick.—A typical illustration of the method of collecting the fat after solution of the casein, without the aid of whirling, is found in the process devised by Patrick.[460]
The solvent employed is a mixture of acetic, sulfuric and hydrochloric acids, saturated with sodium sulfate, in the respective volumetric proportions of nine, five and two. The separation is accomplished in a large test tube drawn out near the top into a constricted neck which is graduated to measure the volume of the separated fat or to give direct percentage results.