In respect of this process there would be danger, on long standing, of the formation of free acids from butter glycerids, and these acids would be removed by the process of washing prescribed. In this case the quantity of fat obtained would be less than in the original sample.
459. Comparison of Methods.—An immense amount of work has been done by analysts in comparing the various types of extraction methods outlined above.[445]
The consensus of opinion is that good results are obtained by all the methods when properly conducted, and preference is given to the two methods finally adopted by the Association of Official Chemists. As solvents, pure ether and petroleum spirit of low boiling point are preferred. The direct extraction gravimetric processes are important, since it is to these that all the other quicker and easier methods must appeal for the proof of their accuracy.
460. Wet Extraction Methods.—It has been found quite impracticable to extract the fat from milk by shaking it directly with the solvent. An emulsion is produced whereby the solvent itself becomes incorporated with the other constituents of the milk, and from which it is not separated easily even with the aid of whirling. The disturbing element which prevents the separation of the solvent is doubtless the colloid casein, since, when this is previously rendered soluble, the separation of the solvent holding the fat is easily accomplished.
The principle on which the methods of wet extraction are based is a simple one; viz., to secure a complete or partial solution of the casein and subsequently to extract the fat with a solvent immiscible with water. The methods may be divided into three great classes; viz., (1) those in which the solvent is evaporated from the whole of the extracted fat and the residual matters weighed; (2) processes in which an aliquot part of the fat solution is employed and the total fat calculated from the data secured; (3) the density of the fat solution is determined at a definite temperature and the percentage of fat corresponding thereto determined from tables or otherwise. Methods (1) and (2) are practically identical in principle and one or the other may be applied according to convenience or to local considerations. The methods may be further subdivided in respect of the reagents used to secure complete or partial solution of the casein, as, for instance, alkali or acid.[446]
461. Solution in an Acid.—A good type of these processes is the method of Schmid.[447] In this process ten cubic centimeters of milk are placed in a test tube of about five times that content, graduated to measure small volumes. An equal quantity of hydrochloric acid is added, the mixture shaken, boiled until it turns dark brown, and cooled quickly. The fat is extracted by shaking with thirty cubic centimeters of ether. After standing some time the ethereal solution separates and its volume is noted. An aliquot part of the solution is removed, the solvent evaporated, and the weight of fat in the whole determined by calculation.
The schmid process has been improved by Stokes,[448] Hill,[449] and Richmond.[450] The most important of these variations consists in weighing instead of measuring the milk employed, thus insuring greater accuracy. Dyer and Roberts affirm that the ether dissolves some of the caramel products formed on boiling condensed milk with hydrochloric acid, and that the data obtained in such cases by the process of Schmid are too high.[451]
Since lactic acid is also slightly soluble in ether, sour milk should not be extracted with that solvent. In these cases petroleum spirit, or a mixture of petroleum and ether, as suggested by Pinette, may be used.[452] Another variation consists in extracting the fat with several portions of the solvent and evaporating all the extracts thus obtained to get the total fat. This method is perhaps the best of those in which the fat is extracted from the residual liquid after the decomposition of the casein by an acid, and may be recommended as both reliable and typical within the limitations mentioned above.
462. Solution in an Alkali.—The casein of milk is not so readily dissolved in an alkali as in an acid, but the solution is sufficient to permit the extraction of the fat. Soda and potash lyes and ammonia are the alkaline bodies usually employed. To promote the separation of the emulsions, alcohol is added with advantage. The principle of the process rests on the observed power of an alkali to free the fat globules sufficiently to allow them to dissolve in ether or some other solvent. When the solvent has separated from the emulsion at first formed, the whole or a part of it is used for the determination of fat in a manner entirely analogous to that employed in the process with the acid solutions described above. There are many methods based on this principle, and some of the typical ones will be given below. Experience has shown that extraction from an alkaline solution is more troublesome and less perfect than from an acid and these alkaline methods are, therefore, not so much practiced now as they were formerly.
463. Method of Short.—Instead of measuring the volume of the separated fat, Short has proposed a method in which the casein is dissolved in an alkali and the fat at the same time saponified. The soap thus produced is decomposed by sulfuric acid and the volume of the separated fat acids noted. This volume represents eighty-seven per cent of the corresponding volume of fat.[453]