SEPTEMBER, 1008. Vol. XXXIII., No. 380. PROCEEDINGS OF THE SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS. THE ESTIMATION OF FAT IN DRY MILKS. BY BASIL G. McLELLAN, F.I.C. SINCE the advent on the market of condensed milk in a dry form, as produced by the Just-Hatmaker and other procesBes, analysts have had to contend with the problem of analyging yet another form of milk. The difficulty of completely extracting the fat from the older varieties of condensed milk-especially in the case of sweetened milks-has been noted from time to time, and this difficulty still presents itself in the case of dry milks. The incompleteness of the extraction is due to the coating of the fat globules in the milk during the process of evaporation with an impermeable substance, which prevents tho solvent from penetrating.This investigation has been made with a view to test the extracting powers of354 THE ANALYST. various solvents which have been recommended for milk analysis, and also the accuracy of one or two rapid methods which have been recently suggested as applicable to this form of milk. The Werner-Schmidt method, which consists in the treatment of 2 grams of milk with warm water, boiling with hydrochloric acid and extracting with successive quantities of ether, gives good results, and is rapid in its application. Many analysts, however, prefer to extract the milk powder direct, using a Soxhlet extraction apparatus, and it is this method which I have specially investigated. As a result of my work I find that it is possible to completely extract the fat from dry milk in this way, but the extraction must be continued for a considerable time.This fact has been noted by many investigators, and is borne out by the accompanying figures : TARLE I. SHOWING LENGTH OF TIME REQUIRED FOR EXTRACTION OF FAT. Time of Extraction using Pure Ether (Sp. Gr., 0.720). Per Cent. Fat Extracted. 1. Four hours ... ... ... ... ... .., 2. Soaked in ether overnight and then for four hours 4. Soaked in ether overnight and then for eight hours 5. True result with longer extraction (see Table 11.) 3. Eight hours ... ... ... ... ... ... 24-84 25-51 25.70 26.14 26.64 The various solvents tried were : 1. Petrol: The petrol used was ordinary motor-car spirit, the fraction which boiled below 50' C. being collected for use.Usually about 70 per cent. of the spirit distils below 50" C. The price of this solvent rectified by the user is about 6d. per pound. 2. Methylated ether of 0.720 sp. gr. 3. Pure ethylic ether of 0.720 sp. gr. 4. Methylated ether of 0.730 sp. gr. 5. Pure ethylic ether, prepared by drying over calcium chloride and then over aodiurn, and finally distilled. This specially prepared ether is recommended in American text-books on milk analysis. I n applying the various tests, 5 grams of the milk preparation were used in each case. This was tLccurately weighed out and made up into a packet with three wrappings of filter-paper, and extracted for a long time with each of the solvents, in a Soxhlet apparatus. I t was customary to allow the packages to soak in the solvent overnight, to extract during the whole of the next day, to again soak in the solvent over the following night, and then, after extracting for an hour or two the following morning, to distil off m d weigh.Thus each solvent acts upon the milk for the same length of time and under the same conditions. Price about 1s. 2d. per pound. Price about 6s. per pound. Price about 10d. per pound.THE ANALYST+ 355 TABLE 11. SHOWINU PERCENTAUES OF FAT EXTRACTED BY THE VARIOUS SOLVENTS. Sample. A . . . . . . B . . . . . . c . . . . . . D . . . . . . E . . . . . . F . . . . . . .a . . . . . . Moisture. Por Cent. 5-70 3.88 5.82 3.90 0.7110 Ethylic Etlicr. - - - 25-00 26.34 25.90 26.64 Pcr Cent. Fat Estrtuted by- 0.720 Methy- lated Ether. - - - 25.12 26.78 25-84 26.62 O * i 3 0 Methy.lated Ether. 22-84 25.46 24.84 25.48 26.72 26.08 26-68 Pure Dry Ethylic Ether. - - - 23.98 24.59 24.64 24.90 Petrol. 20.86 23-92 23.28 23-84 24.39 24.48 24.60 I t will be noted from the table that neither the petrol nor the pure ethylic ether completely extracted the fat in the time mentioned. This fact was further proved by examining the extracted milk under the microscope for fat, which was still present in these two cases. In the case of methylated ether of sp. gr. 0.720 and 0.730 the extraction of fat was practically complete, as judged by microscopical examination, but in neither case was the fat which was extracted pure. These solvents extract more than the fat, this being evident from the fact that the fats, after having been dried in the oven, did not appear clear, but were niixed with a small amount of some semi-charred substance.That the fats extracted by the methylated ethers contained some substance other than pure milk-fat was determined by shaking the fats in the flasks (after weighing) with warm distilled water, cooling so as to solidify the fat, and evaporating the filtered aqueous solution to dryness. No residue was observed with the fats which had been extracted with petrol, 0.720 ethylic ether, and dry ether. A small amount of residue was left in the case of the fat extracted with 0.720 methylated ether and a somewhat larger amount with the fat extracted with 0.730 methylated ether. These residues were too small in quantity for any fuli examination, but were probably composed of lactose or an albuminoid constituent of the milk. In the case of 0.720 pure ether the extraction is complete, as is ascertained from the microscopical examination of the contents of the package ; and further, the fat is obtained clean.If, then, this is adopted as giving correct results, it will be seen that the results Gbtained with the two methylated ethers are too high. I t appears that the small amount of alcohol or moisture which is present in the 0.720 ethylic ether exerts a disintegrating action on the coatings which surround the fat globules, and so make the complete extraction of fat possible. I n the case of pure ether and petrol this disintegration does not go on, and consequently the extraction of fat is not complete. I t is clear, then, that when methylated ethera are used the impurities present This gives a very delicate indication of the presence of fat.356 THE ANALYST.dissolve substances other than fat, this being much more the case with methylated ether of higher specific gravity. The following two methods have been recommended for use in the analysis of dry milks, and although they did not appear to be of value, we have included results obtained by them, which are as follows : 1. One gram of the milk powder is dissolved in about 10 C.C. of warm water in a graduated cylinder. The solution is then mixed with 1 C.C. of ammonia water (of about 10 per cent.), and allowed to stand for about half an hour. Then 10 C.C. of alcohol (96 per cent.), 25 C.C. of dehydrated ether, and 25 C.C. of petroleum ether are successively added, and thoroughly mixed with the milk powder solution.After two or three hours the volume of the ether-petroleum ether-fat solution is read off and 40 C.C. of the solution drawn off into a weighed flask, the ether and petroleum-ether being then evaporated and the fat dried and weighed. 2. Five to 10 grams of the milk are placed in a long-necked flask with 20 C.C. of 50 per cent. alcohol, and boiled for a short time, cooled, and petrol (boiling-point, 40"-50") added, and the liquid again boiled whilst being well shaken. The contents of the flask are then transferred to a 250 C.C. cylinder, and the flask washed out into the cylinder with petrol. The contents of the cylinder are well shaken and allowed to stand till the petrol completely separates. Its volume is then read off, and an aliquot part of this is evaporated in the usual way and weighed. The unreliability of these two methods is shown by the figures given in Table 111. TABLE 111. New Methods tested on Sample G., Table 11. Fat, per Cent. No. 1, two estimations ... ... No. 2, four estimations .. . ... Actual per cent., see Table 11. ... 19.2 20.4 22.44 24-48 { 22.1 24.08 26.64 To summarise : 1. For an accurate estimation of fat in dry milk by direct extraction, ethylic ether (sp. gr. 0.720) must be used. 2. For an estimation accurate to about 0-3 per cent. it is possible to use methylated ether (sp. gr. 0.720). 3. Methylated ether (sp. gr. 0-730), petrol, and specially purified ether give incorrect results. 4. The Werner-Schmidt method gives satisfactory results, whereas the two methods (numbered 1 and 2) given above are useless.