THE ANALYST. 219 THE ANALYSIS OF DRIED MILK. BY H. DROOP RICHMOND, F.I.C. (Read at the Meeting, May 2, 1906.) MILK which has been evaporated to dryness is sold as dried milk or milk-powder, The methods of drying milk vary. The milk in one process is passed over polished steel rollers heated to a temperature above that of boiling water, and the film frOm which most of the water has been evaporated is taken off by knives set in a suitable position ; in another process the milk is condensed in a vacuum pan to one-third or220 THE ANALYST. ' 4-92 27.98 34-16 1-25 24.59 6.24 one-fourth of its bulk, and is then dried in a vacuum chamber in thin layers ; in a third process the milk is, after a preliminary concentration, dried on hot rollers ih vacuo ; and there are other methods which have not come into extensive use.The analysis of these milks requires a few slight modifications of the methods used for ordinary milk. The sample should be ground and well mixed to secure uniformity. Fat cannot be estimated by direct extraction, as the results are always low. The Werner-Schmid method is suitable. If there is no sugar except milk-sugar, the fat, after drying, should be dissolved in petroleum ether, and any residue weighed and subtracted from the total weight. In the presence of much of any other sugar, it is preferable to mix the ethereal solution with an equal bulk of petroleum ether, and shake out with water rendered slightly alkaline with ammonia before the solution is evaporated. fVdk-sugar may be easily and quickly estimated polarimetrically ; 10 grams of milk-powder are ground up in a mortar with sufficient hot water to make it into a paste, which is gradually thinned with hot water, and the solution made up to 100 c.c.; a little ammonia may be added if the milk-powder does not all go into solution.Unless this procedure is followed, incomplete extraction of the sugar may result. Ca?ze-sugar may be conveniently estimated by the method described by Harrison (ANALYST, XXX., 123). Proteids are calculated from the total nitrogen by Kjeldahl's method by the factor 6.87. Ash, Linze, a d Phosphoric Acid as usual. Acidity is estimated by grinding up about 1 to 2 grams with hot water, and The table below gives the analysis of seven samples : Noistwre is estimated by drying about 1 gram in the water-bath.. The usual method is then followed.titrating with & alkali, using phenolphthalein as indicator. Moisture . . . ... ... Fat ... ... ... .. . Milk-sugar , . . ... ... Cane-sugar . . . ... ... Proteids ... ... ... Ash ... . m , ... ... TABLE ~.--COMFOSITION OF DRIED MILK. 6-39 27.35 31.42 27-48 6.00 - _ _ - Water of hydration ... Total ... ... Change of temperature on - - 1.65 100-29 ~ ~~ ____ ____. _ _ .___ __I_ . -__ Total ... ...I 98.64 1 99.14 1-80 100.94 - 3 ._ - - 3.30 23.97 37-32 1.53 26.38 6-19 98.69 1-96 100.65 - 0.2" ~- ~ - ~_ .- - 4 .__ 3-55 2.55 45-60 2-80 35-45 7-89 97-84 -__ 2-40 100.24 - 0.4" 5 4.74 29.16 32.24 26-66 5.63 - 98-43 1-70 100.13 5-15 6-03 19-90 1 25.60 34.96 , 32-83 - ' 2.00 31.10 ' 23-84 7-11 6.44 - 98.22 96-74 1.84 ' 1.73 100*06 98-47 ~- - -THE ANALYST, 221 I t is noticed that none of the analyses add up to 100 per cent., but are con- siderably low ; the milk-sugar has been calculated as anhydrous sugar, and here, I think, is the reason for the deficiency.On shaking the solid residue obtained by drying milk on the water-bath, in which the milk-sugar certainly exists as anhydrous sugar, with water a rise of temperature always takes place ; anhydrous milk-sugar mixed with water always causes a rise of temperature, whilst hydrated milk-sugar causes a fall of 0.55" if more than can be at once dissolved is added. The milk-powders examined, with one exception (No. a), all caused a fall of temperature, and it is seen that the addition of the water of hydration to the total gives figures which are but slightly in excess of 100 per cent.; both the change of temperature and the slight excess over 100 per cent.indicate that the bulk of the milk-sugar, though not all, exists as hydrated sugar. Sample No. 2 differed in appearance from the others, being a heavy powder, instead of being light and flaky, and had doubtless been more dried, and probably contains a considerable proportion of anhydrous sugar ; it is noticed that the addition of the water of hydration would make the total nearly 101 per cent. Sample No. 7 gives a low total, which I believe is accounted for by the presence of invert sugar. I t will be noticed that samples 2, 3, 4 and 7 contain small quantities of cane- sugar; this in sample 2 was admittedly added in the form of saccharate of lime, and was certainly so added, judging from the analytical figures, in No.7. This is used to render the dried milk more soluble, and additions of sodium phosphate and other alkaline substances are made with the same object. I n Table 11. the composition of the original milks, on the assumption that they contain 9.0 per cent. of solids-not-fat, are given. TABLE II.-~OMPOSITION O F ORIGINAL JfILKS. ~ F a t ... ... ... ... Milk-sugar . . . ... ... Proteids . . Q ... ... Ash ... ... ... ... CaO ... ... ... ... P,?,. . . ... ... ... Acidity . .: ... ... 1 3.79 4-36 3.81 0.83 0.19 0.23 8 . 4 O 2 3.88 4.73 3.41 0.87 0.21 0-24 13-2" 3 3-09 4.81 3-40 0.80 0.17 0-23 16.8" 4 0.26 4.62 3.58 0.80 0.17 0.23 16.5" 5 4.07 4-50 3-71 0.79 0.17 0.23 19.6" 6 2-45 4.30 3.82 0.87 0.19 0.29 - 7 3.65 4.68 3.40 0-92 0-27 0.24 11.4" From this table it is seen that No.4 is made from separated milk, and No. 6 from milk deprived of a portion of its cream. The milk used to prepare No. 3 is only just above the Government standard. The normal percentages of lime and phosphoric anhydride in milk are 0.11 per cent. and 0.23 per cent. respectively, but vary some- vhat with the proteids, and the normal acidity is not far from 20". From a con- sideration of the results, it would appear that Nos. 2 and 7 have received an addition of saccharate of lime; and No. 6 has received an addition of a phosphate. Nos. 3 and 4 contain cane-sugar, but there is no evidence of the addition of saccharate of222 THE ANALYST. lime. No. 1 has probably received an addition of sodium carbonate, as the lime is not high enough, considering the high proteids, to indicate an addition of this substance; and No.5 appears to have received no addition whatever. DISCUSSION. The PRESIDENT (Mr. Bevan), in inviting discussion, said that he could not quite understand how it was that, in the case of homogenized milk, the Adams process gave low results as compared with the Werner-Schmid and other methods. He should have thought that the breaking-up of the fat globules would have caused them to be more readily soluble. He should be interested to hear by what process the samples of dried milk referred to were made. Mr. W. PARTRIDGE asked whether any unusual difficulty had been found in extracting the fat with ether in the case of homogenized milk.He had had occasion to examine some samples of condensed homogenized milk in which the degree of homogenization was stated to be very high, and, using the Werner-Schmid process, his results were similar to those of a Norwegian analyst who had examined the same milks; but a London analyst who had also examined them, by Bell's process, had obtained results which were regularly 1 per cent. lower. Mr. HEHNER said that he had made some time ago a few comparative experi- ments with regard to fat extraction in the case of dried milk prepared by the Natmaker process, and had found in every case that a larger proportion of fat was obtained by extracting the powder itself with ether than by any other means. When plaster of Paris and other like materials were used lower results were invariably obtained.I n the Hatmaker process, as in all others, alkali was added in the form of sodium carbonate, which would account for the somewhat high ash. He wondered why Mr. Richmond was so anxious that his analyses should add up to 100, seeing that there must necessarily be a number of undetermined constituents-for instance, there would be about & per cent. of citric acid in the milk, which, of course, would represent about qV per cent. in the dried product. He thought that the reason why, in the case of homogenized milk, the fat was not well extracted from paper was not quite satisfactorily explained by the photographs shown. When milk was mixed with an absorbent it evaporated, and the.fat remained more or less adhering to the surface presented to it.I t could not surely make much difference whether the absorbent was in small or large particles. The question must be whether the surface on which the fat was, so to speak, plastered was large enough to subdivide it completely, and he could not see why the surface presented by one material should be less favourable than that presented by another. Mr. M. WYXTER BLYTH said that, from some experiments which he had made with certain foods for infants, he could confirm what Mr. Richmond had said as to the difficulty of extracting the whole of the fat by the Adams process. I n one case he could only extract about 2 per cent. of fat, whereas a modification of the Werrrer- Schmid method, which he had used on Mr. Richmond's suggestion, yielded 8.24 per cent.of fat. The low results given by the Adams process in such cases were explicable by the fact that, especially with malted foods, a dry mass of sugar wasTHE ANALYST. 223 obtained which was very diflficult to deal with. I n the case, however, of milk in which the fat globules had been broken up, he quite agreed with Mr. Rehner that the reason why the fat could not all be got out was difficult to understand. Dr. LEATHER inquired how long the extraction with ether was carried on. Mr. E. R. BOLTON asked whether the fat extraction was carried out in a Soxhlet apparatus or in one of those extractors in which the solvent dripped down and trickled straight through. He had found that in the case of oil-cakes all the fat was extracted much more quickly in such an extractor than in the Soxhlet apparatus.MY. WYNTER BLYTH said that in his case an ordinary Soxhlet apparatus was used, and the extraction carried on for five hours. Mr. RICHMOND, in reply, said that the milk-powders referred to were obtained by the various processes described in the paper, but he did not know in all cases which process had been used. Some of the samples were very finely ground, and had been dried to a greater extent, and probably contained a larger amount of anhydrous milk- sugar. I t seemed to him that the photographs went a long way towards explaining the lowness of the Adams results. It must be borne in mind that the photographs (not reproduced) were on different scales. Had they been on the same scale the enormous difference in the total surfaces exposed to the milk would have been more apparent.I n the case of kieselguhr the layer was of the same order of thickness as the fat globules, and consequently they were all exposed to the action of the ether. Probably, also, the grinding-up of the milk and kieselguhr had some effect. In the determinations by the Adams method extraction was carried on for at least five hours in a Soxhlet apparatus. I n some cases extraction was carried on for so long as sixteen hours, but the full proportion of fat was still not obtained. He had not found any difficulty with the Werner-Schmid process. He never made more than four extractions, and at the fourth there was generally practically no fat left to take out. He had compared various methods of extracting, and had found four extractions at the outside to be sufficient.This applied also to homogenized condensed milk. Mr. PARTRIDGE asked how much ether Mr. Richmond used for each extraction. He himself used about 15 or 20 C.C. Mr. RICHMOND said that he used perhaps a little more, but not very much. In the case of homogenized milk he thought that the Gottlieb method had advantages over the Werner-Schmid. I t took much less trouble, and the fat came out more cleanly and easily. With regard to the addition of alkali, he thought that sodium phosphate was used quite as often as sodium carbonate. Mr. HEHNER said that he had been consulted as regards the manufacture of dried milk both on the Continent and in America, and had never known any alkali but sodium carbonate. to be used.Mr. RICHMOND said that he knew that sodium phosphate had been used in England. With regard to the milk-sugar, having obtained distinct evidence from the change in temperature on mixing with water that there was some water of hydration, he thought it only right to include this. He did not, of course, suggest that these analyses ought necessarily to add up exactly to 100. As Mr. Hehner had mentioned, there was citric acid present, which, however, was partly compen-224 THE ANALYST. sated for by the fact that a portion of the phosphoric acid was included both in the casein and in the ash. Analyses of dried milk should, however, not be 14 to 24 per cent. low. Mr. FAIRLEY asked what, if the Adams process failed when the milk was spread over a oertain surface, was to prevent its being spread over a larger surface by dilution if necessary, and whether, in the Werner-Schmid determinations, any attempt had been made to find out whether any fixed residue was contained in the fat, because, after all, these discrepancies were minute, and a very small amount of impalpable mineral matter in the fat might make all the difference. Mr. RICHMOND said that there was really nothing to prevent the milk being spread over a larger surface. The fat in the Werner-Schmid determinations was tested to see if it contained anything insoluble in petroleum ether, but was not ignited to see if there was any non-volatile matter. Note.-In the Journal of Analytical Chemistry, vol. iv., part i., I stated : t 6 From the data given by Adams (ANALYST, x., 46), I calculate that on paper the thickness of the film of milk solids is less than 0.0002 inch, and from my own measurements af kieselguhr, I conclude that the film dried up on this medium is about 0.0000005 inch. . . ."I % @ % @ % H. D. R.