Issuer! .hmimry 11, 1909. U. S. DEPARTMENT OF AGRICULTURE, . BUREAU OF ANIMAL INDUSTRY. Rui.irTiN 111. ^ =^- A. !>. Mhl.VIN. CHIEI OP li.-m-.Ai-. ~^= 1 |Bj CHEMICAL AND PHYSICAL STUDY OF " THE LARGE AND SMALL FAT _J GLOBULES IN COWS' MILK. BY R. H. SHAW, Assistant Dairyman , I~>airy Division, AND C. H. ECKLES. I'rofftsor of Dairy /fitslmmfry, (,'rin'crsi/r of Missouri. OF CALIFC LOS ANGELES SEP 23 1952 LIBRARY GOVT. PUBS. ROOM WASHINGTON: GOVERNMENT PRINTING OFFICE. (91 Issued January 11, 1909. U. S. DEPARTMENT OF AGRICULTURE, BUREAU OF ANIMAL INDUSTRY. BULLETIN 111. A. D. iMELVIN, CHIEF OF BUREAU. A CHEMICAL AND PHYSICAL STUDY OF THE LARGE AND SMALL FAT GLOBULES IN COWS' MILK. BY R. H. SHAW, Assistant Dairyman, Dairy Division, AND C. H. ECKLES, Professor of Dairy Husbandry, University of Missouri. WASHINGTON : GOVERNMHNT PRINTING OFFICE. 1909. THE BUREAU OF ANIMAL INDUSTRY. Cl.i.f: \. D. MKI.VI.V Assist'int Chi ft': \. M. FAHRIXGTON. ChufCl-rlc: CHAKIK- < . ( AKKOII.. Biochfinic I>iiisi'>,,: M. DORSET, chief; JAMES A. EMERY, assistant chief . hri.-iiim: Ki>. H. WKHSTKIC, chief; C. B. LANK, assistant chief. Ihi-iifion: KICK P. STEDDOM, chief; MORRIS WOODEN, R.A.RAMSAY, and AI.HERT E. BKHXKK. associate chiefs. ratholoiiiciil I>ii iftion : JOHN R. MOHLER, chief; HENRY J. WASHBURX, assistant chief. Quarantine Itirixinn: UICHAKD \V. IIiCKMAN, chief. Zoological Dirision: B. H. RANSOM, chief. Experiment Station: E. C. SCHROEDER, superintendent; W. E. COTTON, assistant. Animal Husbandman: GEORGE M. ROMMEL. Editor: JAMES M. PICKKS- DAIRY DIVISION. Chief: Ed. H. \\VliM.-r. Assistant Chief: C. B. Lane. Librarian: Miss C. B. Sherman. DAIRY FARMING INVESTIGATION'S. Assistant in charge, B. H. Rawl; assistant, Duncan Stuart. Dairv buildings: J. A. Conover; Architect, K. E. Parks; ventilation experiments, C. R. Potteiger. Herdbook work: Helmer Rabild and William Hart Dexter. Southern dairying: S. E. Barnes, J. E. Dorman, J. T. Eaton, II. P. Lykes, J. H. McClain, A. K.'Risser, H. R. Welch, and T. R. Woodward. DAIRY PRODUCTS INVESTIGATIONS. Assistant in charge, L. A. Rogers. Butter investigations, Albert Lea, Minn., and Washington, D. ('.: Chemist, W. X. Berg; bacteriorologist, S. H. Ayers. Swiss cheese investigations, Albert Lea, Minn.: In charge, C. F. Doane; assistant, T. W. Issajeff. Cheese investigations, Madison, Wis. : Chemist, S. K. Suzuki; bacteriologist, Alfred Larson; cheese maker, J. W. Moore. Cheese investigations, Storrs, Conn.: Mycologist, Charles Thorn; chemist, Arthur W. Dox. Milk secretion investigations, Columbia, Mo.: Chemist, R. II. Shaw; assistants, J. O. Halverson, A. E. Perkins, and G. C. Payne. DAIRY MANUFACTURING INVESTIGATIONS. Assistant in charge, B. D. White; assistant, S. C. Thompson. Creamery records, Albert Lea, Minn.: Creamery practice, John L. Sherk; assistants (rnllaborators), P. W. Noble and J. D. Burk. Creamery practice investigations: J. C. Joslin, Robert McAdam, F. L. Odell, J. C. Winkjer, ana Thomas Corneliuson. Market investigation-: Xe\v York City, C. W. Fryhofer; Chicago, H. J. Credicott; San Francisco, C. L. Mitchell. MARKET MILK INVESTIGATIONS. Assistant chief of division in charge, awistants, Lee II. P. Maynard, Ivan C. Weld, ami George M. Whitaker. RENOVATED BTTTER INSPECTION. Chief in-spec-tor, M. W. Lang. Chicago; as.-istani. Levi Wells, \e\v York. LETTER OF TRANSMITTAL. U. S. DEPARTMENT OF AGRICULTURE, BUREAU OF ANIMAL INDUSTRY, Washington, D. C., October 20, 1908. SIR: I have the honor to transmit herewith, anil to recommend for publication in the bulletin series of this Bureau, the accompanying manuscript of an article entitled "A Chemical and Physical Study of the Large and Small Fat Globules in Cows' Milk," by R. H. Shaw, of the Dairy Division of this Bureau, and C. H. Eckles, of the Missouri Agricultural Experiment Station. This work is a preliminary part of a comprehensive investigation being conducted at the Missouri station by cooperation between that station and the Dairy Division concerning the effects of period of lactation, feed, and other factors on the chemical composition of milk. At the request of the authors acknowledgment is made to Messrs. J. O. Ilalverson and G. C. Payne for assistance in the chemical and microscopical work. . Respectfully, A. D. MELVIN, Chief of Bu reau . Hon. JAMES WILSON, Secretary of Agriculture. 3 CONTENTS. Page. Introduction 5 Previous investigations 5 Experimental work 8 Source of the samples 8 Method of separation 8 Preparation of fat samples for analysis 10 Analytical methods 10 Color determinations 11 Microscopical work 11 Results of experiments 14 Conclusion... 16 ILLUSTRATIONS. Page. FIG. 1 . Showing method of counting globules 12 2. Showing the average relative size of large and small fat globules 13 4 A CHEMICAL AND PHYSICAL STUDY OF THE LARGE AND SMALL FAT GLOBULES IN COWS' MILK. INTRODUCTION. In planning a long investigation on the effect of the period of lactation, feed, and other factors on the chemical composition of milk it was decided that the most practical way of extracting the fat from the milk was by passing it through a small separator and churning the cream. In this method the tendency is for the smaller globules to escape in t>he skim milk and buttermilk; hence it became necessary to ascertain whether or not these small globules which in part escape differ in chemical composition from the large globules, and if such a difference exists, to determine just what it is. Investigations on these points are not lacking, and had these inves- tigations been sufficiently extensive, or had they agreed on the prin- cipal points, the work herein presented would have been unnecessary. However, such is not the case. There is almost a total lack of agree- ment in the work hitherto published, as will be seen later on. When the present investigation was taken up it was the intention to cover only those points necessary to validate certain determinations involved in the long investigation mentioned above. Later it was decided to enlarge it to some extent and publish it as a contribution to the knowledge of the chemistry of butterfat. PREVIOUS INVESTIGATIONS. M tiller (1867) observed that fat extracted from skimmed milk is waxlike, and also that in the analysis of butter where 100 grams or more of fat-free substance is obtained the last particles of fat extracted have more the properties of wax than of butterfat. Schroeder 6 in 1872 published some investigations showing that when milk is fractionally creamed the butterfat obtained from the "Mailer, Alexander. Chemische untermichungen auf dem gebiete der milch- wirthachaft. Die LandwirthHchaftlichen Verauchfl-Stationen. Band 9, pp. 3G4-39G. See p. 390. Chemnitz, 1807. ftSchrueder, G. Beitrage x.ur kenntnifwdea fettgehults der milch. Milch Zeitung, No. 28. pp. 325-327, 1872. 5 6 LARGK AM) SMALL FAT GLOHTLKS IN COWS' MILK. fir>t fraction has a lower melting point and lower specific gravity than the Initterfat obtained from the last fraction. He also stated that the hitter contained no butyric acid. Klusemann in 1893 conducted an investigation with milk from individual cows. lie separated the large and small fat globules by pa>sin_ir t'he whole milk, at a temperature of 5 to 12 C., through a De Laval separator running at low speed. The cream contained the large globules. The skim milk was then warmed somewhat and again passed through the separator running more rapidly. The cream in this instance contained the medium-sized globules. The small glob- ules were obtained by heating the last skim milk to 36 to 38 C. and again passing it through the separator, this time running very rap- idly; the cream was churned in a small dash churn, and the resulting butter was melted and filtered by means of a jacketed filter. He determined the melting and freezing points, the water-insoluble fatty acids, the melting and freezing points of the insoluble fatty acids, the iodin absorption number, the volatile acids, and the specific gravity. In his conclusions he states that the percentage of the insoluble fatty acids, and the melting point and freezing point of the butterfat, and of the insoluble fatty acids are higher in fat from small-globule cream than in fat from large-globule cream, the reverse being true of the percentage of volatile acids and of the iodin number. Gutzeit 6 (1895) published an investigation on the same subject. He chose mixed milk from a herd of 150 cows. He separated the large and small globules as follows: He used 100 liters of milk, 20 liters of which was allowed to rise for six hours in a pan immersed in cold water. The resulting cream contained the large fat globules. The remaining 80 liters was then run through a separator, the cream discarded, and the skim milk coagulated. The whey was then run through a separator, and the resulting whey cream contained the small globules. The two creams were churned separately, and the resulting butters were heated for a time at 60 C. and filtered. He determined the specific gravity, the melting point, the refractive index, insoluble fatty acids, volatile fatty acids, saponification number, and iodin absorption number, and concludes that in homo- geneous milk the fat globules of all sizes have the same chemical- physical composition. a Klusemann, Erich. Die zusammensetzung und die beschaffenheit der ans den grossen und den kleinen fettkugelchen der kuhmilch gewonnenen butter. Inaugural dissertation. Leipzig, 1893. bGutzeit, Ernst. Die schwankungen der mittleren groese der fettkugelchen in der kuhmilch nach laktation, futterung und rasse, sowie fiber den physikalischen und chemischen unterechied der groesten und kleinsten fettkugelchen. Land- virthschaftliche Jahrbucher, Bans 24, pp. 539-668. Berlin, 1895. PREVIOUS INVESTIGATIONS. 7 Lemus a (1902) separated the large and small globules by inter- rupted milking. The first liter of milk drawn he took to represent the small globules and the last fraction of about one-half liter drawn the large globules. After reserving a sample for fat determination and microscopical examination, he extracted the fat according to the method of Soxhlet by shaking with ether after having first made it alkaline with potassium hydroxid. The ethereal layer was then removed and dried at 100 C. to a constant weight. He made some of the same determinations as did Klusemann and Gutzeit, but in his conclusions refers only to the volatile acids and the oleic acid. He states that in the majority of cases the smaller fat globules contained more oleic acid and less volatile acids than did the larger globules. In the three more recent investigations, that is, those by Kluse- mann, Gutzeit, and Lemus, it is seen that no two draw the same conclusions. Gutzeit claims that there is no difference, chemically or physically, between the large and small fat globules in the same milk. Klusemann and Lemus agree in claiming that there is a differ- ence, but disagree on some of the points of difference. Klusemann worked on the milk of five individual cows and on the mixed milk of seven. A study of his results shows that in every case the percentage of insoluble acids in small-globule fat is higher than in the large-globule fat, the difference being from about 0.3 per cent to about 3 per cent; the melting point of the butterfat was higher in the small-globule fat than in the large, the variation being from 0.5 to 2 C. ; the volatile fatty acids in terms of Reichert-Meissl number sho\\vd a decrease in the small-globule fat in six cases and an increase in two cases, the variation being from about 1 c. c. to about 4.5 c. c.; the iodin number in three cases showed practically no difference, in three cases it showed a decrease in the small-globule fat and in one an increase, the largest difference being about 2.5 per cent. Granting that no criticism can be made of his methods and work, it would seem that with the limited number of animals used his conclusions even on the insoluble fatty acids and melting point are unwarranted, and certainly not justified in the case of the other constants. In Lemus's investigation it may be seen that the separation of the large and small globules is not positive. In fact, instead of the relative size of the fat globules being invariably larger in the stripper milk than in the foremilk, the reverse is true in some cases. In 10 cases the iodin number was smaller in the large-globule fat than in the small-globule fat, and in 3 cases it was larger, the difference ranging "Lemufl, Woldemar. t'obor die rheniinrhe IM>JH -haffcnluMt de in don grown mid in den klcinen milchkugelchen enthaltenen fette. Inaugural dissertation. Leipzig, 1902. 8 LARUK ANI SMALL FAT GLOBULES IN COWS* MILK. from 0.11 to 3.o4 JMT cent. With the volatile acids the Reichert- Meissl number was smaller in 16 cases in the large-globule fat than in the small-globule fat, and in 6 cases the reverse obtained, the differ- ences ranging from 0.4n to 13.85 per cent. It is also doubtful whether the fat remained unchanged in the preparation of the samples, where a prolonged heating at 100 C. was necessary to eliminate the ether." Moreover, it would seem that his determinations show the difference between the fat in the foremilk and that in the milk secreted during the process of milking, rather than that between the large and small fat globules in the same milk. EXPERIMENTAL WORK. For our investigation animals were selected which exhibited a wide range of breed, age, and period of lactation. In the earlier part of the investigation the chemical and microscopical work was limited to determinations of the relative size of the fat globules, the iodin num- ber, the Reichert-Meissl number, the Koetstorfer number, and the refractive index. Later it was extended to include the specific gravity, the melting point, the Hehner number, the color, and the percentage of different-sized fat globules. Table 1 gives the data concerning the animals. In Tables 2, 3, and 4 will be found the results of the chemical and microscopical work. SOURCE OF THE SAMPLES. The cows selected to supply the samples represented four breeds Jersey, Holstein-Friesian, Ayrshire, and Shorthorn and were mem- bers of the Missouri Agricultural College herd. Groups supplying the samples were arranged with a view of having certain samples represent milk from cows in the early part of the period of lactation, while others represented milk produced in the latter part of the period. The table shows the number of days in milk, the breed, the age, the date at which samples were taken, and the average percentage of fat in the milk of each co'w. The sample of milk taken was the total product for one day of the group of cows shown in the table. METHOD OF SEPARATION. The first method tried consisted in allowing the milk to stand at room temperature for about two hours until a portion of the cream had come to the^ surface, when it was removed by drawing the remain- der of the milk from a faucet in the bottom of the can. The partially a According to C. A. Brown (Annual Report of the Pennsylvania State College, 1899-1900, p. 208), a temperature much above 50 C. will very soon alter the compo- sition of butterfat. Rutterfat kept at 50 C. for two days showed a loss of over one unit in iodin number. EXPERIMENTAL WORK. 9 skimmed milk secured in this way was again placed in the can and allowed to stand about twelve hours, when the second portion of the cream was removed as before. This skim milk, which still contained approximately 0.3 per cent of fat, was then run through a centrifugal separator at a temperature of 32 C. The cream thus secured contained the smallest fat globules which it is possible to obtain by the centrifugal machine. The cream secured from the first skimming contained the larger fat globules, as they reach the surface first under the influence of gravity, but it also contained some small fat globules. In order to eliminate these this cream was placed in the bottom of a deep, narrow can, which was then filled with the skim milk from the final separation with the centrifugal separator. When most of the cream had again risen it was skimmed off and retained for the sample of the large fat globules. The above method was found to be satisfactory, but equally good results were secured, at a considerable saving of time, by using the centrifugal separator. After several preliminary trials the following plan was adopted: The milk was first run through the centrifugal separator at a tem- perature of 30 to 32 C. and the speed of the machine so regulated by trial that approximately one-half of the fat in the original milk was taken out as cream and one-half remained in the skim milk. It was found that with a hand-power separator, the normal speed of which was 60 revolutions of the crank per minute, this result was obtained when the machine was operated at the rate of 15 to 25 revo- lutions a minute, depending upon the character of the milk separated. The cream secured from this separation contained the larger fat glob- ules, but also a large number of medium sized and some small ones, which were removed as described later. The skim milk secured by this separation contained approximately one-half the fat originally found in the milk, and was again separated at a higher speed from 29 to 40 revolutions of the crank to the minute. The speed was adjusted after trial with each sample of milk so that about 0.2 per cent of fat remained in the skim milk. The cream from this second separation was rejected, as it contained medium sized and some large fat globules which had escaped the first separation. The skim milk containing approximately 0.2 per cent of fat was then heated to 38 C. and again run through the centrifugal separator at a speed of from 65 to 70 revolutions of the crank to the minute. The cream from this separation contained the smallest fat globules which it is possible to secure with a centrifugal separator. It is impossible to obtain all of the fat by any mechanical method of separation. The fat remaining in the skim milk after this final separation was found to be from 0.025 to 0.004 per cent, as shown by the Babcock method, using the double-necked Wagner testing bottle. 10 LARCJK AN1> S.MAI. I. I AT til.OBULES IN COWS 5 MILK. Tho cream from tho first separation containing tho largest fat . globules was added l<> the skim milk from the third and run through the separator again at the speed used in the first separation, to eliminate the smaller fat globules which had remained with the large. The cream from this separation was again mixed with a second lot of the same skim milk from the third separation and the process repeated, this cream being taken as the final sample containing the largest fat globules. The two lots of cream were then churned by shaking in a closed jar as long as any butter could be accumulated. PREPARATION OF FAT SAMPLES FOR ANALYSIS. The butter was immediately melted on a steam bath and allowed to remain in this condition until the curd and water had settled. Special care was taken at this point that the temperature did not rise above 50 or 55 C. and also that it remained, even at this tempera- ture, no longer than was necessary. It was then filtered through a paper filter kept warm by an electrical device, and preserved in corked bottles which were protected from light in a refrigerator until the butter was analyzed. ANALYTICAL METHODS. The methods of the Association of Official Agricultural Chemists were employed whenever possible. Duplicate determinations were made in all cases and in some triplicate or even quadruplicate. Only the briefest description of the official methods used is given below. For a detailed description reference may be made to Bulletin 107 of the Bureau of Chemistry, United States Department of Agriculture. Specific gravity. The specific gravity was determined in small pyk- nometer bottles at the temperature of boiling water. Mdting point. The melting point was determined according to Wiley's method, by placing a disk of the fat in a large test tube con- taining boiled distilled water and boiled alcohol which had been cooled, the tube being placed in a beaker of water which was slowly heated, and noting, by means of a thermometer graduated to 0.1 C., the temperature at which the disk assumed the form of a sphere. Refractive index. The refractive index was determined with a Zciss-Abbe refractometer which had been standardized with distilled water. The fat was kept at a constant temperature above its melt- ing point during the determination by means of a current of warm water circulating through the instrument. The reading was reduced later by means of a factor to 25 C. Volatile adds. The method of Reichert, modified by Mei>sl. was employed in the estimation of the volatile acids, and the results are given as Reichert-Meissl numbers. In saponifying, use was made of the Lefl'man-Beam method, in which the fat is saponified in a flask METHODS OF ANALYSIS. 11 over a naked flame with a mixture of a strong aqueous solution of caustic soda and pure glycerol. The soap obtained in this way is decomposed by sulphuric acid and the liberated volatile acids are distilled over with the steam and titrated with decinormal barium hydroxid solution. Saponification value. The saponification value, or Koettstorfer number, was determined by the regular official method. The fat is saponified in a flask on a steam bath with an alcoholic potash solution, using a long glass tube as reflux condenser. Iodin absorption number. The method of Hubl was employed in this case. The fat is dissolved in chloroform and subjected to the action of a mixture of an alcoholic solution of iodin and of mercuric chlorid in a dark place for three hours, at the end of which time the unabsorbed iodin is determined by titrating with standard sodium thiosuphate solution. The percentage of iodin absorbed by the fat is expressed in the results as the iodin number. Insoluble jatty acids. For this determination the official method, or that of Hehner, as it is called, was employed. It was found, how- ever, that better results were obtained by increasing the amount of' fat to 10 grams, as suggested by Brown." The results are given as Hehner numbers. COLOR DETERMINATIONS. The color determinations were made with the Lovibond tintom- eter, using the standard color glasses. While this instrument is not admirably adapted to the work with milk and cream, so far as the authors' knowledge goes, it is the best that is available. With the melted fat where the light is transmitted through a consider- able layer the determinations can be made quite accurately, but with the opaque milk, cream, and butter, where the light must be reflected from the surface, very slight differences in color can not be deter- mined. The results are given in terms of the numbers assigned to the standard color glasses. MICROSCOPICAL WORK. In the microscopical part of the work the method devised by Dr. S. M. Babcock'' was used. The milk is diluted with distilled water to fifty times and the cream to one hundred times its volume. Fine capillary tubes are drawn out from larger tubes, care being taken that the resulting tubes are round in cross section. They are then broken into pieces 2 to 3 centimeters in length, and by means of forceps each tube is filled by immersing one end in the diluted cream or milk. The tube fills instantly by capillary attraction. Report of Pennsylvania State College. 1899-1900, p. 211. 6 Fourth Annual Report of the New York Agricultural Experiment Station, 1885. 12 LARGE AND SMALL FAT GLOBULES IN COWS' MILK. The two ends are then sealed with vaseline. Three such tubes are u-t'd for each sample, and these are laid parallel on a slide. A drop of glycerin is then placed over the tubes and a cover glass applied. The slide thus prepared is allowed to remain on a leveled slab for half an hour, the purpose of this being to allow the globules to rise to the top of the capillary tubes. At the end of thirty minutes the slide is placed under a microscope provided with an ocular micrometer and a mechanical stage. The number of globules in 50 divisions of the micrometer scale are counted in three places in each tube, the inter- nal diameter of the tube at each place of counting being first deter- Fio. 1. Showing method of counting globules. mined by throwing the ocular micrometer scale across the tube at right angles and carefully counting the divisions within the two walls. The total number of globules in the 50 divisions is recorded, as well as the number under one division in diameter ; also the num- ber between one and two divisions in diameter, and the number having a diameter greater than two divisions. A 1-inch ocular and one-sixth inch objective are employed, and with this combination the value of each division of the ocular micrometer with the instrument used is 0.00258 millimeter. In this way nine determinations are made for each sample. METHOD OF COUNTING GLOBULES. 13 * In order to compare results, the number of globules which would be found in a standard tube 100 divisions in diameter and 50 divisions in height is calculated from the number in each observed tube and the average taken. In making this calculation the formula ~ is d used." With the ocular micrometer used hi this work the standard tube would have a volume of 0.006744 cubic millimeter, and if the milk is diluted to fifty times its volume, 0.006744 divided by 50, or 0.00013488 cubic millimeter, would be the volume of the original milk contained in the standard tube. In the case of cream where it is diluted to one hundred tunes its volume one-half this figure represents the volume of original cream in the standard tube. To find the number of gl.obules present in 0.0001 cubic millimeter it is only necessary to FIG. 2. Showing the average relative size of large and small fat globules. multiply the number in the standard tube by the factor 0.7414, obtained by dividing 0.0001 by 0.00013488. The relative size is found by dividing the percentage of fat by the number of globules in 0.0001 cubic millimeter of the milk or cream. To avoid fractions, the figure obtained in this way is multiplied by 10,000. An example illustrating the method of calculating may be given. A sample of milk containing 4 per cent of fat is chosen. Let 30 repre- sent the number of globules counted in 50 divisions of a tube whose The volume of two cylinders having the same height but different diameter* are to each other as the squares of their diameters. If we let n equal the number of glob- ules in the observed tube, n" the number in the standard tube, d the diameter of the observed tube, and 100 the diameter of the standard tube, then n:n": : SMAI.I. FAT CI.onrLKS IN COWS* MILK. internal diameter has been found to be 40 divisions. Applying the formula M ". we have - - =187, or the number of globules f or 1,600 which would be found in a standard tube whose length is 50 divisions and whose diameter is 100 divisions. Multiplying 187 by the factor 0.7414, we have 138, which would be the number of globules in 0.0001 cubic millimeter. Dividing 4 by 138 and multiplying by 10,000, we have 290 as the relative size of the fat globules in the example taken. As Doctor Babcock states, the relative size is not strictly accurate, since percentages by weight have been confused with volume. To be strictly accurate we must multiply the relative size by a factor obtained by dividing the specific gravity of the milk by the specific gravity of the butterfat. However, since relative values only are required in our work, it was decided that the relative size would answer the purpose, so our calculations have stopped at that point and the results are expressed in terms of relative size. RESULTS OF EXPERIMENTS. The following tables present data as to source of samples and results of the experiments: TABLE I. Data concerning source of samples. Sample No. Date. Breed of cow. Date of last calving. Time in milk. Age of cow. Daily yield of milk when sample was taken. A yerage fat content. 1906 Ayrshire 1906. July 10 Days. 77 Yrs.Mos. 2 5 Pounds. 21.5 Per cent. 4.0 1 Sept. 25 ..do.. Aug. 13 43 3 8 28.9 a9 do July 29 58 3 2 25. 1 4.0 Jersey May 8 1.51 9 2 27.9 4.5 2 Oct 3 ...do Aug. 19 45 9 1 30.3 4. 1 ...do... May 8 148 9 22.1 5.0 ....do Dec. 15 298 8 11 10.3 5.6 . .do Jan. 26 256 2 7 11. 1 5.2 ...do... Feb. 7 244 3 3 10.3 5.8 3 Oct. 9 ...do... Dec. 6 307 8 11 7.4 6.7 .. .do . . . Jan. 25 257 4 7 7.5 5.5 do Feb. 6 345 9 9. i 5.0 Holstein. .. Apr. 18 193 6 4 32.2 a2 do .. . . Apr. 29 182 4 6 20.6 3.0 4 Oct. 28 ...do... May 13 168 4 5 31.9 3.4 ....do Feb. 11 259 4 4 2ai ao 1908. (Shorthorn. . . 1907. Sept. 30 109 4 8 19.7 a9 5 Jan. 17 1 do. . Oct. 13 96 6 3 18.9 4.1 i do Sept. 30 112 4 8 20.5 a9 6 Jan. 20 \....do Oct. 13 99 6 3 19.2 4. 1 Ayrshire Dec. 27 27 3 9 31.9 a9 7 Jan. 23 ..."do .. . .. Sept. 27 118 5 1 22. 1 a 75 | Holstein. .. July 17 192 5 10 24.2 a4 8 Jan. 26 ^. ...do May 31 240 5 9 24.9 as 1 do July 20 190 4 2 23.8 a2 / do Aug. 13 182 7 8 31.2 a i \ do May 7 279 5 5 26.3 ao 1 Jersey June 16 250 6 6 18.5 6.48 Feb. 21 \ do '. June 12 254 4 5 26.1 5.86 RESULTS OF EXPERIMENTS. 15 TABLE 2. Results of microscopical icork. Sample Xo. Size of globules. Relative size of globules. Percentage of globules less than one division in diameter. Percentage of globules between one and two di- visions in diameter Percentage of globules over two di- visions in diameter. Small Ml Large. . 631 Small 91.6 Large. . 741 Small 112 3 Large.. 698 Small 51.7 4 Large.. 316 Small.. 21.7 58.3 40.4 1 3 1 Large 590 15 6 38.4 46.0 Small.. 66.8 45.9 49 3 4 8 6 Large. . . ........ 530 ias 39.3 46.9 Small.. 12.9 73.2 26 6 0.2 7 Large.. .... 703 31.6 61. 1 7.3 Small 44.2 57 8 41 9 0.3 8 Large. . 417 19.3 69 9 10 8 Small 42.3 61.2 38.5 0.3 9 Large.. 299 14 6 80 5.4 Small.. 63.5 67.0 32 3 07 10 Large... 691 7.3 70.9 21.8 TABLE 3. Results of chemical work. Sample size of globules. Specific gravity. Melting point.. Refrac- tive in- dex. Koettstor- fer num- ber. Reichert- Meissl number. lodin number. Hehner number. 1 2 3 4 6 7 8 9 10 ! Small 1.4609 .4609 .46C8 .4608 .4615 .4613 .4614 .4610 .4565 .4563 . 4,"64 .4562 . 4567 .4568 . 4575 .4574 . 4575 . 4573 . 4561 .4560 231.2 230.6 229.8 230.6 234.0 234.6 228.3 228.0 27.29 27. C6 29.76 29.13 25.69 25.49 25.69 26.28 30.91 30.55 31.33 32.29 32.86 32.83 33.37 3145 Large Small Large Small Large Small Large . Small a.. 0.9078 .9044 .9038 .9033 .9038 .9035 .9053 .9044 .9014 .9014 .9059 . MM 33.47 33.33 33.50 33.38 33.37 33.37 32.37 32.90 31.85 32.43 32.43 32.90 Large 232.4 231.0 233.3 229.6 226.6 226.3 227.6 227.9 233.8 234.9 23.92 29.73 29.72 30.20 30.54 31.25 34.56 34.39 35.42 35.44 27.62 87. 20 86.83 86.83 BB -M 87.16 88.53 88.35 85.73 85.50 F6.13 - v. Small <>.. Large 24.15 25.01 26.06 24.03 24.20 24.98 25.90 26.93 27.63 Small Large Small Large... Small Large . . Small Large Sample too small for complete analysis. TAW.K 4. Color of butter and bntterfat from large and small globules. Sample BllttT. ntitterfnt. No. Yellow. Red. Yellow. Red. f Small 15.5 .5 6 \ Large 15.5 .5 (Small... 15.5 .5 ' 1 Large... 15.5 .7 (Small 0.9 0.9 27.2 .1 UATge... .9 .9 27.2 .6 Small .8 .8 20.0 .5 s Large.. . .8 .8 20.0 .5 Small .7 .8 11.0 .5 9 Large. . . .7 .7 10.7 .3 Small 1.0 .8 22.0 .2 Largo .9 .9 21.0 .0 16 LAHGE AND SMALL FAT (il.OBULES IN rows' MILK. CONCLUSION. It is impossible by any known process to effect a complete separa- tion of the large and the small fat globules in milk. The best that can be expected is to secure two creams from the milk, one of which contains a large percentage of large globules and the other a large percentage of small globules, or, in other words, one in which the relative size of the fat globules is comparatively large and another in which it is comparatively small. That such has been obtained in this investigation will be shown by Table 2. A study of Table 3 will fail to reveal any considerable differences between the fat from the large-globule cream and that from the small-globule cream. The differences in all cases are small and in most cases well within the limits of legitimate experimental error. Table 4 also shows little or no variation in color. The investigation, of course, could be extended by introducing more animals and more samples, but the range covered by the animals selected and the number of determinations made would seem to be sufficient to warrant the conclusion that in homogeneous milk the large and small fat globules have identical chemical and physical composition. O UC SOUTHERN REGIONAL LIBRARY FACILITY A 001 102 567 3